Appraisal Clinical Practice Guideline 197 Appraisal of Clinical Practice Guideline: International Consensus Statement on the diagnosis, multidisciplinary management and lifelong care of individuals with achondroplasia Date of latest update: 2022. Date of next update: Not stated. ear-nose-throat conditions (R107 to 109). Some individuals may Patient group: Achondroplasia (skeletal dysplasia with short stat- require long-term respiratory support such as continuous positive ure). Intended audience: Clinicians and consumers. Additional airway pressure (R103). Psychosocial factors should be supported versions: Nil. Funded by: BioMarin Pharmaceutical Inc sponsored (R129 to 136). an initial meeting. Expert working group: Fifty-five experts and patient representatives from 16 countries and five continents. Commentary: Achondroplasia is a genetic disorder that affects bone Consultation with: Consumers with achondroplasia. Approved by: growth, resulting in short stature and other musculoskeletal, neuro- International Achondroplasia Consensus Statement Group. logical and respiratory characteristics. Physiotherapists need to Location: 10.1038/s41574-021-00595-x. Description: Fifty-five in- understand the unique and multifactorial challenges faced by in- ternational experts (including physiotherapists) drafted and dividuals with achondroplasia and be aware of the most up-to-date responded to a two-round Delphi process. Consensus was reached guidelines for their care. for 160 statements, which were collated into 136 recommendations (R) for diagnosis, multidisciplinary management and lifelong care. The strengths of this guideline are that it includes an excellent Assessment: Multidisciplinary follow-up, including physiotherapy overview of current knowledge and comprehensive recommenda- (R47) should be provided for infants, children and adults (R25, 58, tions for the care of individuals with achondroplasia throughout their 67). Achondroplasia-specific charts should be used to assess height, lifespan, covering a wide range of physical, social and psychological weight and head circumference (R26, 44, 45, 71), gross and fine factors. The benefit for physiotherapy practice is that the guideline motor development (R27) and condition-specific adaptive move- provides a clear understanding of the role of the physiotherapist in ment strategies (R46). Individuals should be monitored for devel- the inter-professional team, with specific recommendations for opmental delay (R27, 122, 123), neurological function (R34), assessment, intervention and lifestyle management to help manage cervicomedullary compression (R29), sleep apnoea (R30, 55, 76, 98), acute and chronic musculoskeletal, orthopaedic, neurological, devel- hearing loss (R31, 78, 108), otitis media (R32, 53), thoracolumbar opmental and respiratory issues commonly experienced by these kyphosis (R60, 86), spinal stenosis (R69, 74, 75, 85), limb deformities individuals. A limitation of this guideline is that it does not provide (R90 to 92) and body weight/composition (R127). Common specific details on how to implement the recommendations in clinical specialist referrals include: magnetic resonance imaging (R35, 89), practice, which may pose a challenge for physiotherapists. Moreover, polysomnography (R30, 98), neurosurgical (R42), orthopaedic (R60 the guideline’s recommendations are based on available evidence and to 62, 86) or respiratory evaluation (R103, 104). Intervention: Par- expert consensus, which means that some recommendations may not ents should receive advice on positioning/handling (R28), avoiding be applicable to all individuals with achondroplasia. The guideline early sitting to prevent fixed thoracolumbar kyphosis (R48), avoid- could have benefited from including more details on the potential ing incidental positional death in car seats (R48), and preventing barriers and facilitators to implementing the recommendations and trips and falls (R59). Therapy should address developmental delays strategies to overcome these barriers. Overall, this guideline provides (R57) and provide adaptive equipment, mobility devices and envi- clear and valuable insights for physiotherapists and other healthcare ronmental modifications to maximise independence in home, professionals on best-practice care for individuals with achondro- school, work and community settings (R49, 70, 84, 123). Exercise plasia across their lifespan and how to help improve the health and and physical activity should be prescribed to maintain strength, quality of life for this unique population. body weight and flexibility (R51, 56, 72, 125). Chronic pain (R50, 63, 68, 79, 119), especially back (R74, 120) or knee pain (R94) and Provenance: Invited. Not peer reviewed. fatigue (R63) should be regularly monitored and managed. Post- Leanne M Johnston surgical management may be required for orthopaedic (R87, 88, 92, 96, 97, 121), neurological (R99), respiratory (R100 to 102) or The University of Queensland, Australia https://doi.org/10.1016/j.jphys.2023.05.006 1836-9553/Crown Copyright © 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 196 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Appraisal of Clinical Practice Guideline: National Institute for Health and Care Excellence (NICE) Clinical Practice Guideline for Osteoarthritis in over 16s: diagnosis and management Date of latest update: 19 October 2022. Date of next update: Not dose for the shortest possible time and use should be regularly reviewed stated. Patient group: People aged . 16 years with osteoarthritis. to monitor adverse events and need. Paracetamol or weak opioids Intended audience: Healthcare professionals; commissioners of health should not be routinely offered. Intra-articular corticosteroid injections and social care services; people with osteoarthritis, their families and can be considered for short-term pain relief if other pharmacological carers; and researchers with an interest in osteoarthritis. Additional management is ineffective/unsuitable. Follow-up timing of patients with versions: Updates and replaces NICE guideline CG177 (February 2014). osteoarthritis should be individualised. Referral for hip, knee or shoulder Funded by: The Department of Health and Social Care, UK. Expert replacement should be considered if joint symptoms are substantially working group: Guideline Committee with 15 members including a impacting quality of life and non-surgical management is ineffective. rheumatologist, general practitioner, podiatrist, physiotherapists, or- thopaedic surgeon, lay members, topic advisor, cardiologist (chair) and Commentary: Most recommendations in this guideline are within four co-opted members: radiologist, acupuncturist, manual therapist physiotherapy’s scope of practice (24 of the 35 recommendations); hence, and clinical pharmacist. Consultation with: Stakeholders who regis- physiotherapists have an important role in the best-practice multidisci- tered with NICE after an open call for consultation. This included UK plinary care of people with osteoarthritis. Specifically, physiotherapists organisations for patients using services, carers and the public; UK or- are integral to implementing osteoarthritis recommendations for diag- ganisations representing practitioners; public sector providers and nosis, information, support, non-pharmacological management, follow- commissioners; organisations that fund or carry out research; and pri- up, and review and referral for joint replacement. The strengths of this vate, not-for-profit, voluntary providers of care or services. Approved guideline include that: there is a clearly described scope and purpose, by: Guideline Committee and NICE’s Guidance Executive. Location: including specifically described objectives, research questions (PICO https://www.nice.org.uk/guidance/ng226. Description: Developed by a format) and population; it was developed by a multidisciplinary com- Guideline Committee (GC) based on systematic review evidence and mittee including clinicians and patients; it has clearly defined target experience/opinion of the GC. Thirty-five recommendations are users; it has rigorous methods including a clear description of how rec- included, covering seven domains: diagnosis (n = 2 recommendations); ommendations were formulated, use of systematic literature reviews to information and support (n = 3); non-pharmacological management identify evidence, consideration of benefits and risks, explicit links be- (n = 11); pharmacological management (n = 10); follow-up and review tween recommendations and evidence made, and external review by key (n = 4); referral for joint replacement (n = 4); and arthroscopic pro- stakeholder groups; it has clearly presented, easily identifiable, specific cedures (n = 1). As per the NICE protocol, recommendation wording is and unambiguous recommendations; the GC conflicts of interest are used to reflect strength of recommendations, for example: ‘offer’ (or disclosed and addressed; and tools/resources to facilitate implementation ‘advise’) reflects a strong recommendation and the use of ‘consider’ are provided. Some identified weaknesses include that: there is no pro- reflects a recommendation for which the evidence is less certain. Rec- cedure/schedule for updating the guideline is provided; facilitators and ommendations emphasise that a clinical diagnosis of osteoarthritis barriers to implementation are not described in detail; monitoring and/or should be used rather than routine imaging. Osteoarthritis information auditing criteria are not presented; and strength of recommendations is should be tailored and provided in accessible formats. Patients should not explicitly provided. Based on the identified evidence gaps, 18 rec- be advised where to locate information about specific exercise types and ommendations for future research are provided, four of which are pre- the benefits/limitations of treatments. Patients who are overweight/ sented as key research priorities: clinical and cost-effectiveness of obese should be advised that weight loss improves quality of life, supervised exercise compared with unsupervised exercise; clinical and physical function and reduces pain, and that any amount of weight loss cost-effectiveness of devices (eg, insoles, braces, tape, splints) compared has benefits but losing 10% is likely more beneficial than 5%. Tailored with usual care for the management of foot/ankle osteoarthritis; clinical therapeutic exercise should be offered to all patients with osteoarthritis and cost-effectiveness of topical non-steroidal anti-inflammatory drugs and supervised therapeutic exercise considered. Combining therapeutic and topical capsaicin for osteoarthritis affected joints other than the knee; exercise with an education program and/or behaviour change support and clinical and cost-effectiveness of patient-initiated follow-up should be considered to support exercise adherence. Manual therapy compared with routine follow-up. (eg, manipulation, mobilisation, soft tissue techniques) should only be considered alongside therapeutic exercise for patients with hip or knee Provenance: Invited. Not peer reviewed. osteoarthritis, not as a standalone treatment. Acupuncture, dry needling or electrotherapy should not be offered as treatment. Insoles, braces, Rachel K Nelligan tape and splints should not be routinely offered. Walking aids can be The University of Melbourne, Australia considered for lower limb osteoarthritis. Pharmacological treatments should only be offered alongside non-pharmacological management https://doi.org/10.1016/j.jphys.2023.04.002 and to support exercise; they should be provided at the lowest effective 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 140 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Call for applications for membership of the Editorial Board The Editorial Board currently consists of 14 members: 10 Australian and four international. Applications are invited to fill the following Editorial Board vacancies beginning in 2024: two Australian. All incumbents are entitled to re-apply in the current round. Editorial Board members are given portfolios with substantial responsibilities. This might involve, for example, soliciting submissions and editing contributions for one of the Journal’s ‘Appraisal’ sections. Potential applicants who are not prepared to take on portfolio responsibilities should not apply. The initial term of office commences on 1 January 2024 and expires on 31 December 2026. Editorial Board members are entitled to renominate for a further two successive terms. Selection criteria: 1. a sustained depth and breadth of research experience 2. extensive experience in the review and publication of research 3. prior editorial board experience (highly desirable but not essential) 4. excellent communication skills 5. good working knowledge of the physiotherapy profession and an interest in its future 6. demonstrated international reputation in research relevant to physiotherapy. Eligibility criteria: 7. hold a PhD 8. be a physiotherapist member of the Australian Physiotherapy Association (APA) 9. be a financial member of the APA at the time of application. Responsibilities: contribute to the development of policies that guide the publication of the Journal participate in the activities of the Editorial Board as a voting member manage or co-manage one of the journal portfolios attend regular Editorial Board teleconferences and a two-day face-to-face meeting annually meet and liaise with other members of the Editorial Board and the Scientific Editor as required contribute to the mentoring of Journal of Physiotherapy Editorial Fellows undertake specific tasks from time to time to promote the standing of the Journal. To be considered, physiotherapists applying for positions must submit: 1. a cover letter addressing the numbered criteria, above 2. a brief CV (maximum 3 pages), which includes a clear explanation of the impact of any career interruption(s) over the last 5 years and/or any relative to opportunity considerations. Applicants will be assessed according to the selection criteria listed above, relative to opportunity, and with attention to diversity, equity and inclusion considerations. Applications close Friday, 6 October 2023 and should be directed to Marko Stechiwskyj at [email protected] The Journal of Physiotherapy promotes diversity, equity and inclusion and encourages eligible applicants of all backgrounds to apply. https://doi.org/10.1016/j.jphys.2023.05.019 1836-9553/
Journal of Physiotherapy 69 (2023) 193–194 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Clinimetrics: Childhood Bladder and Bowel Dysfunction Questionnaire Summary Description: The Childhood Bladder and Bowel Dysfunction group (primary, secondary and tertiary healthcare), whereas the Questionnaire (CBBDQ) evaluates the frequency of symptoms of combined childhood bladder and bowel dysfunction in the past Turkish and Brazilian versions were only examined in tertiary month in 5- to 12-year-old children.1 This parent-reported tool con- sists of ten items related to bladder dysfunction and eight items healthcare services targeted at primary urinary incontinence. Face related to bowel dysfunction. Parents respond to each item on a 5- and content validity have been established.1 Regarding structural point Likert scale ranging from 0 (never) to 4 ([almost] daily). The bladder subscale (range 0 to 40) and bowel subscale (range 0 to 32) validity, the CBBDQ is a multidimensional scale, with a two-factor can be used separately or combined into one total score for structure1 showing good internal consistency for the bladder sub- dysfunctional elimination syndrome in children (range 0 to 72); higher scores indicate higher severity. Translations are available in scales (Cronbach’s a, resp. Dutch, Turkish and Brazilian Portuguese Turkish,2 Brazilian Portuguese3 and Dutch.4 A Dutch and English version: 0.74, 0.71 and 0.91), and bowel subscales (Cronbach’s a, resp. version for children aged 12 to 18 years, corresponding to secondary Dutch, Turkish and Brazilian Portuguese version: 0.71, 0.85 and school-aged children, is currently under development. 0.96).1–3 Construct validity of the CBBDQ (total score) was evaluated Feasibility: The CBBDQ is published and available online.1,5 The using relevant subscales of the Dysfunctional Voiding and Inconti- CBBDQ is a brief and simple to use questionnaire that can be completed within 5 to 6 minutes (Dutch) or 5 to 10 minutes (Turk- nence Symptoms Score and the Pediatric Quality of Life inventory as ish).1,2 The items of the CBBDQ are based on conceptual frameworks external criteria.10,11 Convergent validity has been established, as (ICCS-recommendations and Rome IV criteria) and questionnaire development adheres to COnsensus-based Standards for the selection correlations were good (r = 0.64) for the relationship between the of health Measurement Instruments (COSMIN) standards.6–8 During the developmental phase, parents and professionals were involved in CBBDQ and Dysfunctional Voiding and Incontinence Symptoms Score, assessing item relevance and comprehensiveness and comprehensi- bility of the CBBDQ. Pilot testing resulted in adaptations (item whereas divergent validity was established by poor to fair correla- reduction), although no major problems were reported regarding comprehensibility, wording, time to complete and acceptability.1,9 No tions (r = –0.12 to –0.28) between the CBBDQ and three Pediatric training is needed to administer the instrument. Quality of Life inventory subscales, according to expected correlations.2 Reliability, validity and interpretability: Psychometric analyses for the original (Dutch) version were performed for a heterogeneous Reliability has been evaluated for the Turkish and Brazilian Por- tuguese versions of the CBBDQ (total score). Excellent test-retest reliability (time window 1 week) was found for the Turkish and Brazilian versions (ICC 0.97 and 0.94 (CI 0.85 to 0.94; p , 0.001), resp.).2,3 The measurement error for the total CBBDQ score (based on the standard error of measurement) is estimated to be 2.42 and, based on the Bland-Altman plot, the limits of agreement are –6.6 to 5.9 with a low bias of –0.1.3 No floor and ceiling effects have been reported2 and no data on responsiveness to change are available. The minimal detectable change is 6.1.2 Commentary Functional urinary incontinence, constipation and faecal inconti- approach and successful treatment. It is even suggested that nence frequently occur together in the same patient, which is re- the CBBDQ may stimulate collaboration among gastroenterolo- flected in the two subscales of the CBBDQ.12–14 To date, healthcare gists, urologists, psychologists and physiotherapist towards a professionals use a diverse range of (diagnostic) bladder and bowel unified and improved approach to childhood bladder and bowel questionnaires, which are often unpublished and tailored to their dysfunction.9 specific setting, often addressing only bladder or bowel dysfunctions and not always satisfying current developmental standards and not Further sound psychometric analyses are needed (and currently tested on psychometric qualities. The CBBDQ is the first combined under study) to demonstrate the responsiveness and interpretability bowel and bladder questionnaire to be completed by caregivers of (MIC). The CBBDQ needs to be incorporated into future (multidisci- children, with equal attention to bladder and bowel dysfunctions plinary) guidelines. (reflecting daily practice) and regardless of underlying origin or concomitant comorbidities.1,9 Moreover, the frequency grading in the Provenance: Invited. Not peer reviewed. answer categories seems to be more comprehensible in the CBBDQ compared with, for example, the Dysfunctional Voiding and Incon- Esther Bolsa and Emmylou Beekmana,b tinence Symptoms Score.3 a Research Centre for Autonomy and Participation of Persons with a Chronic Illness, Zuyd University of Applied Sciences, The Netherlands The CBBDQ enables physicians and (pelvic) physiotherapists to b Department of Family Medicine, CAPHRI School for Public Health and evaluate and monitor simultaneous assessment of bladder and Primary Care, Maastricht University Medical Centre, The Netherlands bowel dysfunctions in children, which contributes to accurate guidance of therapy (goals), early referral for a multidisciplinary https://doi.org/10.1016/j.jphys.2023.03.003 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
194 Appraisal Clinimetrics References 6. Hayms JS, et al. Gastroenterology. 2016;150:1456–1468. 7. Austin PF, et al. J Urol. 2014;191:1863–1865. 1. Van Engelenburg van Lonkhuyzen ML, et al. J Pediatr Gastroenterol Nutr. 2107;64: 8. COnsensus-based Standards for the selection of health Measurement Instruments 911–917. (COSMIN). Available from: https://www.cosmin.nl/. Accessed April 6, 2023. 2. Van Engelenburg-van Lonkhuyzen ML, et al. Turk J Gastroenterol. 2020;31:482–488. 9. Verhoeven MMR, et al. NVFB Bulletin. 2014;31:10–13. 3. Bernardes RP, et al. J Pediatr. 2020;S0021-7557(20)30246-1. 10. Akbal C, et al. J Urol. 2005;173:969–973. 4. Van Engelenburg-van Lonkhuyzen ML, et al. Fysiopraxis. 2017/2018:22–23. 11. Varni JW, et al. Med Care. 2001;39:800–812. 5. Childhood Bladder and Bowel Dysfunction Questionnaire 5-12 jaar. Available from: 12. Burgers R, et al. J Urol. 2013;189:1886–1891. 13. Van Engelenburg-van Lonkhuyzen ML, et al. Eur J Pediatr. 2017;176:207–216. https://meetinstrumentenzorg.nl/instrumenten/childhood-bladder-and-bowel- 14. Hyman PE, et al. J Pediatr Gastroenterol Nutr. 2017;64:847. dysfunction-questionnaire-5-12-jaar/. Accessed April 6, 2023.
Journal of Physiotherapy 69 (2023) 195 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Clinimetrics: The Prechtl General Movements Assessment Summary Description: General movements (GMs) are endogenously movements can be classified as normal, absent (no fidgety move- ments observed) or abnormal (fidgety movements are exaggerated in generated movement patterns that appear in early infancy and are amplitude and/or speed). General movements are assessed from closely linked with later childhood development;1 they are most ‘birds-eye’ video recordings (approximately 3 to 5 minutes in dura- reliably assessed using the Prechtl General Movements Assessment tion) of infants who are undressed, only wearing a diaper/nappy, (GMA).2,3 The GMA is most commonly used for infants who require care in neonatal intensive care units.4 This includes, but is not limited lying on their backs and moving spontaneously in an alert and calm state.2 The GMA requires certification and training from the to, extreme prematurity (born , 28 weeks), newborn encephalopa- General Movements Trust by undertaking a 3.5-day basic course (see thy, needing surgery in the newborn period or severe infection. These https://general-movements-trust.info/), which costs approximately infants may be at risk of developmental concerns, including cerebral palsy (CP).5 Cerebral palsy is seldom identified contemporaneously to Australian $1,650. brain injury, with 21% of cases of CP diagnosed before an infant is 6 months old.6 Clinimetric properties: The GMA is a reliable tool, particularly for The GMA assesses preterm GMs (preterm age , 37 weeks), assessing fidgety GMs. A systematic review reported reliability values writhing movements (term to 6 to 9 weeks corrected age) or fidgety of kappa 0.84 for writhing movements and 0.84 to 0.92 for fidgety movements (9 to 20 weeks corrected age). Preterm GMs and writhing GMs.7 The absence of fidgety GMs has strong predictive validity for CP, with 97% sensitivity and 89% specificity.3 The predictive validity of movements are similarly classified, although the former are typically writhing movements for CP is weaker, with 93% sensitivity and 59% larger amplitude and faster in speed. Normal writhing movements specificity.3 Given the good predictive validity for CP in early infancy, the GMA is key to driving down the age of CP diagnosis.5 are described as . small to moderate amplitude and slow to moderate speed. Typically, they are ellipsoid in form . (page 21).2 Atypical In terms of prediction for developmental delay other than CP, the writhing movements are described as poor repertoire (movements GMA has good association but weaker predictive validity, with later are monotonous), cramped synchronised (movements are rigid and cognitive, language and motor development.8 Atypical writhing muscles contract and relax simultaneously) or chaotic (abrupt, dis- movements predict cognitive impairment, with sensitivity 80% and specificity 41%.9 Absent or abnormal fidgety movements predict organised movements). Fidgety movements are . small movements of cognitive impairment, with variable sensitivity 21 to 70% and speci- moderate speed and variable acceleration, of neck, trunk and limbs, in all ficity 85 to 95%.8,9 directions, continual in the awake infant . (page 22).2 Fidgety Commentary The GMA is an increasingly common assessment for paediatric Provenance: Invited. Not peer reviewed physiotherapists to have in their toolbox. It is currently recommended Amanda KL Kwong and Alicia J Spittle Department of Physiotherapy, University of Melbourne, alongside brain neuroimaging and the Hammersmith Infant Neuro- Parkville, Australia logical Examination in clinical guidelines for the early diagnosis of high-risk CP.5 Use of the GMA alone to predict CP has reported high References rates of false-positive results in various studies and therefore should 1. Darsaklis V, et al. Dev Med Child Neurol. 2011;53:896–906. not be used in isolation for the diagnosis of CP.3 When the GMA is used 2. Einspieler C, et al. Clin Dev Med. 2004. 3. Kwong AK, et al. Dev Med Child Neurol. 2018;60:480–489. with the Hammersmith Infant Neurological Examination and brain 4. Allinson L, et al. J Paediatr Child Health. 2017;53:578–584. 5. Novak I, et al. JAMA Pediatr. 2017;171:897–907. neuroimaging, the predictive validity is improved (sensitivity 98% and 6. Australian Cerebral Palsy Register Group. 2018. specificity 99%).10 The GMA is being integrated into clinical settings 7. Spittle AJ, et al. Dev Med Child Neurol. 2008;4:254–266. where high-risk infants are monitored early for high-risk CP.11 8. Kwong AK, et al. J Clin Med. 2022;11:1833. 9. Spittle AJ, et al. Pediatrics. 2013;132:e452–e458. Training and ongoing calibration of assessors’ Gestalt perception 10. Morgan C, et al. J Clin Med. 2019;11:1879–1890. of the GMA need to be considered when deciding to use the tool in 11. te Velde A, et al. Brain Sci. 2021;11:1074. clinical and research practice.7 This can sometimes cause difficulty 12. Kwong AK, et al. J Paediatr Child Health. 2019;55:548–554. with timing and access to the GMA, particularly given the short 13. McIntyre S, et al. Dev Disabil Res Rev. 2011;17:114–129. 14. Elliott E, et al. BMJ Open. 2021;11:e041695. window of time when fidgety GMs should be observed. Smartphone apps have been created, such as the Baby Moves app, to allow parents to record infant movements for remote assessment in research set- tings in a timely manner.12 GMA has limited research in the general population, where half of the cases of CP are diagnosed.13 Use of the GMA for population-based screening is likely to be dependent on achieving computer automated assessment of the GMA and is currently under investigation.14 https://doi.org/10.1016/j.jphys.2023.03.004 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 203 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Correspondence: Recovery period should not be overlooked when considering the 1-minute sit-to-stand test to assess exertional desaturation in people with chronic respiratory disease We read with great interest the study conducted by Watson et al1 For these two reasons, evaluation of the SpO2 nadir should not be and congratulate the authors for conducting this high-quality limited to the end of the test but should also include the first minutes research addressing an important issue for clinical practice: an of recovery.3 This has also previously been stressed by Crook et al,5 alternative field test to the 6-minute walk test (6MWT). The authors who identified 20% more patients with desaturation when including compared cardiorespiratory response at the end of the 1-minute sit- the first minute of rest in their analysis. We assume that this is to-stand test (1minSTS) with that recorded after 1 minute and at the sufficiently relevant to change the conclusion from Watson et al1 end of the 6MWT. They concluded that using the oxyhaemoglobin regarding the classification of desaturators compared with the 6MWT. saturation measured via pulse oximetry (SpO2) nadir recorded during the 1minSTS is inappropriate for making decisions about whether Therefore, we believe that the actual conclusion should be limited strategies are needed to prevent exertional desaturation in people to values recorded at the end of the test rather than that ‘the 1- with chronic respiratory disease. minute sit-to-stand test underestimates exertional desaturation compared with the 6-minute walk test’. This may still be true, but it is With this conclusion, the authors give the impression that the our opinion that this would require further investigation. Should the 1minSTS does not perform equally when compared with the 6MWT authors confirm their findings with an additional analysis including to assess exertional desaturation. However, the presented data do the recovery period for people with and without supplemental not capture the whole cardiorespiratory adaptation during the oxygen, it would largely strengthen their conclusion. 1minSTS. The SpO2 was measured at the end of the 1minSTS test, which may not be the most appropriate time point for the following Tristan Bonneviea,b, Fairuz Boujibarc and reasons. First, the 1minSTS test is inherently limited to 1 minute, Francis-Edouard Graviera,b which is shorter than the time required for the cardiorespiratory system parameters to plateau. This means that people stop exer- a ADIR Association, Rouen University Hospital, Rouen, France cising while they are still adjusting to the effort and lactates are still b UR3830 GRHVN, Institute for Research and Innovation in Biomedicine accumulating,2 so that oxygen uptake overshoot may be observed after the end of the test (presumably to compensate for an oxygen (IRIB), Normandy University, UNIROUEN, Rouen, France debt).3 Therefore, ventilatory demand, dynamic hyperinflation and c Department of Thoracic Surgery, Rouen University Hospital, Rouen, ventilation/perfusion mismatching are likely to occur or worsen even after the end of the test for some participants.3 This may France particularly be the case for people with advanced respiratory dis- ease such as those with supplemental oxygen (52% of the cohort). References Second, because values do not plateau, any central alteration in the oxygenation of blood in the lungs that would occur at the end of 1. Watson K, et al. J Physiother. 2023;69:108–113. the test is unlikely to be captured exactly at the same time frame 2. Pornsuriyasak P, et al. COPD. 2018:1–11. with a finger probe (ie, there is a time lag between the central 3. Gephine S, et al. Med Sci Sports Exerc. 2020;52:1441–1448. event and its peripheral measurement).4 4. Wang L, et al. J Clin Monit Comput. 2010;24:149–153. 5. Crook S, et al. Eur Respir J. 2017;49:3. https://doi.org/10.1016/j.jphys.2023.05.007 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 204 Appraisal j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Correspondence: Response to Bonnevie et al We thank Bonnevie et al for their interest in our paper. They raise participants with chronic obstructive pulmonary disease, showing an interesting point that there may be further desaturation following that the extent of desaturation detected by central (forehead) and test completion of the 1-minute sit-to-stand test (1minSTS). peripheral (finger) sensors was identical at each minute during a Although we agree that cardiorespiratory measures do not plateau w6MWT.5 For this reason, we doubt that our data were influenced by during the 1minSTS, we have shown that measures of pulse rate and the peripheral sensor lagging behind the true extent of any desatu- oxygen saturation measured via pulse oximetry (SpO2) also do not ration that occurred during the 1minSTS. plateau during the walking-based tests.1 Therefore, in people with marked limitation in gas transfer, it is possible that desaturation will We believe that the conclusion of our study (ie, ‘the 1-minute sit- also worsen immediately following completion of the 6-minute walk to-stand test underestimates exertional desaturation compared with test (6MWT). Worsening desaturation during recovery from walking the 6-minute walk test’) is reasonable, given that we state in the and sit-to-stand exercise is something we observe in a subgroup of Methods section that the measures of nadir SpO2 were taken from people during pulmonary rehabilitation programs. We acknowledge measures collected during the test.3 Nevertheless, we agree that that in the paper by Crook et al,2 extending the monitoring into the further study is needed to determine whether or not there are dif- recovery period for the 1minSTS resulted in four (of 21) additional ferences in nadirs between tests when measures of SpO2 are people meeting the threshold to be classified as a desaturator. continued into the recovery period. Whether this would be true in our sample is unclear because Crook et al2 excluded those who required long-term oxygen therapy; by Kathryn Watsona, Peta Winshipa,b, Vinicius Cavalheria,c,d, doing so, they are unlikely to have recruited those with the most Caitlin Vicarya, Stephanie Straya, Natasha Beare and Kylie Hilld severe limitations in gas transfer. In contrast, more than half of our a Physiotherapy Department, Fiona Stanley Hospital, Perth, Australia sample used long-term oxygen therapy and so may have had a b National School of Health Sciences and Physiotherapy, Notre Dame greater propensity for ongoing desaturation on completion of both tests.3 University, Perth, Australia c Allied Health, Metropolitan Health Service, Perth, Australia The authors also suggest that the reasons for the disparity in nadir d Curtin School of Allied Health and enAble Institute, Curtin University, SpO2 may reflect the time lag between when a central event occurs (ie, desaturation) and when it is detected by peripheral sensors (ie, a Perth, Australia finger probe). In other words, because the 1minSTS is only 1 minute e Institute for Health Research, Notre Dame University, Perth, Australia long, for this test, the peripheral sensor may lag behind the true extent of the desaturation. If this were true, the effect of this lag References would be reduced in a test of longer duration, such as the 6MWT. The study cited to support this idea collected data in anaesthetised dogs 1. Hill K, et al. Respirology. 2012;17:278–284. and demonstrated that the change in SpO2 measured in the trachea 2. Crook S, et al. Eur Respir J. 2017;49:3. was detected more quickly than desaturation measured using a 3. Watson K, et al. J Physiother. 2023;69:108–113. sensor attached to the tail.4 However, we have published data from 4. Wang L, et al. J Clin Monit Comput. 2010;24:149–153. 5. Wilson S, et al. Respirology. 2013;18:1143–1147. https://doi.org/10.1016/j.jphys.2023.05.005 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 191 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: Cognitive functional therapy produces sustained improvements in chronic, disabling low back pain Synopsis Summary of: Kent P, Haines T, O’Sullivan P, Smith A, Campbell A, Schutze they agreed on with their usual health providers (eg, physiotherapy, R, et al. Cognitive functional therapy with or without movement sensor massage, chiropractic care, medicines or surgery). Outcome measures: biofeedback versus usual care for chronic, disabling low back pain The primary outcome was the 24-point Roland-Morris Disability Ques- (RESTORE): a randomised, controlled, three-arm, parallel group, phase tionnaire (RMDQ) at 13 weeks and the primary economic outcome was 3, clinical trial. Lancet. 2023; https://doi.org/10.1016/S0140-6736(23) quality-adjusted life years (QALYs). Secondary outcomes included patient- 00441-5 specific function, pain self-efficacy, catastrophising and fear avoidance. Results: Both CFT interventions were more effective than usual care: MD Question: Does cognitive functional therapy (CFT) with or without for CFT only –4.6 (95% CI –5.9 to –3.4) and for CFT plus biofeedback –4.6 movement sensor biofeedback improve pain intensity and pain-related (95% CI –5.8 to –3.3) on the RMDQ. Effect sizes were similar at 52 weeks. disability in people with chronic, disabling low back pain? Design: Both CFT interventions were also more effective than usual care for QALYs, Randomised controlled trial with concealed allocation and intention-to- and much less costly in terms of societal costs; direct and indirect costs treat analysis. Setting: Twenty primary care clinics in Australia. Partici- and productivity losses were –AU$5,276 (95% CI –10,529 to –24) and pants: Adults with low back pain for . 3 months with at least moderate –AU$8,211 (–12,923 to –3,500). CFT produced sustained improvements on pain-related physical activity limitation. Spinal fracture, spinal infection, the secondary outcomes listed above. Conclusion: CFT produced large, cancer, pregnancy and inability to be physically active were exclusion sustained improvements in people with chronic low back pain and was criteria. Randomisation of 492 participants allocated 165 to usual care, 164 more cost-effective than usual care. to CFT only and 163 to CFT with movement sensor biofeedback. In- terventions: Both CFT groups received seven treatment sessions over 12 Provenance: Invited. Not peer reviewed. weeks plus a booster session at 26 weeks. These sessions facilitated self- management by targeting pain-related cognitions, emotions and behav- Mark Elkins iours that contribute to pain and disability; it particularly addressed pain- Editor, Journal of Physiotherapy provocative movement patterns such as protective muscle guarding and movement avoidance. In addition, one CFT group also wore a movement https://doi.org/10.1016/j.jphys.2023.05.011 sensor that gave them biofeedback that enhanced their ability to correct unhelpful movement patterns. The control group received whatever care 1836-9553/© 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). Commentary A frustration when managing chronic, disabling low back pain is the lack ineffective and/or harmful interventions for chronic low back pain such as of feasible strategies for primary care. Contemporary guidelines discourage fusion surgery, spinal cord stimulators and opioids. A final challenge is drug therapies, interventional procedures and surgery.1 While guidelines how to ensure quality control for CFT. While the Therapeutic Goods endorse multidisciplinary pain management, this is expensive and difficult Administration in Australia provides oversight for drugs and devices, to access, even in large cities. That leaves the primary care clinician with there is no similar system for complex interventions like physical and options such as exercise and psychological therapies; however, they also psychological therapies and surgery. This last issue is important because have limitations, as illustrated in the 2021 Cochrane review of exercise.2 poorly implemented CFT is unlikely to yield the impressive results seen in The review found that exercise had a clinically important effect at short- RESTORE. term follow-up but not in the medium term or long term. Provenance: Invited. Not peer reviewed. The RESTORE trial is a landmark trial because it reported clinically important and durable effects on a range of important health outcomes. Chris Maher For the primary outcome, the effect at 52 weeks (4.8 Roland Morris units) Institute for Musculoskeletal Health, Sydney, Australia was even slightly larger than at 13 weeks (4.6 units). The program was also cost-effective. RESTORE is an important step in the right direction, https://doi.org/10.1016/j.jphys.2023.05.016 but complex work needs to be done on implementing cognitive functional therapy (CFT) more broadly. References The first challenge is to develop and test more scalable methods to 1. Chiarotto A, Koes BW. N Engl J Med. 2022;386:1732–1740. train clinicians to competently deliver the CFT intervention, as the system 2. Hayden J, et al. Cochrane Database Syst Rev. 2021;9:CD009790. used in the trial would constrain broader rollout. A second challenge is that the national health services in many jurisdictions (including Australia) do not have mechanisms to fund CFT, despite funding https://doi.org/10.1016/j.jphys.2023.05.016 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 191 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: Cognitive functional therapy produces sustained improvements in chronic, disabling low back pain Synopsis Summary of: Kent P, Haines T, O’Sullivan P, Smith A, Campbell A, Schutze they agreed on with their usual health providers (eg, physiotherapy, R, et al. Cognitive functional therapy with or without movement sensor massage, chiropractic care, medicines or surgery). Outcome measures: biofeedback versus usual care for chronic, disabling low back pain The primary outcome was the 24-point Roland-Morris Disability Ques- (RESTORE): a randomised, controlled, three-arm, parallel group, phase tionnaire (RMDQ) at 13 weeks and the primary economic outcome was 3, clinical trial. Lancet. 2023; https://doi.org/10.1016/S0140-6736(23) quality-adjusted life years (QALYs). Secondary outcomes included patient- 00441-5 specific function, pain self-efficacy, catastrophising and fear avoidance. Results: Both CFT interventions were more effective than usual care: MD Question: Does cognitive functional therapy (CFT) with or without for CFT only –4.6 (95% CI –5.9 to –3.4) and for CFT plus biofeedback –4.6 movement sensor biofeedback improve pain intensity and pain-related (95% CI –5.8 to –3.3) on the RMDQ. Effect sizes were similar at 52 weeks. disability in people with chronic, disabling low back pain? Design: Both CFT interventions were also more effective than usual care for QALYs, Randomised controlled trial with concealed allocation and intention-to- and much less costly in terms of societal costs; direct and indirect costs treat analysis. Setting: Twenty primary care clinics in Australia. Partici- and productivity losses were –AU$5,276 (95% CI –10,529 to –24) and pants: Adults with low back pain for . 3 months with at least moderate –AU$8,211 (–12,923 to –3,500). CFT produced sustained improvements on pain-related physical activity limitation. Spinal fracture, spinal infection, the secondary outcomes listed above. Conclusion: CFT produced large, cancer, pregnancy and inability to be physically active were exclusion sustained improvements in people with chronic low back pain and was criteria. Randomisation of 492 participants allocated 165 to usual care, 164 more cost-effective than usual care. to CFT only and 163 to CFT with movement sensor biofeedback. In- terventions: Both CFT groups received seven treatment sessions over 12 Provenance: Invited. Not peer reviewed. weeks plus a booster session at 26 weeks. These sessions facilitated self- management by targeting pain-related cognitions, emotions and behav- Mark Elkins iours that contribute to pain and disability; it particularly addressed pain- Editor, Journal of Physiotherapy provocative movement patterns such as protective muscle guarding and movement avoidance. In addition, one CFT group also wore a movement https://doi.org/10.1016/j.jphys.2023.05.011 sensor that gave them biofeedback that enhanced their ability to correct unhelpful movement patterns. The control group received whatever care 1836-9553/© 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). Commentary A frustration when managing chronic, disabling low back pain is the lack ineffective and/or harmful interventions for chronic low back pain such as of feasible strategies for primary care. Contemporary guidelines discourage fusion surgery, spinal cord stimulators and opioids. A final challenge is drug therapies, interventional procedures and surgery.1 While guidelines how to ensure quality control for CFT. While the Therapeutic Goods endorse multidisciplinary pain management, this is expensive and difficult Administration in Australia provides oversight for drugs and devices, to access, even in large cities. That leaves the primary care clinician with there is no similar system for complex interventions like physical and options such as exercise and psychological therapies; however, they also psychological therapies and surgery. This last issue is important because have limitations, as illustrated in the 2021 Cochrane review of exercise.2 poorly implemented CFT is unlikely to yield the impressive results seen in The review found that exercise had a clinically important effect at short- RESTORE. term follow-up but not in the medium term or long term. Provenance: Invited. Not peer reviewed. The RESTORE trial is a landmark trial because it reported clinically important and durable effects on a range of important health outcomes. Chris Maher For the primary outcome, the effect at 52 weeks (4.8 Roland Morris units) Institute for Musculoskeletal Health, Sydney, Australia was even slightly larger than at 13 weeks (4.6 units). The program was also cost-effective. RESTORE is an important step in the right direction, https://doi.org/10.1016/j.jphys.2023.05.016 but complex work needs to be done on implementing cognitive functional therapy (CFT) more broadly. References The first challenge is to develop and test more scalable methods to 1. Chiarotto A, Koes BW. N Engl J Med. 2022;386:1732–1740. train clinicians to competently deliver the CFT intervention, as the system 2. Hayden J, et al. Cochrane Database Syst Rev. 2021;9:CD009790. used in the trial would constrain broader rollout. A second challenge is that the national health services in many jurisdictions (including Australia) do not have mechanisms to fund CFT, despite funding https://doi.org/10.1016/j.jphys.2023.05.016 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 192 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: Corticosteroid injections combined with exercise therapy are superior to placebo injections and exercise therapy for Achilles tendinopathy symptoms Synopsis Summary of: Johannsen F, Olesen JL, Øhlenschläger TF, Lundgaard-Nielsen M, exercises with 15-repetition maximum over three to four sets with 2 to 3 Cullum CK, Jakobsen AS, et al. Effect of Ultrasonography-Guided Corticosteroid minute rests between sets. The program was supervised by a physiotherapist, Injection vs Placebo Added to Exercise Therapy for Achilles Tendinopathy. with weekly decreased repetitions and increased loading guided by partici- JAMA Network Open. 2022;5:e2219661. pant tolerance. Outcome measures: The primary outcome was between- group difference in change in the total score of the Victorian Institute of Question: Are corticosteroid injections and exercise therapy superior when Sports Assessment-Achilles (VISA-A) questionnaire (range 1 to 100, 100 = no compared with placebo injections and exercise therapy for Achilles tendin- symptoms) at 6 months compared with baseline. Secondary outcomes opathy? Design: Superiority, double-blind, parallel-group randomised included morning pain, pain during exercise, global rating of change and controlled trial. Setting: University clinic, Copenhagen, and a local private tendon thickness. Results: Ninety-one participants completed the 6-month rheumatology clinic. Participants: People aged 18 to 65 years with patient- follow-up. The corticosteroid injections and exercise therapy group had a reported Achilles tendon pain for . 3 months; insidious onset of pain greater improvement in VISA-A score from baseline to the 6-month follow-up aggravated by weight-bearing activities, worse in the morning and/or during compared with the placebo group (MD 18 points, 95% CI 8 to 27; p , 0.001). the initial phases of weight-bearing activities; pain and swelling located 2 to Conclusion: Corticosteroid injections combined with exercise therapy were 6 cm proximal to the Achilles tendon insertion; local tendon thickening (ul- superior when compared with placebo injections and exercise therapy at 6 trasonography) . 7 mm or 20% larger compared with the asymptomatic side. months for the treatment of Achilles tendinopathy. Main exclusion criteria: previous Achilles tendon surgery, injections or rup- tures. Randomisation of 100 participants allocated 48 to corticosteroid in- Provenance: Invited. Not peer reviewed. jections and exercise therapy and 52 to placebo injections and exercise therapy. Interventions: Participants in the corticosteroid group were injected Nina Østerås with 1 mL of methylprednisolone acetate (40 mg/mL) and 1 mL of lidocaine Division of Rheumatology and Research, Diakonhjemmet Hospital, (10 mg/mL). Participants in the placebo group were injected with 1 mL of lipid emulsion and 1 mL of lidocaine (10 mg/mL). Injections were given at the Norway first visit and the participants were offered up to three more injections at intervals of 4 weeks. Both groups received a heavy slow resistance program https://doi.org/10.1016/j.jphys.2023.05.013 three times a week over 3 months including three two-legged heel rise Commentary approach without discontinuing tendon loading sporting activities.3 Taken together this would indicate that a judicious course of exercise, including Exercise is currently deemed to be best level care for tendinopathy, education and pain monitoring to manage overall load,4 should first be whereas corticosteroid injections offer effective short-term pain relief, undertaken and that the likelihood of needing to add corticosteroid although often with poor long-term outcomes.1 Combining the short-term injections is small, as is the relative short-term benefit. efficacy of corticosteroid injections with the effects of exercise, which manifest in the mid-term to longer-term, seems like the best of both worlds. Provenance: Invited. Not peer reviewed. This has not proven to be the case to date, with a study of lateral elbow tendinopathy showing that corticosteroid injections neuter the long-term Bill Vicenzino effectiveness of exercise.2 The Johannsen double-blind randomised trial University of Queensland School of Health and Rehabilitation Sciences: tested corticosteroid injections plus exercise versus placebo injections plus exercise in participants in their mid-forties with Achilles tendinopathy for Physiotherapy, The University of Queensland, Brisbane, Australia an average duration of 23 months. The findings suggested value in adding corticosteroid injections to exercise when considering the primary outcome https://doi.org/10.1016/j.jphys.2023.05.014 of patient-reported pain, function and activity (Victorian Institute of Sports Assessment-Achilles [VISA-A] questionnaire). References Several matters are pertinent to implementing these findings. First, the 1. Vos R-J. Brit J Sport Med. 2021;55:1125–1134. beneficial change in VISA-A was short-term, after which there was no advan- 2. Coombes BK. JAMA. 2013;309:461–469. tage of the corticosteroid injectate (ie, 1 to 2 years), which was mirrored by 3. Silbernagel KG. Am J Sports Med. 2007;35:897–906. morning pain levels and global rating of change. All other measures (eg, pain on 4. Silbernagel KG. J Athl Training. 2020;55:438–447. loading activity) did not show differences at any time point. Second, the au- thors asked participants to refrain from jumping or running activity for 3 months, yet the reported changes in VISA-A were like those demonstrated in another study for exercise alone – when performed with a pain monitoring https://doi.org/10.1016/j.jphys.2023.05.014 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 192 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: Corticosteroid injections combined with exercise therapy are superior to placebo injections and exercise therapy for Achilles tendinopathy symptoms Synopsis Summary of: Johannsen F, Olesen JL, Øhlenschläger TF, Lundgaard-Nielsen M, exercises with 15-repetition maximum over three to four sets with 2 to 3 Cullum CK, Jakobsen AS, et al. Effect of Ultrasonography-Guided Corticosteroid minute rests between sets. The program was supervised by a physiotherapist, Injection vs Placebo Added to Exercise Therapy for Achilles Tendinopathy. with weekly decreased repetitions and increased loading guided by partici- JAMA Network Open. 2022;5:e2219661. pant tolerance. Outcome measures: The primary outcome was between- group difference in change in the total score of the Victorian Institute of Question: Are corticosteroid injections and exercise therapy superior when Sports Assessment-Achilles (VISA-A) questionnaire (range 1 to 100, 100 = no compared with placebo injections and exercise therapy for Achilles tendin- symptoms) at 6 months compared with baseline. Secondary outcomes opathy? Design: Superiority, double-blind, parallel-group randomised included morning pain, pain during exercise, global rating of change and controlled trial. Setting: University clinic, Copenhagen, and a local private tendon thickness. Results: Ninety-one participants completed the 6-month rheumatology clinic. Participants: People aged 18 to 65 years with patient- follow-up. The corticosteroid injections and exercise therapy group had a reported Achilles tendon pain for . 3 months; insidious onset of pain greater improvement in VISA-A score from baseline to the 6-month follow-up aggravated by weight-bearing activities, worse in the morning and/or during compared with the placebo group (MD 18 points, 95% CI 8 to 27; p , 0.001). the initial phases of weight-bearing activities; pain and swelling located 2 to Conclusion: Corticosteroid injections combined with exercise therapy were 6 cm proximal to the Achilles tendon insertion; local tendon thickening (ul- superior when compared with placebo injections and exercise therapy at 6 trasonography) . 7 mm or 20% larger compared with the asymptomatic side. months for the treatment of Achilles tendinopathy. Main exclusion criteria: previous Achilles tendon surgery, injections or rup- tures. Randomisation of 100 participants allocated 48 to corticosteroid in- Provenance: Invited. Not peer reviewed. jections and exercise therapy and 52 to placebo injections and exercise therapy. Interventions: Participants in the corticosteroid group were injected Nina Østerås with 1 mL of methylprednisolone acetate (40 mg/mL) and 1 mL of lidocaine Division of Rheumatology and Research, Diakonhjemmet Hospital, (10 mg/mL). Participants in the placebo group were injected with 1 mL of lipid emulsion and 1 mL of lidocaine (10 mg/mL). Injections were given at the Norway first visit and the participants were offered up to three more injections at intervals of 4 weeks. Both groups received a heavy slow resistance program https://doi.org/10.1016/j.jphys.2023.05.013 three times a week over 3 months including three two-legged heel rise Commentary approach without discontinuing tendon loading sporting activities.3 Taken together this would indicate that a judicious course of exercise, including Exercise is currently deemed to be best level care for tendinopathy, education and pain monitoring to manage overall load,4 should first be whereas corticosteroid injections offer effective short-term pain relief, undertaken and that the likelihood of needing to add corticosteroid although often with poor long-term outcomes.1 Combining the short-term injections is small, as is the relative short-term benefit. efficacy of corticosteroid injections with the effects of exercise, which manifest in the mid-term to longer-term, seems like the best of both worlds. Provenance: Invited. Not peer reviewed. This has not proven to be the case to date, with a study of lateral elbow tendinopathy showing that corticosteroid injections neuter the long-term Bill Vicenzino effectiveness of exercise.2 The Johannsen double-blind randomised trial University of Queensland School of Health and Rehabilitation Sciences: tested corticosteroid injections plus exercise versus placebo injections plus exercise in participants in their mid-forties with Achilles tendinopathy for Physiotherapy, The University of Queensland, Brisbane, Australia an average duration of 23 months. The findings suggested value in adding corticosteroid injections to exercise when considering the primary outcome https://doi.org/10.1016/j.jphys.2023.05.014 of patient-reported pain, function and activity (Victorian Institute of Sports Assessment-Achilles [VISA-A] questionnaire). References Several matters are pertinent to implementing these findings. First, the 1. Vos R-J. Brit J Sport Med. 2021;55:1125–1134. beneficial change in VISA-A was short-term, after which there was no advan- 2. Coombes BK. JAMA. 2013;309:461–469. tage of the corticosteroid injectate (ie, 1 to 2 years), which was mirrored by 3. Silbernagel KG. Am J Sports Med. 2007;35:897–906. morning pain levels and global rating of change. All other measures (eg, pain on 4. Silbernagel KG. J Athl Training. 2020;55:438–447. loading activity) did not show differences at any time point. Second, the au- thors asked participants to refrain from jumping or running activity for 3 months, yet the reported changes in VISA-A were like those demonstrated in another study for exercise alone – when performed with a pain monitoring https://doi.org/10.1016/j.jphys.2023.05.014 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 190 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: In overweight and obese adults with knee osteoarthritis, the addition of telehealth-delivered exercise and diet programs to online education improves pain and function Synopsis Summary of: Bennell KL, Lawford BJ, Keating C, Brown C, Kasza J, Mack- pain measured on a 0 to 10 numerical rating scale and physical function enzie D, et al. Comparing video-based, telehealth-delivered exercise and measured via the function subscale of the Western Ontario and McMaster weight loss programs with online education on outcomes of knee oste- Universities Osteoarthritis Index at 6 months post-randomisation. oarthritis: a randomized trial. Ann Intern Med. 2022;175:198–209. Results: A total of 379 participants completed the study (154, 170 and 55 in the exercise, diet1exercise and control groups, respectively). At 6 Question: In overweight and obese adults with knee osteoarthritis, does months, compared with the control group, the experimental groups the addition of telehealth-delivered exercise and diet programs to online demonstrated better pain (MD exercise –0.8, 95% CI –1.5 to –0.2; MD education improve pain and function? Design: Three-group randomised diet1exercise –1.5, 95% CI –2.1 to –0.8) and function (MD exercise –7.0, controlled trial with concealed allocation. Setting: Single university 95% CI –9.7 to –4.2; MD diet1exercise –9.8, 95% CI –12.5 to –7.0). The diet research centre in Melbourne. Participants: Adults aged between 45 and and exercise program was superior to exercise (MD pain –0.6, 95% CI –1.1 80 years, with knee pain most days for at least 3 months, and a body mass to –0.2; MD function –2.8, 95% CI –4.7 to –0.8). Conclusion: In overweight index between 28 and 41 kg/m2. Exclusion criteria were: type 1 diabetes; and obese adults with knee osteoarthritis, 6-month telehealth-delivered type 2 diabetes requiring medication; stroke, cardiac event, or knee sur- exercise and diet programs improved pain and function. The dietary gery in the past 6 months; fluid intake restriction; and inflammatory intervention conferred modest additional benefits in pain and function arthritis. Randomisation of 415 participants allocated 172 to the exercise over exercise. group, 175 to the diet and exercise group and 68 to the control group. Interventions: All participants were provided with access to a web-based Provenance: Invited. Not peer reviewed. osteoarthritis educational material. In addition, the exercise group received six remotely delivered consultations with a physiotherapist over Vinicius Cavalheri 6 months, hard copy educational booklets, resistance bands for Curtin School of Allied Health, Curtin University, Australia strengthening exercises, and a Fitbit to monitor physical activity. The diet and exercise group received all exercise components plus six individual https://doi.org/10.1016/j.jphys.2023.05.002 consultations with a dietitian over 6 months, with additional dietary and behavioural resources. Outcome measures: The primary outcomes were Commentary Remotely delivered physiotherapy care has increasingly become for various reasons, it is important for clinicians to recognise popular, with promising results observed in improving pain and function evidence-based interventions that are suitable to remotely deliver to for certain populations including osteoarthritis, even before the COVID-19 patients. Overall, these findings provide clinicians with confidence that pandemic.1 Core recommendations for the management of hip and knee delivering an exercise program or a combination of exercise and weight osteoarthritis should include patient education and land-based exercise management program via videoconferencing can improve clinical out- with or without a dietary weight management component.2 In this study, comes such as pain and function in patients with knee osteoarthritis. the authors investigated the effectiveness of two programs (exercise program or exercise program with a dietary intervention) delivered Provenance: Invited. Not peer reviewed. entirely via videoconferencing, based on best-practice recommendations for the management of knee osteoarthritis. The findings showed partici- Luke Davies pants reporting positive improvements in pain and function in both School of Primary and Allied Health Care, Monash University, Melbourne, groups. However, the combination of diet and exercise was found to be superior to exercise alone for primary outcomes at 6 months, with similar Australia results observed at 12 months. Patient adherence to exercise programs is often challenging. An interesting observation in this study was the high https://doi.org/10.1016/j.jphys.2023.05.001 levels of participant self-reported adherence to both the physical activity and weight management plans at the 6-month timepoint. This may be References attributed to the convenience of patients being able to complete these programs in their own environment. Given that we are moving into an 1. Bennell K, et al. Ann Intern Med. 2017;166:453–462. age where both clinicians and patients intend to continue using telehealth 2. Bannuru R, et al. Osteoarthr Cartil. 2019;11:1578–1589. https://doi.org/10.1016/j.jphys.2023.05.001 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 190 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: In overweight and obese adults with knee osteoarthritis, the addition of telehealth-delivered exercise and diet programs to online education improves pain and function Synopsis Summary of: Bennell KL, Lawford BJ, Keating C, Brown C, Kasza J, Mack- pain measured on a 0 to 10 numerical rating scale and physical function enzie D, et al. Comparing video-based, telehealth-delivered exercise and measured via the function subscale of the Western Ontario and McMaster weight loss programs with online education on outcomes of knee oste- Universities Osteoarthritis Index at 6 months post-randomisation. oarthritis: a randomized trial. Ann Intern Med. 2022;175:198–209. Results: A total of 379 participants completed the study (154, 170 and 55 in the exercise, diet1exercise and control groups, respectively). At 6 Question: In overweight and obese adults with knee osteoarthritis, does months, compared with the control group, the experimental groups the addition of telehealth-delivered exercise and diet programs to online demonstrated better pain (MD exercise –0.8, 95% CI –1.5 to –0.2; MD education improve pain and function? Design: Three-group randomised diet1exercise –1.5, 95% CI –2.1 to –0.8) and function (MD exercise –7.0, controlled trial with concealed allocation. Setting: Single university 95% CI –9.7 to –4.2; MD diet1exercise –9.8, 95% CI –12.5 to –7.0). The diet research centre in Melbourne. Participants: Adults aged between 45 and and exercise program was superior to exercise (MD pain –0.6, 95% CI –1.1 80 years, with knee pain most days for at least 3 months, and a body mass to –0.2; MD function –2.8, 95% CI –4.7 to –0.8). Conclusion: In overweight index between 28 and 41 kg/m2. Exclusion criteria were: type 1 diabetes; and obese adults with knee osteoarthritis, 6-month telehealth-delivered type 2 diabetes requiring medication; stroke, cardiac event, or knee sur- exercise and diet programs improved pain and function. The dietary gery in the past 6 months; fluid intake restriction; and inflammatory intervention conferred modest additional benefits in pain and function arthritis. Randomisation of 415 participants allocated 172 to the exercise over exercise. group, 175 to the diet and exercise group and 68 to the control group. Interventions: All participants were provided with access to a web-based Provenance: Invited. Not peer reviewed. osteoarthritis educational material. In addition, the exercise group received six remotely delivered consultations with a physiotherapist over Vinicius Cavalheri 6 months, hard copy educational booklets, resistance bands for Curtin School of Allied Health, Curtin University, Australia strengthening exercises, and a Fitbit to monitor physical activity. The diet and exercise group received all exercise components plus six individual https://doi.org/10.1016/j.jphys.2023.05.002 consultations with a dietitian over 6 months, with additional dietary and behavioural resources. Outcome measures: The primary outcomes were Commentary Remotely delivered physiotherapy care has increasingly become for various reasons, it is important for clinicians to recognise popular, with promising results observed in improving pain and function evidence-based interventions that are suitable to remotely deliver to for certain populations including osteoarthritis, even before the COVID-19 patients. Overall, these findings provide clinicians with confidence that pandemic.1 Core recommendations for the management of hip and knee delivering an exercise program or a combination of exercise and weight osteoarthritis should include patient education and land-based exercise management program via videoconferencing can improve clinical out- with or without a dietary weight management component.2 In this study, comes such as pain and function in patients with knee osteoarthritis. the authors investigated the effectiveness of two programs (exercise program or exercise program with a dietary intervention) delivered Provenance: Invited. Not peer reviewed. entirely via videoconferencing, based on best-practice recommendations for the management of knee osteoarthritis. The findings showed partici- Luke Davies pants reporting positive improvements in pain and function in both School of Primary and Allied Health Care, Monash University, Melbourne, groups. However, the combination of diet and exercise was found to be superior to exercise alone for primary outcomes at 6 months, with similar Australia results observed at 12 months. Patient adherence to exercise programs is often challenging. An interesting observation in this study was the high https://doi.org/10.1016/j.jphys.2023.05.001 levels of participant self-reported adherence to both the physical activity and weight management plans at the 6-month timepoint. This may be References attributed to the convenience of patients being able to complete these programs in their own environment. Given that we are moving into an 1. Bennell K, et al. Ann Intern Med. 2017;166:453–462. age where both clinicians and patients intend to continue using telehealth 2. Bannuru R, et al. Osteoarthr Cartil. 2019;11:1578–1589. https://doi.org/10.1016/j.jphys.2023.05.001 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Journal of Physiotherapy 69 (2023) 189 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: In people with persistent instability after ACL injury, surgical reconstruction is more effective and cost-effective than non-surgical rehabilitation Synopsis Summary of: Beard DJ, Davies L, Cook JA, Stokes J, Leal J, Fletcher H, et al. recreational activities and knee-related quality of life (0 worst to 100 best). Rehabilitation versus surgical reconstruction for non-acute anterior cruciate Secondary outcomes were knee-specific quality of life (ACL QoL), return to ligament injury (ACL SNNAP): a pragmatic randomised controlled trial. Lancet. activity and level of sport participation (Tegner or modified Tegner score), 2022;400(10352):605-615. health-related quality of life (EQ-5D-5L), resource use, intervention-related complications, and satisfaction 18 months after randomisation. Results: A Question: Is surgical reconstruction more effective and cost-effective than total of 316 participants were randomised and 248 (78%) completed the study. rehabilitation in patients with a non-acute anterior cruciate ligament (ACL) At 18 months, KOOS4 scores were higher in the surgical reconstruction group injury and persistent symptoms of instability? Design: Randomised controlled by 8 points (95% CI 3 to 13). Knee-specific quality of life and satisfaction were trial with concealed allocation. Setting: Twenty-nine National Health Service significantly higher in the surgical reconstruction group. The groups did not secondary care hospitals across the UK. Participants: Patients aged 18 years significantly differ in health-related quality of life and intervention-related with ACL injury and persistent (non-acute) symptomatic knee instability. complications. Surgical reconstruction was more cost-effective (£1,017, 95% CI Exclusion criteria were meniscal pathology requiring immediate surgery, evi- 557 to 1,476). Conclusion: Surgical reconstruction is clinically superior and dence of later stage osteoarthritis, and multi-plane multi-ligament instability. cost-effective compared with rehabilitation in patients with non-acute ACL Randomisation of 316 participants allocated 156 to surgical reconstruction and injury and persistent instability. 160 to rehabilitation. Interventions: The surgical reconstruction group received surgery and postoperative rehabilitation as per standard care at the partici- Provenance: Invited. Not peer reviewed. pating hospital. The rehabilitation group received at least six sessions over a 3-month period of physiotherapist-led rehabilitation involving pain and Joshua R Zadro swelling management, range of motion, neuromuscular and muscle strength The University of Sydney, Australia exercises, gait training, and a graded return to function, activity or sport. They were also given the option for surgery later. Outcome measures: The primary https://doi.org/10.1016/j.jphys.2023.03.002 outcome was knee and associated problems, measured by the Knee Injury and Osteoarthritis Outcome Score-4 domain version (KOOS4) score at 18 months after randomisation. The domains were pain, symptoms, difficulty in sports and Commentary Rehabilitation versus surgical reconstruction for non-acute anterior cru- individuals with acute ACL rupture, initial rehabilitation before considering ACL ciate ligament injury (ACL-SNAPP)1 is the third trial comparing surgical and reconstruction is recommended.5 Two high-quality trials found no clinically non-surgical management for ACL injury, providing further evidence to relevant benefit from early ACL reconstruction for acute ACL ruptures compared guide clinical practice. Readers should be aware that ACL-SNNAP is unlike the with rehabilitation and optional delayed surgery,2–4 and early ACL reconstruction previous two trials.2–4 First, ACL-SNNAP included participants with long- is not cost-effective compared with rehabilitation and optional surgery.7 standing knee instability following ACL injury, unlike those with acute ACL rupture as per the previous trials. They also included participants who had Provenance: Invited. Not peer reviewed. not received formal rehabilitation, which does not align with recommendations to commence rehabilitation as soon as possible after ACL Stephanie Filbay injury.5 Second, the trial did not intend to deliver best-practice Department of Physiotherapy, University of Melbourne, Melbourne, rehabilitation. The trial was performed within the UK National Health Service and the volume/duration of rehabilitation (six or more Australia rehabilitation sessions delivered over at least 3 months) may reflect the limited resources in a public health system. It is unclear whether the low https://doi.org/10.1016/j.jphys.2023.03.001 proportion of participants who returned to sport at 18 months (reconstruction 28% versus rehabilitation 24%) is a reflection of the quality References of late-phase rehabilitation where sports-specific exercises, power/agility drills and heavy strength training are recommended.6 1. Beard DJ, et al. Lancet. 2022;400:605–615. 2. Frobell RB, et al. NEJM. 2010;363:331–342. How the findings inform clinical practice 3. Frobell RB, et al. BMJ. 2013;346(jan24,1):f232–f32. 4. Reijman M, et al. BMJ. 2021;9;372:n375. For most patients with non-acute ACL rupture who present within a public 5. Whittaker JL, et al. BJSM. 2022;56:1393–1405. health setting with long-standing knee instability and no prior rehabilitation, 6. Filbay SR, Grindem H. Best Pract Res Clin Rheumatol. 2019;33:33–47. ACL reconstruction can be recommended as first-line treatment. For most 7. Filbay SR, et al. BJSM. 2022;56:1465–1474. https://doi.org/10.1016/j.jphys.2023.03.001 1836-9553/© 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY license (http://creativecommons.org/ licenses/by/4.0/).
Journal of Physiotherapy 69 (2023) 189 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Appraisal Critically appraised paper: In people with persistent instability after ACL injury, surgical reconstruction is more effective and cost-effective than non-surgical rehabilitation Synopsis Summary of: Beard DJ, Davies L, Cook JA, Stokes J, Leal J, Fletcher H, et al. recreational activities and knee-related quality of life (0 worst to 100 best). Rehabilitation versus surgical reconstruction for non-acute anterior cruciate Secondary outcomes were knee-specific quality of life (ACL QoL), return to ligament injury (ACL SNNAP): a pragmatic randomised controlled trial. Lancet. activity and level of sport participation (Tegner or modified Tegner score), 2022;400(10352):605-615. health-related quality of life (EQ-5D-5L), resource use, intervention-related complications, and satisfaction 18 months after randomisation. Results: A Question: Is surgical reconstruction more effective and cost-effective than total of 316 participants were randomised and 248 (78%) completed the study. rehabilitation in patients with a non-acute anterior cruciate ligament (ACL) At 18 months, KOOS4 scores were higher in the surgical reconstruction group injury and persistent symptoms of instability? Design: Randomised controlled by 8 points (95% CI 3 to 13). Knee-specific quality of life and satisfaction were trial with concealed allocation. Setting: Twenty-nine National Health Service significantly higher in the surgical reconstruction group. The groups did not secondary care hospitals across the UK. Participants: Patients aged 18 years significantly differ in health-related quality of life and intervention-related with ACL injury and persistent (non-acute) symptomatic knee instability. complications. Surgical reconstruction was more cost-effective (£1,017, 95% CI Exclusion criteria were meniscal pathology requiring immediate surgery, evi- 557 to 1,476). Conclusion: Surgical reconstruction is clinically superior and dence of later stage osteoarthritis, and multi-plane multi-ligament instability. cost-effective compared with rehabilitation in patients with non-acute ACL Randomisation of 316 participants allocated 156 to surgical reconstruction and injury and persistent instability. 160 to rehabilitation. Interventions: The surgical reconstruction group received surgery and postoperative rehabilitation as per standard care at the partici- Provenance: Invited. Not peer reviewed. pating hospital. The rehabilitation group received at least six sessions over a 3-month period of physiotherapist-led rehabilitation involving pain and Joshua R Zadro swelling management, range of motion, neuromuscular and muscle strength The University of Sydney, Australia exercises, gait training, and a graded return to function, activity or sport. They were also given the option for surgery later. Outcome measures: The primary https://doi.org/10.1016/j.jphys.2023.03.002 outcome was knee and associated problems, measured by the Knee Injury and Osteoarthritis Outcome Score-4 domain version (KOOS4) score at 18 months after randomisation. The domains were pain, symptoms, difficulty in sports and Commentary Rehabilitation versus surgical reconstruction for non-acute anterior cru- individuals with acute ACL rupture, initial rehabilitation before considering ACL ciate ligament injury (ACL-SNAPP)1 is the third trial comparing surgical and reconstruction is recommended.5 Two high-quality trials found no clinically non-surgical management for ACL injury, providing further evidence to relevant benefit from early ACL reconstruction for acute ACL ruptures compared guide clinical practice. Readers should be aware that ACL-SNNAP is unlike the with rehabilitation and optional delayed surgery,2–4 and early ACL reconstruction previous two trials.2–4 First, ACL-SNNAP included participants with long- is not cost-effective compared with rehabilitation and optional surgery.7 standing knee instability following ACL injury, unlike those with acute ACL rupture as per the previous trials. They also included participants who had Provenance: Invited. Not peer reviewed. not received formal rehabilitation, which does not align with recommendations to commence rehabilitation as soon as possible after ACL Stephanie Filbay injury.5 Second, the trial did not intend to deliver best-practice Department of Physiotherapy, University of Melbourne, Melbourne, rehabilitation. The trial was performed within the UK National Health Service and the volume/duration of rehabilitation (six or more Australia rehabilitation sessions delivered over at least 3 months) may reflect the limited resources in a public health system. It is unclear whether the low https://doi.org/10.1016/j.jphys.2023.03.001 proportion of participants who returned to sport at 18 months (reconstruction 28% versus rehabilitation 24%) is a reflection of the quality References of late-phase rehabilitation where sports-specific exercises, power/agility drills and heavy strength training are recommended.6 1. Beard DJ, et al. Lancet. 2022;400:605–615. 2. Frobell RB, et al. NEJM. 2010;363:331–342. How the findings inform clinical practice 3. Frobell RB, et al. BMJ. 2013;346(jan24,1):f232–f32. 4. Reijman M, et al. BMJ. 2021;9;372:n375. For most patients with non-acute ACL rupture who present within a public 5. Whittaker JL, et al. BJSM. 2022;56:1393–1405. health setting with long-standing knee instability and no prior rehabilitation, 6. Filbay SR, Grindem H. Best Pract Res Clin Rheumatol. 2019;33:33–47. ACL reconstruction can be recommended as first-line treatment. For most 7. Filbay SR, et al. BJSM. 2022;56:1465–1474. https://doi.org/10.1016/j.jphys.2023.03.001 1836-9553/© 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY license (http://creativecommons.org/ licenses/by/4.0/).
Journal of Physiotherapy 69 (2023) 160–167 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Research Curl-up exercises improve abdominal muscle strength without worsening inter-recti distance in women with diastasis recti abdominis postpartum: a randomised controlled trial Sandra B Gluppe a, Marie Ellström Engh b,c, Kari Bø a,b a Department of Sports Medicine, Norwegian School of Sports Sciences, Oslo, Norway; b Department of Obstetrics and Gynaecology, Akershus University Hospital, Lørenskog, Norway; c Faculty of Medicine, University of Oslo, Oslo, Norway KEY WORDS ABSTRACT Curl-up Question: What is the effect of a 12-week, home-based, abdominal exercise program containing head lifts and Diastasis recti abdominis abdominal curl-ups on inter-recti distance (IRD) in women with diastasis recti abdominis (DRA) 6 to 12 months Exercise postpartum? What is the effect of the program on: observed abdominal movement during a curl-up; global Postpartum perceived change; rectus abdominis thickness; abdominal muscle strength and endurance; pelvic floor Randomised controlled trial disorders; and low back, pelvic girdle and abdominal pain? Design: This was a two-arm, parallel-group, randomised controlled trial with concealed allocation, assessor blinding and intention-to-treat analysis. Participants: Seventy primiparous or multiparous women 6 to 12 months postpartum, having a single or multiple pregnancy following any mode of delivery, with a diagnosis of DRA (IRD . 28 mm at rest or . 25 mm during a curl-up). Intervention: The experimental group was prescribed a 12-week standardised ex- ercise program including head lifts, abdominal curl-ups and twisted abdominal curl-ups 5 days a week. The control group received no intervention. Outcome measures: The primary outcome measure was change in IRD measured with ultrasonography. Secondary outcomes were: observed abdominal movement during a curl-up; global perceived change; rectus abdominis thickness; abdominal muscle strength and endurance; pelvic floor disorders; and low back, pelvic girdle and abdominal pain. Results: The exercise program did not improve or worsen IRD (eg, MD 1 mm at rest 2 cm above the umbilicus, 95% CI –1 to 4). The program improved rectus abdominis thickness (MD 0.7 mm, 95% CI 0.1 to 1.3) and strength (MD 9 Nm, 95% CI 3 to 16) at 10 deg; its effects on other secondary outcomes were trivial or unclear. Conclusion: An exercise program containing curl-ups for women with DRA did not worsen IRD or change the severity of pelvic floor disorders or low back, pelvic girdle or abdominal pain, but it did increase abdominal muscle strength and thickness. Registration: NCT04122924. [Gluppe SB, Ellström Engh M, Bø K (2023) Curl-up exercises improve abdominal muscle strength without worsening inter-recti distance in women with diastasis recti abdominis postpartum: a randomised controlled trial. Journal of Physiotherapy 69:160–167] © 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction incontinence and pelvic organ prolapse) have been suggested but Diastasis recti abdominis (DRA) is defined as midline separation of not substantiated in studies of women with mild or moderate the two rectus abdominis muscles along the linea alba.1 To diagnose DRA.6–8,10 Keeler et al11 reported that 89% of women’s health DRA, the inter-recti distance (IRD) is measured,2 but there is no physiotherapists used pelvic floor muscle training and 87% used consensus on the cut-off point.3 Ultrasound is the recommended transversus abdominis training in the treatment of DRA. These re- method to measure IRD, with intra-rater and inter-rater intraclass sults were confirmed in a recent study by Gluppe et al.12 Three correlation coefficients . 0.9.4 systematic reviews have evaluated the effect of different exercise The condition has been reported to be highly prevalent in the programs in the treatment of DRA, concluding that due to poor postpartum period, with prevalence rates up to 45% at 6 months and 30% at 12 months postpartum.5 To date, two systematic reviews6,7 methodological quality and small samples sizes of the included tri- have evaluated the consequences of DRA, and DRA has been asso- ciated with weaker abdominal muscles7–9 and more abdominal als, the evidence is insufficient to recommend any specific exercise pain.6,8 Associations with low back pain, pelvic girdle pain and protocol.13–15 Curl-ups have traditionally been discouraged in the pelvic floor disorders (such as urinary incontinence, anal treatment of women with DRA, but a short-term experimental study found that head-lift and abdominal curl-ups reduced the IRD during the exercises.16 Therefore, the research questions for this trial were: https://doi.org/10.1016/j.jphys.2023.05.017 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Research 161 1. What is the effect of a 12-week, home-based, abdominal exercise centres in Norway from January 2020 to December 2022. Post-test program containing head lifts and abdominal curl-ups on IRD in assessment was conducted after 12 weeks of the intervention women with DRA 6 to 12 months postpartum? period by one physiotherapist blinded to group allocation. 2. What is the effect of the exercise program on observed abdominal Inclusion criteria were being a primiparous or multiparous movement during a curl-up, global perceived change, abdominal woman with a diagnosis of DRA 6 to 12 months postpartum, having a muscle strength, abdominal muscle endurance, rectus abdominis single or multiple pregnancy following any mode of delivery, and thickness, pelvic floor disorders, low back pain, pelvic girdle pain being able to understand a Scandinavian language. The diagnosis of and abdominal pain? DRA was based on previous studies, and IRD cut-off was set to 2.8 cm at rest or 2.5 cm during a curl-up 2 cm above or 2 cm below the Method umbilicus.17,18 Diagnosis of DRA was confirmed with ultrasound prior to baseline testing or when signing up for screening. In addition, if a Design protrusion along the linea alba was observed, women were included even if the IRD was less than the cut-off values. This was an assessor-blinded, two-arm, parallel-group rando- mised controlled trial. After baseline testing, all participants were Prior to baseline testing, the participants responded to an elec- randomly allocated equally to either the experimental or control tronic questionnaire gathering information about background group by a person not involved in the assessments. Randomisation variables and comorbidities such as neurological and systemic was computer-generated in blocks of four. Concealed allocation was musculoskeletal diseases or psychiatric diagnoses (Table 1). Height, used. weight and waist circumference were measured at the clinical assessment, and body mass index was calculated for all participants. One trained physiotherapist who had undergone specific training in ultrasound imaging of the abdominal muscles prior to data Intervention collection performed all assessments. Images taken at baseline testing were transferred from the hard disk to a software programa and The focus of the exercise protocol was to strengthen the abdom- analysed offline. The same physiotherapist, blinded for group allo- inal muscles, based on findings from a short-term experimental cation, performed both the ultrasound assessments and the offline study.16 The specific exercises and progression of the program are analyses. All participants were thoroughly informed at the start of the shown in Figure 2. The standardised, individual exercise program was post-test to not reveal group allocation. prescribed for 10 minutes/day, 5 days/week for 12 weeks. Due to the COVID-19 pandemic, the instructions on how to perform the exercises Participants were described by a physiotherapist by telephone or demonstrated on FaceTimeb. In addition, the exercises were captured and described Participants were recruited through women’s health physiother- in detail in a document sent to all participants in the experimental apists, personal trainers, midwives, gynaecologists/obstetricians, group (Figure 2). The experimental group was also provided with a friends and acquaintances and by advertising on social media. The smartphone appc where adherence to training could be registered. numbers of women screened and excluded with reasons for exclusion Furthermore, a daily reminder to exercise and a weekly SMS was are presented in Figure 1. The trial was performed mainly at the given to encourage participants in the experimental group to adhere Norwegian School of Sport Sciences in Oslo and at two physiotherapy to the program. The physiotherapist providing the training program was not involved in any assessments. Assessed for eligibility (n = 247) Excluded (n = 177) • did not meet inclusion criteria (n = 109) • declined to participate (n = 33) • other reasons (n = 35) Week 0 Measured inter-recti distance, abdominal movement, global rating of change, abdominal muscle strength, abdominal muscle endurance, rectus thickness, pelvic floor symptoms, and low back, pelvic and abdominal pain Randomised (n = 70) (n = 35) (n = 35) Lost to follow-up (n = 2) Experimental group Control group • illness (n = 1) • home-based abdominal • no intervention • wanted a more intense exercise program • 12 weeks intervention (n = 1) • 5 sessions/week • 12 weeks Week 12 Measured inter-recti distance, abdominal movement, global rating of change, abdominal muscle strength, abdominal muscle endurance, rectus thickness, pelvic floor symptoms, and low back, pelvic and abdominal pain (n = 33) (n = 35) Figure 1. Flow of participants through the trial.
162 Gluppe et al: Effect of curl-ups on DRA Table 1 classifications from very much worse to fully healed in a numerical Baseline characteristics of the participants. 11-point scale; the instrument has shown good intra-test reliability (ICC = 0.9).19 Variable Total sample Exp group Con group (n = 70) (n = 35) (n = 35) Abdominal muscle strength and endurance: An isokinetic dynamo- metere was used to assess maximal isometric abdominal wall Age (y), mean (SD) 34 (3) 35 (4) 33 (3) strength, limited to trunk flexion. Isometric trunk flexion assessment Height (m), mean (SD) 1.67 (0.10) 1.68 (0.10) 1.66 (0.10) with dynamometers in a sitting position has shown excellent test– Weight (kg), (SD) 69.4 (15.4)a 69.7 (15.6) 69.1 (15.4) retest reliability with ICC . 0.920,21 and strong correlation with the Body mass index (kg/m2), 24.9 (5.6)a 24.8 (5.4) 25.0 (5.8) width of the IRD.9,20 Abdominal muscle endurance was assessed as mean (SD) number of repetitions of a standardised abdominal curl-up to Waist circumference (cm), 85.2 (13.9) 85.3 (13.0) 85.1 (14.9) exhaustion test following the protocol of the American College of mean (SD) Sports Medicine (ACSM) curl-up test.22,23 The protocol for these two Weight gain last pregnancy (kg), 15.5 (5.6) 15.9 (6.1) 15.1 (5.2) strength tests is described in detail elsewhere.8 mean (SD) Heavy lifting at work, n yes (%)b 6 (24) 3 (23) 3 (25) Rectus abdominis thickness: Muscle thickness was defined as the Singleton parity, n (%)c distance between the inside edges of the superior and inferior fascial 14 (22) 5 (16) 9 (27) borders.2 Participants rested in a standardised supine position with 1 41 (64) 21 (68) 20 (61) arms alongside and knees bent with the feet on the bench. Three 2 9 (14) 5 (16) 4 (12) ultrasound images were taken at rest 2 cm above the umbilicus. The 3 transducer was moved transversely over the midpoint of the right Mode of delivery, n (%) 44 (69) 25 (81) 19 (58) rectus abdominis and the average was calculated. This protocol was vaginal 9 (14) 1 (3) 8 (24) modified from a published protocol.2 caesarean 11 (17) 5 (16) 6 (18) both vaginal and caesarean 25 (36) 12 (34) Pelvic floor disorders, low back pain, pelvic girdle pain and abdom- Use of contraceptives, n yes (%) 57 (81) 13 (37) 27 (77) inal pain: Prior to testing at baseline and Week 12, all participants Breastfeeding 1 time/day, n (%) 27 (77) responded to the following questions in the electronic questionnaire: Striae, n (%) 30 (43) 18 (51) ‘Do you have symptoms from your bowel, bladder or pelvic region during teenage 39 (56) 12 (34) 16 (46) that bother you (eg, urinary leakage, bowel leaks or feeling of bulge during pregnancy 14 (20) 23 (66) 7 (20) in the vagina)?’, ‘Do you have low back pain?’, ‘Do you have pelvic postpartum 7 (20) girdle pain?’ and ‘Do you have pain in your abdomen?’. If participants Menstruating, n (%) 38 (54) 21 (60) responded yes to these questions, they were asked to respond to the yes 24 (34) 17 (49) 10 (29) following: the Pelvic Floor Distress Inventory-short form 20,24 the no 8 (11) 14 (40) 4 (11) Oswestry Disability Index25 and the Pelvic Girdle Questionnaire,26 uncertain 136 (118) 4 (11) 120 (121) respectively. If participants responded yes to having abdominal Physical activity (minutes/wk), 150 (115) pain, they were asked to indicate the location of the pain and to what mean (SD) degree (from 0 = not at all to 10 = a lot) the pain affected their ac- tivities of daily living. Each scale score of the Pelvic Floor Distress Con = control group, Exp = experimental group. Inventory-short form 20 ranges from 0 (no disability) to 100 (maximum disability) and the sum score from the three scales Percentages may not sum to 100% due to the effects of rounding. together range from 0 to 300. The sum score in the Oswestry a N = 69; missing data for one woman who did not want to measure weight from Disability Index is calculated in percent from 0 (not disabled) to 100 (disabled) and the Pelvic Girdle Questionnaire is calculated in percent control group (valid percent reported). from 0 (not at all) to 100 (to a large extent). All three instruments b N = 25; 25 women reported to be back to work and therefore responded to this have been validated and are recommended for assessment of symptoms of PFD, functional measure of disability due to low back question. pain, and limitation in activities/participation due to pelvic girdle c N = 64; missing data on six women had a twin delivery (valid percent reported). pain.24–26 The control group received no intervention and was discouraged Data analysis from conducting specific abdominal training/exercises. However, general physical activity and training of other muscle groups was not An a priori power calculation was conducted based on a conser- restricted. vative approach. The effect size for the experimental group was set to 0.8 cm with SD of 0.8, for the control group the effect size was set to Outcome measures 0.2 and SD 0.8. These numbers are based on results from a previously published randomised trial,27 but the effect size was slightly adjusted Primary outcome upwards due to that study’s low power. In addition, the adjustment Change in inter-recti distance: One physiotherapist performed the was based on other relevant studies.28,29 Therefore, we planned to include 58 participants with 29 in each group. To allow for some loss clinical examinations of all participants at baseline and at the end of to follow-up, the final estimation was increased to 70 participants the 12-week intervention period. To diagnose DRA, the IRD was with 35 in each group. assessed with transabdominal ultrasound, a portable two- dimensional ultrasound machine with a linear transducerd. The ul- Data were analysed using SPSS 28. Background variables were trasound imaging protocol has been described in detail elsewhere.16 reported as means with standard deviations (SD) or numbers with percentages. Within-group and between-group comparisons of To standardise the measurement locations, two marks were made continuous and categorical data were analysed by the independent on the skin: one 2 cm above and one 2 cm below the centre of the sample t-test and chi-square test for independence, respectively. For umbilicus (Figure 3).4 Images were captured at rest and during a curl- continuous variables, an ANCOVA was used as a linear regression with up at both measurement locations. During the curl-up, the partici- Week 12 value as the dependent variable and group and the baseline pants were in a standardised supine position with arms crossed over variable as the independent variables. Difference in change between the chest and the curl-up was performed until the shoulder blades groups from baseline to Week 12 are reported with 95% CI. P-values , were off the bench. The end position of the ultrasound assessment of 0.05 were considered statistically significant. Analyses were based on IRD during the curl-up is illustrated in Figure 4. intention to treat. Secondary outcomes Clinical observation of diastasis recti abdominis: The assessor observed the abdomen while the participants performed a curl-up and registered if the following was seen at baseline and Week 12: protrusion, sink-in, no movement, or uncertain. Perceived change of the condition with the Global Rating of Change scale: At Week 12, participants in both groups were asked to report whether they perceived improvement in their DRA compared with baseline on a Global Rating of Change scale. This scale includes
Research 163 Exercise program Weeks 1 to 4: Perform one set of exercises 1, 2 and 3 in the order described below. Weeks 5 to 8: Perform two sets of exercises 1, 2 and 3 in the order described below with a 1-minute pause between sets. Weeks 9 to 12: Perform three sets of exercises 1, 2 and 3 in the order described below with a 1-minute pause between sets. Perform all exercises slowly and check that the exercises are performed correctly without causing a significant protrusion (please observe your abdomen while doing the exercises). You may conduct the exercises while lying on the floor with your child. Exercises with explanation 2) Oblique curl-up (both right and left side) 3) Curl-up Lie on your back with your legs bent and one arm alongside, Lie on your back with your arms crossed over your chest 1) Head lift the other hand behind your head • Slowly inhale and exhale Lie on your back with your legs bent and • Slowly inhale and exhale • After exhaling, lift your head and bend the upper part of your back up arms alongside • After exhaling, lift your head and bend the upper part of • Slowly inhale and exhale until both shoulder blades are free from the floor • After exhaling, lift your head up with your your back obliquely up until one shoulder blade is free • Lower slowly from the floor chin towards your chest • Lower slowly • Lower slowly Progression for weeks 11 and 12 Progression 1 × 8 repetitions 1 × 8 repetitions 1 × 8 repetitions Weeks 1 and 2 1 × 10 repetitions 1 × 10 repetitions 1 × 10 repetitions Weeks 3 and 4 2 × 10 repetitions 2 × 10 repetitions; Holding time 1 second Weeks 5 and 6 2 × 12 repetitions 2 × 12 repetitions; Holding time 2 seconds 2 × 10 repetitions; Holding time 1 second Weeks 7 and 8 3 × 10 repetitions 3 × 10 repetitions; Holding time 2 seconds Weeks 9 and 10 3 × 12 repetitions 3 × 12 repetitions; At the top, tilt forward three times 2 × 12 repetitions; Holding time 2 seconds Weeks 11 and 12 3 × 10 repetitions; Holding time 2 seconds 3 × 12 repetitions; Place your hands in a grip behind your neck (avoid picking up speed with your arms on the way up!) Figure 2. Details of the experimental intervention. Results Twenty-one participants (66%) in the experimental group adhered to 80% of their prescribed exercise sessions. Flow of participants through the study Compliance with the trial protocol A total of 247 women were screened with ultrasound for DRA, with 177 (72%) excluded mostly due to not meeting the IRD inclusion Extra exercise sessions were added for three participants criteria. Of the 70 women who met the eligibility criteria, 35 women where the Week 12 assessment was delayed due to lockdown were randomised to the experimental group and 35 to the control during the pandemic. Eighteen participants did not perform the group (Figure 1). Two participants (4%) dropped out of the experi- maximal isometric strength with the dynamometer due to the mental group and none dropped out of the control group. The reasons lockdown. for drop-out are described in Figure 1. Women in the experimental group completed 74% (SD 26) of their prescribed exercise sessions. Figure 3. Inter-recti distance measurement locations during ultrasound assessment. Figure 4. The end position during curl-up in the ultrasound assessment of IRD. Reproduced with permission of Kristina L Skaug.
164 Gluppe et al: Effect of curl-ups on DRA Baseline characteristics of the participants At baseline, 76% of all participants were unable to perform one curl-up according to the ACSM curl-up test procedure; therefore, the The baseline characteristics of the participants are shown in data on number of curl-ups were strongly skewed in both groups. Table 1 and in the first two columns of data in Tables 2 to 8. Individual Although the few participants in each group who could perform participant data are available in Table 9 on the eAddenda. Among the ACSM curl-ups generally showed a small increase in their number of total sample, 97% of the participants had a college/university educa- curl-ups by week 12, there was no between-group difference in the tion, 97% were of Caucasian genetic origin, 99% were married/ amount of improvement: Hodges–Lehmann MD 0 curl-ups, 95% CI cohabitating and 99% were non-smokers. Six (9%) participants had 0 to 0 (Table 6). one twin delivery each: four participants in the experimental group and two in the control group. Pelvic floor disorders, low back pain, pelvic girdle pain, and abdominal pain Change in inter-recti distance When participants were questioned about the presence of any For all participants, the widest IRD was measured at rest 2 cm pelvic floor disorders or low back, pelvic girdle or abdominal pain, above the umbilicus (Table 2). Mean IRD decreased from baseline to very imprecise estimates of effect were elicited. This is evident in the week 12 for all IRD measurements in both groups, except from a wide confidence intervals that included the possibility of clinically minor increase in IRD measured below the umbilicus at rest in the worthwhile effects in either direction (Table 7). However, when control group. Little to no difference in change occurred between the questioned about the severity of pelvic floor disorders or low back, two groups from baseline to week 12, with the estimates of mean pelvic girdle or abdominal pain, more precise estimates of effect difference and their 95% CIs below the smallest worthwhile effect could be obtained (Table 8). The effects were all so small as to be thresholds of 5 mm for above the umbilicus and 8 mm for below the clearly clinically trivial because the confidence limits were all , 10% umbilicus30 (Table 2). One minor exception was the effect below the of the outcome measure’s scale. umbilicus during a curl-up, where one end of the 95% CI just reached the smallest worthwhile effect, indicating a slight possibility that the Discussion experimental intervention might have a worthwhile benefit on IRD there. Certainly, the curl-ups did not worsen DRA to any clinically The effect of the exercise program on the primary outcome, IRD, important extent. was measured above and below the umbilicus and at rest and during a curl-up. The mean between-group differences and most of their 95% Observed movement of the abdomen CIs were below the smallest worthwhile effect thresholds of 5 mm for above the umbilicus and 8 mm for below the umbilicus.30 One The estimates of the effect of the experimental intervention on exception was the effect below the umbilicus during a curl-up, where observed movement of the abdomen during a curl-up were very one end of the 95% CI just reached the smallest worthwhile effect (8 imprecise (Table 3); therefore, the effect of the curl-ups on movement mm), indicating a slight possibility that the experimental interven- of the abdomen remains uncertain. tion might have a worthwhile benefit on IRD there. Certainly, the curl-ups did not worsen DRA to any clinically important extent. The Global rating of change exercise program clearly improved abdominal muscle strength and thickness, although it is unclear whether these effects were large At week 12, 20 of 33 participants (61%) in the experimental group enough to be worthwhile. The exercise program clearly did not and 15 of 35 participants (43%) in the control group reported change abdominal muscle endurance or the severity of pelvic floor improvement in DRA (RR 1.41, 95% CI 0.88 to 2.27). None of the disorders or low back, pelvic or abdominal pain to any worthwhile participants in either group reported worsening of the condition. extent. The effects on the other secondary outcomes were unclear due to wide confidence intervals. Rectus abdominis thickness The choice of intervention was based on a short-term experi- The experimental group improved the thickness of their rectus mental study showing an immediate IRD reduction during head-lift abdominis more than the control group by a mean difference of 0.7 and abdominal curl-ups.16 This effect was not found when the par- mm. However, the confidence interval spanned worthwhile and ticipants were performing the same exercise for 12 weeks in this trial. trivial effects (95% CI 0.1 to 1.3) as shown in Table 4. The reason for a negligible effect on IRD may be due to the choice of exercises, that only 66% of participants in the experimental group Abdominal muscle strength and endurance adhered to 80% of the prescribed exercise and that the exercises were home-based and unsupervised. It was difficult to make direct The experimental group improved their maximal isometric comparisons between other randomised trials applying abdominal strength more than the control group (Table 5). However, the confi- training to reduce IRD. The studies differ in measurement methods to dence intervals spanned worthwhile and trivial effects: MD 8 nM assess IRD (palpation, calliper, ultrasound), choice of cut-off value for (95% CI 1 to 14) at 30 deg and MD 9 nM (95% CI 3 to 16) at 10 deg. diagnosing DRA, and inclusion of primiparous and/or multiparous women. In addition, onset of the training intervention, type of Table 2 Mean (SD) of groups, mean (SD) within-group difference and mean (95% CI) between-group difference for width of inter-recti distance under various conditions. Inter-recti distance (mm) Groups Within-group difference Between-group difference Week 12 minus Week 0 Week 12 minus Week 0 Week 0 Week 12 Exp (n = 35) Con (n = 35) Exp (n = 32) Con (n = 35) Exp Con Exp minus Con At rest 2 cm above umbilicus 37 (8) 40 (10) 36 (9) 38 (10) –1 (5) –2 (4) 1 (–1 to 4) At rest 2 cm below umbilicus 29 (10) 27 (12) 30 (11) –1 (6) 2 (7) –3 (–6 to 0) During a curl-up 2 cm above umbilicus 27 (10) 28 (10) 26 (12) 28 (12) –2 (7) –2 (6) 0 (–3 to 3) During a curl-up 2 cm below umbilicus 23 (12) 21 (13) 24 (13) –2 (8) 1 (12) –3 (–8 to 2) 30 (13) 24a (13) Con = control group, Exp = experimental group. a n = 34.
Research 165 Table 3 Number (%) of participants with specific observed movements of the abdomen during a curl-up, with relative risk (95% CI) for between-group difference in prevalence at 12 weeks. Observed movement of abdomen Week 0 Week 12 Relative risk (95% CI) at week 12 Exp (n = 35) Con (n = 35) Exp (n = 32) Con (n = 33) Exp relative to Con Protrusion 14 (40) 12 (34) 6 (19) 5 (15) 1.23 (0.42 to 3.65) Sink in 5 (14) 2 (6) 5 (16) 3 (9) 1.72 (0.45 to 6.61) No movement 15 (40) 19 (59) 22 (67) 0.89 (0.61 to 1.30) Uncertain 2 (6) 19 (54) 2 (6) 3 (9) 0.69 (0.12 to 3.85) 2 (6) Con = control group, Exp = experimental group. Table 4 Mean (SD) of groups, mean (SD) within-group difference and mean (95% CI) between-group difference for rectus abdominis thickness. Outcome Groups Within-group difference Between-group difference Week 12 minus Week 0 Week 12 minus Week 0 Week 0 Week 12 Exp Con Exp minus Con Exp (n = 34) Con (n = 34) Exp (n = 31) Con (n = 32) 0.7 (0.1 to 1.3) Rectus abdominis thickness (mm) 8.3 (2.1) 8.5 (1.5) 9.5 (2.1) 8.7 (1.8) 1.0 (1.2) 0.3 (1.3) Con = control group, Exp = experimental group. Table 5 Mean (SD) of groups, mean (SD) within-group difference and mean (95% CI) between-group difference for abdominal strength. Maximal isometric strength (Nm) Groups Within-group difference Between-group difference Week 12 minus Week 0 Week 0 Week 12 Week 12 minus Week 0 Exp minus Con Exp (n = 24) Con (n = 27) Exp (n = 20) Con (n = 24) Exp Con 8 (1 to 14) At 30 deg 119 (27) 114 (28) 128 (28) 116 (23) 9 (11) 1 (11) 9 (3 to 16) At 10 deg 91 (27) 88 (24) 100 (26) 89 (25) 10 (11) 1 (10) Con = control group, Exp = experimental group. Table 6 Median (IQR) of groups, median within-group difference (95% CI) and Hodges–Lehmann median (95% CI) between-group difference for abdominal muscle endurance. Outcome Groups Within-group difference Between-group difference Week 0 Week 12 Week 12 minus Week 0 Week 12 minus Week 0 Exp (n = 31) Con (n = 34) Exp (n = 31) Con (n = 34) Exp Con Exp minus Con ACSM curl-ups (n) 0 (0 to 7.5) 0 (0 to 0) 0 (0 to 11.5) 0 (0 to 9) 0 (0 to 1.5) 0 (0 to 1) 0 (0 to 0) ACSM = American College of Sport Medicine, Con = control group, Exp = experimental group. Table 7 Number (%) of participants reporting the presence of pelvic floor disorders, low back pain, pelvic girdle pain and abdominal pain, and relative risk (95% CI) between groups. Reported symptoms, n (%) Week 0 Week 12 Relative risk (95% CI) at Week 12 Exp (n = 35) Con (n = 35) Exp (n = 32) Con (n = 34) Exp relative to Con Pelvic floor disorders 18 (51) 12 (34) 13 (41) 9 (26) 1.53 (0.76 to 3.09) Low back pain 18 (51) 16 (46) 15 (47) 13 (38) 1.22 (0.70 to 2.16) Pelvic girdle pain 12 (34) 13 (37) 15 (47) 13 (38) 0.71 (0.28 to 1.77) Abdominal pain 6 (17) 4 (11) 6 (19) 2.13 (0.58 to 7.79) 3 (9) Con = control group, Exp = experimental group. Table 8 Mean (SD) of groups, mean (SD) within-group difference and mean (95% CI) between-group difference for reported symptom severity. Symptom severity Groups Within-group difference Between-group difference Week 12 minus Week 0 Week 0 Week 12 Week 12 minus Week 0 PFDI – short form 20 (0 to 300) Exp (n = 35) Con (n = 35) Exp (n = 32) Con (n = 34) Exp Con Exp minus Con Oswestry Disability Index (0 to 100) Pelvic Girdle Questionnaire (0 to 100) 28a (43) 10b (25) 17b (29) 16c (38) –5 (28) 4 (27) –9 (–25 to 8) Abdominal pain severity (0 to 10) 6 (9) 5 (9) 6 (10) 5 (9) –1 (6) –1 (8) 0 (–3 to 3) 11 (19) 4 (10) 8 (15) –6 (14) –3 (16) 1 (2) 10 (19) 1 (2) 0 (1) 0 (2) 0 (1) –4 (–11 to 4) 0 (1) 0 (–1 to 1) Con = control group, Exp = experimental group, PFDI = Pelvic Floor Distress Inventory. a n = 26. b n = 27. c n = 31.
166 Gluppe et al: Effect of curl-ups on DRA exercises and training programs, training dosage, supervision of the huge focus on DRA in social media and young mothers’ concerns exercise interventions and methodological quality of the studies13-15 about their abdomen postpartum.12 Even small and possible negli- gible DRAs may result in fair avoidance and drop-out from physical made comparison between studies challenging. As there is a natural activities both during pregnancy and following childbirth. The World remission of DRA in the postpartum period,5 the effect of in- Health Organization recommends that all women start or continue a regular exercise program during pregnancy and in the postpartum terventions should be compared with untreated controls. Seven period.39 Mota et al17 suggested that the upper limit for normal IRD during pregnancy and postpartum needs to be re-evaluated. The randomised trials have compared abdominal training with an un- current results indicate that women with mild to moderate diastasis treated control group. Two randomised trials28,29 found no clear ef- can perform abdominal crunches with no risk of worsening DRA. fect of their exercise intervention: Keshwani et al28 included exercises However, the effect on abdominal exercises in women with severe for isolated activation of transversus abdominis; and Gluppe et al29 DRA needs further investigations. evaluated the effect of pelvic floor muscle training including weekly Strengths of the present study were the randomised design, a group training involving abdominal exercises. The latter study priori power calculation, low loss to follow-up, blinding of the assessor and the use of an exercise program including three exer- measured IRD with palpation. Hence, comparison with the present cises shown to give an immediate reduction of IRD in a short-term experimental study.8 The exercise program also included progres- results was difficult. Five randomised trials have found a beneficial sion in load and volume every second week. The same assessor effect on IRD over untreated controls.27,31–34 The participants in these performed all assessments, ultrasound was used to assess IRD, the assessment procedures were standardised with the aim of mini- studies were comparable with this study regarding parity and onset mising inaccuracy, and a homogeneous group of parous women 6 to 12 months postpartum was included. Limitations were that of training postpartum. However, the studies were hampered with unsupervised training may have reduced adherence and intensity sample sizes (eg, n = 10)27 and the cut-off values for DRA in some of the training, and that the abdominal muscle strength tests have studies indicated no DRA.34 Only one of the randomised trials27 not been tested for reliability. In addition, this sample consisted of measured IRD with ultrasound and this study was a pilot trial.27 a limited number of participants with severe diastasis, and the results might therefore not be generalisable to women with severe The present study found beneficial effects on maximal isometric DRA. muscle strength and muscle thickness. Only one other randomised trial A 10-minute, standardised, abdominal exercise program con- has compared an exercise group with an untreated control group ducted 5 days a week for 12 weeks is unlikely to decrease IRD. measuring trunk flexion strength and endurance.28 Although the test Furthermore, the exercise program was effective in increasing was different, their results correspond with the current findings. A abdominal muscle strength and rectus abdominis thickness. The limitation of the measurement method used in the present study program, which included head-lift, curl-up and twisted curl-up, did not worsen DRA, so parous women with mild to moderate DRA (Humac NORM) is lack of reliability data. However, reliability studies should not be discouraged from performing such exercises. Further high-quality randomised trials of exercise programs including with comparable dynamometers to measure maximal trunk flexion have women with severe DRA are warranted. shown good to excellent intra-tester reliability.20,21 The present study What was already known on this topic: Diastasis recti also found that the program increased rectus abdominis thickness. This abdominis is common in the postpartum period. Literature re- views have not found conclusive evidence about the effects of corresponds with increased rectus abdominis thickness in both inter- abdominal exercises on the inter-recti distance, but curl-ups have traditionally been discouraged in the treatment of women with vention arms (exercise delivered in person or via ZOOM) in the study by diastasis recti abdominis. Kim et al.35 Participants’ characteristics, time since last birth and the What this study adds: Although an exercise program with ultrasound protocol were comparable with the present study. head lifts and curl-ups is unlikely to reduce the inter-recti dis- tance, it definitely does not worsen it. Furthermore, the exercise The negligible effect of the experimental intervention on program increases abdominal muscle thickness and strength. Parous women should not be discouraged from performing head abdominal muscle endurance assessed with the ACSM curl-up test in lifts and curl-up exercises. the present study did not correspond with the positive effect found Footnotes: a MicroDicom software, MicroDicom, Sofia, Bulgaria. b FaceTime, Apple, Cupertino, USA. with dynamometry. Interestingly, . 70% of the participants were c Athlete Monitoring, Fitstats Technologies Inc, Honiton, UK. unable to perform a single curl-up according to this test. Hence, the d Logic e R7, GE Healthcare, Chalfont St Giles, UK. e Humac NORM isokinetic dynamometer, CSMi, Soughton, USA. ACSM curl-up test might be too difficult for women with DRA post- Ethics approval: The Regional Medical Ethics Committee (REK partum.8 In addition, there is no reliability data on this test and the South-East 2018/2312) and the Norwegian Centre for Research Data (440860) approved this study. All participants gave written informed results should therefore be interpreted with caution. In contradiction consent before data collection began. to this study group, Botla and Saleh36 found that their participants Competing interests: Nil. Source of support: The Norwegian Women’s Public Health Asso- were able to perform 24 curl-ups, but the mean IRD values were ciation fully funded the study: H1/2018. Acknowledgements: The authors thank Associate Professor smaller in their participants compared with our study’s participants Morten Fagerland, PhD, for statistical advice; physiotherapists Kris- and below the normal IRD values in postpartum women according to tina L Skaug and Marte D Gram for randomisations of participants Mota et al.17 and teaching the exercise program and follow-up of the exercise group through Athlete Monitoring; Kristin Benum and Henrik Despite the finding that our participants did not reduce the IRD, they were able to increase maximal abdominal muscle strength and rectus abdominis thickness, with no widening of the IRD. Performing crunches and sit-ups were discouraged for a long time because these exercises were believed to worsen DRA. However, several short-term experimental studies have now found that contracting the trans- versus abdominis and pelvic floor muscles widens the IRD, while crunches narrow the IRD.16,37,38 Our results of no negative effect on IRD by conducting head-lift, crunch and twisted crunches regularly for 3 months indicate that women with mild to moderate diastases can perform these exercises safely; furthermore, these exercises improve abdominal strength. A postulated association between DRA and pelvic floor disorders, low back pain, pelvic girdle pain and abdominal pain has not been substantiated in many studies.6,10 The present study did not find any effect of curl-up training on any of these conditions, and the results support the findings in a systematic review of randomised trials.14 A recent study by Yalfani et al31 found a positive effect of abdominal training on low back pain and disability in women with DRA. Further studies are needed to explore the effects of abdominal training on these conditions in women with DRA. An interesting finding of the present study was that 72% of women screened for participation believed they had a DRA but were excluded due to not meeting our IRD inclusion criteria. This may reflect the
Research 167 Warpe-Kinn for providing physiotherapy rooms for testing at 18. Candido G, Lo T, Janssen PA. Risk factors for diastasis of the recti abdominis. J. Assoc Chart Physiother Women’s Health. 2005;97:49–54. Nøtterøy and in Fredrikstad. 19. Kamper SJ, Maher CG, Mackay G. Global rating of change scales: a review of Correspondence: Sandra B Gluppe, Department of Sports strengths and weaknesses and considerations for design. J Man Manip Ther. 2009;17:163–170. Medicine, Norwegian School of Sports Sciences, Oslo, Norway. Email: 20. Stark B, Emanuelsson P, Gunnarsson U, Strigard K. Validation of Biodex system 4 for [email protected] measuring the strength of muscles in patients with rectus diastasis. J Plast Surg Hand Surg. 2012;46:102–105. References 21. Jensen KK, Kjaer M, Jorgensen LN. Isometric abdominal wall muscle strength 1. Venes D, Taber C. Taber’s Cyclopedic Medical Dictionary. 22nd edition. Philadelphia: assessment in individuals with incisional hernia: a prospective reliability study. FA Davis Co; 2013. Hernia. 2016;20:831–837. 2. Coldron Y, Stokes MJ, Newham DJ, Cook K. Postpartum characteristics of rectus 22. Ferguson B. ACSM’s Guidelines for Exercise Testing and Prescription 9th Ed. 2014. abdominis on ultrasound imaging. Man Ther. 2008;13:112–121. J Can Chiropr Assoc. 2014;58, 328–328. 3. van de Water AT, Benjamin DR. Measurement methods to assess diastasis of the 23. Sparling PB, Millard-Stafford M, Snow TK. Development of a cadence curl-up test rectus abdominis muscle (DRAM): A systematic review of their measurement for college students. Res Q Exerc Sport. 1997;68:309–316. properties and meta-analytic reliability generalisation. Man Ther. 2016;21:41–53. 24. Teig CJ, Grotle M, Bond MJ, Prinsen CA, Engh MA, Cvancarova MS, et al. Norwegian 4. Mota P, Pascoal AG, Sancho F, Bø K. Test-retest and intrarater reliability of 2- translation, and validation, of the Pelvic Floor Distress Inventory (PFDI-20) dimensional ultrasound measurements of distance between rectus abdominis in and the Pelvic Floor Impact Questionnaire (PFIQ-7). Int Urogynecol J. 2017;28: women. J Orthop Sports Phys Ther. 2012;42:940–946. 1005–1017. 5. Sperstad JB, Tennfjord MK, Hilde G, Ellstrom-Engh M, Bø K. Diastasis recti 25. Fairbank JC, Pynsent PB. The Oswestry Disability Index. Spine. 2000;25:2940–2952. abdominis during pregnancy and 12 months after childbirth: prevalence, risk 26. Stuge B, Garratt A, Krogstad Jenssen H, Grotle M. The pelvic girdle questionnaire: a factors and report of lumbopelvic pain. Br J Sports Med. 2016;50:1092–1096. condition-specific instrument for assessing activity limitations and symptoms in 6. Fuentes Aparicio L, Rejano-Campo M, Donnelly GM, Vicente-Campos V. Self-re- people with pelvic girdle pain. Phys Ther. 2011;91:1096–1108. ported symptoms in women with diastasis rectus abdominis: A systematic review. 27. Tuttle LJ, Fasching J, Keller A, Patel M, Saville C, Schlaff R, et al. Noninvasive J Gynecol Obstet Hum Reprod. 2020;50:101995. treatment of postpartum diastasis recti abdominis: a pilot study. J Womens Health Phys Ther. 2018;42:65–75. 7. Benjamin DR, Frawley HJ, Shields N, van de Water ATM, Taylor NF. Relationship 28. Keshwani N, Mathur S, McLean L. The impact of exercise therapy and between diastasis recti of the abdominal muscles (DRAM) and musculoskeletal abdominal binding in the management of diastasis recti abdominis in the early dysfunctions, pain and quality of life: a systematic review. Physiotherapy. post-partum period: a pilot randomized controlled trial. Physiother Theory Pract. 2019;105:24–34. 2019:1–16. 29. Gluppe SL, Hilde G, Tennfjord MK, Engh ME, Bø K. Effect of a postpartum training 8. Gluppe S, Ellström Engh M, Bø K. Women with diastasis recti abdominis might program on prevalence of diastasis recti abdominis in postpartum primiparous have weaker abdominal muscles and more abdominal pain, but no higher preva- women: a randomized controlled trial. Phys Ther. 2018;98:260–268. lence of pelvic floor disorders, low back and pelvic girdle pain than women 30. Keshwani N, McLean L. Ultrasound imaging in postpartum women with diastasis without diastasis recti abdominis. Physiotherapy. 2021;111:57–65. recti: intrarater between-session reliability. J Orthop Sports Phys Ther. 2015;45: 713–718. 9. Hills NF, Graham RB, McLean L. Comparison of trunk muscle function between 31. Yalfani A, Bigdeli N, Gandomi F. Comparing the effects of suspension and isometric- women with and without diastasis recti abdominis at 1 year postpartum. Phys Ther. isotonic training on postural stability, lumbopelvic control, and proprioception in 2018;98:891–901. women with diastasis recti abdominis: a randomized, single-blinded, controlled trial. Physiother Theory Pract. 2022:1–13. 10. Fei H, Liu Y, Li M, He J, Liu L, Li J, et al. The relationship of severity in diastasis recti 32. Pampolim G, Ribeiro dos Santos B, Gadiolli Verzola I, Ferres AM, Rodrigues da abdominis and pelvic floor dysfunction: a retrospective cohort study. BMC Womens Silva GB, Silva Sarmento S. Physiotherapy in the reduction of diastasis of the recti Health. 2021;21:1–8. abdominis in immediate postpartum. Revista de Pesquisa: Cuidado e Fundamental. 2021;13:856–860. 11. Keeler J, Albrecht M, Eberhardt L, Horn L, Donnelly C, Lowe D. Diastasis recti abdominis: 33. Laframboise FC, Schlaff RA, Baruth M. Postpartum exercise intervention targeting a survey of women’s health specialists for current physical therapy clinical practice for diastasis recti abdominis. Int J Exerc Sci. 2021;14:400–409. postpartum women. J Womens Health Phys Ther. 2012;36:131–142. 34. Bobowik PZ_, Da˛ bek A. Physiotherapy in women with diastasis of the rectus abdominis muscles. Adv Rehab. 2018;3:11–17. 12. Gluppe S, Ellström Engh M, Bø K. Primiparous women’s knowledge of diastasis 35. Kim S, Yi D, Yim J. The effect of core exercise using online videoconferencing recti abdominis, concerns about abdominal appearance, treatments, and perceived platform and offline-based intervention in postpartum woman with diastasis recti abdominal muscle strength 6-8 months postpartum. A cross sectional comparison abdominis. Int J Environ Res Public Health. 2022;19:7031. study. BMC Womens Health. 2022;22:428. 36. Botla MA, Saleh MS. Effect of Russian current stimulation on abdominal strength and endurance in postnatal diastasis recti: a randomized controlled trial. Fizjo- 13. Benjamin DR, van de Water AT, Peiris CL. Effects of exercise on diastasis of the terapia Polska. 2020;20:16–22. rectus abdominis muscle in the antenatal and postnatal periods: a systematic re- 37. Depledge J, McNair P, Ellis R. Exercises. Tubigrip and taping: can they reduce rectus view. Physiotherapy. 2014;100:1–8. abdominis diastasis measured three weeks post-partum? Musculoskelet Sci Pract. 2021;53:102381. 14. Gluppe S, Engh ME, Bø K. What is the evidence for abdominal and pelvic floor 38. Theodorsen NM, Strand LI, Bø K. Effect of pelvic floor and transversus abdominis muscle training to treat diastasis recti abdominis postpartum? A systematic review muscle contraction on inter-rectus distance in postpartum women: a cross- with meta-analysis. Braz J Phys Ther. 2021;25:644–675. sectional experimental study. Physiotherapy. 2019;105:315–320. 39. Evenson KR, Mottola MF, Artal R. Review of recent physical activity guidelines 15. Berg-Poppe P, Hauer M, Jones C, Munger M, Wethor C. Use of exercise in the during pregnancy to facilitate advice by health care providers. Obstet Gynecol Surv. management of postpartum diastasis recti: a systematic review. J Womens Health 2019;74:481–489. Phys Ther. 2022;46:35–47. 16. Gluppe SB, Engh ME, Bø K. Immediate effect of abdominal and pelvic floor muscle exercises on interrecti distance in women with diastasis recti abdominis who were parous. Phys Ther. 2020;100:1372–1383. 17. Mota P, Pascoal AG, Carita AI, Bø K. Normal width of the inter-recti distance in pregnant and postpartum primiparous women. Musculoskelet Sci Pract. 2018;35:34–37.
Journal of Physiotherapy 69 (2023) 136–138 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Editorial Economic evaluations of physiotherapy interventions Ianthe Boden a,b, Mark R Elkins c,d a School of Health Sciences, University of Tasmania, Launceston, Australia; b Launceston General Hospital, Launceston, Australia; c Journal of Physiotherapy; d Faculty of Medicine and Health, University of Sydney, Sydney, Australia This Editorial introduces another of Journal of Physiotherapy’s benefit of this specific soccer warm-up,9 which is now recom- article collections, which feature papers in a specific field of research mended for implementation at national levels.10 that were published in the journal within the past decade.1–4 The studies in this article collection relate to economic evaluations of Another paper reporting an impactful cost-benefit analysis was a physiotherapy interventions. High-income countries are spending trial that investigated the benefit of adding six group balance classes progressively more on healthcare every year. This outlay is dispro- to usual care for subacute rehabilitation ward patients.11 This high- portionate to inflation, population growth and increases in gross quality randomised controlled trial of 162 participants found that domestic product.5 Continued healthcare spending at the current rate the intervention was cost-effective in improving functional status at 3 is unsustainable, necessitating health policy changes and fiscal months. Although staffing additional group balance classes on a constraint measures. Initiatives such as Choosing Wisely encourage rehabilitation ward costs a hospital more money initially, down- healthcare providers, policymakers and consumers to select ‘high- stream savings in episodes of care costs and hospital readmissions value’ healthcare, where interventions are both therapeutically resulted in a net cost saving to the hospital of AU$5,698 per patient effective and cost-effective.6 In a tightening fiscal environment, (95% CI 165 to 11,265). This is called a ‘dominant’ strategy, a win-win, physiotherapy interventions will be increasingly required to where clinical benefits are gained at significantly less cost to the demonstrate cost-effectiveness to be maintained as standard practice provider than providing usual care. This is a clear example of ‘high- or newly implemented. value’ physiotherapy care. In 2013, the Consolidated Health Economic Evaluation Reporting It is rare for a cost-benefit analysis of a clinical trial to demon- Standards (CHEERS) statement was released to guide quality report- strate such significant findings. Costs are usually secondary outcomes ing of health economic papers.7 The papers within this article with high heterogeneity and wide variation in values. Clinical trials collection, on average, met 85% of the required reporting items. are frequently underpowered for cost outcomes. Determining whether an intervention has a particular fiscal influence (cost saving, All monetary values have been inflated to 2023 values from the cost neutral or costlier) is typically unlikely within a trial-based cost- original trial data. benefit analysis. Even at meta-analysis level, cost-benefit analyses can be limited by cost value variations. A systematic review inves- Cost-benefit analyses tigating the effect of additional weekend allied health services to subacute rehabilitation, acute surgical wards and medical wards12 A health economic evaluation, as a minimum, involves a cost- included 19 studies, three of which included health economic out- benefit analysis. This compares the difference in mean costs per comes. Despite aggregation of trial data, the meta-analysis of cost- patient between an intervention and alternative treatment, usual benefit was imprecise (MD AU$226, 95% CI –329 to 614). It remains care or control. Costs include amounts required to deliver the unknown whether the addition of weekend allied health services on intervention/control and ongoing costs accured over a defined acute wards provides a cost-benefit to the hospital as payer of the period of time. A cost-benefit analysis is reported in monetary service. A simple cost-benefit analysis is largely an insensitive blunt units. A basic cost-benefit analysis was conducted in the first paper tool to determine the value of a physiotherapy intervention. in this article collection.8 This cluster-randomised controlled trial of 479 male amateur soccer players found a small reduction of 16 less Cost-effectiveness analyses injuries per 1,000 players in teams who conducted a specialised pre-training warm-up compared with usual warm-up. Despite this A new intervention could be costlier than usual care, yet lead to small therapeutic effect, an integrated cost-benefit analysis found better outcomes. The improvement in outcome may outweigh this that this effect was still economically worthwhile, with a net so- additional cost. This is the ‘willingness to pay’ concept in health cietal cost saving of V263 (95% CI 20 to 557) per player. Sporting economics. ‘Willingness to pay’ is an amount a payer (eg, a patient, injuries are associated with high healthcare costs and lost work hospital, government or society) would be prepared to pay for a time, so even a small reduction in injury rates can lead to sub- treatment to achieve an improvement in outcome. A cost- stantial societal cost savings. This intervention could have been effectiveness analysis assesses the difference in costs between in- cast aside as an ineffective therapy; however, the positive cost- terventions relative to the difference in outcomes. benefit findings spurred other trials to confirm the therapeutic Incremental cost-effectiveness ratio The result of a cost-effectiveness analysis is called an incremental cost-effectiveness ratio (ICER). An ICER is expressed in monetary units and gives the decision-maker an understanding of the costs saved, or https://doi.org/10.1016/j.jphys.2023.05.004 1836-9553/© 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).
Editorial 137 needing to be spent, to achieve one-unit improvement in a desired Investigators took 105 people with a chronic health condition who outcome. had just completed a 6-week gym-based exercise program and randomised them into a 12-month ongoing maintenance program ICER ¼ Total costs ðinterventionÞ2Total costs ðcontrolÞ held either in a gym or their home. Cost-utility was calculated for Effects ðinterventionÞ2Effects ðcontrolÞ QALY improvement at 12 months. The gym-based supervised exercise program cost AU$600 more to provide than the home-based pro- An ICER using a common clinical outcome can allow multiple in- gram. This extra outlay did not come with reduced costs overall or an terventions within a field to be compared directly against each other. improvement in QALY. The authors calculated that society would For example, the relative cost-effectiveness of preventing a single case need to be willing to pay . $500,000 to be certain that a gym-based of incontinence was compared across different modes of pelvic floor program would gain one QALY for a person with a chronic health muscle training in a systematic review of 17 trials.13 Investigators found condition. Sensitivity analyses found the ICER remained . $400,000. that individual physiotherapy sessions provided during pregnancy, in As the ICER results were reliably and consistently above an affordable the postpartum period, and group-based classes provided during willingness to pay, the authors did not recommend that a gym-based pregnancy were equally effective in preventing incontinence. However, 12-month maintenance exercise program be implemented as stan- when ICERs for each case of incontinence prevented were compared, a dard care. After completion of a 6-week supervised gym-based ex- difference in cost-effectiveness was found. The ICER for pelvic floor ercise program for people with a chronic health disease, a home- muscle training provided to individual patients in the postpartum based maintenance program provides much better value for money period cost the health service AU$2,281 for each case of incontinence than a gym-based program. prevented, whereas individual training during pregnancy cost $878 for each case of incontinence prevented. Group exercise classes during Another paper assessing the comparative cost-effectiveness of pregnancy provided the best value with a cost saving of $16, but only if different modes of exercise programs set out to determine the effec- eight or more participants attended the classes. With a group size of tiveness and cost-utility of 2 months of home-based, group-based or five or less, the ICER was similar to individual training. This study individually supervised physiotherapy exercise therapy in patients provides decision-makers with unique high-quality information that with sub-acromial pain.17 This multicentre randomised controlled trial can be used to allocate local health service resources. From a healthcare found that shoulder pain improved equally well at 6 months, regard- service perspective, if the aim is to prevent incontinence, delaying less of treatment mode. Individual supervised physiotherapy sessions pelvic floor muscle training to the postpartum period is low-value and were costlier than group exercise sessions, requiring society to expend inefficient care. Better value is gained by providing physiotherapy to V16,500 for an equivalent gain in QALY, whereas a home-based individual patients during pregnancy, with even higher value realised if shoulder exercise program provided the highest value, saving society a minimum of eight patients are taught and trained in a group. This type V26,000 for QALY gain at 6 months compared with group-based of comparative cost-effectiveness analysis was only possible as the therapy. This paper provides consumers and clinicians with the con- included studies used the same clinical outcome (episodes of inconti- fidence that a home-based shoulder exercise program of prescribed nence) and could be compared directly. shoulder exercises and fortnightly face-to-face follow ups with a physiotherapist can provide similar outcomes as more intensive indi- Cost-utility analyses vidual physiotherapy sessions and group-exercise sessions at signifi- cantly less cost to both the consumer and society. To enable a meaningful comparison of economic value between unrelated health interventions and cohorts, the quality adjusted life Most recently published in the Journal of Physiotherapy was a year (QALY) is the most recognised health economic metric of health economic evaluation whose primary health economic measure choice.14 A QALY is a composite score combining a participant’s was a cost-utility approach. This high-quality, three-arm, multicentre quality of life and their survival status at a set time point. A QALY randomised controlled trial assessed the comparative cost to improve score, also known as a ‘utility’, ranges from 0 (death) to 1 (perfect QALY at 12 months of providing the following treatment modes to health). A health economic analysis that considers costs in the 204 people with persistent gluteal tendinopathy: 14 sessions of context of QALYs is called a cost-utility analysis. The cost-utility ICER outpatient physiotherapy (education plus exercise), a single cortico- is expressed as the number of monetary units saved or needed to be steroid injection or wait-and-see.18 The physiotherapy alone spent to achieve a one-unit improvement in QALY. Four papers in this approach was costlier over the year (AU$8,825) than a corticosteroid article collection provided a cost-utility analysis.15–18 injection ($6,354) and wait-and-see approach ($6,527), yet resulted in significantly better QALY gains compared with the other two ap- Oosterhuis et al conducted a multicentre randomised controlled proaches. However, the conclusions of whether this improvement in trial to assess the cost-utility of providing routine early outpatient- utility outcome provides good value depends on whose perspective based physiotherapy in the first 6 to 8 weeks after discharge from the value is judged from and who is paying for the service. From the hospital following lumbar discectomy.15 This was compared with no healthcare system’s perspective (costs directly accrued by the physiotherapy. The investigators found that physiotherapy did not healthcare system in the 12-month period) compared with a wait- improve functional status, pain or quality of life 6 months after and-see approach, the cost-utility for physiotherapy alone to discectomy surgery. Costing data included post-discharge medical manage persistent gluteal tendinopathy was an additional cost of care, absenteeism and productivity costs. The cost-utility ICER found $4,400 per patient, whereas a single corticosteroid injection could that physiotherapy in the first 2 months after lumbar discectomy save the healthcare system $22,200 per patient. However, from a may save society V100,000 to improve a patient’s QALY at 6 months. societal perspective where productivity costs are also included, However, this ICER value was subject to a high degree of fragility, physiotherapy alone and corticosteroid approaches had similar with sensitivity analyses finding the ICER swinging grossly in the overall costs. Both approaches would cost society more money than a opposite direction with a slight change in the underlying model wait-and-see approach ($37,000 and $240,000), yet because physio- assumptions. This intervention could cost society over V1 million therapy alone provided superior QALY gains, only this approach was for a QALY improvement. This demonstrates a key methodological below a reasonable willingness to pay. feature of a high-quality health economic evaluation. Cost- effectiveness models should be tested under different assump- Combined analyses tions, perspectives and conditions to ensure that the findings are robust and stable. This gives decision-makers greater certainty that Comparing different types of healthcare interventions principally the findings are reliable enough to make policy and resource- via a cost-utility and QALY metric may be an insensitive method for allocation decisions. determining the value of some physiotherapy interventions. Many physiotherapy treatments target short-term clinical outcomes rather An example of sensitivity analyses confirming the primary finding than long-term quality of life or prevention of mortality. For example, is a single-centre controlled trial conducted by Jansons et al.16 pelvic floor muscle training is effective at reducing incontinence
138 Editorial episodes, yet arguably less impactful on long-term quality of life or chronic obstructive pulmonary disease.21 The authors skilfully mortality. To manage this tension between reporting a globally extracted health economic data from the included trials to present accepted yardstick cost-utility evaluation versus physiotherapy ICERs for cost-utility for QALYs gained and cost-effectiveness for outcome-specific cost-effectiveness, three papers in this article other discrete outcomes, such as exercise capacity, exacerbations collection conducted both.19–21 and quality of life. These data were provided from both a healthcare provider and societal perspective. Each trial was A multicentre randomised controlled trial of 362 elderly women ranked accordingly as dominant (both effective and cost saving), with existing incontinence compared 3 months of individual pelvic dominated (both ineffective and more expensive) or more effec- floor muscle training or group training for both cost-effectiveness and tive but costlier than usual care. Four studies were dominant cost-utility.19 This trial found that both interventions were equally (both effective and cost saving), two led to worse outcomes at effective in improving continence, with . 70% of women in both higher costs, with the remaining five studies reporting improved groups experiencing a clinically important reduction in daily incon- outcomes but at an increased cost to the provider. For in- tinence episodes. Both were also equally effective in improving terventions that came with extra cost, the ICER amount was quality of life at 1 year. However, group-based pelvic floor muscle presented giving the decision-maker clear information to deter- training was more cost-effective, giving an ICER of CA$50,000 saved mine whether the cost to provide the intervention is within an to the healthcare provider for each unit of continence improvement amount they are willing to pay for a specific outcome. compared with individual training. Yet, when cost-utility was eval- uated from the provider perspective, the group training cost . As physiotherapy researchers continue to increase the integration $900,000 for QALY gain at 1 year compared with individual-based of health economic evaluations within their clinical trials, it is vital training. It would be up to a decision-maker to consider whether that physiotherapy decision-makers simultaneously upskill in the the excellent cost-effectiveness to prevent incontinence is worth knowledge of health economic terms, methodologies and the inter- more than the unfavourable cost-utility to improve QALY. Comparing pretation of findings. Physiotherapists, clinic owners, hospital treatments only with a cost-utility analysis may hide true and department heads and policy-makers will be increasingly required to worthwhile cost-effectiveness for important outcomes and benefits make evidence-informed decisions in determining which high-value that physiotherapy interventions are effective in treating. physiotherapy services to provide and which low-value services should be ceased. The case for measuring cost-effectiveness with clinically relevant outcomes to physiotherapy treatments in addition to cost-utility is Competing interests: Nil. highlighted in the Journal of Physiotherapy’s Paper of the Year for 2020, a Source(s) of support: Nil. health economic evaluation of one of PEDro’s Top 20 Physiotherapy Acknowledgements: Nil. Trials.20 This was a multicentre, international, randomised controlled Provenance: Invited. Not peer reviewed. trial of 441 people awaiting major abdominal surgery. The authors Correspondence: Mark R Elkins, Centre for Education & Workforce assessed the cost-effectiveness from the hospital’s perspective of pre- Development, Sydney Local Health District, Sydney, Australia. Email: operative physiotherapy education and breathing exercise training to [email protected] prevent a postoperative pulmonary complication compared with an information booklet alone. Cost-utility to improve QALY at 1 year from References surgery was also assessed. Preoperative physiotherapy cost AU$120/ patient to provide but halved postoperative pulmonary complications 1. Boujibar F, et al. J Physiother. 2022;68:218–219. and lead to net savings in downstream hospital ward costs of $550, 2. Reubenson A, et al. J Physiother. 2022;68:153–155. equivalent to a return in investment to the hospital of $8 saved for every 3. Zadro JR, et al. J Physiother. 2022;68:86–88. $1 spent on preoperative physiotherapy. To prevent a pulmonary 4. Chalmers JK, et al. J Physiother. 2022;68:5–6. complication, the ICER was a saving to the hospital of $5,500 for every 5. Hensher M, et al. Soc Sci Med. 2020;266:113420. complication prevented. This result was highly stable with applied 6. Landon SN, et al. BMC Health Serv Res. 2022;22:845. sensitivity analyses. To improve QALY, the ICER was a saving of $27,500, 7. Husereau D, et al. Value Health. 2013;16:231–250. although this finding was vulnerable to changes in model assumptions. 8. Krist MR, et al. J Physiother. 2013;59:15–23. The ability for this intervention to significantly impact long-term 9. Gomes Neto M, et al. Clin Rehabil. 2017;31:651–659. quality of life is uncertain, with a cost-utility approach likely to pro- 10. Bizzini M, et al. Br J Sports Med. 2013;47:803–806. vide an insensitive measure of the value of this intervention. The real 11. Treacy D, et al. J Physiother. 2018;64:41–47. value of preoperative physiotherapy is in its strong effectiveness in 12. Sarkies MN, et al. J Physiother. 2018;64:142–158. preventing a postoperative pulmonary complication in the first week 13. Brennen R, et al. J Physiother. 2021;67:105–114. after major surgery. From the hospital perspective, preoperative 14. De Silva S, et al. J Physiother. 2022;68:58–59. physiotherapy is more effective and saves money compared with usual 15. Oosterhuis T, et al. J Physiother. 2017;63:144–153. care. This is a ‘dominant’ health economic finding representing high- 16. Jansons P, et al. J Physiother. 2018;64:48–54. value physiotherapy care where both the patient and payer benefit 17. Christiansen DH, et al. J Physiother. 2021;67:124–131. from implementing the intervention. 18. Wilson R, et al. J Physiother. 2023;69:35–41. 19. Cacciari LP, et al. J Physiother. 2022;68:191–196. Evaluating the cost-effectiveness of physiotherapy in- 20. Boden I, et al. J Physiother. 2020;66:180–187. terventions via both QALY cost-utilities and specific outcome cost- 21. Leemans G, et al. J Physiother. 2022;68:271–283. effectiveness was a task that Leemans et al set themselves when they analysed 11 randomised controlled trials in a systematic Websites review of physiotherapy treatments for the management of PEDro www.pedro.org.au Choosing Wisely Australia www.choosingwisely.org.au
Journal of Physiotherapy 69 (2023) 148–159 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Research High-velocity power training has similar effects to traditional resistance training for functional performance in older adults: a systematic review Robert T Morrison a, Sue Taylor a, John Buckley b, Craig Twist c, Chris Kite a,d,e,f a Chester Medical School, Faculty of Health, Medicine and Society, University of Chester, University Centre Shrewsbury, Shrewsbury, UK; b The School of Allied Health Professions, Keele University, Staffordshire, UK; c School of Sport and Exercise Sciences, Liverpool John Moores University, Liverpool, UK; d School of Public Health Studies, Faculty of Education, Health and Wellbeing, University of Wolverhampton, UK; e Warwickshire Institute for the Study of Diabetes, Endocrinology and Metabolism (WISDEM), University Hospitals Coventry and Warwickshire NHS Trust, Coventry, UK; f Centre for Sport, Exercise & Life Sciences, Coventry University, Coventry, UK KEY WORDS ABSTRACT Meta-analysis Questions: What is the effect of high-velocity power training (HVPT) compared with traditional resistance Resistance training training (TRT) on functional performance in older adults? What is the quality of intervention reporting for Aging the relevant literature? Design: Systematic review and meta-analysis of randomised controlled trials. Physical functional performance Participants: Older adults (aged . 60 years), regardless of health status, baseline functional capacity or Physical therapy residential status. Interventions: High-velocity power training with the intent to perform the concentric phase as quickly as possible compared with traditional moderate-velocity resistance training performed with a concentric phase of 2 seconds. Outcome measures: Short Physical Performance Battery (SPPB), Timed Up and Go test (TUG), five times sit-to-stand test (5-STS), 30-second sit-to-stand test (30-STS), gait speed tests, static or dynamic balance tests, stair climb tests and walking tests for distance. The quality of intervention reporting was assessed with the Consensus on Exercise Reporting Template (CERT) score. Results: Nineteen trials with 1,055 participants were included in the meta-analysis. Compared with TRT, HVPT had a weak-to- moderate effect on change from baseline scores for the SPPB (SMD 0.27, 95% CI 0.02 to 0.53; low-quality evidence) and TUG (SMD 0.35, 95% CI 0.06 to 0.63; low-quality evidence). The effect of HVPT relative to TRT for other outcomes remained very uncertain. The average CERT score across all trials was 53%, with two trials rated high quality and four rated moderate quality. Conclusion: HVPT had similar effects to TRT for functional performance in older adults, but there is considerable uncertainty in most estimates. HVPT had better effects on the SPPB and TUG, but it is unclear whether the benefit is large enough to be clinically worthwhile. [Morrison RT, Taylor S, Buckley J, Twist C, Kite C (2023) High-velocity power training has similar effects to traditional resistance training for functional performance in older adults: a systematic review. Journal of Physiotherapy 69:148–159] © 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction relationship illustrates how angular velocity and torque determine movement at each anatomical joint.17,18 Functional trajectories are The World Health Organization describes healthy ageing as ‘the influenced by a loss in either force or velocity as people age,19,20 with process of developing and maintaining the functional ability that enables wellbeing in old age’.1 Age-related functional decline is movement velocity considered more essential in determining func- characterised by a loss of independence in completing activities of tional performance.21,22 Consequently, muscle power (the product of daily living,2 which has a detrimental impact on quality of life,3 mortality4 and the incidence of injurious falls.5 Because functional force and velocity) is a critical component of functional performance decline falls under the physical domain of frailty syndrome,6 func- in older individuals23,24 that declines faster than strength over tional trajectories influence frailty risk, making people more vulner- time,25,26 with reductions of up to 6% per year in people aged . 70 able to negative health outcomes.7 Whilst acute functional decline years.27 Addressing deficits in strength or power through resistance can be caused by a medical event, progressive functional decline is training may help to mitigate age-related functional decline.28,29 often latent and can begin as early as the fifth decade of life.8–10 Resistance training refers to exercise where muscular contractions Muscular strength, defined as the ability to apply force to an are resisted by external loads.30 High-velocity power training (HVPT) external resistance or object,11 is a predictor of functional decline,12–14 deteriorating by w15% every decade beyond 50 years of age15 and by is typically performed with lighter loads (0 to 60% of one-repetition 3.4% annually in people aged . 75 years.16 The force-velocity maximum (1RM)) at faster concentric speeds ( 1 second); it aims to enhance the rate of force development and peak muscular power.31,32 To improve maximum strength, traditional moderate- velocity resistance training (TRT) typically employs loads , 60% of https://doi.org/10.1016/j.jphys.2023.05.018 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Research 149 1RM with a concentric movement duration of 2 to 3 seconds.33 While Science, CINAHL and SportDiscus (via EBSCOhost) on 5 February 2021. multiple studies support the utility of both TRT and HVPT for All databases were searched using a customised algorithm (see improving functional performance in older adults,29,34–40 the relative Appendix 1 on the eAddenda). Additionally, a search for theses was effect of each training modality on functional performance is likely to performed using ProQuest Dissertations and Theses Globalb. Grey be task specific.41 This is because health-related domains such as literature was identified using Google Scholarc. To increase search cardiorespiratory fitness, muscular strength, muscular endurance, sensitivity, citation tracking for prior and derivative papers was per- flexibility and body composition,42 as well as skill-related domains formed using Connected Papersd. Finally, retraction or errata from the such as power, agility, balance, speed, coordination and reaction time, included studies was searched for using PubMed. Search results were all influence functional performance to varying degrees.43 Although imported into the Rayyan tool55 for deduplication and study both TRT and HVPT can enhance muscular strength, HVPT is regarded selection. as more helpful for activities of daily living, such as rising from a chair or balance recovery for fall prevention,44,45 which are important Each paper retrieved by the searches was evaluated by two re- components of older individuals’ quality of life, independence and viewers. Trials were included if they met the predetermined eligi- health. bility criteria summarised according to the Population, Intervention, Comparator, Outcome, and Study (PICOS) model54 (Box 1). The initial To date, several systematic reviews have made direct comparisons screening of title and abstract classified papers as eligible, ineligible of the effectiveness of these modalities on older adults’ functional or potentially eligible. Where a reviewer could not exclude a paper performance24,37,46–49 but did not fully address the quality of the from the title and abstract or information was missing, two reviewers intervention reporting, which limits the translation and application inspected the full-text version of the paper and discussed relevant into practice. Two previous reviews were narrative syntheses that did points. Any disagreements were resolved through arbitration by a not quantitatively synthesise study data.24,37 Three additional sys- third member of the review team. The selection process was recorded tematic reviews included a small number of studies in their analyses using the PRISMA flow diagram (Figure 1), with the included studies (six studies,47 three studies46 and four studies49); the data were summarised in Table 3; further details are available in Appendix 2 on possibly incomplete and the small sample sizes reduced confidence in the eAddenda. the findings. A more recent systematic review included 14 trials, but the selection criteria excluded studies where external loading was Assessment of characteristics of studies , 60% of 1RM.48 Although heavier loads are superior to improve maximum strength,50 lower resistance training loads (, 60% of 1RM) Risk of bias still provide substantial increases in muscular strength and physical The risk of bias for each included study was evaluated using function in older adults,36,51 meaning that relevant studies may have been omitted. Further, the authors of that review acknowledged that version 2 of the Cochrane risk of bias tool.56 Bias was assessed the exclusion of balance outcomes was a limitation of their review. based on five domains: randomisation process (selection bias), Although balance is not a functional task, it is important to include as deviations from intended interventions (performance bias), it limits movement confidence52 and is part of the widely used Short missing outcome data (attrition bias), measurement of the outcome Physical Performance Battery (SPPB). (detection bias) and selection of the reported result (reporting bias). This systematic review aimed to rate the quality of reporting of the exercise interventions in the included studies using a specific tool. In Each potential source of bias was graded as high, low or unclear, addition, the Grading of Recommendations Assessment, Develop- along with a justification for each decision in a ’Risk of Bias’ table. ment, and Evaluation (GRADE) method was used to rate the quality Summary judgements were categorised as low risk, some concerns or and strength of the evidence. high risk, according to the Cochrane guidelines.56 Blinding of partici- pants is unfeasible in exercise interventions and all studies scored Therefore, the research questions for this systematic review and ‘high’ on this component; however, this was excluded when consid- meta-analysis were: ering overall performance bias risk. 1. What is the effect of HVPT compared with TRT on functional per- Two reviewers assessed each trial independently with discor- formance in older adults? dances arbitrated by a third team member. Cohen’s k was used to determine inter-rater agreement. Where study details were inade- 2. What is the quality of intervention reporting for the relevant quate, study authors were contacted for clarification. literature? Method Reporting quality Intervention reporting quality was assessed using the Consensus This systematic review was prospectively registered on PROS- PERO, conducted in accordance with the Cochrane guidelines for on Exercise Reporting Template (CERT).57 The CERT was developed systematic reviews of interventions,53 and reported according to the to establish a consensus on reporting exercise interventions that Preferred Reporting Items for Systematic Reviews and Meta-Analyses would not only improve transparency of the research, but also allow (PRISMA) 2020 statement.54 The PRISMA-2020 reporting checklist exercise interventions to be replicated and implemented into and PRISMA-S reporting checklist for searches were utilised (see clinical care.58 It provides guidance on 16 items across seven do- Tables 1 and 2 on the eAddenda). mains that are required to make exercise interventions replicable. Specific domains include materials, provider, delivery, location, Identification and selection of studies dosage, tailoring and compliance. Individual trials were scored on this 16-item checklist with a maximum score of 19. The CERT has Eligible studies had to be randomised trials (including parallel, previously been implemented by grading studies based on per- cluster and crossover designs) and reported in English. No limits were centage of total score.59 Studies satisfying . 75% of criteria were placed on the geographical location or date of publication. Endnote considered to have a high level of reporting, 60 to 74% as moderate X9 softwarea was used to manage the study records retrieved. and , 60% as low. Electronic searches Study design All trials that met the inclusion criteria (Box 1) and reported A comprehensive search was performed of the Cochrane Central Register of Controlled Trials (CENTRAL), PubMed (MEDLINE), Web of quantitative data in their analysis were included. Whilst trials must have compared interventions of HVPT with TRT, data were also included where the same additional interventions were completed concurrently in both groups.
150 Morrison et al: Power or resistance training in the elderly Box 1. Inclusion criteria. used with the intent to perform the concentric portion of an exercise Design as quickly as possible, whereas TRT was any training protocol without the explicit intent to maximise concentric velocity. Randomised controlled trial Participants Outcome measures Trials had to report an objective measure of functional Adults aged . 60 years Intervention performance. Primary outcome measures were the Short Physical Performance Battery (SPPB), Timed Up and Go test (TUG), five times High-velocity power training sit-to-stand test (5-STS), 30-second sit-to-stand test (30-STS), gait Concentric phase performed as quickly as possible speed tests, stair climb tests, walking tests for distance, and static and Comparator dynamic balance tests. Traditional moderate-velocity resistance training Concentric cadence 2 seconds Data analysis Outcome measures Objective measures of functional performance Appendix 2 on the eAddenda details the specific items for data Participants extraction. A standardised form was devised in Microsoft Excele to This review included adults aged . 60 years regardless of health extract baseline and post-intervention measures. One reviewer status, baseline functional capacity or residential status (eg, com- extracted data and a second reviewer independently cross-checked munity dwelling or institutionalised). the accuracy of data extraction for all papers. Subsequently, one Interventions reviewer entered study data into the Cochrane Collaboration’s Review Trials were included where they had an intervention group allo- Manager 5.4 software.f cated HVPT and a comparison group allocated TRT. Where other groups existed, only data pertaining to the HVPT and TRT groups were Where multiple publications existed for the same trial, these were considered. For the purposes of this review, HVPT must have been linked together. The initial paper was used as the primary reference; however, data were extracted from all publications to obtain the Figure 1. Flow of studies through the review. maximum information possible. Where change-from-baseline data were not reported, the Cochrane Revman calculatorf was used to calculate change scores. In cases where the standard deviation (SD) of the change score was not reported, a standard equation53 was used to calculate them in Microsoft Excel. Where possible, the effects of in- terventions on all continuous outcome data were presented as a mean difference (MD) and 95% confidence interval (95% CI). Where studies used different scales to measure the same outcome, data were presented as a standardised mean difference (SMD) and 95% CI. A sensitivity analysis was performed in these cases to remove studies with different scales and present the remaining studies as MD. Im- perial units were converted to metric units where applicable. Only data taken directly before and after the intervention were used. For example, de-training phases at follow-up were excluded as the focus of this review was on intervention effectiveness rather than residual effects. In the case of trials reporting multiple time points, the longest period of intervention was used. In the event of missing data, two attempts were made to contact study authors to obtain missing data, with a response waiting time of 6 weeks from the first contact attempt. Where data could not be obtained,60,61 a sensitivity analysis was intended to be performed to assess the impact of these studies, but only if it were thought that this would introduce significant bias. Statistical heterogeneity was assessed by both visual inspection of forest plots and formal statistical tests. For visual inspection, het- erogeneity was ascertained by the overlap of CIs for each study, with little or no overlap indicating substantial heterogeneity. Formal assessment used the chi-square test to determine whether differ- ences in results were due to chance alone. Heterogeneity was then quantified by the I2 statistic, which shows the percentage of variation across the studies resulting from heterogeneity and not chance.53 In accordance with Cochrane guidelines, I2 values of 0 to 40% were interpreted as low, 30 to 60% as moderate, 50 to 90% as substantial and 75 to 100% as considerable heterogeneity.53 In cases of moderate, substantial or considerable heterogeneity, study effects were inspected, and an attempt was made to explain heterogeneity via removal of outlier studies and subgroup analysis. To explore publi- cation and small study biases, a funnel plot and Egger’s test for asymmetry was used where at least 10 different-sized studies were pooled.62 Where studies were judged to be sufficiently similar, a meta- analysis was conducted by grouping continuous outcome data, with results presented as forest plots. A random-effects model was chosen as it was deemed more generalisable to the wider population.63 Change from baseline scores were the primary focus to account for between-group variability. Post-intervention group differences were included to assess whether the intervention produced a
Research 151 Table 3 Characteristics of included studies. Trial Participants Intervention Outcome measures SPPB (composite) Balachandran 2010 75 n = 21/17 Regimen HVPT TRT Bean 2004 76 Age: 71 y (SD 11) Bean 2009 77 BMI: . 30 kg/m2 Modality: Pneumatic machines 3 x 10 to 12 reps 3 x 10 to 12 reps Bottaro 2007 78 % Female: 100 Period: 15 wk 50 to 80% 1RM 70% 1RM Correa 2012 79 Healthy Session: 40 to 60 min AFAP 2s Drey 2011 69,70 Frequency: 2/wk Englund 2017 71 n = 21/20 Gray 2018 73 Age: 78 y (SD 10) Modality: Bodyweight/vests 3 x 10 reps 3 x 10 reps SPPB (5-STS, static balance, Henwood 2006 66–68 BMI: , 30 kg/m2 Period: 12 wk RPE , 16 NR gait speed) Kelly 2016 80 % Female: 100 Session: 30 min AFAP 2s Lopes 2016 72 Healthy Frequency: 3/wk Marsh 2009 81 Mierzwicki 2020 60 n = 138/138 Modality: Bodyweight/vests 2 x 10 to 12 reps 2 x 10 reps SPPB (composite) Miszko 2003 82 Age: 75 y (SD 10) Period: 16 wk RPE 11 to 16 RPE 11 to 16 BMI: , 30 kg/m2 Session: 45 to 60 min AFAP 2s % Female: 69 Frequency: 3/wk Mobility limitation Modality: Machines 3 x 8 to 10 reps 3 x 8 to 10 reps TUG (8 ft) n = 24/20 Period: 10 wk 40 to 60% 1RM 40 to 60% 1RM 30-STS Age: 66 y (SD 8) Session: NR AFAP 2 to 3 s BMI: , 25 kg/m2 Frequency: 2/wk % Female: 0 Healthy Modality: Machines/bodyweight 3 to 4 x 8 to 12 reps 3 to 4 x 8 to 12 reps 30-STS Period: 12 wk b 8RM to 12RM 8RM to 12RM n = 58/58 AFAP 2s Age: 67 y (SD 5) Session: NR BMI: NR a % Female: 100 Frequency: 2/wk Healthy Modality: Bodyspider machine 2 x 6 to 15 reps 2 x 6 to 15 reps SPPB (5-STS, static balance, n = 69/69 Period: 12 wk c RPE 10 to 16 RPE 10 to 16 gait speed) Age: 77 y (SD 10) AFAP 2 to 3 s BMI: , 30 kg/m2 Session: 25 min % Female: 70 Pre-frail Frequency: 2/wk n = 26/26 Modality: Multi-joint dynamometer 3 x 8 reps 3 x 8 reps SPPB (composite) Age: 65 y (SD 6) Period: 6 wk NR NR TUG BMI: , 30 kg/m2 Session: NR 240 deg/s 75 deg/s % Female: 62 Frequency: 3/wk Healthy Modality: Machines/free-weights 3 x 10 reps 3 x 10 reps TUG (8ft) n = 99/53 Period: 48 wk d 50% 1RM 80% 1RM 30-STS Age: 81 y (SD 10) AFAP 2s BMI: , 30 kg/m2 Session: 60 min % Female: 72 Healthy Frequency: 2/wk n = 67/53 Modality: Machines 3 x 8 reps 3 x 8 reps 5-STS Age: 70 y (SD 9) Period: 24 wk 40 to 75% 1RM 75% 1RM Static balance BMI: , 30 kg/m2 Session: 60 min AFAP 3s 6 m walk (3 variation) % Female: 56 Frequency: 2/wk 400 m walk Healthy 3 x 10 reps Stair climb Modality: Machines/bodyweight 3 x 10 reps 50 to 80% 1RM n = 38/38 Period: 7 wk 2s TU Age: 72 y (SD 10) Session: NR 50 to 80% 1RM 6-minute walk BMI: . 30 kg/m2 Frequency: 2/wk AFAP e Stair climb % Female: 63 Total knee arthroplasty Modality: Machines 3 to 4 x 6 to 8 reps 3 x 6 to 8 reps 6-minute walk Period: 12 wk 40% 1RM 60% 1RM 30-STS n = 55/37 Session: 60 min AFAP 2 to 3 s TUG Age: 67 y (SD 11) Frequency: 3/wk Postural control BMI: , 30 kg/m2 % Female: 100 Modality: Machines 3 x 8 to 10 reps 3 x 8 to 10 reps SPPB (composite) Healthy Period: 12 wk 40 to 70% 1RM 40 to 70% 1RM Session: 60 min AFAP 2 to 3 s n = 45/36 Frequency: 2/wk Age: 75 y (SD 10) BMI: . 30 kg/m2 Modality: Pneumatic machines 2 x 15 to 20 reps 2 to 3 x 5 reps TUG % Female: 70 Period: 10 wk 50% 1RM 87 to 93% 1RM 5-STS Healthy Session: 30 to 40 min AFAP Slow and controlled 30-STS Frequency: 2/wk 10 m fast walk n = 18/14 3 x 6 to 8 reps Age: 86 y (SD 10) Modality: Machines 40% 1RM 3 x 6 to 8 reps CS-PFP balance and BMI: NR Period: 16 wk f AFAP 80% 1RM endurance tests % Female: 71 (Pre-)frail Session: NR w4 s Frequency: 3/wk n = 65/39 Age: 73 y (SD 6) BMI: , 30 kg/m2 % Female: 56 Healthy
152 Morrison et al: Power or resistance training in the elderly Table 3 (Continued) Trial Participants Intervention Outcome measures Regimen HVPT TRT Monteiro 2019 83 N = 80/80 Modality: Machines 3 to 4 x 3 to 6 reps 2 to 3 x 8 to 12 reps TUG (8 ft) Age: 75 y (SD 11) Period: 16 wk 40 to 60% 1RM 60 to 80% 1RM BMI: , 30 kg/m2 Session: NR ,1s Slow to moderate % Female: 100 Frequency: 3/wk Healthy Pamukoff 2014 84 n = 20/15 Modality: Machines 2 x 8 to 10 reps 2 x 8 to 10 reps Forward and lateral lean step Age: 71 y (SD 5) Period: 6 wk 1 x AMRAP 1 x AMRAP recovery (dynamic balance) BMI: . 30 kg/m2 Session: 60 min 50% 1RM 50% 1RM % Female: 45 Frequency: 3/wk AFAP 2 to 3 s Healthy Ramírez-Campillo 2014 85 n = 60/45 Modality: Machines/free-weights 3 x 8 reps 3 x 8 reps TUG (8 ft) Age: 67 y (SD 9) Period: 12 wk 45 to 75% 1RM 75% 1RM 30-STS BMI: . 30 kg/m2 Session: 70 min AFAP 3s 10 m fast walk % Female: 100 Frequency: 3/wk Healthy Richardson 2019 74 n = 54/50 Modality: Machines 3 x 14 reps 3 x 7 reps TUG (8 ft) Age: 67 y (SD 12) Period: 10 wk 40% 1RM 80% 1RM 30-STS BMI: , 30 kg/m2 Session: NR AFAP 2s 6-min walk % Female: 50 Frequency: 1 and 2 /wk Healthy Sayers 2003 61 n = 30/25 Modality: Pneumatic machines 3 x 8 reps 3 x 8 reps Gait speed usual Age: 73 y (SD 2) Period: 16 wk 70% 1RM 70% 1RM gait speed fast BMI: , 30 kg/m2 Session: NR AFAP 2s 10-STS % Female: 100 Frequency: 3/wk Stair climb Healthy Dynamic balance Tiggemann 2016 86 n = 30/25 Modality: Machines 2 to 3 x 8 to 15 reps 2 to 3 x 8 to 15 reps TUG (3 m) Age: 65 y (SD 7) Period: 12 wk RPE 13 to 18 RPE 13 to 18 5-STS BMI: , 30 kg/m2 Session: NR AFAP 2s 6-min walk % Female: 100 Frequency: 2/wk Stair climb Healthy Yoon 2017 87 n = 58/30 Modality: Elastic bands 2 to 3 x 12 to 15 reps 2 to 3 x 8 to 10 reps SPPB (composite) Age: 76 y (SD 6) Period: 12 wk BMI: , 25 kg/m2 Session: NR RPE 12 to 13 RPE 15 to 16 TUG (8 ft) % Female: 100 Frequency: 2/wk Mild cognitive impairment AFAP .2s AFAP = as fast as possible, AMRAP = as many repetitions as possible, BMI = body mass index, CS-PFP = continuous scale physical functional performance, HVPT = high-velocity power training, NR = not reported, RCT = randomised control trial, RM = repetition maximum, RPE = rating of perceived exertion, SPPB = short physical performance battery, TRT = traditional resistance training, TUG = Timed Up and Go, 5-STS = five times sit-to-stand, 10-STS = 10 times sit-to-stand, 30-STS = 30-s sit-to-stand. a Assumed . 30 from body composition. b First 6 weeks TRT. c 8-week vitamin D supplementation. d First 24 weeks TRT. e Open chain exercises 1 s. f First 8 weeks TRT. between-group effect where baseline group differences existed that judged as overall high risk of bias or rated low quality by the CERT may bias the estimate of the treatment effect.64 For the primary checklist. comparison, data were pooled from all relevant trials stratified by functional test outcome. The results section includes meta-analyses Results that revealed statistically significant effects, with all remaining ana- lyses found in Appendix 3 on the eAddenda. Flow of studies through the review The Grading of Recommendations, Assessment, Development and The search yielded 4,129 records, with 2,937 remaining after Evaluation (GRADE) approach was used to assess overall certainty of removal of duplicates. After title and abstract screening, 40 records evidence.65 The GRADE included the assessment of study limitations, remained, reducing to 24 after full-text analysis (Figure 1, Appendices consistency of effect, indirectness, imprecision and publication bias. 2 and 5 on the eAddenda). Included studies were based on 21 All outcomes began as high certainty and were downgraded based on trials.60,61,66–87 Henwood et al66–68 and Drey et al69,70 presented the GRADE criteria. Decisions were justified with notes and com- multiple papers from the same cohort. ments. The degree of certainty was classified as high, moderate, low or very low (Table 4). The authors of five trials were contacted to obtain missing outcome data.60,61,71–73 Unfortunately, after two attempts, there was Subgroup analysis no response from two authors.60,61 Neither trial reported favourable Post hoc subgroup analysis was performed for baseline body mass effects of HVPT over TRT on any outcome: TUG test,60 dynamic bal- ance, stair-climb and gait speed.61 index (BMI), training frequency (, three versus three/week), age (, 70 versus 70 years) and intervention duration (, 12 versus 12 The third paper from the Henwood et al trial66 included a subset weeks). See differences between protocol and review in Appendix 4 of participants who were re-trained after a wash-out period. As this on the eAddenda. paper was omitted from the main meta-analyses, a sensitivity anal- ysis was conducted for each relevant outcome. This same trial66 was Sensitivity analysis the only one to report the backwards walking and 400 m walk tests. To determine the stability of the results in this systematic re- Whilst the effect for the backwards walking test was very uncertain, there was a favourable effect for HVPT in the 400 m walk test (MD view, sensitivity analysis was performed by excluding studies
Research 153 Table 4 Summary of findings table for primary outcomes: high-velocity power training versus traditional lower velocity training. Outcomes (Change from baseline) Difference (95% CI) Number of Certainty of the Comments participants (studies) evidence (GRADE) Short Physical Performance Battery (SPPB) SMD 0.27 (0.02 to 0.53) 245 (6 RCTs) 4422 High-velocity power training may improve SPPB low a,b performance slightly compared with lower velocity training but with some uncertainty Timed Up and Go test (TUG) SMD 0.35 (0.06 to 0.63) 305 (10 RCTs) 4422 High-velocity power training may improve TUG low b,c performance slightly compared with lower velocity training but with some uncertainty Five times sit-to-stand (5-STS) SMD 0.00 (–0.70 to 0.69) 127 (4 RCTs) 4422 High-velocity power training may have little to low b,d no effect on 5-STS performance compared with lower velocity training 30-second sit-to-stand MD 0.96 (–0.49 to 2.41) 179 (6 RCTs) 4422 High-velocity power training may have little to test (30-STS) low b,e no effect on 30-STS performance compared with lower velocity training Static balance SMD 0.36 (–0.33 to 1.04) 126 (4 RCTs) 4422 High-velocity power training may have little to low b,f no effect on static balance compared with lower velocity training Dynamic balance MD –0.23 (–0.85 to 0.38) 41 (2 RCTs) 4222 High-velocity power training may have little to very low b,g no effect on dynamic balance compared with lower velocity training Usual gait speed SMD –0.35 (–1.04 to 0.34) 102 (3 RCTs) 4422 High-velocity power training may have little to low b,h no effect on usual gait speed compared with lower velocity training Fast gait speed SMD 0.08 (–0.64 to 0.79) 68 (2 RCTs) 4222 High-velocity power training may have little to very low b,i no effect on fast gait speed compared with lower velocity training Long walking tests SMD 0.17 (–0.15 to 0.49) 153 (5 RCTs) 4422 High-velocity power training may have little to low b,j no effect on long walking test performance compared with lower velocity training Stair climb tests SMD 0.20 (–0.27 to 0.67) 101 (3 RCTs) 4422 High-velocity power training may have little to low b,k no effect on stair climb performance compared with lower velocity training High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect. RCT = randomised controlled trial. a Downgraded once: five trials unclear/high risk of selection bias, four trials unclear/high risk of performance bias, three trials unclear/high risk of detection bias, and all trials unclear risk of reporting bias. b Downgraded once: imprecision. c Downgraded once: all trials unclear/high risk of selection bias, six trials unclear/high risk of performance bias, seven trials unclear/high risk of detection bias, and all trials unclear risk of reporting bias. d Downgraded once: three trials unclear/high risk of selection bias, three studies unclear/high risk of performance bias, two trials unclear/high risk of detection bias, and all trials unclear risk of reporting bias. e Downgraded once: all trials unclear/high risk of selection bias, detection bias, and reporting bias; five trials unclear/high risk of performance bias. f Downgraded once: three trials high risk of selection bias, three trials unclear/high risk of performance bias, one trial high risk of attrition bias, three trials unclear/high risk of detection bias, all unclear risk of reporting bias. g Downgraded twice: unclear or high risk of bias across all domains. h Downgraded once: two trials high risk of selection bias, two trials unclear risk of performance bias, two trials unclear/high risk of detection bias, and all trials unclear risk of reporting bias. i Downgraded twice: both included trials unclear/high risk of selection bias, performance bias, detection bias, and reporting bias. j Downgraded once: all trials unclear/high risk of selection bias and reporting bias; four trials unclear/high risk of performance bias; five trials high risk of detection bias. k Downgraded once: all trials unclear/high risk of selection and reporting bias; and two trials unclear risk of performance bias. 11.36 seconds), although it was unclear whether the effect was large HVPT and one TRT),60,61,71,75–78,80,84,86 eight trials had three groups (two intervention and one control),69,72,73,79,81,82,85,87 two trials had enough to be clinically worthwhile (95% CI 0.26 to 22.46). four groups (three intervention and one control),66,83 and one trial One study74 trained groups at different frequencies; these data were had five groups (four intervention and one control).74 The sample sizes ranged from 1860 to 138 participants77 randomised. Most treated separately for meta-analysis. The SPPB data from another study71 were excluded from meta-analysis as post-intervention scores trials (k = 19, 90%) were conducted in a university setting (gym- were maximum (12 points) for both HVPT and TRT groups, meaning that nasium/therapy centre). An outpatient physical therapy centre was used in one trial80 and one trial was conducted in a community the magnitude of improvement beyond this score could not be ascer- gymnasium.82 tained. Also, the static balance outcome reported by one study72 was omitted due to having multiple discrete measures on different scales. Characteristics of the included studies Risk of bias The characteristics of studies included in this systematic review Details of the risk of bias assessment for all included trials are are summarised in Table 3 with full details in Appendix 2 on the found in Appendix 2 on the eAddenda. Review authors’ judgements eAddenda. This review included 21 trials with 1,055 participants. are shown in Figures 2 and 3. Inter-rater agreement for this assess- All the included studies were randomised controlled trials (RCTs). ment was calculated as high (Cohen’s k = 0.788, 95% CI 0.515 to There was a total of 57 groups; nine trials had two groups (one 1.000). Overall, 16 (76%) of the trials were judged as having a high
154 Morrison et al: Power or resistance training in the elderly Figure 3. Risk of bias graph: review authors’ judgements about each risk of bias item presented as percentages across all included studies. to be high risk.60,66,72,75,76,78,81–83 In terms of performance bias, six trials (29%) were judged as low risk,69,71,72,77,80,83 eight trials (38%) as unclear risk61,66,74–76,78,79,86 and seven trials (33%) as high risk.60,73,81,82,84,85,87 Most trials (k = 16, 76%) were judged to be at low risk of attrition bias,60,61,66,69,71,72,74–81,83,86 one trial as unclear risk87 and four as high risk.73,82,84,85 For detection bias, most trials (k = 13, 62%) were judged as high risk61,66,72–74,78,79,81–85,87 and seven trials (33%) as low risk.60,69,71,75,77,80,86 All trials were judged to have an unclear risk of reporting bias. Quality of reporting The average CERT score across all trials was 53% (Figure 4; Appendices 6 and 7 on the eAddenda). Two trials (10%) were rated ‘high’ quality,77,81 whilst four trials (19%) were rated ‘moderate’.61,76,80,84 Common downgrade criteria included: not reporting qualifications/experience of those delivering the interven- tion, details for supervision, adherence, replication and study adherence/fidelity. Conversely, almost all trials (k = 18, 86%) provided detail on exercise progression and one trial did not provide an adequate description of the intervention.71 Participants Of the 1,055 participants who were randomised, 224 were lost to follow-up, resulting in follow-up data being available for 831 par- ticipants. Across studies that reported participant sex (k = 20), 70% of the sample were female. All participants were female in nine trials61,72,75,76,79,83,85–87 and all male in one trial.78 The mean participant age ranged from 6571 to 86 years.60 Most trials (k = 17, 81%) had participants that were assumed healthy. Three trials incorporated participants who were classified as pre-frail or frail60,69,77 and one trial included participants who were post-total knee arthroplasty.80 Figure 2. Authors’ judgement of methodological quality using the Cochrane Risk of Interventions Bias 2 tool. The intervention period ranged from 671,84 to 48 weeks,73 with a overall risk of bias and five (24%) were deemed as having some concerns. No trial was judged as having a low overall risk of bias. mean period of 14 weeks. However, only part of the intervention period compared HVPT with TRT in three trials.73,79,82 Training fre- One trial was judged as low risk for selection bias,69 11 trials (52%) quency was twice per week in 12 trials,60,66,69,73,75,78–81,86,87 and three as unclear risk61,71,73,74,77,79,80,84–87 and nine trials (43%) were deemed times/week in nine trials.61,71,72,76,77,82–85 One trial had groups training both once and twice/week74 and these were treated sepa- rately for meta-analysis. Where reported (k = 11), session duration ranged from 2569 to 70 minutes,85 with the most common session duration being 60 minutes (k = 6).66,72,73,81,84 Session duration was not reported in 10 trials.61,71,74,78–80,82,83,86,87 Most studies defined HVPT as ‘a concentric velocity as fast as possible’ (k = 18, 86%) and three trials defined HVPT as a 24 deg/ second concentric velocity,71 , 1 second,83 and 1 second,80 respectively. TRT was performed in nine trials at a concentric veloc- ity of 2 seconds,61,73–77,79,80,86 2 to 3 seconds in five trials,69,72,78,81,84 3 seconds in two trials,66,85 4 seconds in one trial,82 . 2 seconds in one trial,87 and 75 deg/second in one trial.71 TRT was defined in two trials
Research 155 Figure 4. Summary graph of results for the reporting quality evaluation: % of trials rated by each Consensus on Exercise Reporting (CERT) domain. as ‘slow and controlled’ and ‘slow to moderate’ concentric velocities, Effects of interventions respectively.60,83 Ten outcomes were assessed when HVPT was compared with TRT for Interventions used weight-stack resistance machines in nine functional performance in older adults (Tables 4 and 5). The estimates of trials,66,72,74,78,81–84,86 pneumatic resistance machines in three effect were very uncertain for the 5-STS, static balance, dynamic bal- trials,60,61,75 elastic resistance bands in two trials,69,87 bodyweight ance, usual gait speed, fast gait speed, long walking, and stair climb with weighted vests in two trials,76,77 a multi-joint dynamometer in outcomes for change from baseline, or comparison of post-intervention one trial,71 and combined interventions in four trials values (Table 5, and Figures 5 to 11 in Appendix 3 on the eAddenda). (Table 3).73,79,80,85 To determine training intensity, trials used a wide Findings for the 30-STS showed comparable change scores although range of % 1RM or RPE scale values (Table 3). Intensity was not re- post-intervention values revealed an effect in favour of HVPT (Figure 12 ported in one trial.71 in Appendix 3 on the eAddenda). A sensitivity analysis to include additional data from Henwood et al66 for the 5-STS, static balance, usual Outcomes gait speed, fast gait and stair climb outcomes did not influence results (Figures 13 to 17 in Appendix 3 on the eAddenda). The removal of Drey Reported outcomes (Table 3) included the TUG test (k = et al,69 who used a different 5-STS scale, did not influence results when 11),60,71–74,78,80,83,85–87 SPPB composite test score (k = 7),69,71,75–77,81,87 pooled as a MD (MD 0.36 seconds, 95% CI –1.27 to 1.98; three trials, 83 5-STS (k = 5),60,66,69,76,86 30-STS (k = 6),60,72–74,78,79,85 various balance participants, I2 = 85%; Figure 18 in Appendix 3 on the eAddenda). Sub- tests (k = 7),61,66,69,72,76,82,84 endurance-oriented walking tests (k = group analyses revealed a favourable effect for HVPT in the TUG, 30-STS 6),66,72,74,80,82,86 gait speed tests (k = 6),60,61,66,69,76,85 and stair climb and static balance outcomes for higher training frequencies (Appendix 8 tests (k = 4).61,66,80,86 on the eAddenda). Table 5 Effect estimates and heterogeneity for change from baseline and post-intervention values. Outcome (test) Trials Change scores from baseline Post-intervention values N SMD 95% CI I2 N SMD 95% CI I2 (%) Lower Upper (%) Lower Upper 0 SPPB 69,75–77,81,87 245 0.27 0.02 0.53 0 245 0.23 –0.02 0.48 3 0.63 0.50 50 TUG 71–74,78,80,83,85–87 305 0.35 0.06 0.69 33 305 0.27 0.04 0.78 0 2.41 1.64 36 5-STS 66,69,76,86 127 0.00 –0.70 1.04 72 127 0.26 –0.25 0.76 0 0.56 0.69 66 30-STS 72–74,78,79,85 179 0.96a –0.49 0.34 73 179 0.86a 0.08 0.61 47 0.79 0.69 0 Static balance 66,69,76,82 126 0.36 –0.33 0.49 70 126 0.30 –0.15 0.64 0 0.67 0.15 Dynamic balance 72,84 15 –0.48 –1.51 0 15 –0.33 –1.36 Usual gait 66,69,76 102 –0.35 –1.04 64 102 –0.09 –0.80 Fast gait 66,85 68 0.08 –0.64 55 68 0.03 –0.63 Long walking tests 72,74,80,82,86 153 0.17 –0.15 0 153 0.32 –0.00 Stair climb 66,80,86 101 0.20 –0.27 29 101 –0.24 –0.64 SPPB = short physical performance battery, TUG = Timed Up and Go test, 5-STS = five times sit-to-stand, 30-STS = 30-second sit-to-stand. a mean difference.
156 Morrison et al: Power or resistance training in the elderly Outcome SMD (95% CI) Study Random A SPPB Figure 27. Funnel plot for the Timed Up and Go test outcome. Balachandran 2014 score analysis. This resulted in the I2 statistic being reduced to 0% Bean 2004 with the effect maintained (SMD 0.42, 95% CI 0.18 to 0.66). There was Bean 2009 no indication of publication and small studies bias. Visual inspection Drey 2011 of a funnel plot did not find marked asymmetry and Egger’s test did Marsh 2009 not show asymmetry (Figure 27). Yoon 2017 Discussion Pooled This systematic review and meta-analysis provides evidence that B TUG HVPT may be as effective as TRT for functional performance in older adults, but there is still considerable uncertainty. Twenty-one RCTs Bottaro 2007 were included and 19 (1,007 participants) were meta-analysed. How- Englund 2017 ever, most trials lacked pre-registration and were judged to have a high Gray 2018 overall risk of bias with poor intervention reporting quality, resulting in Kelly 2016 the quality of evidence being classified as low to very low. Although Lopes 2016 meta-analyses for two global tests of functional performance, the SPPB Monteiro 2019 and TUG tests, showed weak to moderate effects in favour of HVPT, Ramírez-Campillo 2014 these estimates were imprecise with some uncertainty. Richardson 2019-HF Richardson 2019-LF All of the reported outcomes rely on the integration of several Tiggemann 2016 health and skill-related domains,88 thus the results may have been Yoon 2017 influenced by changes in rate of force development, maximal strength or a combination of both.88–90 Accordingly, individual components of Pooled global tests of functional performance should be reported where ef- fects of muscle power and strength are being investigated, particu- –2 –1 0 1 2 larly where tests incorporate disparate movements.91 However, as functional performance is a multidimensional concept, it is not Favours TRT Favours HVPT possible to identify the relative contribution of power or strength to each individual test component.92,93 The majority of included trials Figure 19. Forest plot showing the effect of high-velocity power training (HVPT) reported positive changes from baseline for both interventions, sug- relative to traditional resistance training (TRT) as change scores for (A) the short gesting that each intervention may provide a distinct benefit. This physical performance battery (SPPB) and (B) Timed Up and Go test (TUG). implies that, in terms of strength or power, functional performance cannot be limited to an either/or strategy and that a combined There were weak to moderate effects for change from baseline approach may produce the best results. Notably, this is consistent scores favouring HVPT for the SPPB and TUG outcomes. Figure 19 with the American College of Sport Medicine position stand for depicts simplified forest plots for these meta-analyses. Detailed for- resistance training progression in older adults,33 which recommends est plots are found in Figure 20 in Appendix 3 on the eAddenda. the concurrent performance of both HVPT (one to three sets, 30 to 60% 1RM, 12 to 15 repetitions) and TRT (one to three sets, 60 to 80% For the SPPB, results were based on pooled SMD as one study 1RM, 8 to 12 repetitions). However, current global physical activity utilised a modified scale.75 Meta-analysis revealed a weak to mod- guidelines do not recommend a specific dose of power training, erate effect in favour of HVPT on change from baseline SPPB score although they recognise the role of multicomponent physical activity (SMD 0.27, 95% CI 0.02 to 0.53; six trials, 245 participants, I2 = 0%; involving aerobic, balance and resistance training activities.94–96 low-certainty evidence; Table 5; Figure 19A; Figure 20 in Appendix 3 on the eAddenda). In sensitivity analyses, this effect was not retained following removal of trials rated as having a high overall risk of bias or removal of trials rated ‘low’ (, 60%) on the CERT checklist (Figures 21 and 22 in Appendix 3 on the eAddenda). A sensitivity analysis was performed because Balachandran et al75 used a circuit training protocol for HVPT but not TRT. The removal of this trial, with the results pooled as a MD, revealed an uncertain effect (MD 0.33, 95% CI 0.00 to 0.67; five trials, 228 participants, I2 = 0%; Figure 23 in Appendix 3 on the eAddenda). Two variants of the TUG test were reported, so results were based on pooled SMD. There was a weak to moderate effect in favour of HVPT for TUG change from baseline scores (SMD 0.35, 95% CI 0.06 to 0.63; 10 trials, 305 participants, I2 = 33%; low-certainty evidence; Table 5; Figure 19B, Figure 24 in Appendix 3 on the eAddenda). This was supported by a weak to moderate effect for HVPT in comparison of post-intervention values (SMD 0.27, 95% CI 0.04 to 0.50, I2 = 3%; low-certainty evidence; Figure 24 in Appendix 3 on the eAddenda). Sensitivity analyses revealed no effect for TUG change scores or post-intervention values when trials rated as having a high overall risk of bias or trials rated ‘low’ on the CERT checklist were removed (Figures 25 and 26 in Appendix 3 on the eAddenda). The largest outlier was removed to explore heterogeneity in the TUG change
Research 157 Training specificity, which is often overlooked in trials comparing mean baseline values in all studies reporting the SPPB outcome were HVPT and TRT, was an important factor that limited interpretation of . 7, which is considered a threshold for functional impairment.121 Individuals with greater functional impairment can still increase the findings of this systematic review. While training adaptation muscle strength and power and may benefit the most from resistance occurs at the specific speeds performed,97 this remains a contentious training.122 It is unknown whether TRT and HVPT may have compa- issue.98 Some authors suggest that the intent to move faster, even at rable effects in this population. slower actual velocities (eg, TRT), may provide comparable changes in To reduce the risk of bias in future trials, pre-registration, disclo- rate of force development.97,99 Yet, focusing on training that mirrors sure of the randomisation process and blinding of assessors should be performed. Moreover, the lack of participants with low physical the demands of a certain task improves functional performance to a functioning highlights the need to include a wider range of baseline greater extent.100–102 The specific physiological adaptations that occur functional capacity in trials. Given the diverse interventions, there is a during functional movement training can be attributed to the Specific need to conduct more studies using similar protocols for machines, Adaptation to Imposed Demands principle103 and are affected by Fitts elastic bands or task-specific training, combined with more specific law.104 As a result, muscle power is compromised until task mastery training for outcomes. This would ensure a consistency of training has occurred.105 This concept has been demonstrated by improve- stimulus and enable a better comparison between HVPT and TRT. ments in rate of force development over a single practice session.106 Finally, the value of either modality may not be fully revealed from a focus on specific functional outcomes; therefore, a broader range of Only three trials in this review incorporated some outcome practice outcomes should be considered for future research. Future trials during the intervention period;76,80,85 one of these used specific sit- should compare the effects of HVPT and TRT on quality of life, to-stand training with weighted vests, reporting a large effect size exercise-related perceived exertion and exercise adherence in older in favour of HVPT.76 Moreover, strength can be developed in a adults. generalised manner and transfers the capacity to generate force In conclusion, this review discovered that there is low to very low across tasks, whereas skill transfer (ie, power) is limited.107 Given the quality evidence that the efficacy of HVPT may be equivalent to TRT protocols for functional performance in older adults, but the true lack of training specificity across all outcomes, it is plausible that the effect remains uncertain due to the high degree of imprecision. There transfer of strength developed from either HVPT or TRT explains their is currently insufficient evidence to recommend HVPT over TRT in practice. Across the included studies, there was a lack of training comparable effect. Adaptations to HVPT may have been attenuated by specificity and diverse protocols, which may have diluted the true effect of either modality. Future research is required to determine a lack of task-specific practice to maximise rate of force develop- whether specific and/or combined protocols may be more favourable ment.108 Hypothetically, higher velocity training may have a role for with either of these individual approaches for enhancing functional specific ‘functional training’;109 purpose-driven exercise that mirrors performance in older individuals. activities of daily living.110 This type of application is important for older adults’ quality of life, mobility and independence;111 however, What is already known on this topic: Functional perfor- data on this topic remain scarce.112 mance is a vital component of quality of life, independence and health in older adults. Both higher and lower velocity resistance Strength curves based on the muscles’ force-angle (torque) rela- training are viable options to improve functional performance. tionship govern the mechanical loading of exercise movements,113 What this study adds: Resistance training at both higher and classified as: ascending, descending and bell-shaped.114 Training lower velocities are similarly effective in improving older adults’ protocols used in the trials included pneumatic machines, weight- functional performance, although the estimated difference in effect was uncertain. High-velocity power training had better stack machines, elastic resistance, body weight and free weights. effects than traditional resistance training on the Short Physical Performance Battery and Timed Up and Go test, but it is unclear Elastic resistance and free weights have different strength curves whether these benefits are large enough to be clinically worthwhile. than resistance machines and are likely to provide a unique training stimulus,115,116 with pneumatic machines being shown to improve Footnotes: a Endnote. Version Endnote X9. Philadelphia, PA: movement velocities when compared with free weights.117 Conse- Clarivate Analytics; 2013. quently, it remains unclear whether a particular method of training b ProQuest Dissertations and Theses Global, www.proquest.com. c Google Scholar, https://scholar.google.com/. had a greater influence on the results. d Connected Papers, www.connectedpapers.com. There were some limitations to this systematic review and meta- e Microsoft Excel, https://office.microsoft.com/excel: Micro- soft; 2018. analysis. Study selection criteria meant that other interventions with f Revman. London: The Cochrane Collaboration; 2020. eAddenda: Tables 1 and 2, Appendices 1 to 8 a power component such as plyometric training or jump training were Ethics approval: N/A. Competing interests: Nil. omitted. Despite a comprehensive search, it is possible that some Source(s) of support: Nil. Acknowledgements: Nil. relevant literature was missed. Only studies published in English were Provenance: Not invited. Peer reviewed. Correspondence: Chris Kite, School of Public Health Studies, included, meaning that some publications in other languages may have Faculty of Education, Health and Wellbeing, University of Wolverhampton, UK. Email: [email protected] been missed. Version 2 of the Cochrane risk of bias tool has poor to References modest inter-rater reliability, meaning that the risk of bias findings should be interpreted with caution. There were some differences be- 1. World Health Organization. World report on ageing and health. Geneva: World Health Organization; 2015. tween the study protocol and the review related to the definition of HVPT, and pre-defined subgroup and sensitivity analyses. These changes did not influence the overall results (see Appendix 4 on the eAddenda). Several factors may influence the overall completeness of the evidence. Limited follow-up in primary studies means that this sys- tematic review included only baseline and immediately post- intervention data and did not seek to evaluate the retained effects of either training modality. Factors outside of functional performance outcomes, such as quality of life, were not considered. There is also a paucity of pragmatic studies that examine real-world outcomes outside of a university or therapy clinic. Most studies used supervised group training sessions that may have influenced the participants’ effort as per the Hawthorne effect.118 The generalisability of the findings were limited by most participants being female, reportedly healthy and without functional impairment at baseline. Although a precise threshold has not been identified,119,120 participants were possibly above the threshold required to see meaningful differences from resistance training, as seen in the relatively small change scores. A lack of functional impairment at baseline across outcomes pre- cluded further subgroup analysis in this regard. For example, the
158 Morrison et al: Power or resistance training in the elderly 2. Jonkman NH, Del Panta V, Hoekstra T, Colpo M, van Schoor NM, Bandinelli S, et al. 35. Latham N, Anderson C, Bennett D, Stretton C. Progressive resistance strength Predicting trajectories of functional decline in 60- to 70-year-old people. Geron- training for physical disability in older people. Cochrane Database Syst Rev. tology. 2018;64:212–221. 2003:CD002759. 3. Simone PM, Haas AL. Frailty, leisure activity and functional status in older adults: 36. Raymond MJ, Bramley-Tzerefos RE, Jeffs KJ, Winter A, Holland AE. Systematic re- relationship with subjective well being. Clin Gerontol. 2013;36:275–293. view of high-intensity progressive resistance strength training of the lower limb compared with other intensities of strength training in older adults. Arch Phys Med 4. Yeh K-P, Lin M-H, Liu L-K, Chen L-Y, Peng L-N, Chen L-K. Functional decline and Rehabil. 2013;94:1458–1472. mortality in long-term care settings: Static and dynamic approach. J Clin Gerontol Geriatr. 2014;5:13–17. 37. Rice J, Keogh J. Power training: can it improve functional performance in older adults? A systematic review. Int J Exerc Sci. 2009;2:6. 5. Ward RE, Leveille SG, Beauchamp MK, Travison T, Alexander N, Jette AM, et al. Functional performance as a predictor of injurious falls in older adults. J Am Geriatr 38. Hvid LG, Strotmeyer ES, Skjødt M, Magnussen LV, Andersen M, Caserotti P. Soc. 2015;63:315–320. Voluntary muscle activation improves with power training and is associated with changes in gait speed in mobility-limited older adults - A randomized controlled 6. Khezrian M, Myint PK, McNeil C, Murray AD. A review of frailty syndrome and its trial. Exp Gerontol. 2016;80:51–56. physical, cognitive and emotional domains in the elderly. Geriatrics. 2017;2. 39. Pereira A, Izquierdo M, Silva AJ, Costa AM, Bastos E, González-Badillo JJ, et al. Ef- 7. Morley JE, Vellas B, van Kan GA, Anker SD, Bauer JM, Bernabei R, et al. Frailty fects of high-speed power training on functional capacity and muscle performance consensus: a call to action. J Am Med Dir Assoc. 2013;14:392–397. in older women. Exp Gerontol. 2012;47:250–255. 8. Brown RT, Diaz-Ramirez LG, Boscardin WJ, Lee SJ, Steinman MA. Functional 40. Reid KF, Martin KI, Doros G, Clark DJ, Hau C, Patten C, et al. Comparative effects of impairment and decline in middle age: a cohort study. Ann Intern Med. light or heavy resistance power training for improving lower extremity power and 2017;167:761–768. physical performance in mobility-limited older adults. J Gerontol A Biol Sci Med Sci. 2015;70:372–378. 9. Wloch EG, Kuh D, Cooper R. Is the hierarchy of loss in functional ability evident in midlife? Findings from a British Birth Cohort. PLoS One. 2016;11:e0155815. 41. Barry BK, Carson RG. The consequences of resistance training for movement con- trol in older adults. J Gerontol A Biol Sci Med Sci. 2004;59:M730–M754. 10. Hébert R. Functional decline in old age. CMAJ. 1997;157:1037–1045. 11. Siff M. Biomechanical foundations of strength and power training. London: Blackwell 42. Varela S, Ayán C, Cancela J. Batteries assessing health related fitness in the elderly: a brief review. Eur Rev Aging Phys Act. 2008;5:97–105. Scientific; 2001. 12. Dos Santos L, Cyrino ES, Antunes M, Santos DA, Sardinha LB. Sarcopenia and 43. Corbin CB, Pangrazi RP, Franks BD. Definitions: Health, fitness, and physical activity. President’s Council on Physical Fitness and Sports Research Digest. 2000. https://eric. physical independence in older adults: the independent and synergic role of ed.gov/?id=ED470696 muscle mass and muscle function. J Cachexia Sarcopenia Muscle. 2017;8:245–250. 13. Francis P, McCormack W, Toomey C, Lyons M, Jakeman P. Muscle strength can 44. Hazell T, Kenno K, Jakobi J. Functional benefit of power training for older adults. better differentiate between gradations of functional performance than muscle J Aging Phys Act. 2007;15:349–359. quality in healthy 50–70 y women. Braz J Phys Ther. 2017;21:457–464. 14. Visser M, Deeg DJ, Lips P, Harris TB, Bouter LM. Skeletal muscle mass and muscle 45. Caserotti P. Strength training in older adults: changes in mechanical muscle strength in relation to lower-extremity performance in older men and women. function and functional performance. Open Sports Sci J. 2010;3. J Am Geriatr Soc. 2000;48:381–386. 15. Larsson L. Histochemical characteristics of human skeletal muscle during aging. 46. Steib S, Schoene D, Pfeifer K. Dose-response relationship of resistance training in Acta Physiol Scand. 1983;117:469–471. older adults: a meta-analysis. Med Sci Sports Exerc. 2010;42:902–914. 16. Milanovic Z, Pantelic S, Trajkovic N, Sporis G, Kostic R, James N. Age-related decrease in physical activity and functional fitness among elderly men and women. 47. Tschopp M, Sattelmayer MK, Hilfiker R. Is power training or conventional resis- Clin Interv Aging. 2013;8:549–556. tance training better for function in elderly persons? A meta-analysis. Age Ageing. 17. Bieryla KA, Anderson DE, Madigan ML. Estimations of relative effort during sit-to- 2011;40:549–556. stand increase when accounting for variations in maximum voluntary torque with joint angle and angular velocity. J Electromyogr Kinesiol. 2009;19:139–144. 48. Orssatto LBdR, de la Rocha Freitas C, Shield AJ, Pinto RS, Trajano GS. Effects of 18. Westing S, Seger J. Eccentric and concentric torque-velocity characteristics, torque resistance training concentric velocity on older adults’ functional capacity: A output comparisons, and gravity effect torque corrections for the quadriceps and systematic review and meta-analysis of randomised trials. Exp Gerontol. 2019;127: hamstring muscles in females. Int J Sports Med. 1989;10:175–180. 110731. 19. Pojednic RM, Clark DJ, Patten C, Reid K, Phillips EM, Fielding RA. The specific contributions of force and velocity to muscle power in older adults. Exp Gerontol. 49. Huber A. The Effects of Power Training for Functional Performance as Compared to 2012;47:608–613. Strength Training in Older Adults: A Meta-Analysis. California State University: 20. Alcazar J, Rodriguez-Lopez C, Ara I, Alfaro-Acha A, Rodriguez-Gomez I, Navarro- Fresno; 2020. Cruz R, et al. Force-velocity profiling in older adults: An adequate tool for the management of functional trajectories with aging. Exp Gerontol. 50. Schoenfeld BJ, Wilson JM, Lowery RP, Krieger JW. Muscular adaptations in low- 2018;108:1–6. versus high-load resistance training: A meta-analysis. Eur J Sports Sci. 2016;16:1– 21. Van Roie E, Verschueren SM, Boonen S, Bogaerts A, Kennis E, Coudyzer W, et al. 10. Force-velocity characteristics of the knee extensors: an indication of the risk for physical frailty in elderly women. Arch Phys Med Rehabil. 2011;92:1827–1832. 51. Watanabe Y, Madarame H, Ogasawara R, Nakazato K, Ishii N. Effect of very low- 22. Dalton BH, Power GA, Vandervoort AA, Rice CL. Power loss is greater in old men intensity resistance training with slow movement on muscle size and strength than young men during fast plantar flexion contractions. J Appl Physiol. in healthy older adults. Clin Physiol Funct Imaging. 2014;34:463–470. 2010;109:1441–1447. 23. Reid KF, Fielding RA. Skeletal muscle power: a critical determinant of physical 52. Liu-Ambrose T, Khan KM, Donaldson MG, Eng JJ, Lord SR, McKay HA. Falls-related functioning in older adults. Exerc Sport Sci Rev. 2012;40:4–12. self-efficacy is independently associated with balance and mobility in older 24. Byrne C, Faure C, Keene DJ, Lamb SE. Ageing, muscle power and physical function: a women with low bone mass. J Gerontol A Biol Sci Med Sci. 2006;61:832–838. systematic review and implications for pragmatic training interventions. Sports Med. 2016;46:1311–1332. 53. Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, et al. 25. Petrella JK, Kim JS, Tuggle SC, Hall SR, Bamman MM. Age differences in knee Cochrane Handbook for Systematic Reviews of Interventions. NJ, USA: John Wiley & extension power, contractile velocity, and fatigability. J Appl Physiol. 2005;98:211– Sons; 2019. 220. 26. Foldvari M, Clark M, Laviolette LC, Bernstein MA, Kaliton D, Castaneda C, et al. 54. Page MJ, Moher D, Bossuyt P, Boutron I, Hoffmann TC, Mulrow CD, et al. PRISMA Association of muscle power with functional status in community-dwelling elderly 2020 explanation and elaboration: updated guidance and exemplars for reporting women. J Gerontol A Biol Sci Med Sci. 2000;55:M192–M199. systematic reviews. BMJ. 2021:372. 27. Clark DJ, Pojednic RM, Reid KF, Patten C, Pasha EP, Phillips EM, et al. Longitudinal decline of neuromuscular activation and power in healthy older adults. J Gerontol A 55. Ouzzani M, Hammady H, Fedorowicz Z, Elmagarmid A. Rayyan—a web and mobile Biol Sci Med Sci. 2013;68:1419–1425. app for systematic reviews. Syst Rev. 2016;5:1–10. 28. Valenzuela PL, Morales JS, Pareja-Galeano H, Izquierdo M, Emanuele E, de la Villa P, et al. Physical strategies to prevent disuse-induced functional decline in the 56. Sterne JA, Savovic J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a elderly. Ageing Res Rev. 2018;47:80–88. revised tool for assessing risk of bias in randomised trials. BMJ. 2019:366. 29. Papa EV, Dong X, Hassan M. Resistance training for activity limitations in older adults with skeletal muscle function deficits: a systematic review. Clin Interv Aging. 57. Slade SC, Dionne CE, Underwood M, Buchbinder R. Consensus on exercise 2017;12:955. reporting template (CERT): explanation and elaboration statement. Brit J Sports 30. Campbell MD, Rutherford ZH. The Role of Physical Activity and Exercise in Man- Med. 2016;50:1428–1437. aging Obesity and Achieving Weight Loss. In: Weaver JU, ed. Practical Guide to Obesity Medicine. Elsevier; 2018:215–230. 58. Slade SC, Dionne CE, Underwood M, Buchbinder R, Beck B, Bennell K, et al. 31. Earles DR, Judge JO, Gunnarsson OT. Velocity training induces power-specific ad- Consensus on exercise reporting template (CERT): modified Delphi study. Phys aptations in highly functioning older adults. Arch Phys Med Rehabil. 2001;82:872– Ther. 2016;96:1514–1524. 878. 32. Fielding RA, LeBrasseur NK, Cuoco A, Bean J, Mizer K, Singh MAF. High-velocity 59. Charlton PC, Drew MK, Mentiplay BF, Grimaldi A, Clark RA. Exercise interventions resistance training increases skeletal muscle peak power in older women. J Am for the prevention and treatment of groin pain and injury in athletes: a critical and Geriatr Soc. 2002;50:655–662. systematic review. Sports Med. 2017;47:2011–2026. 33. Ratamess N, Alvar B, Evetoch T, Housh T, Kibler W, Kraemer W. Progression models in resistance training for healthy adults [ACSM position stand]. Med Sci Sports 60. Mierzwicki JT, Fox MA, Griffith KR, Harrison KM, Holstay DK, Singley NM. Com- Exerc. 2009;41:687–708. parison of high-intensity resistance and power training programs in pre-frail and 34. Liu Cj, Latham NK. Progressive resistance strength training for improving physical frail older adults. Phys Occup Ther Geriatr. 2020;38:271–282. function in older adults. Cochrane Database Syst Rev. 2009:CD002759. 61. Sayers SP, Bean J, Cuoco A, LeBrasseur NK, Jette A, Fielding RA. Changes in function and disability after resistance training: does velocity matter?: a pilot study. Am J Phys Med Rehabil. 2003;82:605–613. 62. Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau J, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of rand- omised controlled trials. BMJ. 2011;343. 63. Borenstein M, Hedges LV, Higgins JP, Rothstein HR. A basic introduction to fixed- effect and random-effects models for meta-analysis. Res Synth Methods. 2010;1:97–111. 64. Clifton L, Clifton DA. The correlation between baseline score and post-intervention score, and its implications for statistical analysis. Trials. 2019;20:1–6. 65. GRADE-Working-Group. Grading quality of evidence and strength of recommen- dations. BMJ. 2004;328(7454):1490. 66. Henwood TR, Taaffe DR. Short-term resistance training and the older adult: the effect of varied programmes for the enhancement of muscle strength and func- tional performance. Clin Physiol Funct Imaging. 2006;26:305–313.
Research 159 67. Henwood TR, Taaffe DR. Detraining and retraining in older adults following long- 90. Wages NP, Simon JE, Clark LA, Amano S, Russ DW, Manini TM, et al. Relative term muscle power or muscle strength specific training. J Gerontol A Biol Sci Med contribution of muscle strength, lean mass, and lower extremity motor function in Sci. 2008;63:751–758. explaining between-person variance in mobility in older adults. BMC Geriatr. 2020;20:1–11. 68. Henwood TR, Riek S, Taaffe DR. Strength versus muscle power-specific resistance training in community-dwelling older adults. J Gerontol A Biol Sci Med Sci. 91. Balachandran A, Signorile JF. How to improve reporting of the Short Physical 2008;63:83–91. Performance Battery scores. J Gerontol A Biol Sci Med Sci. 2015;70:1595–1596. 69. Drey M, Zech A, Freiberger E, Bertsch T, Uter W, Sieber CC, et al. Effects of strength 92. Rutkauskiene L, Lendraitiene E. Batteries assessing functional fitness of older training versus power training on physical performance in prefrail community- people: A brief review. J Complex Health Sci. 2019;2:64–69. dwelling older adults. J Am Geriatr Soc. 2011;59:S66–S67. 93. Beaudart C, Rolland Y, Cruz-Jentoft AJ, Bauer JM, Sieber C, Cooper C, et al. 70. Zech A, Drey M, Freiberger E, Hentschke C, Bauer JM, Sieber CC, et al. Residual Assessment of muscle function and physical performance in daily clinical practice. effects of muscle strength and muscle power training and detraining on physical Calcif Tissue Int. 2019;105:1–14. function in community-dwelling prefrail older adults: a randomized controlled trial. BMC Geriatrics. 2012;12:68–68. 94. Piercy KL, Troiano RP, Ballard RM, Carlson SA, Fulton JE, Galuska DA, et al. The physical activity guidelines for Americans. JAMA. 2018;320:2020–2028. 71. Englund DA, Sharp RL, Selsby JT, Ganesan SS, Franke WD. Resistance training per- formed at distinct angular velocities elicits velocity-specific alterations in muscle 95. Department of Health and Social Care. Physical Activity Guidelines: UK Chief strength and mobility status in older adults. Exp Gerontol. 2017;91:51–56. Medical Officers’ Report. 2019. 72. Lopes PB, Pereira G, Lodovico A, Bento PCB, Rodacki ALF. Strength and power 96. Sims J, Hill K, Hunt S, Haralambous B. Physical activity recommendations for older training effects on lower limb force, functional capacity, and static and dynamic Australians. Australas J Ageing. 2010;29:81–87. balance in older female adults. Rejuvenation Res. 2016;19:385–393. 97. Behm D, Sale D. Velocity specificity of resistance training. Sports Med.1993;15:374–388. 73. Gray M, Powers M, Boyd L, Garver K. Longitudinal comparison of low- and high- 98. Kawamori N, Newton RU. Velocity specificity of resistance training: Actual move- velocity resistance training in relation to body composition and functional fitness of older adults. Aging Clin Exp Res. 2018;30:1465–1473. ment velocity versus intention to move explosively. Strength Cond J. 2006;28:86. 99. Behm D. Velocity specificity in resistance training is determined by intended rather 74. Richardson DL, Duncan MJ, Jimenez A, Juris PM, Clarke ND. Effects of movement velocity and training frequency of resistance exercise on functional perfor- than the actual contraction velocity. Toronto: McMaster University; 1990. mance in older adults: a randomised controlled trial. Eur J Sport Sci. 100. Manini T, Marko M, VanArnam T, Cook S, Fernhall B, Burke J, et al. Efficacy of 2019;19:234–246. resistance and task-specific exercise in older adults who modify tasks of everyday 75. Balachandran A, Krawczyk SN, Potiaumpai M, Signorile JF. High-speed circuit life. J Gerontol A Biol Sci Med Sci. 2007;62:616–623. training vs hypertrophy training to improve physical function in sarcopenic obese 101. Morrissey MC, Harman EA, Johnson MJ. Resistance training modes: specificity and adults: a randomized controlled trial. Exp Gerontol. 2014;60:64–71. effectiveness. Med Sci Sports Exerc. 1995;27:648–660. 102. Reilly T, Morris T, Whyte G. The specificity of training prescription and physio- 76. Bean JF, Herman S, Kiely DK, Frey IC, Leveille SG, Fielding RA, et al. Increased Ve- logical assessment: A review. J Sports Sci. 2009;27:575–589. locity Exercise Specific to Task (InVEST) training: a pilot study exploring effects on 103. Fox EL, Bowers RW, Foss ML. The physiological basis of physical education and leg power, balance, and mobility in community-dwelling older women. J Am athletics. USA: Wm. C. C Brown Pub Dubuque; 1989:204–209. Geriatr Soc. 2004;52:799–804. 104. Fitts PM. The information capacity of the human motor system in controlling the amplitude of movement. J Exp Psychol Gen. 1954;47:381. 77. Bean JF, Kiely DK, LaRose S, O’Neill E, Goldstein R, Frontera WR. Increased velocity 105. Corbin CB, Janz KF, Baptista F. Good health: The power of power. J Phys Educ exercise specific to task training versus the National Institute on Aging’s strength Recreat Dance. 2017;88:28–35. training program: changes in limb power and mobility. J Gerontol A Biol Sci Med Sci. 106. Hinder MR, Schmidt MW, Garry MI, Carroll TJ, Summers JJ. Absence of cross-limb 2009;64A:983–991. transfer of performance gains following ballistic motor practice in older adults. J Appl Physiol. 2011;110:166–175. 78. Bottaro M, Machado SN, Nogueira W, Scales R, Veloso J. Effect of high versus low- 107. Issurin VB. Training transfer: scientific background and insights for practical velocity resistance training on muscular fitness and functional performance in application. Sports Med. 2013;43:675–694. older men. Eur J Appl Physiol. 2007;99:257–264. 108. Maffiuletti NA, Aagaard P, Blazevich AJ, Folland J, Tillin N, Duchateau J. Rate of force development: physiological and methodological considerations. Eur J Appl 79. Correa CS, LaRoche DP, Cadore EL, Reischak-Oliveira A, Bottaro M, Kruel LF, et al. 3 Physiol. 2016;116:1091–1116. different types of strength training in older women. Int J Sports Med. 2012;33:962– 109. Stenger L. What is functional/neuromotor fitness? ACSM’s Health Fit J. 2018;22: 969. 35–43. 110. Thompson WR. Worldwide survey of fitness trends for 2020. ACSMs Health 80. Kelly MA, Finley M, Lichtman SW, Hyland MR, Edeer AO. Comparative analysis of Fit J. 2019;23:10–18. high-velocity versus low-velocity exercise on outcomes after total knee arthro- 111. De Vriendt P, Peersman W, Florus A, Verbeke M, Van de Velde D. Improving health plasty: a randomized clinical trial. J Geriatr Phys Ther. 2016;39:178–189. related quality of life and independence in community dwelling frail older adults through a client-centred and activity-oriented program. A pragmatic randomized 81. Marsh AP, Miller ME, Rejeski WJ, Hutton SL, Kritchevsky SB. Lower extremity controlled trial. J Nutr Health Aging. 2016;20:35–40. muscle function after strength or power training in older adults. J Aging Phys Act. 112. Hazell T, Kenno K, Jakobi J. Functional benefit of power training for older adults. J 2009;17:416–443. Aging Phys Act. 2007;15:349–359. 113. Kulig K, Andrews JG, Hay JG. Human strength curves. Exerc Sport Sci Rev. 1984; 82. Miszko TA, Cress ME, Slade JM, Covey CJ, Agrawal SK, Doerr CE. Effect of strength 12:417–466. and power training on physical function in community-dwelling older adults. 114. McMaster DT, Cronin J, McGuigan M. Forms of variable resistance training. J Gerontol A Biol Sci Med Sci. 2003;58:171–175. Strength Cond J. 2009;31:50–64. 115. Melchiorri G, Rainoldi A. Muscle fatigue induced by two different resistances: 83. Monteiro AM, Bartolomeu RF, Forte P, Carvalho J. The effects of three different Elastic tubing versus weight machines. J Electromyogr Kinesiol. 2011;21:954–959. types of training in funtional fitness and body composition in older women. J Sport 116. Flanagan SP, Salem GJ. Lower extremity joint kinetic responses to external resis Health Res. 2019;11:289–304. tance variations. J Appl Biomech. 2008;24:58–68. 117. Frost DM, Cronin JB, Newton RU. A comparison of the kinematics, kinetics 84. Pamukoff DN, Haakonssen EC, Zaccaria JA, Madigan ML, Miller ME, Marsh AP. The and muscle activity between pneumatic and free weight resistance. Eur J Appl effects of strength and power training on single-step balance recovery in older Physiol. 2008;104:937–956. adults: a preliminary study. Clin Interv Aging. 2014;9:697–704. 118. Sedgwick P, Greenwood N. Understanding the Hawthorne effect. BMJ. 2015:351. 119. Willen C, Sunnerhagen KS, Ekman C, Grimby G. How is walking speed related to 85. Ramírez-Campillo R, Castillo A, Carlos I, Campos-Jara C, Andrade DC, Álvarez C, muscle strength? A study of healthy persons and persons with late effects of et al. High-speed resistance training is more effective than low-speed resistance polio. Arch Phys Med Rehabil. 2004;85:1923–1928. training to increase functional capacity and muscle performance in older women. 120. Rantanen T, Avela J. Leg extension power and walking speed in very old people Exp Gerontol. 2014;58:51–57. living independently. J Gerontol A Biol Sci Med Sci. 1997;52:M225–M231. 121. Brown JD, Lo-Ciganic W-H, Shao H, Pahor M, Manini TM. trajectories of short 86. Tiggemann CL, Dias CP, Radaelli R, Massa JC, Bortoluzzi R, Schoenell MC, et al. Effect physical performance battery are strongly associated with future major mobility of traditional resistance and power training using rated perceived exertion for disability: Results from the LIFE Study. J Clin Med. 2020;9:2332. enhancement of muscle strength, power, and functional performance. Age. 122. Lopez P, Pinto RS, Radaelli R, Rech A, Grazioli R, Izquierdo M, et al. Benefits of 2016;38:42. resistance training in physically frail elderly: a systematic review. Aging Clin Exp Res. 2018;30:889–899. 87. Yoon DH, Kang D, Kim Hj, Kim JS, Song HS, Song W. Effect of elastic band-based high-speed power training on cognitive function, physical performance and muscle strength in older women with mild cognitive impairment. Geriatr Gerontol Int. 2017;17:765–772. 88. Freiberger E, de Vreede P, Schoene D, Rydwik E, Mueller V, Frändin K, et al. Per- formance-based physical function in older community-dwelling persons: a sys- tematic review of instruments. Age Ageing. 2012;41:712–721. 89. Orssatto LB, Bezerra ES, Schoenfeld BJ, Diefenthaeler F. Lean, fast and strong: Determinants of functional performance in the elderly. Clin Biomech. 2020;78, 105073.
Journal of Physiotherapy 69 (2023) 168–174 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Research Pain catastrophising and kinesiophobia mediate pain and physical function improvements with Pilates exercise in chronic low back pain: a mediation analysis of a randomised controlled trial Lianne Wood a,b,c, Geronimo Bejarano d, Ben Csiernik e, Gisela C Miyamoto f, Gemma Mansell g, Jill A Hayden h, Martyn Lewis b, Aidan G Cashin i,j a Spinal Surgical Division, Nottingham University Hospitals NHS Trust, Nottingham, UK; b School of Medicine, Keele University, Newcastle-under-Lyme, UK; c Faculty of Health and Life Sciences, University of Exeter, Exeter, UK; d University of Texas Health Science Center (UTHealth), Austin, Texas, USA; e Department of Health Sciences, Ontario Tech University, Oshawa, Ontario, Canada; f Masters and Doctoral Program in Physical Therapy, Universidade Cidade de São Paulo, São Paulo, Brazil; g School of Psychology, College of Health & Life Sciences, Aston University, Aston Triangle, Birmingham, UK; h Department of Community Health and Epidemiology, Dalhousie University, Halifax, Canada; i Centre for Pain IMPACT, Neuroscience Research Australia, Sydney, Australia; j School of Health Sciences, Faculty of Medicine & Health, University of New South Wales, Sydney, Australia KEY WORDS ABSTRACT Exercise Question: How much are the reductions in pain intensity and improvements in physical function from Pilates Low back pain exercise mediated by changes in pain catastrophising and kinesiophobia? Design: This was a secondary Mediation analysis causal mediation analysis of a four-arm randomised controlled trial testing Pilates exercise dosage (once, Pilates twice or thrice per week) against a booklet control. Participants: Two hundred and fifty-five people with Physical therapy chronic low back pain. Data analysis: All analyses were conducted in R software (version 4.1.2) following a preregistered analysis plan. A directed acyclic graph was constructed to identify potential pre-treatment mediator-outcome confounders. For each mediator model, we estimated the intervention-mediator effect, the mediator-outcome effect, the total natural indirect effect (TNIE), the pure natural direct effect (PNDE), and the total effect (TE). Results: Pain catastrophising mediated the effect of Pilates exercise compared with control on the outcomes pain intensity (TNIE MD –0.21, 95% CI –0.47 to –0.03) and physical function (TNIE MD –0.64, 95% CI –1.20 to –0.18). Kinesiophobia mediated the effect of Pilates exercise compared with control on the outcomes pain intensity (TNIE MD –0.31, 95% CI –0.68 to –0.02) and physical function (TNIE MD –1.06, 95% CI –1.70 to –0.49). The proportion mediated by each mediator was moderate (21 to 55%). Conclusion: Reductions in pain catastrophising and kinesiophobia partially mediated the pathway to improved pain intensity and physical function when using Pilates exercise for chronic low back pain. These psychological components may be important treatment targets for clinicians and researchers to consider when prescribing exercise for chronic low back pain. [Wood L, Bejarano G, Csiernik B, Miyamoto GC, Mansell G, Hayden JA, Lewis M, Cashin AG (2023) Pain catastrophising and kinesiophobia mediate pain and physical function improvements with Pilates exercise in chronic low back pain: a mediation analysis of a randomised controlled trial. Journal of Physiotherapy 69:168–174] © 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction Exercise is a complex intervention that may exert its effects through many plausible biological, psychological and social mechanisms;13 Chronic low back pain (CLBP) is the leading cause of impaired physical function worldwide;1,2 approximately 577 million people are however, the exact mechanisms are not fully understood and require affected with CLBP at any one time.2 CLBP is the most common reason further investigation.14,15 Pilates is one type of exercise therapy that is that people seek healthcare intervention in the United States and United Kingdom.3–7 It is important to provide more effective and increasingly being used and has demonstrated clinically important cost-effective interventions to reduce the socioeconomic impact of CLBP. International guidelines consistently recommend exercise improvements in pain and physical function compared with other therapy as a first-line treatment;8–11 it is cost-effective and provides exercise types.16 Developed by Joseph Pilates in the 1920s, Pilates is a moderate improvements in pain and physical function compared mind-body exercise originally named ‘Contrology’.17 The exercises can with usual care.8,10–12 be performed on a mat or with specialised equipment following six basic principles: centring, concentration, control, precision, flow and breathing.18,19 These six principles may target several psychological mechanisms (eg, kinesiophobia, pain catastrophising)20 or biological https://doi.org/10.1016/j.jphys.2023.05.008 1836-9553/© 2023 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Research 169 mechanisms (eg, muscle strength and control).18,21 Wun et al14 pro- 95% CI 26.4 to 23.0 for twice weekly; MD 23.3, 95% CI 25.0 to 21.6 posed that exercise may work through neuromuscular, psychosocial, for thrice weekly). The full description and results of the RCT have neurophysiological, cardiometabolic and tissue-healing mechanisms. been described elsewhere.20 However, the exact mechanism(s) for how Pilates exercise affects patient-important outcomes are unclear. Interventions Due to their clinical importance to participants and providers, pain Pilates treatment and physical function are agreed core outcomes for randomised All participants in the Pilates arms received supervised individu- controlled trials (RCTs) of exercise therapy for CLBP.22 However, most exercise therapies tested in RCTs are not designed to directly target alised one-to-one mat and apparatus exercises for 6 weeks with pain or physical function (primary outcomes) but likely target medi- varying dosages. Treatment sessions lasted for 1 hour. The three ating variables or mechanisms that are assumed to improve pain or different Pilates groups compared dosages of once weekly, twice physical function.23 The United Kingdom Medical Research Council’s weekly and thrice weekly. There was no clinically significant differ- guidance on the evaluation of complex interventions recommend the ence between these three groups so they were collapsed into one use of process evaluation to explore how interventions create group for this analysis. Most participants adhered to the treatment change.24 Psychological variables (eg, pain catastrophising and kine- dosage, with 85% attendance in the once and twice weekly arms, and siophobia) have increasingly been acknowledged as important medi- 82% attendance in the thrice weekly arm. TIDieR and CERT assess- ating effects in explaining how interventions create change.14,15 These ments were conducted (see Appendices 1 and 2, respectively, on the mediating effect(s) require further evaluation through the use of eAddenda).20,37 mediation analysis to quantify the relationship between the inter- vention, proposed mediator and primary outcome. This is achieved by Control arm deconstructing the total effect, which is the entire effect of the inter- The control group did not receive any additional treatment other vention on the outcome, into indirect effect(s), which operate through the mediator(s), and direct effects, which do not.25 Understanding than an information booklet. The booklet contained recommenda- these mediating variables is important to improve future design, de- tions related to posture and movements of activities of daily living, livery and evaluation of exercise therapy in clinical settings and in RCT information on low back pain and anatomy of the spine and pelvis.20 design.26 Few mediation studies have been performed on exercise They were informed that they would receive Pilates after the 12- therapy RCTs for LBP.27–31 The dataset20 included in this study is one of month follow-up. seven studies that were identified in a review that pre-specified their exercise intervention with potential mechanisms of effect for how Data collection Pilates effected change in pain and physical function outcomes.32 All questionnaires and scales used to assess outcomes and medi- This study aimed to use causal mediation analysis to determine ators were translated and adapted to Brazilian Portuguese and had whether the effect of Pilates exercise on pain intensity and physical acceptable measurement properties,35,38–42 with equivalent results to function was mediated by changes in pain catastrophising and the original versions. All outcomes and mediators were measured at kinesiophobia. baseline and at 6 weeks and 6 months after randomisation. Therefore, the research question for this mediation analysis was: Outcome measures The primary outcomes for the causal mediation analyses were How much are the reductions in pain intensity and improvements in physical function from Pilates exercise mediated by changes in pain intensity, rated on an 11-point Numeric Rating Scale (0 to 10 pain catastrophising and kinesiophobia? points), and physical function on the 24-item Roland-Morris Disability Questionnaire (0 to 24 points) at 6 months after random- Method isation.35,38,39 Higher scores indicated worse outcomes in both instruments. Design and participants Putative mediators This study is reported according to AGReMA (A Guideline for The putative mediators were pain catastrophising (Pain Cata- Reporting Mediation Analyses)33 and is a secondary causal mediation analysis of a four-arm RCT that assessed Pilates exercise dosage (once, strophising Scale: 0 to 52 points) and kinesiophobia (Tampa Scale of twice or thrice per week) in addition to advice compared with advice Kinesiophobia: 17 to 68 points) measured at 6 weeks after random- only. In brief, this RCT was set in a physiotherapy clinic in Sao Paulo, isation. Higher scores demonstrated worse outcomes in both scales. Brazil, with blinding of research staff and statisticians. Participants Both measures are common in research and clinical practice.40–44 were recruited through community advertisements. Randomisation was performed after baseline assessment using computer-generated Confounders random numbers concealed in sealed, opaque envelopes. All groups No confounding of the intervention-mediator and intervention- received advice through an educational booklet and participants were allowed to use their normal medication. Two hundred and ninety-six outcome relationships was assumed due to random allocation of participants with CLBP were included. The primary trial was powered participants. A directed acyclic graph (DAG) was constructed to for 296 participants, but no formal sample size was calculated for the identify potential pre-treatment confounders of the mediator- mediation analysis. Due to the nature of a secondary mediation outcome relationship for each mediator and outcome of interest analysis, patient and public involvement was not possible. (see Appendix 3 on the eAddenda). This was modified according to literature and peer feedback.45,46 Our DAG implied that the following Statistically significant differences in the primary outcomes of pre-treatment confounders required adjustment: sex, age, duration of pain intensity and physical function were found in favour of all Pilates LBP, educational level, use of pain medication, feeling depressed arms at 6 weeks after randomisation in comparison with advice only. (mood) and income level. The analysis included: age and duration of Pain intensity was measured on the Numerical Rating Scale from 0 ‘no LBP (each of which was measured as a continuous variable in years); pain’ to 10 ‘the worst possible pain’34 (MD 21.2, 95% CI 22.2 to 20.3 and previous treatment, use of pain medicine and feeling depressed for once weekly; MD 22.3, 95% CI 23.2 to 21.4 for twice weekly; (each of which was recorded as a dichotomous variable in response to MD 22.1, 95% CI 23.0 to 21.1 for thrice weekly). Physical function was a single yes/no question at baseline). measured with the Roland Morris Disability Questionnaire35,36 using 24 dichotomous questions; higher summed scores indicated greater Data analysis disability (MD 21.9, 95% CI 23.6 to 20.1 for once weekly; MD 24.7, A preregistered analysis plan registered on Open Science Frame- work was followed for the mediation analysis. All analyses were conducted in free softwarea using the ‘mediation’ package.47 The
170 Wood et al: Mediation analysis of Pilates for low back pain Age, symptom duration, previous treatment, medication, depression, pain at baseline, catastrophising at baseline Pain catastrophising at 6 weeks Intervention-mediator effect (Path a) Mediator-outcome effect (Path b) Pilates exercise (once, twice Pure Natural Direct Effect Pain intensity at 6 or thrice weekly) (PNDE) months Figure 1. Single mediator model for pain intensity outcome. The total natural indirect effect (TNIE) is represented by the orange and yellow lines through the mediator pain catastrophising (the combination of paths a and b); the pure natural direct effect (PNDE) is represented by the purple line; and the total effect (TE) is the combination of the orange, yellow and purple lines. The influence of possible confounders is represented by the green lines. primary aim of identifying single mediator mechanisms through but as only small proportions of mediation were found in the single kinesiophobia and pain catastrophising to physical function and pain mediator models, this was not performed. intensity was estimated from single mediator models. A single mediator model was constructed for each outcome (pain intensity Missing data and physical function) and mediator combination, with four models in total. Missing data did not exceed 15% so post hoc sensitivity analyses were not conducted to assess the possible impact of missing data. All For each single mediator model, we estimated the intervention- analyses were conducted on complete cases using listwise deletion. mediator effect, the mediator-outcome effect, the total natural indi- rect effect (TNIE), the pure natural direct effect (PNDE) and the total Interpretation of results effect (TE) were estimated (see Figure 1). The TNIE is the average intervention effect through the mediator; the PNDE is the average To assist in interpreting the size of the mediated effects, the intervention effect that works through all other mechanisms, proportions mediated were classified as: 0 to 20% small, . 20 to 50% excluding the selected mediator; and the TE is the average effect of moderate and . 50% large.48 the intervention on the outcome. The TE is the sum of the TNIE and PNDE on the additive scale (see Figure 1 for example). The proportion Results mediated is the fraction of TE that is explained by TNIE. Participants For each single mediator model, we fitted two regression models: a mediator model and an outcome model. Linear regression models Participants (n = 255) were predominantly female (n = 201, 75%), were used for all analyses because in each case the outcome variable middle-aged (47 years, SD 15) and had a long duration of LBP was measured on a continuous scale.47 We ran the mediator models symptoms (mean 6.44 years, SD 6.91) (Table 1). A total of 6.42% of using linear regression, with treatment allocation as the independent missing data was identified in both the mediation variables at 6 variable and the mediator as the dependent variable, and the baseline weeks (pain catastrophising and kinesiophobia) and 10.47% was values of the mediator as a covariate. Each of the outcome models for identified in the 6-month post-randomisation data for both pain in- physical function and pain intensity used linear regression. The tensity and physical function. Complete case analysis was used, outcome models were constructed with the mediator as the inde- resulting in the reduction of sample size from 296 to 255 (see pendent variable; the outcome as the dependent variable; and the Appendix 4 on the eAddenda). There was no difference between the treatment allocation, baseline values of the meditator and outcome excluded cases in the mediation dataset and the original dataset. variables in addition to the set of observed pre-treatment con- founders as covariates. To improve model flexibility, we included an Table 1 interaction term (treatment allocation with mediator) in the outcome Baseline variables of the meditation dataset compared with the original dataset. models.47 The regression outputs of each model were checked for posterior predictive checks, linearity, homogeneity of variance, Variable Original dataset Mediation dataset collinearity, influential observations and normality of residuals. The ‘mediate’ function47 was used to compute TE, TNIE and PNDE. We (n = 296) (n = 255) used 1,000 bootstrapped simulations to generate 95% confidence in- tervals. Modelling assumptions for linear regression models (linearity Age (y), mean (SD) 48 (15) 47 (15) and normally distributed residuals) were checked using graphical Female, n (%) 224 (76) 201 (76) methods. The ‘Tmint’ function was used to assess the statistical sig- Pain intensity (0 to 10), mean (SD) 6.2 (1.9) 6.3 (1.9) nificance of the intervention-mediator interactions.47 Physical function (0 to 24), mean (SD) 11.7 (5.1) 11.9 (5.1) Possible confounders We conducted sensitivity analyses to determine the robustness of 6.48 (6.93) 6.44 (6.91) the TNIE to bias introduced by residual confounding.47 The ‘medsens’ Duration of symptoms (y), mean (SD) 142 (48) 133 (50) function was used to estimate the magnitude of residual confounding Previous treatment, n (%) 160 (54) 148 (52) that would cause the point estimate of the TNIE to be zero.47 We also Medication use, n (%) 176 (59) 160 (60) repeated the causal mediation analyses using the same four single Feeling depressed, n (%) 25 (11) 25 (11) mediator models comparing the Pilates twice weekly arm only and Pain catastrophising (0 to 52), the control arm. We had planned to use multiple mediator models, 40 (8) 40 (8) mean (SD) Kinesiophobia (17 to 68), mean (SD)
Research 171 Table 2 exercise therapies, explaining around a half of the improvement in Causal mediation analysis of pain intensity at 6 months after randomisation. physical function. Variable Pain catastrophising Kinesiophobia Pain intensity and physical function are two of the agreed core (n = 255) (n = 255) outcomes for LBP research.50 These are the most commonly used primary outcomes in RCTs of exercise and CLBP.23,32,51 In this sec- Mean difference (95% CI) ondary analysis, pain catastrophising and kinesiophobia were both mediators on the pathway to changing pain intensity and physical Intervention-mediator effect –4.17 (–7.17 to –1.17) –4.65 (–6.70 to –2.60) function.20 Increasingly, CLBP is recognised as a condition with (path a) multidimensional effects, with an interaction between physical, Mediator-outcome effect –0.03 (–0.09 to 0.03) –0.00 (–0.09 to 0.09) psychological, social, lifestyle, comorbid health states and non- (path b) modifiable factors (genetics, sex and life stage).52 The fear- TNIE –0.21 (–0.47 to –0.03) –0.31 (–0.64 to –0.05) avoidance model is well-established, and suggests that fear of pain PNDE –0.75 (–1.62 to 0.07) –0.67 (–1.50 to 0.19) drives persistent pain states and pain-related disability (conceptual Proportion mediated (treated) 0.30 (0.03 to 1.45) theory). Kinesiophobia (fear of movement) is a construct included Total effect 0.20 (0.03 to 1.08) –0.98 (–1.79 to –0.14) within this model. The ‘activity’ avoidance model suggests that when –0.96 (–1.75 to –0.17) fear of pain exists, this leads to a conditioned response of increased fear, anxiety and muscle tension.53 More recently, the common-sense TNIE = total natural indirect effect; PNDE = pure natural direct effect. model recognised fear of movement as a natural response to CLBP and suggested that it is a key treatment target for physiotherapists.54 We calculated the correlation between baseline variables and Many interventions for CLBP, such as mind-body, exercise and found kinesiophobia and pain catastrophising to be correlated at 59%, cognitive behavioural treatments, appear to share similar mediating pathways via shared psychological mechanisms such as pain cata- which was insufficiently collinear to prevent the planned study strophising, kinesiophobia, self-efficacy and distress.55 Woby et al56 proceeding (see Appendix 5 on the eAddenda).49 found, in an observational study, that changes in fear-avoidance be- liefs and increased perceptions of control over pain were predictive of Causal mediation analysis changes in physical function in those with CLBP. Previous studies have demonstrated the mediating role of pain catastrophising in Tai Chi The causal mediation analysis showed that changes in pain cata- exercise and aerobic exercise on pain intensity and physical function strophising and kinesiophobia could partially explain how Pilates ex- in comparison with waitlist controls.27,28 Fear-avoidance beliefs ercise reduced pain intensity and improved physical function outcomes. mediated the effect of physical function when individualised exer- Specifically, significant indirect effects were found for pain cata- cises were prescribed in comparison with usual care.31 Other studies strophising (MD –0.21, 95% CI –0.47 to –0.03) and kinesiophobia (MD have found mediating effects of pain self-efficacy in cognitive func- –0.31, 95% CI –0.64 to –0.05) on pain intensity (see Table 2). Significant tional therapy compared with group exercise and education on indirect effects were also found for pain catastrophising (MD –0.64, 95% physical function outcomes.57 Many of the mediation analyses per- CI –1.21 to –0.20) and kinesiophobia (MD –1.06, 95% CI –1.74 to –0.46) formed to date on studies of exercise and psychological interventions on physical function (see Table 3). The proportion of TE mediated by suggest that these psychological mechanisms may explain around 20 pain catastrophising and kinesiophobia was 0.20 (95% CI 0.03 to 1.08) to 30% of the pathways to changes in outcomes.58 This is in contrast and 0.30 (95% CI –0.03 to 1.45) on pain intensity and 0.34 (95% CI 0.07 to the results of this study, which demonstrated that kinesiophobia to 1.61) and 0.55 (95% CI –0.20 to 2.88) on physical function, respec- appeared to moderate half (55%) of the pathway of Pilates exercise to tively. Intervention-mediator interactions were statistically significant improved physical function. All other combinations of mediator- for both pain catastrophising and kinesiophobia for physical function outcome relationships in this study only mediated 20 to 34% of the but not for pain intensity (see Appendix 6 on the eAddenda). Sensitivity pathways to improved pain and physical function. This study adds analyses demonstrated that these effects were likely to be robust to further evidence that kinesiophobia may have an important media- residual confounding (see Figure 2 - middle and far right panels), as tion role in changing the outcome of physical function. However, the moderate confounding (0 to 5 points) would be required to reduce the proportion mediated is strongly influenced by the control group, such TNIE effect to 0. The results in the Pilates twice weekly group only in that if the control group is also likely to change the mediator slightly, comparison with the control arm demonstrated a partial mediating this will result in a smaller proportion mediated; this is in contrast to effect of pain catastrophising and kinesiophobia on physical function a scenario where the control group does not address the mediator at but not on pain intensity (see Appendix 6 on the eAddenda). all and the proportion mediated is likely to be larger. Discussion Although Pilates exercises seem to be more effective than other exercise types for patients with CLBP, there are no studies that have This study aimed to investigate how Pilates exercise reduced pain investigated mediators that may contribute to the overall effects of intensity and improved physical function compared with an educa- Pilates exercise on pain intensity and physical function. Thus, the tional booklet control. Through causal mediation analysis, it found results of this study provide novel results to fill this gap in the liter- that both pain catastrophising and kinesiophobia were treatment ature. Furthermore, this mediation study is a secondary analysis of a mediators explaining a small to moderate proportion of the effect of high-quality RCT with a large sample size, concealed allocation, Pilates exercise on pain intensity and physical function. Changing intention-to-treat analysis, , 15% missing data, and adherence to kinesiophobia appears to be an important treatment target of treatment of . 82%. This analysis provides exploratory findings that warrant further evaluation, as the original RCT was not powered to Table 3 provide definitive evidence on mediators in the original RCT analysis Causal mediation analysis on physical function at 6 months after randomisation. plan. We preregistered the analysis plan for the mediation analysis, and the analysis and results were reported according to AGReMA Variable Pain catastrophising Kinesiophobia recommendations. This study had limitations in that the RCT was not (n = 255) (n = 255) designed to conduct mediation analyses, but proposed mediators were identified and measured a priori, which is a strength of the trial Mean difference (95% CI) design and this analysis. Although the sample size was reduced due to the presence of missing data, this was likely completely at random Intervention-mediator effect –4.18 (–7.19 to –1.17) –4.73 (–6.78 to –2.67) and unlikely to bias results. There may have been residual con- (path a) founding through unmeasured confounders, but sensitivity analyses Mediator-outcome effect –0.00 (–0.12 to 0.11) –0.00 (–0.19 to 0.15) were performed to ensure robustness of these results. (path b) TNIE –0.64 (–1.21 to –0.20) –1.06 (–1.74 to –0.46) PNDE –1.23 (–2.92 to 0.44) –0.82 (–2.42 to 0.83) Proportion mediated (treated) Total effect 0.34 (0.07 to 1.61) 0.55 (0.20 to 2.88) –1.87 (–3.45 to –0.35) –1.88 (–3.34 to –0.39) TNIE = total natural indirect effect; PNDE = pure natural direct effect.
172 Wood et al: Mediation analysis of Pilates for low back pain Figure 2. Summary plots of causal single mediation models. Within the sensitivity plots, the average mediation effects are plotted as a function of the sensitivity parameter (magnitude of residual confounding). A sensitivity parameter of 0 represents null hypothesised levels of residual confounding, and the extremes of –1 and 1 represent maximum hypothesised levels of residual confounding. Grey zones represent 95% confidence limits of the estimated mediation effect across a range of hypothesised levels of residual confounding. CLBP is a condition with multidimensional effects, and it is still the body to movement, and importance of physical activity in re- unclear which components of intervention are needed to improve covery to help further reduce pain catastrophising and fear of patient’s symptoms. The importance of focusing on the bio- movement.64–66 These educational messages could be verbally psychosocial nature of CLBP in treatment prescription has become a communicated during exercise or non-verbally through environ- priority in recent years.3,52,59,60 The findings from this study can help mental cues such as posters in the clinic. clinicians to optimise the provision of Pilates exercise through a biopsychosocial lens by identifying and directly targeting the iden- This mediation analysis contributes to the initial understanding of tified psychological components. Doing so may lead to reductions in the underlying positive effect of the addition of Pilates exercise to pain intensity and improvements in physical function in patients with advice on pain intensity and physical function when compared with CLBP. Clinicians could consider targeting their Pilates exercise treat- advice only. Further research identifying the best methods to target ments to better address these identified psychological components by these components may be useful for improving treatment delivery. first conducting a biopsychosocial assessment,61,62 including the use However, a limited number of potential mediating factors were of patient-reported outcome measures (eg, Tampa Scale of Kinesi- investigated in this study. Future research could also investigate other ophobia or Pain Catastrophising Scale), to identify the importance of key factors that may be important in the process of changing clinical these psychological components in CLBP. A greater understanding outcomes (such as pain self-efficacy or pain-related distress) in of the patient’s biopsychosocial presentation, including primary Pilates exercise-based treatment, as well as physical measures (such contributor factors, will provide greater opportunity for exercise as motor control, strength, range of motion) given the hypothesised prescription and communication to be individualised to the patient mind-body effects. The non-specific effects of therapeutic engage- and their respective goals.62,63 Second, clinicians could consider ment and alliance and their effect on overall outcomes and psycho- optimising their clinical encounter by prescribing or supervising logical aspects of CLBP are increasingly being recognised and may Pilates exercise to better reduce fear of movement and worrisome account for part of the unmeasured mediation effect.67–70 Other thoughts about pain. This could be achieved through both verbal and mediators that may also contribute to the effect on pain intensity and non-verbal communication that emphasises safety and confidence in physical function include personal components such as exercise self- the person’s ability to perform the exercise.62 Clinicians could further efficacy (measured with the exercise self-efficacy scale71) or the pa- support their exercise prescription by providing additional education tient’s locus of control.72 Trials wishing to prospectively capture these regarding the benefit of exercise and physical activity, robustness of data need to ensure that there is a justifiable theoretical basis for the proposed mediators, there are sufficient measurement points to allow
Research 173 evaluation of change, and that the mediator is temporally measured problèmes de santé les plus fréquents dans les soins primaires. Can Fam Physician. in comparison with the outcome variable, such that the impact of the 2018;64:832–840. mediator on a later outcome time point can be assessed.26 Under- 7. Dieleman JL, Cao J, Chapin A, Chen C, Li Z, Liu A, et al. US Health Care Spending by standing the discrete components that mediate pathways of exercise Payer and Health Condition, 1996-2016. JAMA. 2020;323(9):863–884. to changes in pain intensity and physical function will allow clinicians 8. Qaseem A, Wilt TJ, McLean RM, Forceia M. Noninvasive treatments for acute, to more accurately target their interventions to greatest effect. subacute and chronic low back pain: a clinical practice guideline from the Amer- ican College of Physicians. Ann Intern Med. 2017;166(7):514–530. Reductions in pain catastrophising and kinesiophobia mediated 9. Oliveira CB, Maher CG, Pinto RZ, Traeger AC, Lin CWC, Chenot JF, et al. Clinical some of the pathway to improved pain intensity and physical func- practice guidelines for the management of non-specific low back pain in primary tion when using Pilates exercise for CLBP. Kinesiophobia particularly care: an updated overview. Eur Spine J. 2018;27:2791–2803. appeared to mediate a significant proportion of the pathway to 10. NICE. Low back pain and sciatica in over 16s: assessment and management. improved physical function. These psychological components may be Assessment and non-invasive treatments Low back pain and sciatica in over 16s. important treatment targets to consider when tailoring exercise for London: National Institute for Health and Care Excellence; 2016. CLBP. Clinicians should focus on the psychological elements associ- 11. Stochkendahl MJ, Kjaer P, Hartvigsen J, Kongsted A, Aaboe J, Andersen M, et al. ated with CLBP through assessment and targeting of treatments to National Clinical Guidelines for non-surgical treatment of patients with recent these factors, as they may be important intermediate factors to create onset low back pain or lumbar radiculopathy. Eur Spine J. 2018;27:60–75. change in the outcomes pain intensity and physical function. This 12. Miyamoto GC, Lin CWC, Cabral CMN, Van Dongen JM, Van Tulder MW. Cost- study provides exploratory results that suggest further prospective effectiveness of exercise therapy in the treatment of non-specific neck pain and evaluation in a fully-powered RCT is warranted. low back pain: A systematic review with meta-analysis. Br J Sports Med. 2019;53:172–181. What was already known on this topic: Pilates is one type 13. Ferreira ML, Smeets RJEM, Kamper SJ, Ferreira PH, Machado LAC. Can we explain het- of exercise therapy that has been increasingly used and has erogeneity among randomized clinical trials of exercise for chronic back pain? a meta- demonstrated clinically important improvements in low back pain regression analysis of randomized controlled trials. Phys Ther. 2010;90:1383–1403. and its associated dysfunction compared with other exercise 14. Wun A, Kollias P, Jeong H, Rizzo RR, Cashin AG, Bagg MK, et al. Why is exercise types. prescribed for people with chronic low back pain? A review of the mechanisms of What this study adds: Reductions in pain catastrophising and benefit proposed by clinical trialists. Musculoskelet Sci Pract. 2021;51:102307. kinesiophobia partially mediated the pathway to improved pain https://doi.org/10.1016/j.msksp.2020.102307 intensity and physical function when using Pilates exercise for 15. Wood L, Bishop A, Lewis M, Smeets RJEM, Bronfort G, Hayden JA, et al. Treatment chronic low back pain. These psychological components may be targets of exercise for persistent non-specific low back pain: a consensus study. important treatment targets for clinicians and researchers to Physiotherapy. 2021;112:78–86. https://doi.org/10.1016/j.physio.2021.03.005 consider when prescribing exercise for chronic low back pain. 16. Hayden JA, Ellis J, Ogilvie R, Stewart SA, Bagg MK, Stanojevic S, et al. Some types of exercise are more effective than others in people with chronic low back pain: a Footnotes: a R software V4.1.2, R Core Team, Vienna, Austria. network meta-analysis. J Physiother. 2021;67:252–262. eAddenda: Appendices 1 to 6 can be found online at https://doi. 17. Latey P. The Pilates Method: history and philosophy. J Bodyw Mov Ther. org/10.1016/j.jphys.2023.05.008 2001;5:275–282. Ethics approval: No ethical approval was required for this sec- 18. Yamato TP, Maher CG, Saragiotto BT, Hancock MJ, Ostelo RW, Cabral CM, et al. ondary analysis of existing data. All participants signed informed Pilates for low back pain. Cochrane Database Syst Rev. 2015:CD010265. consent documents and ethical approval was granted by the Research 19. Wells C, Kolt GS, Bialocerkowski A. Defining Pilates exercise: A systematic review. Ethics Committee of the Universidade Cidade de São Paulo Complement Ther Med. 2012;20:253–262. (CAAE:29303014.7.0000.0064) for the original trial. 20. Miyamoto GC, Franco KFM, van Dongen JM, Franco YRDS, de Oliveira NTB, Competing interests: The authors have no conflicts of interest to Amaral DDV, et al. Different doses of Pilates-based exercise therapy for chronic low declare. back pain: a randomised controlled trial with economic evaluation. Br J Sports Med. Source(s) of support: LW’s time was funded with an Orthopaedic 2018;52:859–868. Research UK Early Career Fellowship and an NIHR Post-doctoral 21. Hodges PW, Richardson C. Inefficient muscular stabilisation. Spine. 1996;21:2640– bridging fellowship during the time this project was conceived and 2650. undertaken. 22. Chiarotto A, Deyo RA, Terwee CB, Boers M, Buchbinder R, Corbin TP, et al. Core Acknowledgements: Prof Cristina Maria Nunes for sharing the outcome domains for clinical trials in non-specific low back pain. Eur Spine J. trial dataset with LW. 2015;24:1127–1142. Data sharing: The authors are happy to share the code used on 23. Wood L, Foster NE, Lewis M, Bishop A. Exercise interventions for persistent direct request. non-specific low back pain – does matching outcomes to treatment targets Correspondence: Lianne Wood, Faculty of Health and Life Sci- make a difference? A systematic review and meta-analysis. J Pain. 2021;22:107– ences, University of Exeter, UK. Email: [email protected] 126. 24. Skivington K, Matthews L, Simpson SA, Craig P, Baird J, Blazeby JM, et al. A new References framework for developing and evaluating complex interventions: Update of Medical Research Council guidance. BMJ. 2021;374(2018):1–11. 1. Buchbinder R, van Tulder M, Öberg B, Costa LM, Woolf A, Schoene M, et al. Low 25. Cashin AG, Lee H. An introduction to mediation analyses of randomized controlled back pain: a call for action. Lancet. 2018;391(10137):2384–2388. trials. J Clin Epidemiol. 2021;133:161–164. 26. Whittle R, Mansell G, Jellema P, Windt D Van Der. Applying causal mediation 2. James SL, Abate D, Abate KH, Abay SM, Abbafati C, Abbasi N, et al. Global, regional, methods to clinical trial data: What can we learn about why our interventions and national incidence, prevalence, and years lived with disability for 354 Diseases (don’t) work? Eur J Pain. 2017;21:614–622. and Injuries for 195 countries and territories, 1990-2017: A systematic 27. Hall AM, Kamper SJ, Emsley R, Maher CG. Does pain-catastrophising mediate the analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392(10159): effect of tai chi on treatment outcomes for people with low back pain? Complement 1789–1858. Ther Med. 2016;25:61–66. 28. Smeets RJEM, Vlaeyen JWS, Kester ADM, Knottnerus JA. Reduction of pain cata- 3. Hartvigsen J, Hancock MJ, Kongsted A, Louw Q, Ferreira ML, Genevay S, et al. What strophizing mediates the outcome of both physical and cognitive-behavioral low back pain is and why we need to pay attention. Lancet. 2018;391(10137):2356– treatment in chronic low back pain. J Pain. 2006;7:261–271. 2367. 29. Stevens ML, Boyle E, Hartvigsen J, Mansell G, Søgaard K, Jørgensen MB, et al. Mecha- nisms for reducing low back pain: a mediation analysis of a multifaceted intervention 4. Chou R. Clinician’s corner. Will this patient develop persistent disabling low back in workers in elderly care. Int Arch Occup Environ Health. 2019;92:49–58. pain? JAMA. 2015;303(13):1295–1302. 30. Mansell G, Hill JC, Main C, Vowles KE, van der Windt D. Exploring what factors mediate treatment effect: example of the STarT Back study high-risk intervention. 5. Dagenais S, Caro J, Haldeman S. A systematic review of low back pain cost of illness J Pain. 2016;17:1237–1245. studies in the United States and internationally. Spine J. 2008;8:8–20. 31. Mansell G, Hill JC, Main CJ, Von Korff M, van der Windt D. Mediators of treatment effect in the back in action trial. Clin J Pain. 2017;33:811–819. 6. Finley CR, Chan DS, Garrison S, Korownyk C, Kolber MR, Campbell S, et al. What are 32. Wood L, Ogilvie R, Hayden JA. Specifying the treatment targets of exercise in- the most common conditions in primary care? Une revue systématique Les terventions: do we? Br J Sports Med. 2020;54:1235–1236. 33. Lee H, Cashin AG, Lamb SE, Hopewell S, Vansteelandt S, Vanderweele TJ, et al. A guideline for reporting mediation analyses of randomised trials and observa- tional studies: the AGReMA Statement. JAMA. 2021;326(11):1045–1056. 34. Farrar JT, Young JP, LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain. 2001;94:149–158. 35. Nusbaum L, Natour J, Ferraz MB, Goldenberg J. Translation, adaptation and vali- dation of the Roland-Morris questionnaire - Brazil Roland-Morris. Braz J Med Biol Res. 2001;34:203–210. 36. Roland M, Morris R. A study of the natural history of back pain: part I: develop- ment of a reliable and sensitive measure of disability in low-back pain. Spine. 1983;8:141–144. 37. Miyamoto GC, Costa LOP, Galvanin T, Cabral CMN. Efficacy of the addition of modified Pilates exercises to a minimal intervention in patients with chronic low back pain: a randomized controlled trial. Phys Ther. 2013;93:310–320.
174 Wood et al: Mediation analysis of Pilates for low back pain 38. Costa LOP, Maher CG, Latimer J, Ferreira PH, Ferreira ML, Pozzi GC, et al. Clinimetric analysis of a Cognitive Functional Therapy clinical trial. Eur J Pain. testing of three self-report outcome measures for low back pain patients in Brazil: 2020;24:1765–1774. Which one is the best? Spine. 2008;33:2459–2463. 58. Lee H, Hübscher M, Moseley GL, Kamper SJ, Traeger AC, Mansell G, et al. How does pain lead to disability? A systematic review and meta-analysis of mediation studies 39. Costa LOP, Maher CG, Latimer J, Ferriera PH, Pozzi GC, Ribeiro RN. Psychometric in people with back and neck pain. Pain. 2015;156:988–997. Characteristics of the Brazilian-Portuguese Versions of the Functional Rating 59. Cowell I, O’Sullivan P, O’Sullivan K, Poyton R, McGregor A, Murtagh G. Perceptions Index and the Roland Morris Disability Questionnaire. Spine. 2007;32: of physiotherapists towards the management of non-specific chronic low back pain 1902–1907. from a biopsychosocial perspective: A qualitative study. Musculoskelet Sci Pract. 2018;38:113–119. 40. Souza FS de, Marinho C da S, Siqueira FB, Maher CG, Costa LOP. Psychometric 60. Grimus J, Horler C, Hebron C. Building bespoke exercise: The clinical reasoning testing confirms that the Brazilian-Portuguese adaptations, the original versions of processes of physiotherapists when prescribing exercise for persons with muscu- the Fear-Avoidance Beliefs Questionnaire, and the Tampa Scale of Kinesiophobia loskeletal disorders. Musculoskeletal Care. 2022. have similar measurement properties. Spine. 2008;33:1028–1033. 61. Fu Y, Yu G, McNichol E, Marczewski K, Closs SJ. The association between patient– professional partnerships and self-management of chronic back pain: A mixed 41. Sehn F, Chachamovich E, Vidor LP, Dall-Agnol L, de Souza ICC, Torres ILS, et al. methods study. Eur J Pain. 2018;22:1229–1244. Cross-Cultural Adaptation and Validation of the Brazilian Portuguese Version of 62. Cashin AG, Booth J, Mcauley JH, Jones MD, Traeger AC. Making exercise count: the Pain Catastrophizing Scale. Pain Med. 2012;13:1425–1435. Considerations for the role of exercise in back pain treatment. 2022;20:259–270. 63. Holopainen R, Piirainen A, Heinonen A, Karppinen J, O’Sullivan P. From 42. Siqueira FB, Teixeira-Salmela LF, Magalhaes L de C. Analysis of the psychometric “Non-encounters” to autonomic agency. Conceptions of patients with low back properties of the Brazilian version of the Tampa scale for kinesiophobia. Acta pain about their encounters in the health care system. Musculoskelet Care. Ortopaedica Brasil. 2007;15:19–24. 2018;16:269–277. 64. Slade SC, Kent P, Patel S, Bucknall T, Buchbinder R. Barriers to primary care clinician 43. Sullivan MJ. The pain catastrophizing scale: user manual1. Montreal: McGill Uni- adherence to clinical guidelines for the management of low back pain: A systematic versity; 2009:36. review and metasynthesis of qualitative studies. Clin J Pain. 2016;32:800–816. 65. Friedrich M, Cermak T, Maderbacher P. The effect of brochure use versus therapist 44. Roelofs J, Goubert L, Peters ML, Vlaeyen JWS, Crombez G. The Tampa Scale for teaching on patients performing therapeutic exercise and on changes in impair- Kinesiophobia: further examination of psychometric properties in patients with ment status. Phys Ther. 1996;76:1082–1088. chronic low back pain and fibromyalgia. Eur J Pain. 2004;8:495–502. 66. Wand BM, Cashin AG, McAuley JH, Bagg MK, Orange GM, Moseley GL. The Fit-for- Purpose Model: Conceptualizing and managing chronic nonspecific low back pain 45. Wood L, Mansell G, Miyamoto GC, Cabral CMN, Lewis M. Understanding how as an information problem. Phys Ther. 2023:103. Pilates exercise creates an effect on pain and physical function in patients with low 67. Miller CT, Owen PJ, Than CA, Ball J, Sadler K, Piedimonte A, et al. attempting to back pain using secondary mediation analysis. In: International Forum for Back and separate placebo effects from exercise in chronic pain: a systematic review and Neck Pain Research. 2021. meta-analysis. Sports Med. 2022;52:789–816. 68. Meade LB, Bearne LM, Sweeney LH, Alageel SH, Godfrey EL. Behaviour change tech- 46. Imai K, Keele L, Yamamoto T. Identification, Inference and Sensitivity Analysis for niques associated with adherence to prescribed exercise in patients with persistent Causal Mediation Effects. Stat Sci. 2010;25:51–71. musculoskeletal pain: Systematic review. Br J Health Psychol. 2019;24:10–30. 69. Hall AM, Ferreira PH, Maher CG, Latimer J, Ferreira ML. The influence of the 47. Tingley D, Yamamoto T, Hirose K, Keele L, Imai K. mediation : R Package for Causal therapist-patient relationship on treatment outcome in physical rehabilitation: a Mediation Analysis. J Stat Softw. 2014;59:1–38. systematic review. Phys Ther. 2010;90:1099–1110. 70. Ferreira PH, Ferreira ML, Maher CG, Refshauge KM, Latimer J, Adams RD. The 48. Cohen J. Statistical Power Analysis. Curr Dir Psychol Sci. 1992;1:98–101. therapeutic alliance between clinicians and patients predicts outcome in chronic 49. Kim JH. Statistical Results. Korean J Anesthesiol. 2019;72:558–569. low back pain. Phys Ther. 2013;93:470–478. 50. Chiarotto A, Terwee CB, Deyo RA, Boers M, Lin CWC, Buchbinder R, et al. A core 71. Mcauley E, Lox C, Duncan TE. Long-term maintenance of exercise, self-efficacy, and physiological change in older adults. J Gerontol. 1993;48:218–224. outcome set for clinical trials on non-specific low back pain: study protocol for the 72. Härkäpää K, Järvikoski A, Mellin G, Hurri H, Luoma J. Health locus of control beliefs development of a core domain set. Trials. 2014;15:511. and psychological distress as predictors for treatment outcome in low-back pain 51. Gianola S, Frigerio P, Agostini M, Bolotta R, Castellini G, Corbetta D, et al. patients: results of a 3-month follow-up of a controlled intervention study. Pain. Completeness of outcomes description reported in low back pain rehabilitation 1991;46:35–41. interventions: a survey of 185 randomized trials. Physiother Can. 2016;68:267– 274. Websites 52. O’Sullivan P, Caneiro JP, O’Keeffe M, O’Sullivan K. Unraveling the complexity of low back pain. J Orthop Sports Phys Ther. 2016;46:932–937. Open Science Framework protocol https://osf.io/jt6p4/ 53. Vlaeyen JWS, Linton SJ. Fear-avoidance and its consequences in chronic musculo- skeletal pain: a state of the art. Pain. 2000;85:317–332. 54. Bunzli S, Smith A, Schütze R, Clinical M, Lin I, Sullivan PO. Making sense of low back pain and pain-related fear. J Orthop Sports Phys Ther. 2017;47:628–636. 55. Alaiti RK, Castro J, Lee H, Caneiro JP, Vlaeyen JWS, Kamper SJ, et al. What are the mechanisms of action of cognitive-behavioural, mind-body, and exercise-based interventions for pain and disability in people with chronic primary musculo- skeletal pain? Clin J Pain. 2022;38:502–509. 56. Woby SR, Watson PJ, Roach NK, Urmston M. Are changes in fear-avoidance beliefs, catastrophizing, and appraisals of control, predictive of changes in chronic low back pain and disability? Eur J Pain. 2004;8:201–210. 57. O’Neill A, O’Sullivan K, O’Sullivan P, Purtill H, O’Keeffe M. Examining what factors mediate treatment effect in chronic low back pain: A mediation
Journal of Physiotherapy 69 (2023) 175–181 j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s Research Physiotherapists perceive hosting clinical placements in private practice as an investment in future physiotherapists: a mixed-methods study Casey L Peiris a, Vidya Lawton b, Ruth Dunwoodie c, Alison Francis-Cracknell d, Alan Reubenson e, Cherie Wells f a School of Allied Health, Human Services and Sport, La Trobe University, Melbourne, Australia; b Faculty of Medicine, Health and Human Sciences, Macquarie University, Sydney, Australia; c School of Health and Rehabilitation Sciences, The University of Queensland, Brisbane, Australia; d Faculty of Medicine, Nursing and Heath Sciences, Monash University, Melbourne, Australia; e Curtin School of Allied Health, Curtin University, Perth, Australia; f School of Community Health, Charles Sturt University, Port Macquarie, Australia KEY WORDS ABSTRACT Physical therapy Questions: What do private practitioners perceive to be the benefits, barriers, costs and risks of hosting Clinical education physiotherapy students on clinical placement? What models of placement are used and what support would Private practice private practitioners like to enable them to continue hosting students? Design: A national mixed-methods Mixed methods study comprising a survey and four focus groups. Participants: Forty-five private practitioners from six states and territories who host on average 208 students per year (approximately one-third of all physio- therapy private practice placements in Australia) completed the survey. Fourteen practitioners participated in focus groups. Results: Participants reported that hosting placements helped to recruit graduates and assisted private practitioners in developing clinical and educator knowledge and skills. Cost (both time and financial) and difficulties securing a sufficient caseload for students were perceived barriers to hosting placements. Hosting placements was perceived to be low risk for clients due to supervised care, but there was potential for risk to business reputation and income when hosting a poorly performing student. Participants mostly described a graded exposure placement model whereby final-year students progressed from observation to shared care to providing care under supervision. Participants perceived that they could be assisted in hosting placements if they were to receive additional financial and personalised support from universities. Conclusion: Private practitioners perceived hosting students to be beneficial for the practice, the profession, staff and clients; however, they did report them to be costly and time-consuming. Universities are perceived to play a vital role in providing training, support and communication with educators and students for ongoing placement provision. [Peiris CL, Lawton V, Dunwoodie R, Francis-Cracknell A, Reubenson A, Wells C (2023) Physiotherapists perceive hosting clinical placements in private practice as an invest- ment in future physiotherapists: a mixed-methods study. Journal of Physiotherapy 69:175–181] Crown Copyright © 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction university clinics) has increased more than four-fold in the past decade to 44%.3,4 Nevertheless, private practice placements still seem A recent review of the Australian health workforce indicated that approximately 42 to 50% of all physiotherapists in Australia work in underused in physiotherapy clinical education, because 91% of all private practice.1 To ensure work readiness of physiotherapy gradu- physiotherapy placements are in hospital settings4 despite only 31% ates in this setting, the university curriculum should include work- of physiotherapists working in hospital settings.1 integrated learning opportunities for students within private prac- tice services (ie, a private practice placement).2 Physiotherapists in The lack of clinical placements in private practice settings may private practice provide first contact services within the community, where clients (or third-party funders) pay directly for physiotherapy have contributed to research findings that new graduate physio- services that span musculoskeletal, aged care, neurological, cardio- therapists in Australia may be underprepared for private practice respiratory, gender-specific health and paediatric domains. Physio- employment.5–7 This may in turn: influence the quality and efficiency therapists may also elect to host students in their practice and of client care; lead to reduced client satisfaction, retention and provide supervision as clinical educators. The proportion of students who experience a private practice placement in Australia (excluding business income; and result in graduate stress, fatigue and risk of burnout.5 Meanwhile, providing private practice placements for physiotherapy students may give opportunities for physiotherapy graduates to gain the contextual skills to be successful in this setting upon graduation.2 However, previous research in private practice https://doi.org/10.1016/j.jphys.2023.05.009 1836-9553/Crown Copyright © 2023 Published by Elsevier B.V. on behalf of Australian Physiotherapy Association. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
176 Peiris et al: Hosting students in private practice physiotherapy settings has identified barriers such as time and cost associated with (CEMANZ) contact list. The university clinical education managers clinical supervision, limited practitioner willingness, lack of supervi- were asked to email an invitation, which included participant infor- sory training and availability, and space restrictions.8–11 mation, consent and a survey hyperlink to their private practice clinical partners who met the inclusion criteria. A reminder email was A recent study exploring university clinical education manager sent 8 weeks later. Survey participants indicated their willingness to perspectives on physiotherapy clinical placements in private practice participate in focus groups within the survey. Purposive sampling was in Australia outlined a range of private practice placement models, used for focus groups to ensure representation from a mix of owners and suggested that placements in private practice are beneficial and and clinical educators and a mix of metropolitan and rural practices. safe for students, private practices and universities.4 However, uni- Focus group recruitment continued until no new themes were versity clinical education managers identified that there was not identified (data saturation).20 All participants provided informed enough supply for all physiotherapy students to undertake a place- consent prior to participation. ment in private practice, and that they were costly to run for both private practitioners and universities. It was also perceived that Data collection placement quality and experiences in private practice were incon- sistent.4 To enhance placement quality and capacity, university clin- Data were collected through online, self-administered surveys ical education managers felt that they needed sufficient time and using SurveyMonkey softwarea circulated between November 2018 resources to adequately engage and support private practitioners.4 and April 2019. The survey was made up of Likert-response questions, multiple-choice questions and open-ended questions. Participants Current research focusing on the provision of clinical placements were asked to provide their demographic details (age, qualifications, for physiotherapy students in private practice in Australia is experience), private practice characteristics (location, size and pri- limited.8,10,12,13 Researchers have previously interviewed8,12,13 or mary scope of practice), student characteristics (number of student surveyed10 practitioners in specific geographical areas to provide placements provided, types of clinical placement) and the model of valuable insight into benefits, challenges and how to integrate stu- supervision they use. dents into private practice. However, the findings of these studies may not be representative of the experiences of private practice Four semi-structured focus groups were conducted via conference physiotherapists Australia-wide. call to explore private practice manager/clinical educator experiences, as they are less formal than individual interviews and may allow With increasing physiotherapy student numbers and global con- participants to explore and clarify their views in a supportive group.21 cerns over health professional placement shortages,14–18 private Focus groups were facilitated by one researcher (CW) who followed a practice placements provide a viable solution to enhance student flexible interview schedule (Appendix 1 on the eAddenda). All focus preparedness and increase placement capacity in physiotherapy group interviews were audio recorded and transcribed verbatim. programs.2 Wider exploration of the benefits, barriers and enablers influencing private practice placement provision from private prac- Data analysis titioner perspectives may therefore inform recommendations for enhancing future private practice placement capacity and quality For closed survey questions, descriptive statistics were calculated improvement activities. This research aimed to explore the perspec- using frequencies, percentages, means and standard deviations or tives of physiotherapy private practice practitioners who host stu- 95% confidence intervals (CI) for normally distributed data or me- dents on placements. dians and 95% CIs for non-normally distributed data using Statistical Package for Social Sciences (SPSS)b. Therefore, the research questions for this national mixed-methods study were: Focus group data collection and data analysis occurred almost simultaneously so that questions could be added and new topics 1. What models of placement are currently being used in private explored. An interpretive description methodology was followed practice? because this approach is particularly relevant in clinical contexts and health professional education as it moves beyond theory to generate 2. What are the benefits, risks and costs associated with providing knowledge to inform practice.22 Interpretive description is a flexible private practice placements? approach that can help researchers highlight occurrences that stakeholders need to understand and directs analysis to answer 3. What are the barriers and enabling factors associated with questions that enhance the understanding of complexities that will providing private practice placements? assist clinical educators, private practice owners and universities achieve their goals.22 Method Nvivo softwarec was used to manage qualitative data and under- Design take analyses. Three researchers (CW, CP, VL) independently read transcripts to code and organise the data. Researchers examined data A mixed methods study was conducted, including a nationwide and assigned codes to portions of text, then looked for interactions survey and focus group interviews of physiotherapy private practi- and links between codes to develop themes and sub-themes. Re- tioners who have hosted physiotherapy students on clinical place- searchers then met to discuss and resolve any differences in inter- ment. University Human Research Ethics Committee approval was pretation and categorisation of themes, and collaborated to come to received prior to commencement of the study (H18214). consensus on generating interpretations to address each research question in a process of inductive reasoning using constant compar- Participants ative analysis. Participants were physiotherapists registered with the Australian As researchers cannot avoid taking their own experiences into Health Practitioner Regulation Agency (AHPRA), working in private qualitative research, the researchers’ backgrounds were described to practice as a practitioner who had supervised one or more entry-level enhance reflexivity. All researchers had a background in physio- physiotherapy students on a private practice clinical placement therapy and were either current or previous university clinical edu- within the past 3 years (ie, a clinical educator) or was an owner of a cation managers. Four researchers had previously worked in private practice that hosts students. Complete sampling was used to ensure a practice (CP, VL, AR, CW) and two had PhD qualifications (CP, CW). representative sample within the bounds of willingness to partici- The researchers’ technical knowledge of clinical education and pri- pate. Snowballing was used to maximise response rates.19 Clinical vate practice brought insight to the interpretive description analysis education managers from all Australian universities that offered process. entry-level physiotherapy degrees in 2018 (n = 22) were identified through the Clinical Educator Managers Australia and New Zealand
Research 177 Results Table 1 n = 45 Participant characteristics. Participants 25:20 Surveys 40 (9) Forty-five private practitioners from six different states and ter- 7:38 ritories in Australia completed the survey (Table 1). Practitioners had Gender, n male:female a mean of 17 (SD 9) years of experience and reported that they (or Age (y), mean (SD) 15 (33) their practice/group of practices) hosted between one and 37 (median Role in practice, n manager:clinical educator 9 (20) two) physiotherapy students each year. Altogether, these practi- State, n (%) 1 (2) tioners have hosted approximately 208 physiotherapy students each 9 (20) year for the previous 3 years. Fourteen practitioners participated in New South Wales/Australian Capital Territory 11 (24) focus groups, with six from Western Australia (WA), four from Vic- Queensland 17 (9) toria (VIC) and four from New South Wales/Australian Capital Terri- Tasmania tory (NSW/ACT). Victoria 29 (64) Western Australia 5 (11) Placement characteristics and models Years of experience in physiotherapy, mean (SD) Type of practice, n (%) 2 (4) Most placements were for final-year students (78%) and ran for 5 Musculoskeletal/sports 4 (9) weeks (Table 1) and most (86%) practices hosted one student at a Mixed musculoskeletal/sports/paediatrics/aged/ 4 (9) time (Table 2). Not all practices charged patients for student con- 1 (2) sultations; some offered free or reduced cost services, while other gender-specific/neurological 31 (69) practices charged the usual consultation cost (Table 2). Neurological 2 (1 to 37) Men’s and women’s health Student placement experiences in private practice consisted of Aged care 35 (78) graded exposure to client care, including observation, shared client Paediatrics 5 (11) consultations and providing client consultation under supervision Completed any training, n (%) 5 (11) (Table 2) with students progressed based on competence: Students hosted (n/yr), median (range) Year of students, n (%) n = 14 As they get more and more competent and as they develop their skills Year 4/final year only during that placement, they will start to take over the educator’s Year 4/final year and penultimate year placements 7:7 caseload. But because it’s always the educator’s caseload, it’s always Years 1 and 2 (introductory/observational placements) only one-on-one, and the educator is always there with that student for 4 (29) that patient for every single consultation. (NSW/ACT2) Focus group interviews 4 (29) 6 (43) Participants suggested a range of ideas for improving placement Gender, n male:female 37 (6) experiences, such as having longer placements so that students can State, n (%) 14 (5) become ingrained in the team, see a range of clients, and have more 2 (1 to 37) time to settle in and become familiar with the practice and processes. New South Wales/Australian Capital Territory Other suggestions included exposing students to private practice Victoria earlier via observation and simulation and ensuring that educators Western Australia have clear guidelines of how to host and supervise a student on Age (y), mean (SD) placement. Experience (y), mean (SD) Students hosted (n/year), median (range) Benefits for practice, staff and clients Participants highlighted how hosting students on placement Survey participants identified that hosting students benefited the connected them with local universities and assisted graduate practice by facilitating staff professional development and client care recruitment: (Figure 1). Participants explained how supervising students improved an educator’s clinical reasoning, teaching and research skills: There are so many good experts in the field that, all of a sudden, just being involved with the university gets you access to. (NSW/ACT2) By actually having to explain what we’re doing there, we’re devel- oping our own clinical reasoning skills . And I saw a major devel- We’ve had a lot of our previous students end up being full-time opment in their skill set, their clinical reasoning, their knowledge physiotherapists at our company. (WA5) base very, very quickly and quicker than I’ve expected from staff who aren’t taking students. (NSW/ACT2) Participants indicated that their motivation for hosting students was altruistic, and that benefits were primarily for the students and the profession: I’m doing it as an investment in future physiotherapists. (NSW/ACT4) Barriers to providing placements Most survey respondents agreed that time and financial costs associated with supervision were barriers to providing clinical placements (Figure 1): The largest cost is our time, and our time is expensive as business owners and as practitioners. (NSW/ACT2) [Staff] have to rethink what they are doing and why they are doing it Of 45 survey participants, 30 (67%) commented that supervision and I think that ultimately leads to a better service. (WA1) time caused loss of productivity and 11 (24%) mentioned that prep- aration time (liaising with universities, training, preparing tutorials) Participants also identified benefits for clients where students was a cost. Focus group participants commented: provided additional attention or treatments, especially for clients with complex needs: The amount of time that staff are putting in and the time that that takes away from their ability to do clinical work is where a lot of the Clients like it, they feel like they are getting two for the price of one. costs comes in with taking students. (VIC2) (WA4) However, eight (18%) survey participants said there were no costs The student has been really beneficial to have in there, you know, as a associated with hosting students and some focus group participants second set of ears and a second set of hands. (VIC3) agreed, saying they can enhance efficiency:
178 Peiris et al: Hosting students in private practice physiotherapy Table 2 n = 45 Supervisors are with them 100% of the time, so we make sure there Characteristics of private practice placement models. are no concerns before we start to let them do a little bit more on 7 (1 to 80) their own. (WA1) Placement characteristics 56 (49 to 62) Participants perceived that the main risks of hosting students Staffing 30 (30 to 45) related to having a poorly performing student. Underperforming full-time equivalent staff (n), median (95% CI) students were perceived to pose a risk to the reputation of their 4.9 (4.7 to 5.0) business, required additional time and support from the educator, Placement modela 37.5 (35 to 38) and thereby lead to reduced earning capacity and business profits: initial student consult time (minutes), mean (95% CI) follow-up student consult time (minutes), median 5 (4 to 7) I’m still running a clinical list and running a business, and doing all (95% CI) the other things that I do on my day to day, and if you suddenly get a length of placement (wks), mean (95% CI) 30 (5 to 40) student that is really not up to date, that just stuffs everything up. placement schedule (hr/wk), median (95% CI) (NSW/ACT4) average student caseload (patients/d), median (95% CI) 30 (20 to 30) 30 (25 to 50) Support needed to facilitate private practitioners to host students Time use by students (% time), median (95% CI)a observing client consultations provided by a 10 (5 exclusively) All survey participants agreed that university training, support physiotherapist 15 (7 exclusively) and financial contribution enhanced their capacity to host students sharing client consultations with a physiotherapist 14 (6 exclusively) (Figure 1). All participants reported access to clinical educator providing client consultations under supervision of a training and support from universities, while 47% (21/45) reported physiotherapist 15 (6 exclusively) receiving financial contributions towards placement costs. However, almost one-third of survey participants (14/45) reported that no one How clients are billed when they see a student, na,b 10 (1 exclusively) at their practice had attended training and there was confusion about client consultations are provided free of charge inconsistency of payments across universities. Participants explained client consultations are provided at a lower cost 5 (1 exclusively) that they couldn’t attend training because of time, distance and lost client consultations are provided at usual physiotherapist productivity, and given that supervising students already took up a cost 24 (55) considerable amount of their time they were unlikely to dedicate any client consultations are provided at usual physiotherapist 14 (32) more time to educator training: cost but are longer 6 (14) additional student consultations are provided free of It gets pretty hard to get anyone to go out and do a general profes- charge sional development, they are actually really focused on their clinical additional student consultations are provided at lower skills and remuneration, and I think therefore, people will probably cost prioritise those in their free time and not the training that’s going to help them be clinical mentors, which is already taking up a consid- Most common educator:student ratios used on private erable amount of their time. (WA5) practice placement, n (%)a Participants suggested that there may be more uptake of univer- 1:1 sity training if financial reimbursement for time was provided, and/or 2:1 free professional development opportunities (not related to clinical 1:2 to 1:5 education) were provided as an incentive to attend. Options for individualised online and face-to-face after-hours training was also a Missing data n = 1. valued by participants. Developing good relationships and commu- b More than one option could be selected. nication with the university was also considered essential for ongoing placement provision and made managing students easier: The costs are minimal . our practice doesn’t lose out on any revenue or any ability to see patients. If anything, that can potentially be more I think one of the most positive things I’ve ever had in my experience efficient. (WA4) is that I actually have a direct contact to one person who knows students well . That made my life very easy. (NSW/ACT2) Difficulties sourcing enough clients for students to see was perceived to be a barrier by most (75%) survey participants. They Discussion explained that some clients declined to be treated by a student because they were paying for and seeking treatment from an expert This mixed methods study provides insight from 45 private not a novice. practice physiotherapists in six states and territories in Australia regarding hosting students on placement. Participants perceived Most survey participants (82%) also identified third-party funding wide-ranging benefits for staff, practice and clients, with minimal restrictions as a barrier to hosting students (Figure 1), as they limited risks of harm for clients due to close supervision by educators. The the ability of the practitioner to bill for clients when a student was time and financial costs of hosting placements and difficulties involved in their care: securing sufficient caseloads for students were perceived as barriers, while university training, support and financial reimbursement I don’t agree with the legislation, because there’s no reason, I don’t assisted practitioners in hosting students. Poorly performing students think, that a student can’t be giving a good enough treatment, were perceived to introduce reputational risk, and therefore need to especially if they’re being supervised. (NSW/ACT3) be carefully managed and supported. This was a larger problem for practices who reported most of their The perceived benefit of placements for student learning and clients to be funded by third-party payers: transition to private practice employment has been suggested in previous research.5,6 This study offers insight into perceptions of key Students can’t be involved in assessment or treatment of any Medi- benefits for the practice and its staff. However, not all clients were care funded patients . and as a result of that, it reduces the number reported to be willing to have student input, which may have of students that we can take on. (WA4) contributed to the challenge of ensuring that students had a sufficient . which really then just isolates it to a privately paying patient, which you know, we only have a small percentage, compared to those other patient groups. (VIC3) Practitioners perceived that changes to legislation were needed to facilitate increased placement capacity and more hands-on oppor- tunities for students. Hosting students was perceived as low risk to clients but potentially a risk to the business The clinical risk of hosting students on placement was perceived to be low, as students are supervised when providing client services:
Search
