Hong Kong Physiotherapy Journal

Letter to the Editor Hong Kong Physiotherapy Journal Vol. 43, No. 1 (2023) 1–2 DOI: 10.1142/S101370252375001X Hong Kong Physiother. J. 2023.43:1-2. Downloaded from worldscientific.com Hong Kong Physiotherapy Journal by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. https://www.worldscientific.com/worldscinet/hkpj Accessible lab manual for physical therapy students Michelle Reinink1,*, James Suchy2,† and Ward Glasoe1,‡ 1Division of Physical Therapy — MMC 388 420 Delaware Street SE, University of Minnesota Minneapolis, MN 55455, USA 2Allied Health Director of Boynton Health 410 Church Street SE, University of Minnesota Minneapolis, MN 55455, USA *[email protected] †[email protected] ‡[email protected] Revised 18 July 2022; Accepted 18 July 2022; Published 10 September 2022 To the Editor: °exibility, and anthropometrics. For the training of physical therapists, we created a manual that e±- Physical therapy students must acquire a sub- ciently instructs these foundational examination stantial body of knowledge and numerous technical and treatment skills. In the spirit of academic col- skills in the classroom in preparation for working legiality, this letter provides educators a direct with patients. Early in the curriculum of physical download link (https://hdl.handle.net/11299/ therapy training, educators are responsible for 227802) to the Clinical Assessment lab manual laying a foundation of basic skills for students to hosted by the University of Minnesota (USA) build upon throughout the curriculum and apply in Digital Conservancy. No registration is required, clinical practice. Because physical therapists rely the download is free, and the PDF ¯le can be heavily upon strong hands-on examination skills, reproduced or adapted for educational use. physical therapy students require early and com- prehensive training on how to assess the basics: The Hong Kong Physiotherapy Journal has an range of motion, strength, muscle length and international scope that reaches a vast audience *Corresponding author. Copyright@2023, Hong Kong Physiotherapy Association. This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 (CC BY- NC-ND) License which permits use, distribution and reproduction, provided that the original work is properly cited, the use is non-commercial and no modi¯cations or adaptations are made. 1

2 M. Reinink et al. Hong Kong Physiother. J. 2023.43:1-2. Downloaded from worldscientific.com within the profession of physical therapy. Its active learning laboratories with clear instructions by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. readership includes physical therapy educators and for learning the basic assessment skills and patient- clinicians who mentor physical therapy students handling techniques. It is not intended to be used in across all areas of clinical practice. As the profes- isolation but instead was developed to be used in sion of physical therapy continues to grow and conjunction with assigned textbook readings, lec- evolve in scope and prominence globally, we tures, and demonstrations. The manual serves as a believe sharing teaching resources stands to bene¯t framework to guide our laboratory sessions, pro- physical therapy educators, their students, and the viding simple-to-follow instructions that allow fac- broader physical therapy community. ulty more time to teach. For students, it is a succinct and helpful resource in learning these skills The manual is authored by three physical and a quick reference for them in their early clinical therapists with clinical backgrounds in orthopedics experiences and in subsequent coursework in the and decades of experience teaching physical thera- curriculum. While all physical therapist training py students at the graduate level. It has been programs teach these skills in the way they ¯nd written for PT6280 Clinical Assessment, a course most e®ective, we o®er our lab manual as a resource Doctor of Physical Therapy students at the Uni- for the instruction of basic hands-on skills for the versity of Minnesota take in the ¯rst year of the examination and treatment of movement system curriculum with anatomy taught as a prerequisite. impairments commonly seen in physical therapy. Thus, the clinical skills in this lab manual were written in such a way to build upon the students' We have no con°icts of interest to disclose. study of human anatomy. This lab manual outlines

Research Paper Hong Kong Physiotherapy Journal Vol. 43, No. 1 (2023) 3–17 DOI: 10.1142/S101370252350004X Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com Hong Kong Physiotherapy Journal by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. https://www.worldscientific.com/worldscinet/hkpj E®ectiveness of exercise programmes in improving physical function and reducing behavioural symptoms of community living older adults with dementia living in Asia, and impact on their informal carers: A systematic review and meta-analysis Yulisna Mutia Sari1,*, Keith D. Hill1, Den-Ching A. Lee1,2 and Elissa Burton3,4 1Rehabilitation, Ageing and Independent Living (RAIL) Research Centre School of Primary and Allied Health Care, Monash University, Melbourne, Victoria, Australia 2School of Primary and Allied Health Care Monash University, Melbourne, Victoria, Australia 3Curtin School of Allied Health, Curtin University Perth, Western Australia, Australia 4enAble Institute, Curtin University Perth, Western Australia, Australia *[email protected] Received 12 July 2022; Accepted 27 October 2022; Published 28 January 2023 Background: There is a growing evidence on the bene¯ts of exercise for older people living with dementia in developed countries. However, cultural, health-care systems and environmental di®erences may impact on the uptake of exercise and outcomes in di®erent regions of the world. Objective: This study synthesised the available evidence examining the e®ectiveness of exercise interven- tions on improving physical function and reducing behavioural symptoms in community-dwelling older people living with dementia in Asia, and the impact on their informal carers. Methods: Six databases were searched to November 2021. Randomised controlled trials (RCTs) or qua- si-experimental studies evaluating exercise interventions for community-dwelling older people with dementia *Corresponding author. Copyright@2023, Hong Kong Physiotherapy Association. This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 (CC BY- NC-ND) License which permits use, distribution and reproduction, provided that the original work is properly cited, the use is non-commercial and no modi¯cations or adaptations are made. 3

Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com 4 Y. M. Sari et al. by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. living in Asia were included. The Cochrane risk-of-bias tool for randomised trials and Downs and Black checklist had been used to assess methodological quality of the studies. Meta-analyses using a ¯xed e®ects model assessed the e®ects of exercise interventions where su±cient data were available. Mean di®erence (MD) with 95% con¯dence interval (CI) was used to pool results. Results: Nine studies (¯ve RCTs) were included (Hong Kong-4, China-1, South Korea-2, Taiwan-1, Indo- nesia-1). Exercise improved dynamic balance [Functional Reach (2 studies, n ¼ 111 people with dementia), MD ¼ 2:61, 95% CI (1.55, 3.67)], but not for the Berg Balance Scale (MD ¼ 1:10, 95% CI [À2.88, 5.07]), Timed Up and Go (MD ¼ À3:47, 95% CI [À7.27, 0.33]) and 5 times sit to stand tests (MD ¼ À1:86, 95% CI [À5.27, 1.54]). Single studies where data could not be pooled showed no e®ect of exercise on behavioural symptoms or impact on informal carers. Conclusion: Exercise appeared to have a bene¯cial e®ect on improving balance performance among older people with dementia living in Asia, however, this evidence is limited and inconsistent, and should be interpreted with caution. Further high-quality large RCTs are necessary for advancing the evidence base of exercise interventions for this population. Keywords: Asia; caregivers; dementia; exercise; physical functional performance. Introduction environment, knowledge/preferences of consumers and health-care systems between non-Asian and Dementia is a worldwide public health concern Asian countries may limit these interventions that primarily a®ects older people. It was the ¯fth being directly implemented with similar e®ects in leading cause of death globally in 2016, contrib- Asian countries. For example, cross-cultural issues uting 28.8 million Disability Adjusted Life Years in°uencing translation of research have been lost.1 The number of people living with dementia reported in reviews of other health conditions,19 worldwide is expected to increase from 55.2 million including falls.20 Therefore, it is likely that research in 2020 to 152 million by 2050.2 More than two- evidence from non-Asian countries may need to be thirds of people living with dementia live in Asia, explored in the Asian context to be directly ap- and countries in Asia are rapidly ageing.3 These plicable and be able to be e®ective in translation factors have seen many of these countries facing into Asian countries. challenges, particularly in providing health-care services that are only starting to focus on ageing There were several additional limitations asso- populations, and more speci¯cally on management ciated with these previous reviews, including of dementia care.4 grouping of studies regardless of the di®erences in settings (long term/residential care, hospital and As well as the cognitive decline in people living community settings), and severity of cognitive with dementia, dementia can lead to impairments impairment (with or without presence of demen- of gait,5 disability and dependency,6 at least 50% tia). The di®erence in settings may in°uence what increase in risk of falls,7 and also increased burden interventions can be undertaken and resources and psychological stress for their family and car- available to support the interventions within a ers,8,9 and increased costs for health-care and so- particular setting. Grouping of people with mild cial-care systems.10,11 All of these factors highlight cognitive impairment (MCI) with those who had a the importance of programmes that can address diagnosed dementia may create substantial het- these issues for people with dementia and their carers. erogeneity in the sample, particularly as MCI does not necessarily lead to dementia.21 In light of these Previous systematic reviews and meta-analyses limitations and the growing need for evidence to have reported signi¯cant positive e®ects of exercise guide dementia care planning in the ageing com- on physical function of older adults with dementia munity of Asia, it is important to evaluate exercise (e.g., balance, mobility and endurance), beha- e®ects focussing on people with dementia who are vioural symptoms and impact on informal car- living in the community, and in an Asian country ers.12–17 Sixty percent of the world's population so that the evidence will be more readily applicable live in Asia,18 however, these reviews have included to this population. predominantly studies from non-Asian countries. Moderate di®erences in the contexts of culture,

Review of exercise programmes in Asia to improve outcomes for people with dementia 5 Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com The aims of this review were to synthesise the commentaries, systematic reviews, case studies, by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. available evidence on the e®ectiveness of exercise qualitative papers, poster abstracts or dissertations. interventions on physical function, and beha- vioural symptoms of community-dwelling older Procedure people with dementia living in Asia, and the im- pact of these programmes on their informal carers. The following electronic databases were searched from inception to November 2021 to identify studies Method meeting the inclusion criteria: CINAHL, Ovid Medline, PubMed, Ovid Embase, PsycINFO and Design Scopus. Grey literature was not included due to the variability of scienti¯c rigour and lack of guidelines This study was a systematic review and meta- in performing grey literature searches systematical- analysis. ly.22–24 Medical Subject Heading (MeSH) terms where appropriate were used in the search. An ex- Inclusion criteria ample of the search strategy used for the Ovid da- tabase is presented in Appendix A. In addition, the The inclusion criteria were that the study (1) was reference lists of all selected articles and excluded conducted in a country within Asia; (2) recruited review articles were checked for relevant articles. community-dwelling people aged 60 years or above (because many parts of Asia have a \\young old\" Three stages were conducted in selecting papers: population, therefore the commonly used age cut- stage one included initially screening the titles by point for older people is de¯ned as people aged ! the ¯rst author (YMS) based on the eligibility 60 years); (3) involved participants with a medical criteria to identify relevant papers. Stage two in- diagnosis of any type of dementia; (4) used an ex- volved (YMS) fully screening the abstracts. This ercise or physical activity programme as an inter- was followed by stage three which involved (YMS) vention for which the e®ect from the intervention screening the full articles then discussing with could be clearly demonstrated; (5) had a control other authors (KDH and EB) on ¯nal papers to group that received usual care, no routine inter- identify whether they met the inclusion criteria vention, placebo; or comparison group(s) that re- and to achieve consensus. The Preferred Reporting ceived other interventions; (6) reported outcomes Items for Systematic Reviews and Meta-Analyses of physical function (the ability to perform activi- (PRISMA) guidelines were adhered to, to ensure ties of daily living, and measures of balance, mo- the results were reported systematically.25 After bility, endurance and strength), behavioural the selection of studies, data were extracted by symptoms (the behavioural and psychological (YMS) using a data extraction form for the char- symptoms experienced by people with dementia acteristics of study design, purpose, intervention, including hallucination, aggression, agitation and participants, sample size, programme dropout rate, inhibition), and/or impact on informal carers dementia severity, outcome measures, ¯ndings/ef- (feelings of burden experienced by informal carers fect of the intervention, intervention duration and of people with dementia related to their physical follow-up (FU) assessment time points. functioning, emotional, ¯nancial and social well- being) and (7) design was a randomised controlled The Cochrane risk-of-bias tool for randomised trial (RCT), pilot RCT or quasi-experimental trials26 was used to assess methodological quality study. of the RCT articles and Downs and Black check- list27 was used to assess the quasi-experimental Exclusion criteria studies. Two researchers assessed each paper in- dependently. Care was taken to ensure the author Studies were excluded if they (1) published in (s) did not review the quality of any co-authored languages other than English; (2) had a study paper. For the Cochrane risk-of-bias tool, the key sample mean age < 60 years, or more than 50% of domains assessed by the tool included (1) sequence the sample aged < 60 years; (3) were letters, generation; (2) allocation concealment; (3) blind- ing of participants, personnel and outcome asses- sors; (4) incomplete outcome data; (5) selective

Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com 6 Y. M. Sari et al. Results by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. outcome reporting and (6) other sources of bias.24 Four thousand, two hundred and ¯fty-two papers Articles were identi¯ed (across domains) as \\low were generated by the search strategy from the six risk of bias\" if they had low risk of bias for all databases. Following removal of duplicate articles, key domains; \\unclear risk of bias\" if they had screening of title, abstract and full text (see Fig. 1, unclear risk of bias for one or more key domains; or study °owchart), nine articles were included in the \\high risk of bias\" if they had high risk of bias for review. The included studies had 511 people with one or more key domains.24 The Downs and dementia and 82 carers; with 482 (94.3%) people Black checklist provides an overall numeric score with dementia and 73 (89%) carers completing out of 27 points based on ¯ve themed sections: post-programme testing. Five studies were study quality (overall quality), external validity RCTs,32–36 one was a pilot study that used a (ability to generalize ¯ndings), study bias (in clustered randomised design37 and three quasi-ex- interventions and outcome measures), confounding perimental studies38–40 were included. Four studies and selection bias (in sampling) and power were conducted in Hong Kong,32,33,35,37 two from (sample size).27 South Korea,34,39 one from China,36 Taiwan40 and Indonesia,38 respectively. Data analysis Sample sizes across the studies ranged from 26 Physical function and behavioural symptoms of to 90 participants with dementia [mean (SD) of people with dementia, and impact on informal 56.8 (Æ21:3) participants)] (Table 1). Participants carers of older people with dementia were assessed with dementia had an average age of 78.0 (Æ6:1) by various measures in the included studies. In- (one study did not report the mean/SD for age),38 verse variance using the DerSimonian and Laird's and 60% were female. Two studies also included method28 and ¯xed e®ect meta-analyses29 was carers as participants, with a mean (SD) sample performed for outcomes where data were available size of carers being 41 (Æ21:2).33,37 Ninety percent to be pooled for meta-analyses. The consistency in of the carer samples were female, and the average results between studies in the meta-analyses was age of carers in one study was 51.3 (Æ19:0),37 while assessed using I 2 where a value of 0% indicated no the Law and Kwok study reported that 28.5% of heterogeneity and larger values showed increasing their sample of carers were aged over 50 years (did heterogeneity.30 The mean di®erence (MD) and not report mean age data).33 95% con¯dence intervals (CI) were used to pool continuous outcome measures that used the same Four of the ¯ve authors replied to our email units of measurement at the end of the intervention contacts and provided missing information re- regardless of the type of exercise or physical ac- quired for meta-analyses. One author did not pro- tivity intervention. Where quasi-experimental vide the SD of an outcome measure and therefore studies were present and there were su±cient this was calculated by a review author. Liu et al.37 studies, a sensitivity analysis was planned to assess clari¯ed that their data analyses had accounted for if studies of lower scienti¯c quality a®ected the the clustering e®ect due to the study design. meta-analyses results. Sub-group analysis was also planned if there were su±cient studies to perform Types of exercise used in the studies sub-group analysis.29 There were three types of exercise used in the The corresponding authors of included studies included studies including multimodal exercise were contacted for further information where data programmes (5 studies),32–35,39 aerobic exercise in the published articles were insu±cient for meta- (3 studies)36,38,40 and tai chi (1 study).37 analysis. Where standard deviations (SD) were unavailable from the study or from author contact, Multimodal exercise programmes it was calculated with pffitffiffihe formula of SD ¼ Standard Error ðSEÞ Â n. Review Manag- Five trials tested multimodal exercise programmes er (RevMan) version 5.4.131 was used to pool data that comprised more than one exercise type.32–35,39 where two or more studies reported the same out- The intervention duration varied between 5 and come for the meta-analyses. Statistical signi¯cance 52 weeks with a re-assessment at the end of the was set at p < 0:05 for all analyses. intervention. One study had an additional FU

Review of exercise programmes in Asia to improve outcomes for people with dementia 7 Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Fig. 1. Study selection °ow chart. assessment 12 weeks after the programme programmes which included a discrete exercise ended.32 The multimodal exercise interventions comprised of limb mobilisation exercise,32,33 limb component being reported, but also included an strengthening exercise,32 balance exercise,32 walking,32,33 aerobic exercise,33,35 shoulder pulley additional element of non-exercise intervention exercise,33 °exibility exercise35 and the Otago such as self-management support;33 whole-body exercise programme.34,39 Three studies provided vibration (WBV) training;32 handicraft and music activities39 and a cognition pro- gramme.34,39 However, for these studies, the

Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Table 1. Summary of included articles. 8 Y. M. Sari et al. Sample size; % female; age (years) MMSE score (SD) or rating Reference Study design Study purpose Intervention (SD); dropout of dementia Outcome measure Findings FU Liu et al., A pilot cluster To evaluate feasibility IG: 16 week 10-step simpli¯ed Tai-chi 26 dyads of PLWD and carers (13 5-min Montreal Cognitive Feasibility assessment included Recruitment rate-58% (26/45 IL ¼ 16 2018 Hong RCT Kong and preliminary e®ects training programme, 1 h, 2 Â =week dyads/group); female PLWD IG: Assessment score IG: 15.8 recruitment, attrition, adherence assessed dyads), high attendance weeks; of a simpli¯ed 10-step in the community centre and 30 min, 61% (8/13), CG: 69% (9/13), female ( Æ 5:37), CG: 11.8 ( Æ 3:07) and engagement. Outcome rate-81% (25.8/32 sessions). No FU ¼ NA Tai-chi programme 3 Â =week at home. CG: recreational carers IG: 100% (11/11), CG: 77% measures were motor adverse incidents reported. High activities organised by the centres, (10/13); age PLWD IG: 79.8 performance, including mobility: attrition rate of 38%, and mean 1 h 2 Â =week ( Æ 8:16), CG: 80.5 ( Æ 6:94) years; TUG; functional leg muscle home practice time decreased age carers IG: 46.2 ( Æ 17:62), CG: strength: Timed-Chair-Stand; between weeks 8 and 16. Small 55.6 ( Æ 19:67); dropouts, IG: 5 dynamic bilateral stance: FR; but insigni¯cant improvement participants CG: 2 participants and dynamic standing balance: was observed for most motor Step Test performance outcomes for IG compared to CG Lam et al., RCT To evaluate e®ects of IG: routine day activity programme 54 PLWD (27/group); female IG: CMMSE score IG: 13.6 Primary outcome was functional Signi¯cant improvement in IL ¼ 9 weeks; 2018 Hong Kong adding WBV to a combined with WBV training (WBV 70.3% (19/27), CG: 77.7% (21/27); ( Æ 4:7), CG: 15.6 ( Æ 4:5) mobility (TUG). Secondary TUG, BBS and Tinetti balance FU ¼ 12 routine activity at 30 Hz, 2 mm peak-to-peak age IG: 79.7 ( Æ 5:5), CG: 79.9 outcomes were: BBS, Tinetti assessment for both groups, weeks programme among amplitude) ranged between 4 and ( Æ 6:7) years; dropouts, IG: 1 balance assessment, 5 times sit however, no signi¯cant group by community-dwelling 6 min/training session, 2 Â =week; participant, CG: 2 participants to stand time, QoL in time interactions. High individuals with mild– CG: the routine programme only Alzheimer's disease attendance rate for training moderate dementia without WBV. Routine programme: questionnaire, and Activities- (86.0%), low adverse event 30–60 m active limb mobilization speci¯c Balance Con¯dence incidence-7.4% (2/27) in the IG exercises, leg strengthening exercises, scale. Also recorded attendance reporting mild knee pain walking and balance exercises. Also rate and adverse events included social and cognitive activities Law and RCT To explore feasibility of Participants in both groups 56 dyads of PLWD and their carers CMMSE score IG: 19 (16– Attendance rate and non- Attendance rate was high IL ¼ 8 weeks; Kwok, 2019 Hong Kong a multicomponent underwent an individual (28 dyads in each group); female 22); CG: 20 (15–23) attendance reasons were (94.4%). Compared with the FU ¼ NA intervention physiotherapy session (personalised PLWD IG: 82.1% (23/28), CG: recorded for feasibility. Carer CG, IG carers signi¯cantly programme and to home exercises with 5–10 repetitions 78.6% (2/28); female carers IG: self-e±cacy and distress: the improved in three domains of evaluate its e®ects on of each set of knee exercises within 92.9% (26/528), CG: 89.3% (25/28); Chinese versions of RSCSE and RSCSE scores (P 0:005) and behavioural and 15–20 min and care education) 7 Â age PLWD IG: 82.4 ( Æ 5:8), CG: neuropsychiatric symptoms: the carers' distress (P ¼ 0:004) post- psychological =week and 8 weeks of a 1-h- 82.7 ( Æ 5:3) years; dropouts IG: 1 NPI-Q intervention. No e®ects were symptoms of dementia structured group exercise session dyad, CG: 1 dyad observed for BPSD severity in (BPSD) and weekly at a day care centre. The IG PLWD. No adverse events or psychological health of received a multicomponent falls were reported carers programme, combining knee OA- speci¯c therapeutic exercise for participants with dementia and self- management support for the carers, the CG attended the routine group exercise programme only Yang et al., RCT To evaluate e®ect of IG: cycling training, moderate 50 PLWD (25 participants in each MMSE score, IG: 21.33 MMSE, QoL Alzheimer's Signi¯cant improvement by the IL ¼ 12 2015 China moderate intensity intensity, 70% maximum heart rate group); female: 66% (33/50); age 71.9 ( Æ 2:24), CG: 20.00 disease, plasma Apo-a1 level, the IG compared to CG for MMSE, weeks; aerobic exercise on for 40 min (5 min warm-up, 30 min ( Æ 6:9) years ( Æ 3:50) Alzheimer's Disease Assessment QoL Alzheimer's Disease score FU ¼ NA older people with mild target intensity exercise, 5 min Scale-cognition score and NPI-Q and the plasma Apo-a1 level Alzheimer's disease reorganisation movement) (P < 0:05), the Alzheimer's 3 Â =week. training time of initial Disease Assessment Scale- stage was 25–30 min, and exercise cognition score and NPI-Q score load was 0.5 kg. CG: health education was signi¯cantly decreased (P < 0:05) baseline to post- intervention in IG compared to CG

Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Table 1. (Continued ) Sample size; % female; age (years) MMSE score (SD) or rating Reference Study design Study purpose Intervention (SD); dropout of dementia Outcome measure Findings FU Miu et al., RCT To explore e®ect of IG: aerobic exercise training with 85 PLWD (IG ¼ 36 and CG ¼ 49); MMSE score, IG: 19.52 Physical performance: 6-min Greater signi¯cant IL ¼ 52 2008 Hong Kong exercise in older people treadmill, bicycle, arm ergometry and female IG: 42% (15/36), CG: 63% ( Æ 4:59), CG: 19.69 ( Æ 4:0) walking distance, FR, BBS and improvements in physical weeks; with dementia °exibility exercises for 45–60 min and (31/49); age IG: 75 ( Æ 7), CG: 78 SF12 QoL questionnaire. performance in IG than CG for FU ¼ NA 10-min °exibility training, 2 Â =week. ( Æ 6) years; dropouts IG: none, Cognitive function: MMSE and the 6-min walking distance (12- CG: conventional medical treatment CG: 3 ADAS-Cognitive Subscale. month mean treatment Depressive symptoms: Cornell di®erence 29.75, p ¼ 0:014) and Scale for Depression in Dementia FR (12-month mean treatment and Carer stress: Zarit Burden di®erence 2.63, p ¼ 0:009). No Interview statistically signi¯cant di®erences between groups in cognitive function, and no e®ect on depression Lee and Don RCT To evaluate e®ects of a IG: cognition programme for 30 min 60 older persons with mild dementia CDR ¼ 1 Cognitive function: LOTCA-G; Cognitive function and ADLs IL ¼ 8 weeks; Review of exercise programmes in Asia to improve outcomes for people with dementia 9 Kim, 2018 South Korea physical activity and physical activity programme for (30/group); female IG: 60% (18/30), and ADLs: FIM signi¯cantly improved in both FU ¼ NA Shih et al., Quasi- programme in older 30 min 3 Â =week for 8 weeks (Otago CG: 53.3% (16/30); age IG: 76.27 groups, however, e®ects in IG 2019 Taiwan experimental adults with mild exercise programme with 1–6 kg ( Æ 3:86), CG: 75 ( Æ 4:98) years were signi¯cantly greater than study dementia ankle weight); CG: cognition CG (i.e., statistically signi¯cant programme only between group di®erence) To evaluate e®ects of 3 groups (2 IG and 1 CG); IG: 60 PLWD (20/group); female IG CDR score all groups, 1 ¼ Neuropsychiatric symptoms Neuropsychiatric symptoms IL ¼ 24 walking on sundown morning or afternoon walking (morning): 73.3% (11/15), IG 54:3% (25/46), 2 ¼ 37% (sundown syndrome): The signi¯cantly decreased on two weeks; syndrome in older programme with carer supervision for (afternoon): 40% (6/15), CG: 62.5% (17/46), 3 ¼ 8:7% (4/46) Chinese version of C-CMAI (morning and afternoon) IG FU ¼ NA adults with 30 min 4 Â =week; CG: usual daily (10/16); age IG (morning): 75.1 compared to CG. No statistically Alzheimer's disease activities ( Æ 8:3), IG (afternoon): 79.9 ( Æ 8:3), signi¯cant di®erences between CG: 78.1 ( Æ 7:3); dropouts IG morning and afternoon IG (morning): 5, IG (afternoon): 5, CG: 4 Kim et al., Quasi- To explore e®ects of an IG: occupation-centred activity 30 PLWD (15/group); female IG: MMSE-K score, IG: 15.5 Cognitive function: MMSE-K Cognitive function improved IL ¼ 5 weeks; 2017 South experimental Korea study occupation-centred programme (physical activities, 13.3% (2/15), CG: 26.7% (4/15); age ( Æ 2:9), CG: 15.6 ( Æ 2:4) and GDS. Fall-related factors: signi¯cantly in both IG and CG, FU ¼ NA activity programme in cognitive activities, daily life IG: 82 ( Æ 4:6); CG: 80.9 ( Æ 3:4) FES-K, Chair stand test, 244 cm fall-related factors and the community-dwelling activities, instrumental daily life Up and Go Test, one leg quality of life signi¯cantly older people with activities, handicraft, traditional standing test. QoL: KQOL-AD improved only in the IG dementia Korean music activities and other music activities) for 60 min 5 Â =week and dementia medications; CG: only taking medications for dementia symptoms Juniarti Quasi- To investigate the IG: physical exercise (low-impact 90 PLWD (45/group); female IG: MMSE score, IG: 22.5 Cognitive function: MMSE; and Cognitive function and physical IL ¼ 4 weeks; et al., 2021 experimental Indonesia study e®ect of exercise and aerobic exercise) and reading therapy 77.8 (35/45), CG: 66.7 (30/45); age ( Æ 2:7), CG: 20.93 ( Æ 3:4) physical activity level: PASE activity level signi¯cantly FU ¼ NA learning therapy on the for 60 min 3 Â =week; CG: daily IG, 60–70 years: 71% (32/45), > 70 improved in IG but not in the cognitive functions and activities programme years: 28.9% (13/45); CG, 60–70 CG. There were statistically daily physical activities years: 62.2% (28/45), > 70 years: signi¯cant di®erences between of people living with 37.8% (17/45) groups in cognitive function and dementia physical activity level Notes: RCT, Randomised Controlled Trial; PLWD, People Living with Dementia; IL, intervention length; FU, follow up; IG, intervention group; CG, control group; MMSE, Mini Mental State Examination; CMMSE, Cantonese Mini Mental State Examination; SD, Standard Deviation; TUG, Timed Up and Go; LOTCA-G, Loewenstein Occupational Therapy Cognitive Assessment for Geriatric Population; FIM, Functional Independence Measure; NPI-Q, neuropsychiatric inventory questionnaire; BBS: Berg Balance Scale; QoL, Quality of Life; FR, Functional Reach; RSCSE, Revised Scale for Caregiving Self-E±cacy; ADAS, Alzheimer Disease Assessment Scale-Cognitive subscale; ADLs, Activity of Daily Livings; C-CMAI, Cohen-Mans¯eld Agitation Inventory-Chinese version; MMSE-K, Mini Mental State Examination-Korea; GDS, Global Deterioration Scale; FES-K, Korean Falls E±cacy Scale for the Elderly; KQOL-AD, Korean version of the Quality of Life-Alzheimer's Disease Scale; PASE, Physical Activity Scale for the Elderly.

Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com 10 Y. M. Sari et al. Tai chi by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. exercise intervention e®ect was clearly able to be Liu et al.37 investigated the e®ectiveness of Tai-chi identi¯ed compared to the control group. exercise. The programme was delivered by a Tai- chi master at a community centre over a 16-week Three studies reported dropout rates between period with re-assessment at the end of the inter- 3.5%35 to 5.5%32 for people with dementia, vention. Carers were actively engaged to provide and 5%33 for carers. Reasons for participants with supervision and support in undertaking the Tai-chi dementia withdrawing were hospitalisation,32,33 programme at home. The dropout rates were 38% declined to continue35 and death.35 Two studies in the Tai-chi group. Reasons for participants with reported exercise adherence rates of 86%32 and dementia withdrawing were hospitalisation, de- 94.4%.33 Reasons for non-participation included clined to continue and carer health issue. Adher- knee pain onset, holiday, did not like the vibration ence rate was 81% for Tai-chi practice sessions, training, regular medical FU32,33 and medical recruitment rate was 58% and positive engagement conditions.33 No adverse events were reported. in the exercise programme was reported. No ad- verse events were reported. Aerobic exercise Quality of studies Three studies investigated the e®ectiveness of aerobic exercise including cycling training,36 The risk of bias assessment for each study is walking40 and low-impact aerobic exercise.38 The reported in Tables 2 and 3. Based on the assess- intervention duration varied between 4 and 24 ment criteria in Cochrane risk-of-bias tool, four weeks with re-assessment at the end of the inter- RCT studies did not provide su±cient information vention. One study reported a dropout rate of and were scored as having an unclear risk of 23.3%, and reasons for participants with dementia bias.32,34–36 Two studies were assessed as having a withdrawing were hospitalisation, carer changed and refused to participate.40 No adverse events were reported. Table 2. The Cochrane risk-of-bias tool for randomised trials. Study Selection bias Performance bias Attrition bias Reporting bias Other bias Sequence Allocation Blinding of participants, Incomplete Selective outcome Free of generation concealment personnel and outcome assessors outcome data reporting other bias Liu et al.  x  x x Lam et al.  x   xx Law and Kwok x x   xx Yang et al. x x x x xx Miu et al. x x x x xx Lee and Don Kim x x x x xx Notes: Bias was scored as low risk (), unclear (x) or high risk (). Selective outcome reporting domain was scored with unclear for the studies if the study protocol is not available. Table 3. Down and Black quality checklist for quasi-experimental studies. Internal validity Reporting External Internal confounding Su±ciently Subtotal Quality Study (10) validity (3) validity bias (7) selection bias (7) powered? score (27) interpretation Shih et al. 9 2 5 4 Yes 20 Good Kim et al. 4 1 3 Juniarti et al. 7 0 5 1 No 9 Poor 3 Yes 15 Fair Notes: Values in parentheses indicate total score available. Adapted Downs and Black (1998) quality appraisal checklist: 27 total points possible ratings for poor ( 14), fair (14–19), good (20–25), excellent (26–28).

Review of exercise programmes in Asia to improve outcomes for people with dementia 11 Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com high risk of bias33,37 due to lack of blinding and Mobility Assessment (POMA) balance score.32 by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. selection bias attributable to small sample size and Pooled results from Liu et al.37 and Miu et al.35 recruitment strategy used, respectively. For the quasi-experimental studies, one study was assessed indicated a bene¯t of their exercise interventions as having good quality40 with a total score of 20, another was assessed as having poor quality39 on FR performance compared to controls (MD (total score 9) and the third study was assessed as [95% CI] ¼ 2.61 [1.55 to 3.67], heterogeneity: having fair quality38 (total score 15). I 2 ¼ 0%, p ¼ 0:95) (Fig. 2). In contrast, pooled results from Lam et al.32 and Miu et al.35 indicated E®ectiveness of exercise programmes no bene¯t of their exercise interventions on BBS Various outcome measures were used across the nine studies in this review. Assessment tools used performance compared to controls (MD [95% in more than one study were the Timed Up and Go CI] ¼ 1.10 [À2.88 to 5.07], heterogeneity: I 2 ¼ 0%, (TUG) test (3 studies),32,37,39 Functional Reach p ¼ 0:45) (Fig. 3). (FR) (2 studies),35,37 Berg Balance Scale (BBS) (2 studies),32,35 5 times sit to stand (2 studies)32,37 There were two single studies where the balance and the Neuropsychiatric Inventory-Questionnaire outcomes could not be pooled. Liu et al.37 reported (NPI-Q) (2 studies).33,36 Because of the small number of studies which could be included in each a bene¯t of their exercise intervention on Step Test meta-analysis, sensitivity analysis, sub-group performance [E®ect size (Cohen's d) ¼ 0.24] com- analysis and publication bias detection (either pared to controls (p-value was not reported). In using funnel plot asymmetrical testing or Egger's contrast, Lam et al.32 reported no bene¯t of their regression method) were not able to be performed.29,41,42 exercise intervention on the Tinetti POMA balance score [E®ect size ¼ 0:006, 95% CI (À0.1, 1.3), Physical function p ¼ 0:705] compared to controls. Balance. Balance outcomes were reported in three Functional mobility. Functional mobility was studies, using the FR test,35,37 the Step Test,37 the reported by three studies using the TUG test32,37,39 BBS32,35 and the Tinetti Performance Oriented and the Tinetti POMA gait score.32 Pooled results from Lam et al.32 and Liu et al.37 indicated no bene¯t of their exercise interventions on TUG test performance compared to controls (MD ½95%CIŠ ¼ À3:47 [À7.27 to 0.33], heterogeneity: I 2 ¼ 0%, p ¼ 0:57) (Fig. 4). Kim et al.39 were not included in the meta-analysis as they used a di®erent distance of walking for their TUG test, but did report a bene¯t of exercise on TUG (p < 0:05) Fig. 2. Forest plot of comparison: intervention versus control for FR. FR ¼ Functional Reach, SD ¼ standard deviation, CI ¼ confidence interval, IV ¼ inverse variance. Fig. 3. Forest plot of comparison: intervention versus control for BBS. BBS ¼ Berg Balance Scale, SD ¼ standard deviation, CI ¼ confidence interval, IV ¼ inverse variance.

12 Y. M. Sari et al. Fig. 4. Forest plot of comparison: intervention versus control for TUG Test. TUG ¼ Timed Up and Go, SD ¼ standard deviation, CI ¼ confidence interval, IV ¼ inverse variance. Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com (actual e®ect size and p-value were not reported). Balance con¯dence. Lam et al.32 reported no by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Data from Lam et al.32 also could not be pooled, bene¯t of their exercise intervention on balance con¯dence using the Activities-speci¯c Balance however, they reported no bene¯t of their exercise Con¯dence scale [E®ect size ¼ 0:005, 95% CI (À6.1, 7.0), P ¼ 0:757]. intervention on the Tinetti POMA gait score [Ef- fect size ¼ 0:034, 95% CI (À0.6, 0.4), p ¼ 0:178]. Behavioural symptoms Endurance. Miu et al.35 separated the e®ects of Three studies reported behavioural symptom out- comes.33,35,36 A meta-analysis was not able to be exercise into diagnostic groups (Alzheimer's dis- performed even though two studies33,36 reported the NPI score. Law and Kwok33 reported individ- ease group, and a mixed dementia group [Alzhei- ual scores for each domain of the NPI-Q, while mer's disease þ vascular dementia, or vascular Yang et al.36 reported the total score of the scale. dementia]) and reported a bene¯t of their exercise Law and Kwok33 reported a bene¯t of exercise on behavioural symptoms as measured by the NPI-Q intervention in the Alzheimer's disease group and score for carer stress (p ¼ 0:004) (e®ect size was the mixed dementia group at 3 (p < 0:001) and 6 not reported). Yang et al.36 reported a signi¯cant months (p ¼ 0:009) measured by the 6-min walk di®erence within the exercise intervention group as test compared to controls (e®ect size was not measured by the NPI-Q (p ¼ 0:004) (e®ect size was not reported), however, they did not report dif- reported). ferences between the intervention and control groups. Miu et al.35 reported no bene¯t of their Lower limb strength. Three studies reported exercise intervention on behavioural symptoms as measured by the Cornell depression scale compared lower limb strength outcomes using the 5 times sit to controls (p-value and e®ect size were not to stand test,32,37 and 30 s sit to stand test.39 reported). Pooled results from Lam et al.32 and Liu et al.37 Impact on informal carers indicated no bene¯t of their exercise interventions Two studies reported impact on informal carer on lower limb strength on the 5 times sit to stand outcomes, although the results were not be able to test compared to controls (MD ½95%CIŠ ¼ À1:86 be pooled as these scales measured di®erent [À5.27 to 1.54], heterogeneity: I 2 ¼ 0%, p ¼ 0:99) (Fig. 5). Data from Kim et al.39 could not be pooled, however, they reported improved 30 s sit to stand test performance (p < 0:05) (actual e®ect size and p-value were not reported). Functional independence. Lee and Don Kim34 reported no bene¯t of their exercise inter- vention on functional independence as measured by the Functional Independent Measure (FIM) (p > 0:05) (e®ect size was not reported). Fig. 5. Forest plot of comparison: intervention versus control for 5 times sit to stand test, SD ¼ standard deviation, CI ¼ confidence interval, IV ¼ inverse variance.

Review of exercise programmes in Asia to improve outcomes for people with dementia 13 Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com constructs of impact on carers.33,35 Miu et al.35 reviews where research evidence in Asia relative to by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. reported no e®ect of their exercise intervention on the rest of the world are emerging, for example the the impact on informal carers using the Zarit e®ectiveness of fall prevention programmes on burden scale (p-value and e®ect size were not community-dwelling older people in Asia.20 reported), while Law and Kwok33 reported a ben- e¯t of their exercise intervention measured by the The studies included in this paper varied sub- Revised Score for Caregiving Self-E±cacy stantially in terms of the nature of the exercise [(p ¼ 0:0001, 0.001, 0.005) for three domains of the programmes, dosage (session length (minutes), Revised Score for Caregiving Self-E±cacy] (e®ect frequency, duration of the exercise programme) sizes were not reported). and participants characteristics (age, gender, de- mentia severity). This, and the limited ability to Discussion pool study results due to varied outcomes evalu- ated between studies, limits conclusions able to be Despite the growing and promising evidence that made related to these factors that may in°uence exercise programmes can improve physical func- outcomes. Five out of nine studies28–31,35 used tion and reduce behavioural symptom outcomes of multimodal exercise as the intervention (which older adults with dementia and the impact on their combine more than one exercise type) or included informal carers, this systematic review identi¯ed an additional element of a non-exercise interven- that little of this research occurred in Asia (where tion (where the exercise intervention e®ect was 60% of the world population live)18 and where the clearly able to be identi¯ed compared to the control ageing trajectory is increasing rapidly.3 Overall, group). Multimodal exercises usually included a only four meta-analyses were able to be performed, balance training component, which may help to spread over three of the nine studies included in reduce falls risk. However, a systematic review of this review. The results need to be interpreted with exercise programmes to reduce falls risk in caution due to the small number of studies, small cognitively intact older people,43 which highlighted sample sizes and low or unknown risk of bias in the the importance of balance training to reduce falls majority of overall methodological quality ratings. risk, also concluded that programmes needed to A meta-analysis of two studies shows that exercise include three hours of exercise/week to be likely to interventions may improve balance performance reduce falls risk. Only one of the multimodal pro- when measured by a single task measurement (i.e., grammes in this review met this exercise dosage.33 FR). However, this bene¯t was not demonstrated Several studies focussed on participants with mild in other meta-analyses when balance was measured severity dementia.34,36,38 Exercise interventions using a multi-task measurement scale (i.e., BBS), that commence early after dementia diagnosis, or or for functional mobility or lower limb strength. when people have mild dementia severity may be Single studies where results could not be pooled more likely to have e®ective outcomes than those reported that their exercise interventions had no targeting people with more advanced dementia, al- e®ect on endurance, functional independence and though further research is needed to con¯rm this.13 balance con¯dence. Single studies however did re- port signi¯cant bene¯ts for lower limb strength Overall, the studies indicated that community- (30 s sit to stand) and balance (Step Test). Mixed based exercise programmes appear to be safe and e®ects were reported on behavioural symptoms and feasible for older people with dementia living in impact on informal carers between the single Asia. Most studies in this review utilised exercise studies reporting these outcomes. programmes limited to short, time limited or epi- sodic bursts (between 4 and 52 weeks (Median ðIQR Compared to other recent systematic reviews ½25; 75ŠÞ ¼ 9 [6.5, 20]), whereas guidelines and not limited to Asia, where a substantially greater recommendations highlight the need for exercise as number of research studies have been undertaken an approach to care utilised for people with de- and reported the bene¯ts of exercise to improve mentia needs to become a lifestyle change and a physical function, behavioural symptoms, and to sustained behaviour.14,44 This review identi¯ed reduce impact on informal carers,14,15,17 these only one study32 that reported a 12-week FU while bene¯ts were not consistently identi¯ed in our re- the other eight studies did not conduct a FU as- view. This is similar to ¯ndings of other systematic sessment. It is unknown if participants continued to participate in exercises beyond the intervention period to sustain any bene¯ts that were gained.

Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com 14 Y. M. Sari et al. is bene¯cial for improving physical function, reducing by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. behavioural symptoms and carer impact for this Exercise load and intensities were reported for population. There is also a need for greater standar- some studies included in this review. However, in disation of measures across studies to increase pool- other studies, aspects of intensity and load that ing of data in future research to establish e®ects. In may have an e®ect on the ¯ndings, were reported addition, future studies should include longer term inconsistently. Based on the American College of FU to evaluate sustained exercise participation be- Sports Medicine (ACSM)45 exercise prescription yond the formal exercise programme duration. guidelines, increasing exercise intensity may lead to the positive response of health and ¯tness bene¯ts, Conclusion and exercise below a minimum intensity or threshold will likely not challenge the body su±- Our review is a starting point to building the evi- ciently to result in physiologic adaptive changes dence base of exercise for older people with de- (the overload principle). mentia living in Asian countries. Although one meta-analysis (n ¼ 2 studies) identi¯ed improved This review focussed on Asia because of the balance performance among older people with de- cultural and health system factors in the Asian mentia living in Asia, overall ¯ndings were incon- context that may in°uence outcomes relative to sistent and limited by the small number of studies, studies conducted in non-Asian countries.19,20 small sample sizes and risk of bias in the method- Factors that could possibly in°uence e®ective ological quality. There is a need to strengthen the translation of exercise interventions in Asian research evidence investigating e®ectiveness of ex- countries may include: (1) family role including ercise interventions for people living with dementia ¯lial piety; (2) health system and service including across Asia, including consideration of local con- people's expectations of health practitioners; (3) text factors that may in°uence uptake, sustained lifestyle factors; (4) di®erent understanding of in- participation and outcomes. tervention approach and prevention19,46 and (5) barriers and facilitators to exercise participation Con°ict of Interest among older Asians such as cultural and family values.46 It is also important to note that there is The authors have no con°icts of interest relevant considerable diversity between Asian countries to this paper. e.g., country income status. Most studies in this review were from the more economically developed Funding/Support Asian countries. More research and assistance to conduct research in developing countries would Yulisna Mutia Sari was funded by an Australia provide a more comprehensive evidence base for Awards Scholarship and Dr. Elissa Burton was exercise interventions for people with dementia in supported by an NHMRC Investigator Grant. Asia and enhance the translation of e®ective re- search between Asian countries. Author Contributions There were several limitations of this review. We Conception and design of study were done by Y. did not search the grey literature and have only in- M. Sari, K. D. Hill and E. Burton. Acquisition of cluded studies published in English. It is possible data was done by Y. M. Sari. Analysis and/or in- that if other studies exist, they may modify the terpretation of data were done by Y. M. Sari, K. D. results. The small number of studies, small sample Hill, D.-C.A. Lee, E. Burton. This manuscript was size of individual studies likely resulting in some drafted by Y. M. Sari, K. D. Hill, D.-C.A. Lee, outcomes being underpowered, unclear risk of bias E. Burton and was revised critically for important and only a few studies that were able to be pooled for intellectual content by Y. M. Sari, K. D. Hill, D.-C. meta-analysis may have over-estimated the result of A. Lee, E. Burton. Finally, approval to publish this meta-analysis and potentially limited the evidence manuscript was given by Y. M. Sari, K. D. Hill, D.- of this review. Another limitation is the absence of C. A. Lee and E. Burton. FU assessments in most of the studies beyond the formal intervention period to evaluate sustained, longer term exercise participation. Future research for large RCTs of community- based exercise interventions for people with de- mentia in Asia is needed to establish whether exercise

Review of exercise programmes in Asia to improve outcomes for people with dementia 15 Appendix A. Search Keywords disease and other dementias, 1990–2016: A sys- (According to tematic analysis for the Global Burden of Disease Medline Study 2016. Lancet Neurol 2019;18(1):88–106. doi: Terminology) 10.1016/S1474-4422(18)30403-4. 2. World Health Organization. Global status report Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com 1 Dementia/[MeSH] on the public health response to dementia. 2022. by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. 2 Dement à ti.ab Available at https://www.who.int/publications/i/ 3 Alzheimer disease/[MeSH] item/9789240033245. 4 Alzheimer à ti.ab 3. World Bank. Live Long and Prosper: Aging in East 5 Cognitive à impair à ti.ab Asia and Paci¯c. Washington, DC, 2016. 6 Cognition disorders/[MeSH] 4. Alzheimer's Disease International. Dementia in the 7 1 or 2 or 3 or 4 or 5 or 6 Asia Pasi¯c Region. 2018. Available at https:// 8 Exercise/[MeSH] www.alz.co.uk/adi/pdf/Dementia-Asia-Paci¯c- 9 Exercise à ti.ab 2014.pdf. 10 Exercise therapy/[MeSH] 5. Bridenbaugh SA, Kressig RW. Quantitative gait 11 Exercise training ti.ab disturbances in older adults with cognitive 12 Exercise movement techniques/[MeSH] impairments. Curr Pharm Des 2014;20(19):3165– 13 Physical activit à ti.ab 72. doi: 10.2174/13816128113196660688. 14 Physical exerc à ti.ab 6. Cipriani G, Danti S, Picchi L, Nuti A, Di Fiorino 15 Physical therapy ti.ab M. Daily functioning and dementia. Dement Neu- 16 Physical therapy modalities/[MeSH] ropsychol 2020;14(2):93–102. doi: 10.1590/1980- 17 Physiotherapy ti.ab 57642020dn14-020001. 18 8 or 9 or 10 or 11 or 12 or 13 or 14 or 7. Allan LM, Ballard CG, Rowan EN, Kenny RA. Incidence and prediction of falls in dementia: A 15 or 16 or 17 prospective study in older people. PLoS One 2009;4 19 Physical function ti.ab (5):e5521. doi: 10.1371/journal.pone.0005521. 20 Functional status ti.ab 8. Walter E, Pinquart M. How e®ective are dementia 21 Physical ¯tness/[MeSH] caregiver interventions? An updated comprehen- 22 Physical ¯tness ti.ab sive meta-analysis. Gerontologist 2020;60(8):E609– 23 Balance à ti.ab 19. doi: 10.1093/geront/gnz118. 24 Postural balance/[MeSH] 9. Cheng S-T. Dementia caregiver burden: A research 25 Mobility à ti.ab update and critical analysis. Curr Psychiatry Rep 26 Mobility limitation/[MeSH] 2017;19(9):1–8. doi: 10.1007/s11920-017-0818-2. 27 Strength à ti.ab 10. Alzheimer's Disease International. World Alzhei- 28 Muscle strength/[MeSH] mer report 2019. Attitudes to dementia. 2020. 29 Endurance ti.ab Available at https://www.alz.co.uk/research/ 30 Physical endurance/[MeSH] WorldAlzheimerReport2019.pdf. 31 Behaviour ti.ab 11. World Health Organization. Global action plan on 32 Health behaviour/[MeSH] the public health response to dementia 2017–2025. 33 Neuropsychiatric symptom à ti.ab 2019. Available at https://www.who.int/mental 34 Mental disorders/[MeSH] health/neurology/dementia/action plan 2017 35 Mental disorder à ti.ab 2025/en/. 36 Caregiver à burden ti.ab 12. Burton E, Cavalheri V, Adams R, et al.. E®ec- 37 Carer à burden tiveness of exercise programs to reduce falls in older 38 Caregiv à ti.ab people with dementia living in the community: A 39 Caregivers/[MeSH] systematic review and meta-analysis. Clin Interv 40 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 Aging 2015;10:421–34. doi: 10.2147/CIA.S71691. 13. Forbes D, Forbes S, Blake C, Thiessen E, Forbes S. or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or Exercise programs for people with dementia (Re- 36 or 37 or 38 or 39 view). Cochrane Database Syst Rev 2015;(4): 41 Communit à ti.ab CD006489. doi: 10.1002/14651858.CD006489. 42 Home ti.ab pub4. 43 41 or 42 14. Lam F, Huang M-Z, Liao L-R, Chung RC, Kwok 44 7 and 18 and 40 and 43 TC, Pang MY. Physical exercise improves strength, balance, mobility, and endurance in peo- References ple with cognitive impairment and dementia: A 1. Nichols E, Szoeke CEI, Vollset SE, et al.. Global, regional, and national burden of Alzheimer's

Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com 16 Y. M. Sari et al. bias in randomised trials. BMJ (Clin Res ed.) by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. 2011;343:d5928. doi: 10.1136/bmj.d5928. systematic review. J Physiother 2018;64(1):4–15. 27. Downs SH, Black N. The feasibility of creating a doi: 10.1016/j.jphys.2017.12.001. checklist for the assessment of the methodological 15. Law C, Lam FM, Chung RC, Pang MY. Physical quality both of randomised and non-randomised exercise attenuates cognitive decline and reduces studies of health care interventions. J Epidemiol behavioural problems in people with mild cognitive Community Health 1998;52(6):377–84. doi: impairment and dementia: A systematic review. J 10.1136/jech.52.6.377. Physiother 2020;66(1):9–18. doi: 10.1016/j. 28. DerSimonian R, Laird N. Meta-analysis in clinical jphys.2019.11.014. trials. Control Clin Trials 1986;7(3):177–88. doi: 16. Lewis M, Peiris CL, Shields N. Long-term home 10.1016/0197-2456(86)90046-2. and community-based exercise programs improve 29. Deeks JJ, Higgins JPT, Altman DG. Analysing function in community-dwelling older people with data and undertaking meta-analyses. In: Higgins cognitive impairment: A systematic review. J JPT, Thomas J, Chandler J et al.., eds., Cochrane Physiother 2017;63(1):23–29. doi: 10.1016/j. Handbook for Systematic Reviews of Interventions, jphys.2016.11.005. Chap. 10. Cochrane, 2022. www.training.cochrane. 17. Orgeta V, Miranda-Castillo C. Does physical ac- org/handbook. tivity reduce burden in carers of people with de- 30. Higgins JPT, Thompson SG, Deeks JJ, Altman mentia? A literature review. Int J Geriatr DG. Measuring inconsistency in meta-analyses. Br Psychiatry 2014;29(8):771–83. doi: 10.1002/ Med J 2003;327(7414):557–60. doi: 10.1136/ gps.4060. bmj.327.7414.557. 18. Population Reference Bureau. World population 31. Review Manager (RevMan). Version 5.3. The data sheet 2018. 2020. Available at https://www. Cochrane Collaboration, 2014. prb.org/2018-world-population-data-sheet-with- 32. Lam FM, Liao LR, Kwok TCY, Pang MYC. E®ects focus-on-changing-age-structures/. of adding whole-body vibration to routine day ac- 19. Chongsuvivatwong V, Phua KH, Yap MT et al.. tivity program on physical functioning in elderly Health and health-care systems in southeast Asia: with mild or moderate dementia: A randomized Diversity and transitions. Lancet 2011;377 controlled trial. Int J Geriatr Psychiatry 2018;33 (9763):429–37. doi: 10.1016/S0140-6736(10)61507- (1):21–30. doi: 10.1002/gps.4662. 3. 33. Law W, Kwok TCY. Impacts of a multicomponent 20. Hill K, Suttanon P, Lin S et al.. What works in falls intervention programme on neuropsychiatric prevention in Asia: A systematic review and meta- symptoms in people with dementia and psycho- analysis of randomized controlled trials. BMC logical health of caregivers: A feasibility pilot Geriatr 2018;18(1):1–21. study. Int J Geriatr Psychiatry 2019;34(12):1765– 21. Pandya SY, Clem MA, Silva LM, Woon FL. Does 75. doi: 10.1002/gps.5191. mild cognitive impairment always lead to demen- 34. Lee HJ, Don Kim K. E®ect of physical activity on tia? A review. J Neurol Sci 2016;369:57–62. doi: cognition and daily living activities of the elderly 10.1016/j.jns.2016.07.055. with mild dementia. J Phys Ther Sci 2018;30 22. Paez A. Gray literature: An important resource in (3):428–33. doi: 10.1589/jpts.30.428. systematic reviews. J Evid Based Med 2017;10 35. Miu DKY, Szeto P, Mak Y. A randomised con- (3):233–40. doi: 10.1111/jebm.12266. trolled trial on the e®ect of exercise on physical, 23. Mahood Q, Eerd DV, Irvin E. Searching for grey cognitive and a®ective function in dementia sub- literature for systematic reviews: Challenges and jects. Asian J Gerontol Geriatr 2008;3(1):8–16. bene¯ts. Res Synth Methods 2014;5(3):221–34. doi: 36. Yang S-Y, Shan C-L, Qing H et al.. The e®ects of 10.1002/jrsm.1106. aerobic exercise on cognitive function of Alzhei- 24. Bellefontaine SP, Lee CM. Between black and mer's disease patients. CNS Neurol Disord Drug white: Examining grey literature in meta-analyses Targets 2015;14(10):1292–7. doi: 10.2174/ of psychological research. J Child Fam Stud 1871527315666151111123319. 2013;23(8):1378–88. doi: 10.1007/s10826-013- 37. Liu JYW, Kwan RYC, Lai CK, Hill KD. A sim- 9795-1. pli¯ed 10-step Tai-chi programme to enable people 25. Liberati A, Altman DG, Tetzla® J et al.. The with dementia to improve their motor performance: PRISMA statement for reporting systematic A feasibility study. Clin Rehabil 2018;32(12):1609– reviews and meta-analyses of studies that evaluate 23. doi: 10.1177/0269215518786530. health care interventions: Explanation and elabo- 38. Juniarti N, Al'Adawiyah MZI, Sari CWM, Haroen ration. Ann Intern Med 2009;151(4):W65. doi: H. The e®ect of exercise and learning therapy on 10.7326/0003-4819-151-4-200908180-00136. cognitive functions and physical activity of older 26. Higgins JPT, Altman DG, Gøtzsche PC et al.. The Cochrane Collaboration's tool for assessing risk of

Review of exercise programmes in Asia to improve outcomes for people with dementia 17 Hong Kong Physiother. J. 2023.43:3-17. Downloaded from worldscientific.com people with dementia in Indonesia. J Aging Res of a new publication bias test. Res Synth Methods by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. 2021;2021:1–9. doi: 10.1155/2021/6647029. 2020;11(4):522–34. doi: 10.1002/jrsm.1414. 39. Kim K-U, Kim S-H, Oh H-W. The e®ects of occu- 43. Hopewell S, Adedire O, Copsey BJ et al.. Multi- pation-centered activity program on fall-related factorial and multiple component interventions for factors and quality of life in patients with demen- preventing falls in older people living in the com- tia. J Phys Ther Sci 2017;29(7):1188–91. doi: munity. Cochrane Database Syst Rev 2016;2016 10.1589/jpts.29.1188. (6):CD012221. doi: 10.1002/14651858.CD012221. 40. Shih YH, Pai MC, Lin HS, Sung PS, Wang JJ. 44. Guideline Adaptation Committee. Clinical Practice E®ects of walking on sundown syndrome in com- Guidelines and Principles of Care for People with munity-dwelling people with Alzheimer's disease. Dementia, 2016. Int J Older People Nurs 2020;15(2):e12292. doi: 45. American College of Sports Medicine. ACSM's 10.1111/opn.12292. Guidelines for Exercise Testing and Prescription. 41. Borenstein M, Higgins JPT. Meta-analysis and 9th ed. Philadelphia: Wolters Kluwer Health/Lip- subgroups. Prev Sci 2013;14(2):134–43. doi: pincott Williams & Wilkins, 2014. 10.1007/s11121-013-0377-7. 46. Karuncharernpanit S, Hendricks J, Toye C. Per- 42. Doleman B, Freeman SC, Lund JN, Williams JP, ceptions of exercise for older people living with Sutton AJ. Funnel plots may show asymmetry in dementia in Bangkok, Thailand: An exploratory the absence of publication bias with continuous qualitative study. Int J Older People Nurs 2016;11 outcomes dependent on baseline risk: Presentation (3):166–75. doi: 10.1111/opn.12091.

Research Paper Hong Kong Physiotherapy Journal Vol. 43, No. 1 (2023) 19–31 DOI: 10.1142/S1013702523500014 Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com Hong Kong Physiotherapy Journal by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. https://www.worldscientific.com/worldscinet/hkpj E®ects of early mobilisation program on functional capacity, daily living activities, and N-terminal prohormone brain natriuretic peptide in patients hospitalised for acute heart failure. A randomised controlled trial Ahmad Mahdi Ahmad1,*, Aya Ibrahim Elshenawy2, Mohammed Abdelghany3 and Heba Ali Abd Elgha®ar1 1Department of Physical Therapy for Cardiovascular and Respiratory Disorders, Faculty of Physical Therapy, Cairo University, Giza, Egypt 2Department of Physiotherapy, Kasr Alainy Hospital, Cairo, Egypt 3Department of Cardiology, Faculty of Medicine, Cairo University, Giza, Egypt *[email protected] Received 28 February 2022; Accepted 20 July 2022; Published 10 October 2022 Background: Patients hospitalised for acute decompensated heart failure (ADHF) show reduced functional capacity, limited activities of daily living (ADL), and elevated N-terminal prohormone of brain natriuretic peptide (NT-proBNP). The management of these patients focuses mainly on medical therapy with little consideration for in-patient cardiac rehabilitation. There has been a growing interest in evaluating the e±cacy of early mobilisation, as the core for in-hospital rehabilitation, in ADHF patients in the last decade; however, the randomised trials on this topic are few. Objective: This randomised-controlled study, therefore, aimed to further test the hypothesis that early supervised mobilisation would have bene¯cial e®ects on functional capacity, ADL, and NT-proBNP in sta- bilised patients following ADHF. *Corresponding author. Copyright@2023, Hong Kong Physiotherapy Association. This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 (CC BY- NC-ND) License which permits use, distribution and reproduction, provided that the original work is properly cited, the use is non-commercial and no modi¯cations or adaptations are made. 19

Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com 20 A. M. Ahmad et al. by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Methods: This is a single-centered, randomised-controlled, parallel-group trial in which 30 patients hospi- talised for ADHF were randomly assigned to two groups; the study group (age ¼ 55:4 Æ 5:46 years, n1 ¼ 15) and the control group (age ¼ 55:73 Æ 5:61 years, n2=15). Inclusion criteria were ADHF on top of chronic heart failure independent of etiology or ejection fraction, clinical/hemodynamic stability, age from 40 to 60 years old, and both genders. Exclusion criteria were cardiogenic shock, acute coronary ischemia, or signi¯cant arrhythmia. Both groups received the usual medical care, but only the study group received an early structured mobilisation protocol within 3 days of hospital admission till discharge. The outcome measures were the 6-min walk distance (6-MWD) and the rating of perceived exertion (RPE) determined from the 6- min walk test at discharge, the Barthel index (BI), NT-proBNP, and the length of hospital stays (LOS). Results: The study group showed signi¯cantly greater improvements compared to the controls in the 6- MWD (252:28 Æ 92:32 versus 106:35 Æ 56:36 m, P < 0:001), the RPE (12:53 Æ 0:91 versus 15:4 Æ 1:63, P < 0:001), and the LOS (10:42 Æ 4:23 versus 16:85 Æ 6:87 days, p ¼ 0:009) at discharge. Also, the study group showed signi¯cant improvements in the BI compared to baseline [100 (100–100) versus 41.87 (35–55), p ¼ 0:009] and the controls [100 (100–100) versus 92.5(85–95), p ¼ 0:006]. The mean value of NT-proBNP showed a signi¯cant reduction only compared to baseline (786:28 Æ 269:5 versus 1069:03 Æ 528:87 pg/mL, p ¼ 0:04) following the intervention. The absolute mean change (Á) of NT-proBNP showed an observed di®erence between groups in favor of the study group (i.e., Á ¼ #282:75 Æ 494:13 pg/mL in the study group versus #26:42 Æ 222:21 pg/mL in the control group, p ¼ 0:077). Conclusion: Early structured mobilisation under the supervision of a physiotherapist could be strongly suggested in combination with the usual medical care to help improve the functional capacity and daily living activities, reduce NT-proBNP levels, and shorten the hospital stay in stabilised patients following ADHF. Trial registration number: PACTR202202476383975. Keywords: Acute heart failure; cardiac rehabilitation; early mobilisation; 6-min walk test; daily living activities; NT-proBNP. Introduction treatment which mainly targets the adequacy of hemodynamic stability and organ perfusion. Not to Heart failure (HF) is considered a major clinical mention that the referral to cardiac rehabilitation problem and public health concern.1 Recent epi- programs is poor in clinical practice and has further demiological studies suggest that the occurrence of reduced after the COVID-19 pandemic.5,6 There- heart failure did not increase globally, but the fore, the need for in-hospital cardiac rehabilitation mortality and hospitalisation associated with the in addition to standardised medical care should be disease have continued to increase despite enor- emphasised. mous management e®orts.1 Acute decompensated heart failure (ADHF) can be de¯ned as acute de- Early mobilisation is the main component of compensation of cardiac function and/or worsen- early cardiac rehabilitation and is closely related to ing of heart failure signs and symptoms with an favourable clinical outcomes.7 Early mobilisation urgent need for hospitalisation.2 Hospitalised refers to the gradual changing of patients' positions patients with ADHF most commonly su®er from from a supine or slumped position in bed to an severe breathlessness, fatigue, lower limb muscle upright sitting in a bedside chair, which is then weakness, and increased weight from peripheral progressed to standing, walking and stair climb- edema, resulting in reduced functional capacity, ing.8,9 This should be started as soon as the clinical poor tolerance to physical activity, and im- stability of the patient is ensured.8,9 In the past, pairment in activities of daily living (ADL).2,3 The most of the rehabilitation-based research work in- physical dysfunction in these patients worsens cluded chronic heart failure patients and excluded further from prolonged hospital stay and bed rest,4 hospitalised patients with recently acute HF.10,11 and can persist even after the restoration of the In the last decade, there has been an increased in- symptoms and signs of acute HF. The physical terest in the evaluation of the safety and e±cacy disabilities in ADHF patients are probably not of early mobilisation programs in ADHF patients fully managed by routine medical/pharmacological by a considerable number of studies.7,9,12–25

E®ects of early mobilisation in patients hospitalised for acute HF 21 Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com Collectively, these studies demonstrated that early function. We hypothesised that early mobilisation by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. structured mobilisation in stabilised patients hos- could be e®ective in improving these clinical out- pitalised for acute HF was safe and e®ective in comes in this patient population based on the increasing functional capacity, improving ADL, ¯ndings from the previous studies. The results of shortening the hospital stay time, and reducing the this study could help highlight the clinical signi¯- re-hospitalisation rate. Nevertheless, apart from cance of early mobilisation in stabilised patients the recent randomised study by Kitzman et al.24 following ADHF and aid strengthen the evidence there have been only four randomised controlled base for the in-patient phase of cardiac rehabilita- trials on the topic of early mobilisation in patients tion in this population. with acute HF,12,14,16,21 based on a recent system- atic review by Babu et al.26 They concluded that Methods the research work in this ¯eld is still growing, and further intervention studies are to be suggested.26 This study was reported as per the CONSORT 2010 Statement for reporting randomised trials. Several outcome measures can be of importance when assessing the clinical bene¯ts of early mobi- Study design and settings lisation in hospitalised acute HF patients. The most important measure is the patients' functional This is a single-centered, randomised controlled, capacity which could be a better prognostic indi- parallel-group intervention study. This study was cator of disease progression or re-hospitalisation conducted at the Cardiology Department of Kasr than cardiac function.27 A reduced functional ca- Alainy Hospital from March 2021 to September pacity prohibits patients from being physically 2021. The study obtained ethical approval from active which leads to a further reduction in their the Ethics Committee of Scienti¯c Research of the functional capacity in \\a vicious circle\" manner. main author's institution (approval No. 012/ The six-minute walk test (6-MWT) is the most 003090) and followed the principles laid down by frequently used test to assess patients' functional the declaration of Helsinki. Informed consent was capacity in heart failure patients.28 The distance obtained from patients before the intervention. covered during the 6-MWT is the primary outcome of the test, but also dyspnea response to the test Randomisation and concealed measured by the rating of perceived exertion allocation (RPE) scale represents another important mea- sure. To add, the Barthel index (BI) is a standar- Simple randomisation using a randomisation table dised tool used to assess the capabilities of ADL in designed by a computer software program was used hospitalised acute HF patients and to predict the in this study with an allocation ratio of 1:1. treatment outcomes.18,29,30 Furthermore, the Sequentially numbered opaque sealed envelopes N-terminal pro-B-type natriuretic peptide (NT- (SNOSE) were used to conceal the allocation se- proBNP) is one of the gold standard biomarkers for quence so that neither the researcher nor the par- the presence and severity of cardiac hemodynamic ticipant was aware of the upcoming assignment. overload and failure.31 NT-proBNP can help eval- uate the response to therapy and predict the clin- Sample size calculation ical outcome in patients with HF.32 The sample size was calculated from the previously Considering the few numbers of randomised published data,16 for the primary outcome measure controlled trials evaluating the e±cacy of early (i.e., 6-MWD). At a p-value of less than 0.05 and a mobilisation in hospitalised patients for acute HF, power of 80%, the sample size (n) was calculated as as previously mentioned, this randomised-con- n ¼ 2 SD2 ðZ=2 þ ZÞ2/d2, according to Charan trolled study aimed to further evaluate the e®ects and Biswas,33 where Z=2 ¼ 1:96 for two tailed of early physiotherapist-supervised mobilisation results at p < 0:05; Z ¼ 0:84 for a power of 80%; program on functional capacity assessed by SD ¼ Standard Deviation (pooled SD of the 6- 6-MWT, ADL assessed by BI, and cardiac hemo- MWD) ¼ 29.15 m;16 d ¼ expected e®ect size ¼ dynamic function, assessed by NT-proBNP (as mean change of the 6-MWD in the intervention primary outcomes); as well as the hospital stay time (as a secondary outcome) in stabilised HF patients following acute decompensation of heart

Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com 22 A. M. Ahmad et al. therapy. Eligible patients were randomly assigned by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. either to a study or a control group. Both groups group (Á 73 m) – mean change of the 6-MWD in were under the usual medical care, but only the the control group (Á 45 m) ¼ 28 m.16 Accordingly, study group received a supervised mobilisation n ¼ 2 (29.15)2 Â ð1:96 þ 0:84Þ2 Ä ð28Þ2 ¼ 17. Ac- program. The °ow of subjects throughout the cordingly, the minimum sample size was estimated study can be shown in Fig. 1. to be 17 patients per group. But, to account for a 10–20% drop-out rate, we recruited a total of 20 Evaluations patients per group. Clinical examination and history taking Subjects These were done by a specialised cardiologist. The Thirty hospitalised acute HF patients, out of 40 age, weight, and other clinical characteristics of patients, underwent the ¯nal analysis of the study. patients were reported. They were recruited in this study by referral from a cardiologist. Eligibility criteria were patients hos- Six-minute walk test pitalised for acute decompensation of chronic heart failure independent of etiology or ejection fraction, The 6-MWT was conducted for all patients under age from 40–60 years old, both gender, and he- the standardised procedure reported by the Euro- modynamic/clinical stability. Exclusion criteria pean Respiratory Society and the American Tho- were unstable vital signs, cardiac arrhythmia, racic Society,34 at discharge.21 The 6-MWT was critical illness, ongoing cardiogenic shock or ino- performed indoors along a °at, straight, enclosed tropic therapy, post-surgical patients, acute myo- corridor with a hard surface. The walking track cardial ischemia, locomotor/neurological was 30 m in length, marked every 10 m. The limitations to ambulation, cognitive impairment, equipment used was mainly a countdown timer, and patients on high FiO2, or continuous oxygen Fig. 1. The °owchart of the study.

E®ects of early mobilisation in patients hospitalised for acute HF 23 Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com the RPE scale, and a ¯ngertip pulse oximetry to Interventions by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. record heart rate and oxygen saturation (SpO2) before, during, and after the test. Patient prepa- Usual medical care ration included comfortable clothing and proper footwear, avoidance of exhausting physical activity All patients in the two groups received standard within 2 h before the test, and adherence to the usual medical treatment as per the guidelines.35 Medi- medications. Before the test, the instructions given to cations were prescribed by a cardiologist and in- patients were to walk (not to jog or run) as far as cluded diuretics, ACE Inhibitors/ARBs, Beta- possible for 6 min, to walk back and forth, to slow Blockers, calcium channel blockers, digoxin, anti- down or stop and rest when becoming exhausted or platelets agents, anti-hyperglycemic agents, and out of breath, to lean against the wall or sit down statins. The medications in the two groups are while resting, and to resume walking as soon as pos- listed in Table 2. Selected patients in both groups sible. During the test, standardised phrases of en- received chest physiotherapy (i.e., respiratory couragement were utilised every minute as per the exercises and hu±ng/coughing) upon need. guidelines,34 and the heart rate and oxygen satura- tion were continuously monitored by a ¯ngertip pulse Mobilisation program oximeter (Granzia, Pulsox-304, Italy). The test was terminated upon the patient request to stop or when Patients in the study group received in-patient 6 min passed. At this point, the RPE was recorded cardiac rehabilitation as an early structured and the distance covered during the test [i.e., the 6- mobilisation program supervised closely by a min walk distance (6-MWD)] was measured. physiotherapist in combination with the usual medical care within the ¯rst three days of hospital Barthel index admission. The mobilisation program started as soon as the clinical stability of patients allowed and The BI was used as a standardised tool to assess the continued gradually throughout the hospitalisation ADL30 in the two groups at baseline and discharge. period until discharge. The mobilization protocol BI comprises 10 items divided into two categories: was designed individually according to the (Fre- those assessing self-care (i.e., feeding, bathing, quency, Intensity, Time, Type (FITT) principle) grooming, dressing, bowel & bladder care, and toilet reported by the recent guidelines of the European use,) and others assessing mobility (i.e., transfers, Society of Cardiology36 (Table 1). The intensity of ambulation, and stair climbing),30 as shown in Ap- mobilisation activities was low to moderate at a pendix A. Each item was scored as described in target heart rate that equaled the resting heart Appendix A and the total sub-scores of the 10 items rate plus 20–30 beats/min,37 monitored objectively were summated to get a total score out of 100. A by a ¯nger pulse oximeter (Granzia, Pulsox-304, score of zero represents completed dependency and a Italy). Also, the targeted intensity was guided by score of 100 represents complete independence. the patient's perceived exertion and set at a RPE of 11–13 on a 20-point Borg scale.36,37 The mobi- N-terminal pro-brain natriuretic peptide lisation program, including the frequency & dura- (NT-proBNP) tion of the sessions and the type of physical activities at each stage, can be shown in detail in At baseline and before the discharge, venous blood Table 1. The criteria for terminating the mobili- samples were taken from patients and the NT- sation sessions were any symptoms or signs sug- proBNP concentrations were assessed using the en- gesting postural hypotension, intolerance to zyme-linked immunosorbent assay (ELISA) tech- physical activity, or poor central/peripheral per- nique according to the manufacturer's instructions fusion such as dizziness, blurred vision, confusion, (SinoGeneClon Biotech Co., Ltd, Human NT- severe breathlessness or fatigue, chest pain, palpi- proBNP Elisa Kit, China, No: SG-10015). tations, leg cramp, cold sweating, pallor, > 4% decrease in O2 saturation, cyanosis, or exaggerated Length of hospital stays (LOS) heart rate response.8,17 The mobilisation session progressed to the next stage once the patient was This was the hospitalisation period in days from clinically stable during and after the premier stage the day of admission to the day of discharge. The (i.e., absence of the above-mentioned signs and LOS was used in this study as a quality metric. symptoms plus: absence of new-onset arrhythmia,

Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Table 1. The prescription of the structured mobilisation program. 24 A. M. Ahmad et al. FITTa Concept Stage 1 Stage 2 Stages Stage 4 Stage 5 Stage 3 Frequency Sessions/week Daily Twice/day Daily Twice/day Daily Twice/day Daily Twice/day Daily Twice/day Sessions/day THR ¼ Resting HRd þ 20 THR ¼ Resting HR þ 20 THR= Resting HR þ 20 THR ¼ Resting THR ¼ Resting Intensity, monitored by: (beats/min) (beats/min) (beats/min) HR þ 20-30 HR þ 30 (beats/min) THR b (beats/min) RPEc on a 20-point scale RPE ¼ 11–12 RPE ¼ 11–12 RPE ¼ 11–12 RPE 13 RPE 13 Time/Duration < 10 min < 10 min 3–5 min 5–10 min 5–15 min Type/mode 1-Active free dynamic 1-Active free dynamic Endurance exercise: Endurance exercise: 1-Endurance exercise: exercises for the upper exercises for the upper Walking Walking Walking and lower limbs from and lower limbs (i.e., the 2- Endurance & strength supine & sitting [1–3 sets, same as in stage 1 but 5–10 reps/set]. -For the from standing) exercise: stair climbing upper limbs: elbow 2-Balance exercises (one °oor) °exion & extension, From sitting & standing. shoulder °exion & (e.g. patient tried to extension, shoulder maintain balance while in abduction & adduction, sitting or standing Scapular retraction. positions against a gentle -For the lower limb: push in all directions dorsi°exion & plantar from the therapist. °exion, knee °exion & extension, hip °exion & extension, hip abduction & adduction. 2-Strength exercise: sit-to- stand exercises (1 set, 10 reps/set) Notes: aFrequency, Intensity, Time, and Type. bTarget Heart rate. cRating of perceived exertion. dheart rate

E®ects of early mobilisation in patients hospitalised for acute HF 25 Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com no worsening of edema, no worsening of dyspnea or signi¯cantly greater improvements in the 6-MWD by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. fatigue, and no reduction of urine output), as (252:28 Æ 92:32 m in the study group versus reported by Kakutani et al.17 It is worth noting 106:35 Æ 56:36 m in the control group, p < 0:001, that the mobilisation program was designed based Unpaired t-test), the RPE (12:53 Æ 0:91 in the on stages and not days, as shown in Table 1. study group versus 15:4 Æ 1:63 in the control group, p < 0:001, Unpaired t-test), and the LOS Statistical analysis (10:42 Æ 4:23 days in the study group versus 16:8 5 Æ 6:87 days in the control group, p ¼ 0:009, Data underwent descriptive statistics and pre- Welch's t-test) following the intervention. The sented as means Æ standard deviations, frequencies study group also showed signi¯cant improvements & percent distributions, and medians & inter- in the BI scores compared to baseline [100 (100– quartile range. For continuous variables, the Kol- 100) versus 41.87 (35–55), p ¼ 0:009, Wilcoxon mogorov–Smirnov test assessed the normality of signed-rank test] and the controls [100 (100–100) distribution, and the Levene's test evaluated the versus 92.5 (85–95), p ¼ 0:006, Mann–Whitney U homogeneity of variance between groups. Data test]. In addition, the study group showed a with normal distribution underwent parametric signi¯cant reduction in the mean value of NT- statistics, and the Paired t-test analysed the proBNP compared to baseline only (786:28 Æ changes within each group post-intervention. For 269:5 pg/mL versus 1069:03 Æ 528:87 pg/mL, normally distributed data with equal variance be- p ¼ 0:04, Paired t-test). Further, this study showed tween groups, the unpaired t-test analysed the an observed di®erence in the absolute mean change di®erence in the mean values of variables between (Á) of NT-proBNP between the two groups in the two groups. For normally distributed data with favor of the study group with a tendency towards unequal variance between groups (i.e., LOS), signi¯cance (i.e., Á ¼ #282:75 Æ 494:13 pg/mL in Welch's t-test analysed the di®erence in the means the study group versus #26:42 Æ 222:21 pg/mL in between the two groups. Data with an abnormal the control group, p ¼ 0:077, Unpaired t-test). distribution that failed the normality test after data transformation (i.e., BI data) underwent non- Discussion parametric statistics. Wilcoxon signed-rank test analysed the changes within each group, and the In agreement with our hypothesis, this study Mann–Whitney U test analysed the di®erence be- showed that early supervised mobilisation in com- tween the two groups pre- and post-intervention. bination with the usual medical care resulted in For categorical variables (i.e., gender and clinical more improvements in the 6-MWD and RPE de- characteristics), the Fisher exact test assessed the rived from the 6-MWT, more enhancement in ADL di®erence between the two groups at baseline. assessed by BI, and shorter hospital stay compared Absolute mean change (Á) from baseline, in the to usual medical care alone, in stabilised patients NT-proBNP, was analysed as an independent who were hospitalised for acute decompensated variable and compared between the two groups HF. This study also showed that early mobilisation using the Unpaired t-test. A signi¯cance level of combined with medical care led to statistically and p < 0:05 was utilised in all statistical tests. The clinically signi¯cant reductions in the NT-proBNP statistical analysis was conducted by the Statistical compared to pre-intervention levels in these Package of Social Science (SPSS) statistics soft- patients. ware program version 25 for Windows (SPSS, Inc., Chicago, IL). Concerning the functional capacity, this study showed that the early mobilisation combined with Results the usual medical care led to signi¯cantly more improvements in the 6-MWD and the RPE post At baseline, there were non-signi¯cant di®erences 6-MWT compared to the medical care alone. In between the two groups in the patients' demo- agreement with these ¯ndings, Babu et al.12 found graphic, anthropometric, or clinical characteristics that the 6-MWD increased signi¯cantly at dis- and medications, as shown in Table 2. Compared charge in hospitalised HF patients who underwent to the control group, the study group showed phase I cardiac rehabilitation compared to patients in the control group. Also, Delgado et al.21

26 A. M. Ahmad et al. Table 2. Baseline characteristics. Study group Control group Characteristics (n1 ¼ 15) (n2 ¼ 15) p-Value Age (years) 55.4 Æ 5.46 55.73 Æ 5.61 0.871 0.245 Gender Males 12 (80%) 8 (53%) Females 3 (20%) 7 (47%) 0.693 0.540 Body mass index (kg/m2) 27.64 Æ 4.05 26.79 Æ 7.21 > 0.99 Ejection fraction (%) 34.2 Æ 12.43 37.2 Æ 14.05 > 0.99 NYHA Classi¯cation Class III 9 (60%) 8 (53%) Class IV 6 (40%) 7 (47%) 0.710 Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com 0.462 by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Heart failure Left-sided 13 (86.6%) 13(86.6%) 0.449 2 (13.3%) 2(13.3%) 0.715 Right-sided 8 (53.3%) 10 (66.7%) > 0.99 7 (46.7%) 10 (66.7%) > 0.99 Diabetes 7 (46.7%) 4 (26.7%) > 0.99 6 (40%) 8 (53.3%) Hypertension 4 (26.7 %) 3 (20%) > 0.99 4 (26.7%) 5 (33.3%) 0.214 Smoking history 9 (60%) 8 (53.3%) > 0.99 > 0.99 Chest infection 11 (73.3%) 11 (73.3%) 0.651 2 (13.3%) 6 (40%) 0.710 Atrial ¯brillation 7 (46.7%) 7 (46.7%) 0.715 3 (20%) 4 (26.7%) Dilated cardiomyopathy 13 (86.7%) 8 (53.3%) 11 (73.3%) Ischemic cardiomyopathy 8 (53.3%) 10 (66.7%) Medications 6 (40%) Diuretics ACE Inhibitors/ARBs Beta-blockers/calcium channel blockers Digoxin Anti-platelets/Anti-coagulants Anti-hyperglycemic Statins Notes: Data are expressed as means Æ standard deviations and as frequencies and percent distributions. The Unpaired t-test was used to analyze continuous variables between groups. The Fisher exact test was used to analyze the unpaired proportions between the two groups. reported that the 6-MWD showed signi¯cantly considered a clinically meaningful outcome for more improvement at discharge following an early hospitalised HF patients and signi¯cantly related mobilisation protocol in hospitalised patients for to a reduced risk for a 3-year mortality post-dis- acute HF. They also found that the breathlessness charge.38 On the contrary, according to a recent that limits daily activities in these patients, asses- meta-analysis by Fuentes-Abola¯o et al.39 HF sed by the London chest activity of daily living patients with poor 6-MWT performance are at a (LCADL) scale, has reduced signi¯cantly in the higher risk for all-cause mortality on the whole and intervention group compared to the control cardiovascular mortality in speci¯c. It is worth group.21 In addition, Oliveira et al.16 showed that noting that dyspnea was associated with unfavor- low-intensity exercise using unloaded in-bed cycle able clinical outcomes in patients with acute HF.20 ergometer resulted in more signi¯cant improve- So, we can suggest that the reduction in the RPE ments in 6-MWD compared to the controls and following early mobilisation in our study could be more reduction in dyspnea compared to baseline in associated with other favorable outcomes such as acute HF patients. Further, Takada et al.22 found better BI and shorter LOS. that early rehabilitation (within three days of ad- mission) led to higher capability for unassisted Another ¯nding in this study was the signi¯cant walking in hospitalised patients with acute HF improvement in the ADL, evidenced by higher BI compared to those who received rehabilitation scores, following the early mobilisation program lately. Notably, an increase of > 50 m in the 6- compared to baseline and the controls. Per this, MWD following in-patient rehabilitation has been Motoki et al.18 found that the median BI scores increased signi¯cantly compared to baseline after

E®ects of early mobilisation in patients hospitalised for acute HF 27 Table 3. Results of the two groups pre- and post-interventions. Outcome measures Study group Control group p-value (n1 ¼ 15) (n2 ¼ 15) 6-MWT a 6-MWDb (meters) At discharge < 0.001** RPE c At discharge 252.29 Æ 95.81 106.36 Æ 58.49 <0.001** 12.53 Æ 0.91 15.4 Æ 1.63 (a 20-point Borg scale) Baseline 0.155 At discharge 41.87 (35–55) 30 (20–40) 0.006 ‡ Barthel index (0–100) 100 (100-100) 92.5 (85-95) p-value 0.536 Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com NT-proBNPd (pg/ml) Baseline 0.009 ‡ 0.031 ‡ 0.163 by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. At discharge 1069.03 Æ 528.87 973.63 Æ 261.45 0.077 LOSe (days) 786.28 Æ 269.5 947.21 Æ 341.91 Á À282.75 Æ 494.13 À26.42 Æ 222.21 0.009 § p-value At discharge 0.043* 0.652 10.43 Æ 4.40 16.86 Æ 7.13 Notes: Data are expressed as means Æ standard deviations and as medians and inter-quartile range. a6-min walk test; b6-min walk distance; crating of perceived exertion. dN-terminal prohormone of brain natriuretic peptide; elength of hospital stays. ÃÃsigni¯cant p-value based on the Unpaired t-test. Ãsigni¯cant p-value based on Paired t-test. ‡signi¯cant p-value based on the Wilcoxon signed-rank test. ‡signi¯cant p-value based on the Mann–Whitney U test. §signi¯cant p-value based on Welch's t-test. Á: absolute mean change. an in-patient cardiac rehabilitation program in NT-proBNP did not return to its normal levels patients following acute HF. Also, Kakutani et al.17 following the mobilisation program in this study, showed that a progressive mobilisation program its mean value showed a signi¯cant reduction from resulted in a signi¯cantly higher BI score than the 1069:03 Æ 528:87 pg/mL to 786.28 Æ 269.5 pg/mL. control group in hospitalised acute HF patients. In This reduction can be of clinical signi¯cance in our addition, Delgado et al.21 recorded a more observed patients aged 55:4 Æ 5:46 years. As per Januzzi increase in the BI score in hospitalised HF patients et al.40 the NT-proBNP levels below 900 pg/mL for who received early mobilisation than in the con- patients ! 50 indicate stabilised cardiac function- trols. Furthermore, in a large retrospective study by ing in acute HF patients. Not to forget that the Suzuki et al.19 early rehabilitation showed a nega- positive changes in the NT-proBNP concentrations tive association with BI deterioration and positively are more clinically meaningful when supported by related to maintenance of ADL. To be mentioned, other positive changes in a functional outcome since the decline in the ADL by acute HF has been such as 6-MWD.41 an independent risk factor for major cardiovascular events,3 it is reasonable to assume that the im- Unsurprisingly, this study revealed that early proved ADL following the mobilisation program in mobilisation led to shorter hospital stays in this study may have a long-term protective e®ect patients hospitalised for acute HF. Fleming et al.15 against these events. found that the earlier the ambulation is, the less the hospitalisation period and the lower the inci- This study also showed that the mean value of dence of hospital readmission. Also, Oliveira NT-proBNP has reduced signi¯cantly in the study et al.16 found that patients in the intervention group compared to baseline only. In agreement group had an earlier hospital discharge compared with this ¯nding, Oliveira et al.16 showed that to the controls. In addition, Kakutani et al.17 bedside low-intensity exercises induced a signi¯- showed that acute HF patients who received early cant reduction in the NT-proBNP compared to progressive mobilisation had a signi¯cantly shorter baseline and not to the controls. In addition, this hospital stay than other patients who did not. study revealed that the Á NT-proBNP was Likewise, Kaneko et al.23 reported that acute-phase more in the study group than in the control (within two days of admission) rehabilitation re- group and was somewhat close to the signi¯cance lated to a shorter hospital stay in patients with level (p ¼ 0:077). Interestingly, although the acute HF.

Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com 28 A. M. Ahmad et al. Conclusion by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. To be mentioned, in the study group, one pa- An early mobilisation program combined with the tient experienced dehydration, and another patient usual medical care can help enhance the functional had a drug-induced long QT syndrome. Both capacity (i.e., \" 6-MWD & # RPE score), improve patients were admitted to the cardiac care unit, ADL, and reduce the LOS in stabilised patients and then, they resumed the mobilisation activities following acute HF to a greater extent than the after being clinically stable as per the cardiologist's usual medical care alone. Early mobilisation could decision. Otherwise, no major cardiovascular also lead to statistically and clinically meaningful events, adverse e®ects, or falls were reported dur- reductions in the NT-proBNP concentrations ing or after the early mobilisation program. Pre- in these patients if combined with routine vious studies also reported no signi¯cant cardiac medical care. The ¯ndings of this study may be of events following mobilisation programs in stabi- clinical signi¯cance for the physiotherapists, lised acute HF patients,15–18,21 indicating that physicians, and other health professionals involved mobilisation can safely start for these patients at in the acute care of hospitalized HF patients. the earliest provided that strict patient selection Future research work on this topic may be and close monitoring are ensured. warranted. The clinical implications of this study are sev- Con°ict of Interests eral. This study extends knowledge about the e±cacy of early physiotherapist-supervised mobili- The authors state they have no con°ict of interests sation to clinicians dealing with acute HF patients. The limited knowledge or the uncertainty about its Funding/Support bene¯ts or safety can be potential reasons for the poor referral to in-patient cardiac rehabilitation No funding was received for this research (CR).42 Also, since limited resource is another major cause for not referring patients to CR in low- Author Contributions and middle-income countries,6 supervised mobili- sation appears to be a cost-e®ective intervention in Ahmad AM has contributed to the concept and such circumstances requiring minimal equipment. the design of the research, the supervision of the Furthermore, as a simple intervention, early mobi- research work, the analysis of data and interpre- lisation under the supervision of experienced phy- tation of the results, writing the whole paper and siotherapists can have good applicability to real- revising it critically for important intellectual world practice. Moreover, not only the early mobi- content, ensuring the scienti¯c accuracy and in- lisation is bene¯cial in terms of improved functional tegrity of the paper, the ¯nal approval of the paper ability and independence of HF patients at dis- before submission, and the second revision of the charge, but also it provides a good opportunity for paper as per the reviewers' instructions. Elshe- engaging in out-patient exercise-based cardiac re- nawy AI contributed to the concept and design of habilitation programs with a satisfactory level of the research, application of physiotherapy sessions, baseline functional capacity at the beginning. evaluation of the outcome measures, data acqui- sition, as well as the revision and ¯nal approval of The limitations of this study include the lack of the paper. Abdelghany M contributed to the con- long-term follow-up after hospital discharge. Also, cept of the research, the medical supervision of the the drop-outs exceeded the pre-set allowed per- research work, as well as the revision and ¯nal centage for the drop-outs. Nevertheless, this study approval of the paper. Abdel-Gha®ar HA con- has several strengths. Our study is one of the few tributed to the concept and the design of the re- randomised controlled studies in the ¯eld of early search, the supervision of the research work, as physical rehabilitation for in-patients with acute well as the revision and the ¯nal approval of the HF, which could help future meta-analyses in lay- paper. ing down a foundation for an evidence-based physiotherapy practice for this patient population. Also, in this study, the use of 6-MWT in collabo- ration with other intermediate endpoints could have produced a consistent evaluation of the clin- ical bene¯ts of the early mobilisation program in hospitalised HF patients.

E®ects of early mobilisation in patients hospitalised for acute HF 29 Appendix A Barthel Index Activity Score Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com Feeding by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. 0 = unable 5 ¼ needshelp cutting, spreading butter, etc. or requires modi¯ed diet 10 ¼ independent Bathing 0 ¼ dependent 5 ¼ independent (or in shower) Grooming 0 ¼ needs to help with personal care 5 ¼ independent face/hair/teeth/shaving (implements provided) Dressing 0 ¼ dependent 5 ¼ needs help but can do about half unaided 10 ¼ independent (including buttons, zips, laces, etc.) Bowels 0 ¼ incontinent(or needs to be given enemas) 5 ¼ occasional accident 10 ¼ continent Bladder 0 ¼ incontinent, or catheterized and unable to manage alone 5 ¼ occasional accident 10 ¼ continent Toilet Use 0 ¼ dependent 5 ¼ needs some help, but can do something alone 10 ¼ independent (on and o®, dressing, wiping) Transfers (bed to chair and back) 0 ¼ unable, no sitting balance 5 ¼ major help (one or two people, physical), can sit 10 ¼ minor help (verbal or physical) 15 ¼ independent Mobility (on level surfaces) 0 ¼ immobile or < 50 yards 5 ¼ wheelchair independent, including corners, > 50 yards 10 ¼ walks with help of one person (verbal or physical) > 50 yards 15 ¼ independent (but may use any aid; for example, stick) > 50 yards Stairs 0 ¼ unable 5 ¼ needs help (verbal, physical, carrying aid) 10 ¼ independent Total (0–100) Higher scores indicate greater functional independence Source: Mahoney FI, Barthel DW. Functional evaluation: The Barthel index. Md State Med J 1965;14:61–65. References 3. Takabayashi K, Kitaguchi S, Iwatsu K, Morikami Y, Ichinohe T, Yamamoto T, et al. A decline in activi- 1. Roger VL. Epidemiology of Heart Failure. A con- ties of daily living due to acute heart failure is an temporary perspective. Circul Res 2021;128:1421– independent risk factor of hospitalization for heart 34. failure and mortality. J Cardiol 2019;73:522–29. 2. Arrigo M, Jessup M, Mullens W, Reza N, Shah 4. Robson, A. A tailored cardiac rehabilitation for AM, Sliwa K, et al. Acute heart failure. Nat Rev patients with acute heart failure. Nat Rev Cardiol Dis Primers 2020;6:16. 2021;18:544.

Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com 30 A. M. Ahmad et al. mobilization program for patients with acute heart by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. failure reduces hospital stay and improves clinical 5. Chun KH, Kang SM. Cardiac rehabilitation in outcome. Circul Rep 2019;1:123–30. heart failure. Int J Heart Fail 2021;3:1–14. 18. Motoki H, Nishimura M, Kanai M, Kimura K, Minamisawa M, Yamato S. Impact of inpatient 6. Taylor RS, Dalal HM, McDonagh STJ. The role of cardiac rehabilitation on Barthel Index score and cardiac rehabilitation in improving cardiovascular prognosis in patients with acute decompensated outcomes. Nat Rev Cardiol 2022;19:180–94. heart failure. Int J Cardil 2019;293:125–30. 19. Suzuki S, Momosaki R, Watanabe T, Abo M. 7. Kono Y, Izawa H, Aoyagi Y, Ishikawa A, Sugiura E®ectiveness of early rehabilitation for acute heart T, Mori E, et al. Predictive impact of early mobi- failure: A retrospective cohort study. J Cardiopulm lization on rehospitalization for elderly Japanese Rehabil Prev 2019;39:E23–25. heart failure patients. Heart Vessels 2020;35:531– 20. Yoshimura K, Hiraoka A, Saito K, Urabe Y, Maeda 36. N, Yoshida T, et al. Dyspnea during in-hospital rehabilitation as a predictor of rehospitalization 8. Ahmad AM. Essentials of Physiotherapy after and mortality in patients with acute heart failure, thoracic surgery: What physiotherapists need to J Cardiopulm Rehabil 2019;39:E24–27. know. A narrative review. Korean J Thorac Car- 21. Delgado BM, Lopes I, Gomes B, Novo A. Early diovasc Surg 2018;51:293–07. rehabilitation in Cardiology – Heart Failure: The eric-HF protocol, a novel intervention to decom- 9. Semsar-Kazerooni K, Dima D, Valiquette J, pensated heart failure patients rehabilitation. Eur J Berube-Dufour J, Goldfarb M. Early mobilization Cardiovasc Nurs 2020;19:592–99. in people with acute cardiovascular disease. Can J 22. Takada S, Kondo T, Yasunaga M, Watanabe S, Cardiol 2021;37:232–40. Kinoshita H, Fukuhara S, et al. Early rehabilita- tion in older patients hospitalized with acute 10. Morris JH, Chen L. Exercise training and heart decompensated heart failure: A retrospective co- failure: a review of the literature. Cardiac Fail Rev hort study. Am. Heart J. 2020;230:44–53. 2019;5:57–61. 23. Kaneko H, Itoh H, Kamiya K, Morita K, Sugimoto T, Konishi M, et al. Acute-phase initiation of car- 11. Cops J, Haesen, S, De Moor B, Mullens W, Hansen diac rehabilitation and clinical outcomes in hospi- D. Exercise intervention in hospitalized heart fail- talized patients for acute heart failure. Int J ure patients, with emphasis on congestion-related Cardiol 2021;340:36–41. complications: A review. Heart Fail Rev 24. Kitzman DW, Whellan DJ, Duncan P, Pastva AM, 2020;25:257–68. Mentz RJ, Reeves GR, et al. Physical rehabilitation for older patients hospitalized for heart failure. New 12. Babu AS, Maiya AG, George MM, Padmakumar Engl J Med 2021;385:203–16. R, Guddattu V. E®ects of combined early in- 25. Murray EM, Whellan DJ, Chen H, Bertoni AG, patient cardiac rehabilitation and structured home- Duncan P, Pastva AM, et al. Physical rehabilita- based program on function among patients with tion in older patients hospitalized with acute heart congestive heart failure: A randomized controlled failure and diabetes: Insights from REHAB-HF. trial. Heart Views 2011;12:99–103. Am J Med 2022;135:82−90. 26. Babu AS, Arena R, Satyamurthy A, Padmakumar 13. Scrutinio D, Passantino A, Catanzaro R, Farinola R, Myers J, Lavie CJ. Review of trials on exercise- G, Lagioia R, Mastropasqua F, et al. Inpatient based rehabilitation interventions following acute cardiac rehabilitation soon after hospitalization for decompensated heart failure: Observations from the acute decompensated heart failure: A propensity WHO International Clinical Trial Re¯stry Plate- score study. J Cardiopulm Rehabil Prev form. J Cardiopulm Rehabil Prev 2021;41:214–23. 2012;32:71–77. 27. Tanaka H. Physical exercise in older patients with heart failure. J Cardiovasc Aging 2022;2:2. 14. Reeves GR, Whellan DJ, O'Connor CM, Duncan 28. Yamamoto S, Yamaga T, Nishie K, Sakai Y, Ishida P, Eggebeen JD, Morgan TM, et al. A Novel re- T, Oka K, et al. Impact of physical performance on habilitation intervention for older patients with prognosis among patients with heart failure: sys- acute decompensated heart failure: The REHAB- tematic review and meta-analysis. J Cardiol HF Pilot Study. JACC Heart Fail 2017;5:359–66. 2020;76:139–46. 29. Uemura Y, Shibata R, Takemoto K, Koyasu M, 15. Fleming LM, Zhao X, Devore AD, Heidenreich PA, Ishikawa S, Murohara T, et al. Prognostic impact Yancy CW, Fonarow GC, et al. Early ambulation among hospitalized heart failure patients is associ- ated with reduced length of stay and 30-day read- missions. Circulation: Heart Failure 2018;11: e004634. 16. Oliveira MF, Santos RC, Artz SA, Mendez VMF, Lobo DML, Correia EB, et al. Safety and e±cacy of aerobic exercise training associated to non-invasive ventilation in patients with acute heart failure. Arq Bras Cardiol 2018;110:467–75. 17. Kakutani N, Fukushima A, Kinugawa S, Yokota T, Oikawa T, Nishikawa M, et al. Progressive

E®ects of early mobilisation in patients hospitalised for acute HF 31 Hong Kong Physiother. J. 2023.43:19-31. Downloaded from worldscientific.com of the preservation of activities of daily living on cardiology and exercise in patients with cardio- by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. post-discharge outcomes in patients with acute vascular disease: The Task Force on sports cardi- heart failure. Circ J 2018;82:2793–99. ology and exercise in patients with cardiovascular 30. Mahoney FI, Barthel DW. Functional evaluation: disease of the European Society of Cardiology The Barthel index. Md State Med J 1965;14:61–65. (ESC). Eur Heart J 2020;42:17–96. 31. Mueller C, McDonald K, de Boer RA, Maisel A, 37. Mytinger M, Nelson RK, Zuhl M. Exercise pre- Cleland JGF, Kozhuharov N, et al. Heart failure scription guidelines for cardiovascular disease association of the European Society of Cardiology patients in the absence of a baseline stress test. J Practical Guidance on the use of natriuretic pep- Cardiovasc Dev Dis 2020;7:15. tide concentrations. Eur JHeart Failure 38. Scrutinio D, Guida P, Passantino A, Scalvini S, 2019;21:715–31. Bussotti M, Forni G, et al. Association of im- 32. Rørth R, Jhund PS, Yilmaz MB, Kristensen SL, provement in functional capacity after rehabilita- Welsh P, Desai AS, et al. Comparison of BNP and tion with long-term survival in heart failure. Int J NT-proBNP in patients with heart failure and re- Cardiol 2022;352:92–97. duced ejection fraction. Circ: Heart Fail 2020;13: 39. Fuentes-Abola¯o IJ, Stubbs B, Perez-Belmonte e006541. LM, Bernal-López MR, Gómez-Huelgas R, Cuesta- 33. Charan J, Biswas T. How to calculate sample size Vargas AI. Physical functional performance and for di®erent study designs in medical research? prognosis in patients with heart failure: a system- Indian J Psychol Med 2013;35:121–26. atic review and meta-analysis. BMC Cardiovasc 34. Holland AE, Spruit MA, Troosters T, Puhan MA, Disord 2020;20:512. Pepin V, Saey D, et al. An o±cial European Re- 40. Januzzi JL, van Kimmenade R, Lainchbury J, spiratory Society/American Thoracic Society Bayes-Genis A, Ordonez-Llanos J, Santalo-Bel M, Technical Standard: ¯eld walking tests in chronic et al. NT-proBNP testing for diagnosis and short- respiratory disease. Eur Respir J 2014;44:1428–46. term prognosis in acute destabilized heart failure: 35. McDonagh TA, Metra M, Adamo M, Gardner RS, An international pooled analysis of 1256 patients: Baumbach A, B€ohm M, et al. ESC Scienti¯c Doc- the international collaborative of NT-proBNP ument Group, 2021 ESC Guidelines for the diag- Study. Eur Heart J 2006;27:330–7. nosis and treatment of acute and chronic heart 41. Ferreira JP, Duarte K, Graves TL, Zile MR, failure: Developed by the Task Force for the diag- Abraham WT, Weaver FA, et al. Natriuretic nosis and treatment of acute and chronic heart Peptides, 6-Min Walk Test, and quality-of-life failure of the European Society of Cardiology questionnaires as clinically meaningful endpoints in (ESC) With the special contribution of the Heart HF Trials. J Am Coll Cardiol 2016;68:2690–707. Failure Association (HFA) of the ESC. Eur Heart J 42. Elsakr C, Bulger DA, Roman S, Kirolos I, Khou- 2021;42:3599–726. zam RN. Barriers physicians face when referring 36. Pelliccia A, Sharma S, Gati S, Bäck M, B€orjesson patients to cardiac rehabilitation: A narrative re- M, Caselli S, et al. 2020 ESC Guidelines on sports view. Ann Transl Med 2019;7:414.

Research Paper Hong Kong Physiotherapy Journal Vol. 43, No. 1 (2023) 33–41 DOI: 10.1142/S1013702523500026 Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com Hong Kong Physiotherapy Journal by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. https://www.worldscientific.com/worldscinet/hkpj Reliability and validity of the Hindi version of international physical activity questionnaire-long-form (IPAQ-LF) Sukhada S. Prabhu* and Anuprita M. Thakur School of Physiotherapy, D.Y. Patil University Nerul, Navi Mumbai, India *[email protected] Received 15 June 2021; Accepted 20 October 2022; Published 23 December 2022 Background: IPAQ-LF is a widely used tool for subjective assessment of physical activity. It has been translated, cross-culturally adapted into many languages and tested in many countries around the world. However, no Hindi version of the long-form of this questionnaire exists till date. Objective: To cross-culturally adapt the IPAQ-LF from English to Hindi language and to evaluate its reliability and validity. Methods: The guidelines by IPAQ Committee were followed for cross-cultural adaptation process. The Test–retest reliability was assessed on 60 participants by administering Hindi IPAQ-LF twice within two- week time frame. The construct validity was assessed by comparing with seven-day pedometer recording. Results: Excellent reliability was observed between total physical activity scores on repeated Hindi IPAQ- LF administrations, with interclass correlation coe±cient of 0.963 at 95% con¯dence interval. The ICC for job, transport, Housework and Leisure domain was calculated to be 0.923, 0.839, 0.862 and 0.939, respec- tively, suggesting excellent reliability. The Cronbach's alpha computed (0.82) suggests good internal con- sistency. The Hindi Version of IPAQ-LF also demonstrated good construct validity with Spearman correlation coe±cient of 0.783. Bland–Altman analyses were performed to evaluate the level of agreement between two constructs. Conclusion: The study demonstrates that Hindi version of IPAQ-LF is a reliable and valid tool for assessing physical activity levels for Hindi speaking population. Keywords: Hindi version of International physical activity questionnaire; long-form; cross-cultural adap- tation; reliability; validity. *Corresponding author. Copyright@2023, Hong Kong Physiotherapy Association. This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 (CC BY- NC-ND) License which permits use, distribution and reproduction, provided that the original work is properly cited, the use is non-commercial and no modi¯cations or adaptations are made. 33

Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com 34 S. S. Prabhu & A. M. Thakur like occupational/job (related to work), domestic by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. (household chores), transportation (walking, pub- Introduction lic transportation) and leisure time (recreational activities). It also di®erentiates between usual sit- Physical activity is de¯ned as any bodily move- ting time on a week day and a weekend day.8 ment produced by skeletal muscles that require energy expenditure. Physical inactivity (lack of The IPAQ has been translated and cross- physical activity) ranks as the fourth leading risk culturally adapted into many languages and factor for global mortality. Moreover, physical in- assessed in many countries around the world. Hindi activity is estimated to be the main cause for ap- is a widely spoken in India as it is the prime and proximately 27% of diabetes, 21–25% of breast and o±cial language. About 41% of total population of colon cancers, and approximately 30% of ischaemic India are Hindi speakers. However, no Hindi ver- heart disease burden.1 Physical activity (PA) has sion of the long-form of this questionnaire exists to been proved to be e®ective in preventing numerous this date. Thus, in order to make the IPAQ ap- lifestyle-related chronic disorders such as cardio- plicable for research among non-English speaking vascular diseases, diabetes, and hypertension.2 To populations in India, it needs to be translated, improve the ¯tness and endurance levels and to culturally adapted and properly evaluated for reduce the risk of non-communicable diseases, the psychometric properties. In India, the sociocultural World Health Organization (WHO) recommends and physical environment di®ers from other parts practicing at least 150 min of moderate-intensity or of world so; mere translation of the IPAQ may be 75 min of vigorous-intensity aerobic PA through- insu±cient to maintain content validity. Hence to out the week, or an equal combination of both.1 maintain the conceptual equivalence, a cross cul- Physical activity is generally classi¯ed as Low, tural adaptation of a questionnaire is necessary. Moderate and High levels of Physical activity and is measured in METs. Hence, the purpose of this study was to translate and cross-culturally adapt the English IPAQ-LF, According to Ainsworth,3 there are various and to evaluate aspects of the reliability and va- objectives and subjective methods used for PA lidity of the Hindi-translated and culturally adap- assessment. Objective methods consist of motion ted version of the IPAQ-LF. sensors like accelerometers, pedometers, heart rate monitors, oxygen consumption meters, wearable Subjects and Methods monitors or other measures of energy expenditure like doubly labelled water, direct calorimetry The study was approved by Institutional Ethics whereas the subjective methods include PA diaries Committee for Biomedical and Health Research. and questionnaires which are the most broadly The sample size estimation was done at 95% of adopted tracking tools.4 In the epidemiological con¯dence interval considering large magnitude of studies, the questionnaires are often used as they e®ect size (0.631) which was derived from our pilot are cost-e®ective and can be easy to administer on study and accordingly 60 participants for reliabil- large population. There are various scales for ity and validity analysis within the age group of assessing the PA such as International Physical 20–69 years were randomly selected. activity questionnaire (IPAQ), Global Physical Activity Questionnaire (GPAQ) and Physical Ac- The informed consent was taken from partici- tivity readiness (PAR), Duke's. pants for study. Participants who were able to move independently, willing to participate in the An international consensus group formulated study and were well versed in Hindi language were the IPAQ in 1998 for young to middle-aged included in the study, whereas individuals with adults.5–7 It is a self-administered questionnaire severe chronic diseases likely to hinder physical which has acceptable validity and reliability when activities, individuals diagnosed with psychiatry or evaluating the levels and patterns of physical ac- cognitive diseases, individuals undergone recent tivity in healthy adults ranging from 15 to 69 years surgery were excluded from the study. of age. There are two forms of IPAQ—long-form and short-form with a reference period of either Methods \\the last seven days\" or \\the usual week.\"5,6 There are seven questions in short form and 27 questions The translation and cultural adaptation of the in long-form. The IPAQ long-form provides speci¯c IPAQ-LF were performed in several steps following details on PA intensity levels in the four domains

Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com the guidelines prescribed by the IPAQ core group Reliability and validity of the Hindi version of IPAQ-LF 35 by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. and WHO which consists of forward translation of questionnaire followed by its backward transla- IPAQ-LF. The pre-¯nal Hindi version was then tion.9 A pilot testing of pre-¯nal synthesised Hindi reviewed by the expert panel for conceptual, lin- Version of IPAQ-LF was done and ¯nal version of gual and metric equivalence. Hindi IPAQ-LF was obtained (Fig. 1). Pilot Testing Forward Translation into Hindi The pilot testing of pre-¯nal Hindi version of After obtaining permission from the original au- IPAQ-LF was carried out on 64 participants who thor of English IPAQ-LF, the English version of could read, write and understand Hindi language. IPAQ-LF was forward translated into Hindi lan- The following questions were asked to them by guage by two bilingual translators who were fa- principal investigator: miliar with the concept of physical activity and were °uent in both the languages. The main focus (1) Did you understand all words? of this phase was to achieve sematic equivalence. (2) How clear was the intent of questionnaire? The translated questionnaire was reviewed by an (3) Do you have any questions about it? expert panel committee which consisted of the (4) How could wordings be clearer? principal investigator, original translator, a (5) Did any question make you feel uncomfortable? healthcare professional knowledgeable about the (6) Were there activities that we missed? concept of physical activity and a lay person. The goal of this expert panel was to retain the con- Based on pilot testing, the appropriate and neces- ceptual, lingual and metric equivalence between sary changes were made in the questionnaire and the two questionnaires and to obtain a synthesised ¯nal Hindi version of IPAQ-LF was obtained. Hindi version of IPAQ-LF. Test– retest reliability: Sixty participants were Backward Translation into English asked to complete the Hindi version of IPAQ-LF The synthesised Hindi version of IPAQ-LF was on two occasions within time frame of two weeks then backward translated into English language by (Day 1 and Day 14). Data obtained from these two two di®erent translators. The synthesised Hindi administrations was analysed by 2-way mixed version of IPAQ-LF and two back translations had model interclass correlation coe±cient. been merged into one pre-¯nal Hindi version of Internal consistency: It was done by correlating items with each other by using Cronbach's alpha value. Fig. 1. Flowchart of translation process and study protocol.

Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com 36 S. S. Prabhu & A. M. Thakur Results by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Construct validity: Omron HJ-325 Pedometer Pilot study of Pre-Final Hindi IPAQ-LF was allocated to all participants. They were A total of 64 participants (43.7% males, 56.2% instructed how to wear the pedometer and to re- females) with mean age of 40:89 Æ 13:79 years and move it only during sleeping and bathing. They BMI of 26:8 Æ 4:79 kg/m2 participated in the were also instructed to record the daily step counts study. 93.7% of participants could understand all for seven-day period (Day 1-Tuesday to Day 7- words whereas intent of questionnaire was very Monday). On eighth day, participants submitted clear to 95.26%. All the participants agreed that no the pedometers to researchers, and were advised to questions made them feel uncomfortable. Various complete Hindi version of IPAQ-LF. The average suggestions on activities to be added/deleted were of daily steps walked for seven days was counted given. These activities (Fig. 2) were computed for and this count was correlated with total physical its MET values based on \\Compendium of Physi- activity score recorded on Hindi version of IPAQ- cal Activities 2011\" and were incorporated to LF. Data obtained was analysed using Spearman prepare the ¯nal Hindi version of IPAQ-LF. correlation coe±cient test of statistical analysis. Fig. 2. Activities added/removed in Hindi Version of IPAQ-LF based on pilot testing.

Table 1. Characteristics of study population. Reliability and validity of the Hindi version of IPAQ-LF 37 Characteristics Mean Æ SD Test–retest Reliability A total 60 participants (46.7% males, 53.3% Age 40:22 Æ 14:15 years females) with mean age of 40:22 Æ 14:15 years took Height 157:97 Æ 8:58 cm part in Test–retest reliability study. Their char- Weight 66:66 Æ 10:58 kg acteristics are described in Table 1. The statistical BMI analysis was performed by using IBM SPSS ver- 26:79 Æ 4:42 kg/m2 sion 23 software. For total physical activity scores, excellent reliability was observed between Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com Table 2. Test–retest reliability of the Hindi version of IPAQ-LF. by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. First administration Second administration Interclass correlation Hindi-IPAQ-LF domains (Mean Æ SD) (Mean Æ SD) coe±cient (ICC) (95% CI) Total PA 6930:48 Æ 7650:01 7363:28 Æ 7147:24 0.985 Job-related PA 2795:19 Æ 6403:71 3205:49 Æ 6140:28 0.988 Transportation-related PA 757:21 Æ 1376:76 988:83 Æ 1475:45 0.955 Housework, House maintenance-related PA 1611:75 Æ 1730:10 0.946 Leisure-related PA 1886:28 Æ 2261 1557:22 Æ 2242:28 0.949 Time spent in sitting 1492:31 Æ 2493:25 2443:46 Æ 1375:91 0.962 2502:69 Æ 1414:12 Table 3. Internal consistency using Cronbach's alpha. Hindi-IPAQ-LF domains Cronbach's alpha Cronbach's alpha Level of agreement No. of items if item deleted Job-related PA (7 items) 0.994 Excellent Q1 0.034 Q2 0.583 Transportation-related PA (6 items) 0.496 Poor Q3 0.578 Poor Q4 0.383 Housework, House 0.518 Fair Q5 0.577 maintenance-related PA (6 items) Good Q6 0.352 Good Q7 0.577 Leisure-related PA (6 items) 0.7 Q8 0.492 Q9 0.454 Time spent in sitting (2 items) 0.849 Q10 0.251 Total PA 0.820 Q11 0.453 Q12 0.196 Q13 0.369 Q14 0.458 Q15 0.026 Q16 0.188 Q17 0.026 Q18 0.184 Q19 0.398 Q20 0.440 Q21 0.480 Q22 0.433 Q23 0.269 Q24 0.430 Q25 0.184 Q26 0.016 Q27 0.022 —

Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com 38 S. S. Prabhu & A. M. Thakur noted in transport, housework and leisure-related by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. domains (Cronbach's alpha value ranging between repeated administrations of Hindi IPAQ-LF, 0.4 to 0.8). with interclass correlation coe±cient (ICC) of 0.963 at 95% con¯dence interval (Table 2). The Construct Validity ICC for job, transport, Housework and Leisure The Spearman Rho value of 0.783 was computed domain was calculated to be 0.923, 0.839, 0.862 (Table 4), indicating strong positive correlation and 0.939, respectively, suggesting excellent between total physical activity levels and daily reliability. steps walked. The Bland–Altman plot in Fig. 3 demonstrates a satisfactory level of agreement Internal consistency reliability between two constructs (mean ¼ 126:12, 95% The Cronbach's alpha computed for internal con- limits of agreement ¼ 5128:25; À4876). Further- sistency was 0.82, suggesting good internal consis- more, a regression analysis was performed and tency among items of questionnaire (Table 3). con¯rmed the absence of statistical bias (0.410; Also, the excellent agreement was found in internal p > 0:05). consistency of job domain with Cronbach's alpha value of 0.994 whereas poor to fair agreement was Table 4. Spearman correlation between the Hindi version of IPAQ-LF and pedometer. Pedometer average steps IPAQ-H total score Spearman's rho Pedometer average steps Correlation Coe±cient 1.000 0.783** IPAQ-H total score Sig. (2-tailed) — 0.000 N 60 60 Correlation Coe±cient 0.783** 1.000 Sig. (2-tailed) 0.000 N — 60 60 Note: **Correlation is signi¯cant at the 0.01 level (2-tailed). Fig. 3. Bland–Altman plot of the total PA measured on Hindi version of IPAQ-LF and average pedometer steps count.

Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com Discussion Reliability and validity of the Hindi version of IPAQ-LF 39 by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. To our best knowledge, this is the ¯rst study to researches, Craig et al. demonstrated reliability of translate and cross-culturally adapt the English IPAQ in 12 countries with ICC of 0.8 indicating IPAQ-LF to Hindi IPAQ-LF, and to evaluate its very good repeatability.13 The reliability studies reliability and validity in Hindi speaking popula- undertaken in Norway and Switzerland showed tion. The guidelines recommended by IPAQ core lowest ICC ranging from 0.30–0.62.14,15 As the group were used for translation and adaptation above-mentioned studies demonstrated great vari- process.9 This method is cost e±cient and widely ability in reliability results, the reason for this can applied in small budget studies. The pre-¯nal ver- be attributed to di®erence in intervals between ¯rst sion of Hindi IPAQ-LF was tested on 64 volun- administration and second administration of the teers, which allowed the researcher to detect the tests. A period shorter than eight days is associated discrepancies and to make necessary changes to the with a higher reliability coe±cient whereas a peri- adapted questionnaire. For instance, transporta- od of three weeks signi¯cantly reduces the reli- tion by tram is invalid for Indians, so this word was ability. In this study, a time frame of two weeks replaced by most frequently used modes of trans- between ¯rst administration and second adminis- port like metro, bus, car, autorickshaw and scooter. tration was decided. This time frame was consid- Also, the activity of snow shoveling is uncommon ered to be the su±cient period to eliminate the in this population, hence it was suggested to be in°uences of the ¯rst response on the results and replaced by other activities with similar MET too short period for any substantial changes to values like washing clothes and utensils, cooking, occur in PA.16 ¯lling of water. Stair climbing activity was added in job domain as this was more common than use of The internal consistency value analysed was 0.82 elevators and escalators which may be unavailable indicating high agreement within questionnaire. at all work places and thus population may take to This result goes in accordance with study done by walking and stair climbing rather than waiting for Khalil et al. which shows high internal consistency elevators to move around at workplace. Recrea- reliability with Cronbach's alpha ranging from 0.76 tional activities like dancing, playing sports like to 1.00 in an adapted Arabic version of IPAQ-LF.17 cricket, badminton, doing power yoga or sur- The Cronbach's alpha if item deleted represents the yanamaskar were some of the commonly performed scale's Cronbach alpha reliability co-e±cient for physical activities in domain of recreation, sports internal consistency if individual item is removed and leisure time. In the last domain which is time from the scale.18 The scale's Cronbach alpha is spent in sitting, words like Monday to Friday/ noted if a particular item is deleted. This value is Saturday were added to clarify the terms weekdays then compared to alpha co-e±cient value to see if and weekends. item wants to be deleted.18 The Hindi version of IPAQ-LF has seven items clustered under job- The Hindi version of IPAQ-LF was tested for related PA domain, two items under time spent in test–retest and internal consistency reliability. The sitting domain and six items each under transpor- ICC computed for total PA, job, transport, tation, Housework- and Leisure-related PA housework and leisure domain was high, ranging domains. In this study, it is seen that Cronbach's from 0.83 to 0.96 suggesting excellent reliability. alpha if item deleted for each item is lower than its The least reliability was established in transport corresponding overall Alpha value for that partic- domain (0.83) and the most reliability in the leisure ular domain. For job-related domain, the Cron- domain (0.93). Our results are in accordance to bach's alpha if item deleted ranges from 0.034 to other studies indicating that various adapted ver- 0.583 which is less than overall Cronbach's alpha for sions of IPAQ-LF have good excellent reliability. this domain which is 0.994 (Table 3). Similar results The Belgium and Chinese versions of IPAQ-LF can be noted for other domains as well like for showed good to excellent reliability of the adapted transportation-related PA, Housework-related PA, versions on test–retest reliability testing with ICC Leisure-related PA and time spent in sitting ranging from 0.52–0.81 and 0.74 to 0.97, respec- (Table 3). Hence, the Cronbach's alpha if item de- tively.10,11 Whereas the Serbian versions of IPAQ- leted for all these items cluster under particular LF demonstrated moderate to excellent reliability domain is lower than overall Cronbach's alpha with ICC from 0.53 to 0.91.12 In the previous value, it can be interpreted that all these items should be retained and cannot be deleted from the domain of the scale in which the item belongs to.

Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com 40 S. S. Prabhu & A. M. Thakur the study. Prabhu S. Sukhada was involved with by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. analysis/interpretation of data whereas Thakur M. Pedometers are e®ective body motion sensors Anuprita was involved with critical revision of that respond to vertical movement created by ac- manuscript for important intellectual content. All tivities such as walking, jogging and running.19 authors were involved in the acquisition of data, They provide precise, convenient, low cost, and drafting of manuscript, and approval of manu- objective measure of physical activity.19 Ped- script to be published. ometers have proven to have su±cient evidence with other sophisticated measures of PA.20 They References are utilised as an objective measure of physical activity and have been used in validation of self- 1. World Health Organization. Global recommenda- report measures including the IPAQ.21,22 The tions on physical activity for health. 2010. Omron HJ-325 pedometer used in this study is most reliable and valid for counting the number of 2. Warburton DE, Nicol CW, Bredin SSD. Health steps.23 A strong positive correlation coe±cient bene¯ts of physical activity: The evidence. Can (r ¼ 0:783) was obtained between total PA from Med Assoc J 2006;174:801–9. Hindi version of IPAQ-LF and pedometer step count in this study. Streiner et al. suggested that for 3. Ainsworth BE, Bassett DR, Strath SJ. Comparison a validity result, the correlation coe±cient should be of three methods for measuring the time spent in situated between 0.4–0.8 for it to accept.24 Our physical activity. Med Sci Sports Exerc 2000;32(9): study ¯ndings are similar to other researchers who S457–S464. used pedometer values to validate IPAQ, for in- stance Bassett et al. (r ¼ 0:47),21 Deng et al. 4. Strath SJ, Kaminsky LA, Ainsworth BE, Ekelund (r ¼ 0:33),22 Craig et al. (r ¼ 0:33)7 and Vandela- U, Freedson PS, Gary RA, et al. Guide to the as- notte et al. (r ¼ 0:38).25 The Bland–Altman plots, sessment of physical activity: Clinical and research. plotted in this study, show more than 95% of the BMC Publ Health 2018;18:49. point lying between upper and lower limits of agreement which leads to the conclusion that Hindi 5. Kim Y, Park I, Kang M. Convergent validity of the Version of IPAQ-LF should be considered as inter- international physical activity questionnaire changeable with the Pedometer and thus the two (IPAQ): Meta-analysis. Publ Health Nutr scales exhibited alternate form of validity. 2013;16:440–52. This study has provided some preliminary evi- 6. Hagstromer M, Oja P, Sjostrom M. The interna- dence for reliability of Hindi version of IPAQ-LF. tional physical activity questionnaire (IPAQ): A study of concurrent and construct validity. Publ In conclusion, our analysis results demonstrate Health Nutr 2006;9(6):755–62. good to excellent reliability and strong evidence on validity of Hindi version of IPAQ-LF that can be 7. Craig CL, Marshall AL, Sjostrom M, Bauman AE, used for assessing physical activity levels in Hindi Booth ML, Ainsworth BE, et al. International speaking population. physical activity questionnaire: 12-country reli- ability and validity. Med Sci Sports Exerc Con°ict of Interest 2003;35:1381–95. The authors have no con°ict of interest relevant to 8. The IPAQ group. Guidelines for data processing this paper. and analysis of the international physical activity questionnaire (IPAQ)—short and long forms. 2005. Funding/Support 9. International Physical Activity Questionnaire [In- This research received no speci¯c grant from any ternet]. Cultural Adaptation. Available at https:// funding agency in the public, commercial or not- sites.google.com/site/theipaq/cultural-adaptation. for-pro¯t sectors. 10. Van Holle V, et al. Assessment of physical activity Author Contributions in older Belgian adults: Validity and reliability of an adapted interview version of the long Interna- Prabhu S. Sukhada and Thakur M. Anuprita were tional Physical Activity Questionnaire (IPAQ-L). responsible for research conception and design of BMC Public Health 2015;15:433. 11. Macfarlane D, et al. Examining the validity and reliability of the Chinese version of the Interna- tional Physical Activity Questionnaire, long form (IPAQ-LC). Publ Health Nutr 2015;14(3): 443–50. 12. Milanović Z. Reliability of the Serbian version of the International Physical Activity Questionnaire for older adults. Clin Interv Aging 2014;9:581–87.

Hong Kong Physiother. J. 2023.43:33-41. Downloaded from worldscientific.com 13. Craig CL et al. International Physical Activity Reliability and validity of the Hindi version of IPAQ-LF 41 by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Questionnaire: 12-Country Reliability and Validi- ty. Med Sci Sports Exerc 2003;35(8):1381–95. 20. Tudor-Locke C, Williams JE, Reis JP, Pluto D. Utility of pedometers for assessing physical activi- 14. Kurtze N, Rangul V, Hustvedt B-E. Reliability and ty: Convergent validity. Sports Med 2002; validity of the international physical activity 32(12):795–08. questionnaire in the Nord-Trøndelag health study (HUNT) population of men. BMC Med Res 21. Bassett DR, Schneider PL, Huntington GE. Phys- Methodol 2008;8:63. ical activity in a old order Amish community. Med Sci Sports Exerc 2004;36(1):79–85. 15. Sember V, Meh K, Sorić M, Starc G, Rocha P, Jurak G. Validity and Reliability of International 22. Deng HB, MacFarlane DJ, Thomas GN, et al. Physical Activity Questionnaires for Adults across Reliability and validity of the IPAQ-Chinese: The EU Countries: Systematic Review and Meta- Guangzhou Biobank cohort study. Med Sci Sports Analysis. Int J Environ Res Publ Health 2020; Exerc 2008;40(2):303–7. 17(19):7161. 23. Lee JA, Williams SM, Brown DD, Laurson KR. 16. Prot F, Sporiš G, Bosnar K. Research methods in Concurrent validation of the Actigraph gt3xþ, kinesiology. Res Methodology 2009; 9:581–587. Polar Active accelerometer, Omron HJ-720 and Yamax Digiwalker SW-701 pedometer step counts 17. Khalil H. Validity and reliability of an adapted in lab-based and free-living settings. J Sports Sci Arabic version of the long international physical 2015;33(10):991–1000. activity questionnaire. BMC Publ Health 2018;18:49. 24. Streiner DL, Norman GR. In: Health Measurement Scales: A Practical Guide to their Development and 18. Glient JA, Rosemary GR. Calculating, Interpreting Use. 2nd ed. Oxford, England: Oxford University & Reporting Cronbach's alpha Reliability Co- Press, 1996. e±cient for Likert-type scales. Midwest Research to Practice Conference in Adult, Continuing & Com- 25. Vandelanotte C, De Bourdeaudhuij I, Philippaerts munity Education. 2003. R, Sj€ostro€m M, Sallis J. Reliability and validity of a computerized and Dutch version of the Interna- 19. Tudor-Locke C, Bassett DR. How Many Steps/ tional Physical Activity Questionnaire (IPAQ). J Day Are Enough? Preliminary pedometer indices Phys Act Health 2005;2:63–75. for public health. Sports Med 2004;34(1):1–8.

Research Paper Hong Kong Physiotherapy Journal Vol. 43, No. 1 (2023) 43–51 DOI: 10.1142/S1013702523500038 Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com Hong Kong Physiotherapy Journal by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. https://www.worldscientific.com/worldscinet/hkpj The seventh cervical vertebra is an appropriate landmark for thoracic kyphosis measures using distance from the wall Arpassanan Wiyanad1,2,*, Sugalya Amatachaya1,2,†,††, Pipatana Amatachaya2,‡,††, Patcharawan Suwannarat2,3,§, Pakwipa Chokphukiao2,4,¶, Thanat Sooknuan2,5,|| and Chitanongk Gaogasigam2,6,** 1School of Physical Therapy, Faculty of Associated Medical Sciences, Khon Kaen University, Khon Kaen 40002, Thailand 2Improvement of Physical Performance and Quality of Life (IPQ) Research Group, Khon Kaen University, Khon Kaen 40002, Thailand 3Department of Physical Therapy, School of Integrative Medicine, Mae Fah Luang University, Chiang Rai, Thailand 4Department of Physical Therapy, Faculty of Allied Health Sciences, Naresuan University, Phitsanulok, Thailand 5Department of Electronics Engineering, Faculty of Engineering and Architecture, Rajamangala University of Technology Isan, Nakhon Ratchasima, Thailand 6Department of Physical Therapy, Faculty of Allied Health Sciences, Chulalongkorn University, Bangkok, Thailand *[email protected] †[email protected] ‡[email protected]; [email protected] §[email protected] ¶[email protected] ||[email protected] **[email protected] Received 22 January 2022; Accepted 20 October 2022; Published 25 January 2023 ††Corresponding authors. Copyright@2023, Hong Kong Physiotherapy Association. This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 (CC BY- NC-ND) License which permits use, distribution and reproduction, provided that the original work is properly cited, the use is non-commercial and no modi¯cations or adaptations are made. 43

Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com 44 A. Wiyanad et al. by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Background: Hyperkyphosis is frequently found nowadays due to the change in current lifestyles and age- related system decline. A simple hyperkyphosis measurement can be made easily using the perpendicular distance from the landmark to the wall. However, the existing evidence applied two di®erent landmarks [occiput and the seventh cervical vertebra (C7)] and the measurement using rulers was susceptible to error due to their misalignment. Objective: To assess an appropriate landmark for thoracic kyphosis measurement using distance from the wall (KMD), by comparing between occiput and C7, as measured using rulers and veri¯ed using data from a specially developed machine, the so-called infrared-gun kyphosis wall distance tool (IG-KypDisT), and the Cobb angles. Methods: Community-dwelling individuals with a risk of thoracic hyperkyphosis (age ! 10 years, n ¼ 43) were cross-sectionally assessed for their thoracic hyperkyphosis using the perpendicular distance from the landmarks, occiput and C7, to the wall using rulers and IG-KypDisT. Then the Cobb angles of these participants were measured within seven days. Results: The outcomes from both landmarks di®ered by approximately 0.8 cm (p ¼ 0.084). The outcomes derived from C7 were more reliable (ICCs>0.93, p <0.001), with greater concurrent validity with the ra- diologic data (r ¼ 0.738, p <0.001), with the overall variance predicted by the regression models for the Cobb angles being higher than that from the occiput (47–48% from C7 and 38–39% from occiput). The outcomes derived from rulers and IG-KypDisT showed no signi¯cant di®erences. Conclusion: The present ¯ndings support the reliability and validity of KMD assessments at C7 using rulers as a simple standard measure of thoracic hyperkyphosis that can be used in various clinical, community, and research settings. Keywords: Round back; spine; Dowager's hump; radiology; Cobb angle. Introduction Of all the existing hyperkyphosis measures, the occiput–wall distance (OWD) is a practical mea- Thoracic hyperkyphosis is a progressive condition sure that can be executed easily using two rulers to with backward deviation of the thoracic spine and measure the perpendicular distance from the a Cobb angle exceeding 40.1 This condition can be landmark to the wall while standing against it.11,12 found in those aged 10 years and over, approxi- However, the OWD faces some criticisms, includ- mately one-third of all teenagers and adults, and ing the following: using the occiput as a landmark 40% of older adults.2,3 The change in current life- could confound data interpretation regarding tho- styles whereby many people spend considerable racic hyperkyphosis and increase the chance of time in an excessively °exed posture using compu- errors due to individuals moving their heads, hav- ters and social media, combined with the reduction ing forward head posture, and the varying shapes in physical activities, may hasten the development of the skulls.10,13,14 Moreover, the data taken from of hyperkyphotic spine in young people.2–4 two rulers were susceptible to errors due to their misalignment, e.g., not parallel to the landmark The thoracic hyperkyphosis can introduce vari- and not perpendicular to the wall. Furthermore, ous adverse health consequences depending upon the outcomes in terms of distance could limit data the severity of the condition, such as diminished comparison with other hyperkyphosis measures self-con¯dence, musculoskeletal pain, impaired that are mostly reported in terms of spinal pulmonary and physical functions, elevated resting angles.11,12,14 Consequently, some recent studies energy consumption, and increased risk of falls and have applied the bony prominence of the seventh vertebral fractures that further accelerate the progression of thoracic curvature and its negative cervical vertebra (C7) as a landmark for the consequences.5,6 Therefore, a standard practical measurement because it is a stable point over the measure plays crucial roles in the early detection superior limit of the thoracic spine.15,16 Nevertheless, and periodic follow-up of abnormality, and in there was no evidence to con¯rm the validity and promoting treatment e®ectiveness and minimizing the harmful consequences that may occur due to reliability of the outcomes from both landmarks, thoracic hyperkyphosis.7–10 particularly covering the ages of all of those at a risk of thoracic hyperkyphosis (10 years old and

Seventh cervical vertebra for thoracic kyphosis measurements 45 Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com over). Therefore, this study investigated the intra- that was approved by the Institutional Ethics by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. and inter-rater reliabilities and concurrent validity Committees for Human Research (HE592270) prior of kyphosis measurement using distance from the to participation in this study. wall (KMD) when measured using rulers at the occiput and C7 and veri¯ed using data from a Research protocols specially developed machine, the so-called infrared- gun kyphosis wall distance tool (IG-KypDisT), and The eligible participants were involved twice in the radiologic data (Cobb's method). In addition, this observational study within seven days. At the the study explored a predictive equation for the ¯rst time, they were interviewed for their demo- overall variance, including KMD from each land- graphics and assessed for their severity of thoracic mark, from the regression models to determine the hyperkyphosis using KMD. Within the following Cobb angles. The ¯ndings would provide clear ev- seven days, participants were informed to arrive at idence to con¯rm an appropriate landmark for the hospital to complete a lateral spinal radio- KMD using rulers for screening and monitoring graphic examination (Cobb's method). Details of thoracic hyperkyphosis in various clinical, com- the measurements were as follows. munity, and research settings, especially in this COVID-19 pandemic period with limited access to Kyphosis measured using the people in a hospital. distance from the wall Methods It has been reported that the KMD data measured with the rulers were susceptible to errors due to Participants their misalignment.10,11,14 Thus, the KMD data measured with rulers in this study were veri¯ed Community-dwelling participants with or without using IG-KypDisT [Patent Application No. signs or symptoms associated with thoracic struc- 1701004049; Fig. 1(a)]10 that was specially devel- tural hyperkyphosis, i.e., those who are unable to oped to ensure perpendicular distance from the consciously straighten the spinal column,3 were landmarks to the wall using two-plane gyroscopes recruited by directly contacting the community (up to 1 mm). Outcomes of the tool showed excel- leaders. The inclusion criteria were as follows: lent correlation to the Cobb angles (r ¼ 0.92, age ! 10 years, body mass index < 29.9 kg/m2, and p <0.001).11 Therefore, IG-KypDisT was used to OWD ! 6.5 cm while attempting to straighten the verify the KMD derived from rulers. Details of the spinal column when measured using rulers by an measurements were as follows. experienced assessor.12,17 The exclusion criteria in- cluded any signs and symptoms that might a®ect Starting position. Participants stood upright participation in this study and confound data inter- with bare feet — as tall as possible — with their pretation, i.e., having abnormal fat mass in the upper heels, sacrum, and back against the wall, and with back or winged scapula that could increase the dis- the head in a neutral position as determined tance from the wall when standing against it, and using the inferior orbital margin and the superior scoliosis as determined using Adam's test. Additional margin of the acoustic meatus in a horizontal plane exclusion criteria were any conditions with contra- [Fig. 1(b)].1,11,12,15 indications for X-ray examination (i.e., pregnancy or cancer),18 di±culty in maintaining a standing pos- Methods of measurement. We used the fol- ture, and inability to understand and follow a com- lowing methods of measurement: mand used in the study. The sample size calculation (1) Rulers: Two rulers were used to quantify the for a primary objective (concurrent validity) using outcomes of each landmark (occiput or C7). The the correlation coe±cient (r) from a pilot study ¯rst ruler was placed on the landmark and another (n ¼ 20, r ¼0.42), with the signi¯cance level set at ruler was used to measure the perpendicular dis- 0.05 and the power level set at 0.8, indicated that tance from the landmark to the wall [Fig. 1(c)]. the study required at least 42 participants. Previous (2) IG-KypDisT: Prior to the measurement, studies relating to the reliability of kyphosis mea- IG-KypDisT was calibrated with a °at upright surements applied at least 15 participants.8,15,18,19 surface. Then, an experienced rater placed the Participants provided a written informed consent measuring bar of IG-KypDisT in contact with the landmark [Fig. 1(d)] and adjusted the tool per- pendicularly with the wall until they noticed a white line on the monitor in a horizontal plane

46 A. Wiyanad et al. Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com Fig. 1. Kyphosis measurements using the distance from the wall: (a) IG-KypDisT and its components. (b) Starting position by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. for KMD. (c) KMD using rulers. (d) KMD using IG-KypDisT. [Fig. 1(a)]. Then the rater pressed the measuring the raters measured the KMD at the occiput and button to obtain the outcomes. C7 using rulers on participants who agreed to be involved in the reliability study for 3 trials/land- The sequences of measurement using rulers and mark. The outcomes over the three trials of each IG-KypDisT were randomly ordered, and the av- rater from each landmark were used to analyze the erage ¯ndings over 3 trials/landmark were used for intra-rater reliability, and the average data over the data analysis. three trials of all raters were used to report the inter- rater reliability.15,19 Reliability study Kyphosis measured using the Cobb's Raters were three physiotherapists who had clini- method cal experience ranging from 2 years to 9 years. Prior to the measurement, the raters were trained All participants were ¯lmed covering the area from for a standard KMD protocol, including the starting the ¯rst to the 12th thoracic vertebra (T1–T12). position, instructions, and measurement methods, for Participants stood upright with their shoulders approximately 20 min. Then they practiced the mea- and elbows °exed at 90 to prevent the thoracic surements on four non-participants. Subsequently,

Seventh cervical vertebra for thoracic kyphosis measurements 47 Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com curvature from overlapping with the upper Results by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. limbs.12,18 Then, the radiologic image was uploaded for calculation of the Cobb angles by an experienced Forty-three participants completed the validity rater using the Surgimap Spine v1.2 software study (average Cobb angle = 40.5, ranging from (Nemaris, Inc., New York, NY, USA).19,20 The rater 11.3 to 67.7), and 17 participants were also in- drew the ¯rst line along the upper surface of the volved in the reliability study (average Cobb fourth thoracic (T4) vertebral end-plate, and the angle = 35, with a range of 28.1– 42) (see second line along the lower surface of the 12th (T12) Table 1 and Supplementary Material). Most par- vertebral end-plate. Then the software automati- ticipants did not have any signs and symptoms of cally generated the Cobb angles based on the in- thoracic hyperkyphosis, except for two older par- tersection of the two lines.12,19,20 The method was ticipants who had vertebral fractures (radiologic repeated for three trials, and the average angle was data). used for data analysis. Statistical analysis Validity of KMD when measured at The Kolmogorov–Smirnov test was applied to as- occiput and C7 sess the normality of the data distribution. De- scriptive statistics were used to explain the The KMD outcomes at each landmark showed no demographics and ¯ndings of the study. With signi¯cant di®erences between the tools (p >0.05). normal data distribution, the Bland–Altman plots However, the KMD outcomes from rulers and IG- and dependent samples t-test were used to indicate KypDisT measured at the occiput showed a wider the outcome di®erences between the rulers and IG- range of di®erences than those at C7 (Fig. 2). In KypDisT for each landmark. The Pearson corre- addition, the KMD outcomes measured at C7 were lation coe±cient was then applied to quantify the longer than those at the occiput by approximately levels of correlation between the KMD data at each 0.8 cm (8.5 Æ 1.8 cm and 9.3 Æ 2.7 cm, p ¼0.084). landmark and the Cobb angles. The levels of cor- Furthermore, the KMD outcomes measured using relation coe±cient were interpreted as negligible rulers at C7 showed signi¯cant correlation with the (below 0.30), low (0.30–0.50), moderate (0.50– data from IG-KypDisT (r ¼ 0.872, p <0.001) and 0.70), and high (above 0.70).21,22 Then, the step- Cobb angles (r ¼ 0.738, p <0.001), and were wise multiple linear regression analysis was utilized higher than those derived at the occiput (r ¼ 0.636 to formulate a predictive equation from all possible with IG-KypDisT and r ¼ 0.442 with Cobb angle, demographic and KMD variables relating to the p <0.003, Fig. 2). Among all possible predictive presence of thoracic kyphosis (i.e., age, gender, variables, only age and KMD data were signi¯cant bodyweight, height, and KMD) to determine the for determining the Cobb angles (p <0.001), Cobb angles.3 Adjusted R2 was used to indicate whereby the overall variance predicted by the re- the most appropriate equation, with an R2 value of gression models when measured at C7 (47–48%, 1 indicating that the data perfectly ¯t the linear Table 2) was higher than that derived at the model.23 The intra-class correlation coe±cients occiput (38–39%, Table 2). (ICCs),3 i.e., participants were assessed by raters who were only the raters of interest, were used to Table 1. Demographics of the participants. quantify the reliability of KMD outcomes at each landmark. The reliability was considered as poor Validity study Reliability study (below 0.50), moderate (0.50–0.75), and high (above 0.75).20,21 Moreover, the absolute reliability Variable (n ¼ 43) (n ¼ 17) of the KMD at each landmark was further assessed and reported in terms of the standard error of Age† (years) 52.8 Æ 25.2 48.8 Æ 22.1 mpeaffisffiffiuffiffiffirffiffieffiffimffiffiffiffieffiffinffiffiffit (SEM), using the equation SD Body mass index† (kg/m2) (45–60.5) (37.4–60.2) Â 1 À ICCs, and the minimal detectable chpaffinffi ge Cobb angles† (deg) 23.5 Æ 3.5 22.6 Æ 2.9 (MDC), using the equation SEM Â1:96 Â 2.15 Gender: female: number (%) (22.4–24.6) (21.2–24.1) The p < 0.05 indicated signi¯cant di®erences. 40.5 Æ 13.3 35 Æ 13.5 (36.5–44.6) (28.1–42) 21 (48) 7 (41) Note: †The data are presented in the form of mean Æ stan- dard deviation (95% con¯dence interval).

48 A. Wiyanad et al. Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com (a) (b) by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Fig. 2. Bland–Altman plots for kyphosis measurement using distance from the wall. Table 2. Predictive equations for kyphosis measurement using distance from the wall when measured at the occiput and C7 (n ¼ 43). Landmark Tool Predictive equation* Adjusted R2 Occiput Rulers 2.06(KMD)+0.26(age)+9.44 0.39 C7 IG-KypDisT 1.48(KMD)+0.24(age)+15.44 0.38 2.49(KMD)+0.13(age)+10.43 0.47 Rulers 2.11(KMD)+0.15(age)+14.13 0.48 IG-KypDisT Note: IG-KypDisT: The infrared-gun kyphosis wall distance tool. KMD: Kyphosis measurement using distance from the wall. *The equations for pre- dicting the Cobb angle were derived from multiple linear regressions. Reliability of KMD when measured raters, with excellent intra- and inter-rater reli- using rulers at the occiput and C7 abilities and small SEM and MDC, particularly when measured at C7 (ICCs ! 0.933, p <0.001, The KMD data measured using rulers showed Table 3). no signi¯cant di®erences among the three Table 3. Rater reliability of kyphosis measurement using distance from the wall using rulers (n ¼ 17). Landmark Variable Rater 1 Rater 2 Rater 3 p-Value Occiput Distance† (cm) 8.16 Æ 1.48 7.45 Æ 1.07 8.06 Æ 1.82 0.337 ‡ C7 Intra-rater reliability (95% CI) 0.995 0.999 <0.001* 0.995 SEM/MDC (cm) (0.989–0.998) (0.988–0.998) (0.997–0.999) <0.001* Inter-rater reliability (95% CI) 0.10/0.29 0.06/0.16 0.839 ‡ 0.08/0.21 <0.001* SEM/MDC (cm) 8.32 Æ 1.40 0.811 (0.576–0.926) 8.64 Æ 1.74 Distance † (cm) 0.994 0.998 <0.001* 0.65/1.80 Intra-rater reliability (95% CI) (0.986–0.998) 8.36 Æ 1.83 (0.995–0.999) 0.11/0.30 0.08/0.22 SEM/MDC (cm) 0.996 Inter-rater reliability (95% CI) (0.992–0.999) SEM/MDC (cm) 0.12/0.32 0.933 (0.850–0.974) 0.42/1.18 Notes: CI: Con¯dence interval; C7: the seventh cervical vertebra; SEM: standard error of measurement; and MDC: minimal detectable change. †Data are presented in the form of mean Æ standard deviation. ‡The p-value from one- way analysis of variance (ANOVA), comparing means among the three raters. *The p-value from the ICCs.

Seventh cervical vertebra for thoracic kyphosis measurements 49 Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com Discussion from C7 or occiput to the wall. Nevertheless, the by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. variation of KMD outcomes when using occiput as This study sought to identify a more valid and a landmark could confound the validity and pre- reliable landmark, C7 or occiput, in determining dictive ability for thoracic angles. Therefore, the KMD as a simple standard clinical screening mea- validity and overall variance predicted of the KMD sure of individuals at risk of structural thoracic measured at C7 by the regression model were found hyperkyphosis. The ¯ndings suggest that, after to be better than those of the occiput (Tables 2 proper training, the KMD measured at C7 using and 3). Therefore, after proper training, the KMD two rulers is more valid and reliable than that at C7 using rulers could be used as a standard derived from the occiput when veri¯ed using the simple and practical measure for screening and data from IG-KypDisT and Cobb angles (Fig. 2, data transferring among the various clinical, com- Tables 2 and 3). munity, home-based, and research settings. The wider range of di®erences in KMD out- Nonetheless, the levels of correlation of KMD comes from rulers and IG-KypDisT when mea- and Cobb angles in this study were lower than sured at the occiput as compared to those at C7 those reported previously (r ¼ 0.902, p <0.001 (Fig. 2) re°ect the characteristics of both land- compared to Flexicurve).15 These di®erences may marks. The KMD measurement at the occiput be associated with the (1) age of the participants, could increase the chance of errors due to the (2) standard method used to verify the outcomes, participants moving their heads,13–15 hair thick- and (3) starting position of the measurements. The ness, and varying shapes of the skulls.1,11 Martinez- concurrent validity of KMD was previously asses- Abadias24 explored the phenotypic and genetic sed in participants who were aged 60 years and patterns of human skulls, and found substantial over using Flexicurve, an indirect standard ky- genetic variations in human skull sizes and shapes. phosis measure.10,15 Older adults have rather ¯xed In contrast, the bony prominence of C7 is a stable spinal columns that result in structural hyperky- point over the superior limit of the thoracic phosis and reliable outcomes over trials even when spine.11,15,16 Therefore, the KMD outcomes when changing the postures, or lifting the arms upward measured at C7 using rulers were similar to those for Cobb's measurements. However, indirect mea- from IG-KypDisT (Fig. 2). surements using Flexicurve are susceptible to errors due to palpation skill for a proper bony The slight di®erences of KMD data between the landmark (i.e., spinous processes of T12) and the occiput and C7 re°ect normal body structure when changes in the shape of Flexicurve once it is re- considered in a sagittal plane. The bony promi- moved from the spine.15 On the contrary, this nence of C7 was closer to the body midline, and study recruited participants aged 10 years and over thus farther from the wall than the occiput when to capture those at a risk of thoracic hyperkyphosis standing against the wall. Thus, the KMD data due particularly to the change in the current life from C7 was longer than that at the occiput by styles, age-related postural changes and system approximately 0.8 cm (p ¼ 0.084), which was decline, including those aged 11–14 years (30%), greater than the SEM or errors due to the mea- 15–18 years (22%), 20–50 years (38%), and > 60 surements, from all raters (0.06–0.42, Table 3). years (22–44%).1–3 Being younger individuals with The ¯ndings implied the true di®erence when an average age of 50.5Æ 26 years (ranging from 11 measuring the KMD using di®erent landmarks, i.e., years to 88 years, Table 1), some participants the occiput or C7, and suggested using the same might have a relatively °exible spine. Although the landmark for data comparisons over trials for the researchers attempted to recruit participants who KMD. were unable to consciously straighten the spinal column, i.e., having structural hyperkyphosis,3 The signi¯cant correlation between KMD and some teenager participants might be unable to straighten their spinal columns due to muscle Cobb angles may re°ect the e®ects of spinal kine- weakness. However, the radiologic assessments with lifting the arms forward through 90 may matic linkage. The change of spinal curvature in a alter their thoracic kyphosis and a®ect the levels of correlation found in this study. This assumption is particular area in°uences the change of other spi- nal areas to compensate and improve balance of the spinal column.14 Therefore, increased thoracic curvature could increase cervical lordosis that subsequently enhanced the perpendicular distance

50 A. Wiyanad et al. Con°ict of Interest associated with a previous study that recruited The authors have no con°icts of interest relevant participants aged 18 years and above (average age to this paper. 54.5 Æ 17.5 years) and also found similar correla- tion as that reported in this study (r ¼ 0.72).25 Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com Limitations Funding/Support by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Our ¯ndings provide strong evidence to support This work was supported by the Grant for Post- the use of C7 as an appropriate landmark for KMD Doctoral Training Program from Khon Kaen using rulers. However, there are some limitations of University (grant no. PD2665-04). the ¯ndings. First, this study recruited totally 43 participants covering all those with a risk of Author Contributions hyperkyphosis (aged 10 years and over). However, with the COVID-19 crisis, the researchers were All authors were responsible for the research con- unable to recruit the participants according to the ceptualization, study design, and ¯nal approval of proportion reported in previous studies, aged 11– the manuscript. Arpassanan Wiyanad, Patchar- 14 years (30%), 15–18 years (22%), 20–50 years awan Suwannarat and Pakwipa Chokphukiao were (38%), and > 60 years (22–44%).1–3 Moreover, also involved in data collection. Arpassanan time of the day for the measurements might a®ect Wiyanad also took part in statistical analysis, data the ¯ndings of the study. Therefore, all partici- interpretation, and drafting of the manuscript. pants were measured for KMD using rulers and Pipatana Amatachaya and Thanat Sooknuan were IG-KypDisT altogether in a single visit. However, additionally responsible for IG-KypDisT develop- the radiologic assessments were scheduled accord- ments and calibration. Sugalya Amatachaya was ing to hospital appointments within the following also responsible for research management, funding seven days. The variation in the assessment sche- application, data interpretation, and ¯nalization of dules could in°uence the correlation found between the manuscript. KMD and the Cobb angles. Furthermore, the overall variance predicted by the regression model Acknowledgments of KMD for the Cobb angles might be in°uenced by other factors, such as the apex of hyperkyphosis, The researchers sincerely thank Mr. Patiphan bone morphology, and system functions. The Tochang and Ms. Kanjana Kramkrathok for their assessments of these variables need further invasive kind support in the equipment used in the study. methods and other mobility tests that would in- crease the complexity and are not in line with the References original aim of a simple measure of KMD. There- fore, a further study may recruit the participants 1. Antonelli-Incalzi R, Pedone C, Cesari M, Iorio AD, according to proportion reported previously to Bandinelli S, Ferrucci L. Relationship between the promote generalizability of the ¯ndings. occiput-wall distance and physical performance in the elderly: A cross sectional study. Aging Clin Exp Conclusions Res 2007;19(3):207–12. The KMD is a simple and practical measure for 2. Eslami S, Hemati J. Prevalence lordosis and dorsal thoracic hyperkyphosis. However, the existing ev- kyphosis deformity among girls 23-11 years and its idence applied two di®erent landmarks and the relationship to selected physical factors. Int J Sport measurement using rulers was susceptible to errors Stud 2013;3(9):924–9. due to their misalignment. Our ¯ndings support the reliability and validity of KMD assessments at 3. Kado DM, Prenovost K, Crandall C. Narrative C7 using rulers as a simple and practical measure review: Hyperkyphosis in older persons. Ann Intern of thoracic hyperkyphosis that can be used in Med. 2007;147(5):330–8, doi: 10.7326/0003-4819- various settings, such as hospitals, clinics, schools, 147-5-200709040-00008. o±ces, or research laboratories. 4. Griegel-Morris P, Larson K, Mueller-Klaus K, Oatis CA, et al. Incidence of common postural abnormalities in the cervical, shoulder, and tho- racic regions and their association with pain in two

Seventh cervical vertebra for thoracic kyphosis measurements 51 Hong Kong Physiother. J. 2023.43:43-51. Downloaded from worldscientific.com age groups of healthy subjects. Phys Ther Validity and reliability of a thoracic kyphotic as- by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. 1992;72:425–31. sessment tool measuring distance of the seventh 5. Ailon T, Sha®rey CI, Lenke LG, Harrop JS, Smith cervical vertebra from the wall. Hong Kong Phy- JS, et al. Progressive spinal kyphosis in the aging siother J 2016;35:30–6. population. Neurosurgery 2015;77(4):S164–72. 16. Mac-Thiong JM, Roussouly P, Berthonnaud E, 6. Kado DM. The rehabilitation of hyperkyphotic Guigui P. Sagittal parameters of global spinal posture in the elderly. Eur J Phys Rehabil Med balance. Spine 2010;35(22):E1194–8. 2009;45(4):583–93. 17. Miladi L. Round and angular kyphosis in paediatric 7. Lewis JS, Valentine RE. Clinical measurement of patients. Orthop Traumatol Surg Res 2013;99 the thoracic kyphosis: A study of the intra-rater (1 Suppl):S140–9, doi: 10.1016/j.otsr.2012.12.004. reliability in subjects with and without shoulder 18. Teixeira FA, Carvalho GA. Reliability and validity pain. BMC Musculoskelet Disord 2010;11:39. of thoracic kyphosis measurements using °exicurve 8. Sedrez JA, Candotti CT, Rosa MIZ, Medeiros method. Rev Bras Fisioter, São Carlos 2007;11 FS, Marques MT, Loss JF. Test-retest, inter- and (3):173–7. intra-rater reliability of the Flexicurve for evalua- 19. Suwannarat P, Wattanapan P, Wiyanad A, tion of the spine in children. Braz J Phys Ther Chokphukiao P, Wilaichit S, Amatachaya S. Reli- 2016;20(2):142–7. ability of novice physiotherapists for measuring 9. van der Jagt-Willems HC, de Groot MH, van Cobb angle using a digital method. Hong Kong Campen JP, et al. Associations between vertebral Physiother J 2017;37:34–8. fractures, increased thoracic kyphosis, a °exed 20. Wu W, Liang J, Du Y, Tan X, Xiang X, Wang W, posture and falls in older adults: A prospective et al. Reliability and reproducibility analysis of the cohort study. BMC Geriatr 2015;15:34. Cobb angle and assessing sagittal plane by com- 10. Wongsa S, Amatachaya S. Kyphosis assessments: puter-assisted and manual measurement tools. Review article. J Med Technol Phys Ther 2014;26(2): BMC Musculoskelet Disord 2014;15:33. 106–16. 21. Mukaka MM. Statistics corner: A guide to appro- 11. Suwannarat P, Amatachaya P, Sooknuan T, priate use of correlation coe±cient in medical re- Tochaeng P, Kramkrathok K, Thaweewannakij T, search. Malawi Med J 2012;24(3):69–71. et al. Hyperkyphotic measure using distance from the 22. Portney LG, Watkins MP. Foundations of Clinical wall to indicate the risk for thoracic hyperkyphosis Research: Application to Practice. 3rd ed. New and vertebral fracture. Arch Osteoporos 2018;13:25. Jersey: Julie Levin Alexander, 2008. 12. Wiyanad A, Chokphukiao P, Suwannarat P, Tha- 23. Hamilton DF, Ghert M, Simpson AH. Interpreting weewannakij T, Wattanapan P, Gaogasigam C, et regression models in clinical outcome studies. Bone al. Is the occiput-wall distance valid and reliable to Joint Res 2015;4:152–3, doi: 10.1302/2046- determine the presence of thoracic hyperkyphosis? 3758.49.2000571. Musculoskelet Sci Pract 2018;38:63–8. 24. Martinez-Abadias N. Evolution patterns of the 13. Allanson JE, Cunni® C, Hoyme HE, McGaughran human skull: A quantitative genetic analysis of J, Muenke M, Neri G. Elements of morphology: craniofacial phenotypic variation. PhD disserta- Standard terminology for the head and face. Am J tion, University of Barcelona, Barcelona, 2005. Med Genet 2009;149A(1):6–28. 25. Siminoski K, Warshawski RS, Jen H, Lee KC. The 14. Ray B, Kalthur S, Kumar B, Bhat MRK, D' souza accuracy of clinical kyphosis examination for de- AS, Gulati HS, et al. Morphological variations in tection of thoracic vertebral fractures: comparison the basioccipital region of the South Indian skull. of direct and indirect kyphosis measures. J Mus- Nep J Med Sci 2015;3(2):124–8. culoskelet Neuronal Interact 2011;11(3):249–56. 15. Amatachaya P, Wongsa S, Sooknuan T, Thawee- wannakij T, Laophosri M, Manimanakorn N, et al.

Research Paper Hong Kong Physiotherapy Journal Vol. 43, No. 1 (2023) 53–60 DOI: 10.1142/S1013702523500075 Hong Kong Physiother. J. 2023.43:53-60. Downloaded from worldscientific.com Hong Kong Physiotherapy Journal by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. https://www.worldscientific.com/worldscinet/hkpj Muscle contraction exercise for low back pain Azzam Alarab1,*, Ratib Abu Shameh1 and Muntaser S. Ahmad2 1Department of physiotherapy, Faculty of Allied Medical Sciences Palestine Ahliya University, Dheisha, Bethlehem, Palestine 2Department of Medical Imaging, Faculty of Allied Medical Sciences Palestine Ahliya University, Dheisha, Bethlehem, Palestine *[email protected] Received 26 March 2019; Accepted 14 February 2023; Published 31 March 2023 Background: Low-back pain (LBP) continues to be one of the main problems for which su®erers seek treatment in primary care. It can be treated with di®erent physiotherapy mechanisms. Objective: The purpose of the study is to compare the e®ect of isotonic and isometric exercise on the reported pain of patients with low back pain. Methods: Thirty participants, 16 males and 14 females aged between 22 and 50 years su®ering from nonspeci¯c low back pain were included. The sample was divided randomly into two groups, group A isometric exercises and group B isotonic exercises, both groups received conservative therapy of TENS and infrared (IR) therapy. The following outcome measures were used: Visual analogue scale, modi¯ed Oswestry disability index (MODI) and Endurance Test Measurement were administered pre-treatment and at the end of four weeks of treatment. Results: Both groups were comparable in terms of demographic data, except for weight. Inter group analysis was done using the Mann–Whitney test. When comparing pre- and post-treatments using VAS scores, there were no signi¯cant di®erences between group A and group B (pre-test: P ¼ 0:285; Post-test: P ¼ 0:838). Mann–Whitney test was used to calculate the P -value test between pre-treatment and post- treatment for MODI and there was no signi¯cant di®erence between group A and group B, where the pre-test P -value was 0.061, and post-treatment was 0.077. Comparing between groups, pre- and post-abdominal endurance scores were done using the Mann– Whitney test. The pre-treatment scores revealed P value of 0.345, and the post-treatment scores revealed P value of 0.305. Therefore, there is no statistically signi¯cant di®erence between group A and group B in endurance scores. Conclusion: There was no di®erence between the use isotonic and isometric exercises on LBP patients. *Corresponding author. Copyright@2023, Hong Kong Physiotherapy Association. This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 (CC BY- NC-ND) License which permits use, distribution and reproduction, provided that the original work is properly cited, the use is non-commercial and no modi¯cations or adaptations are made. 53

54 A. Alarab, R. A. Shameh & M. S. Ahmad Keywords: Isometric exercises; isotonic exercises; core muscle; low-back pain. Hong Kong Physiother. J. 2023.43:53-60. Downloaded from worldscientific.com Introduction status, compared to the patients who received no by Horizon College Physiotherapy on 07/29/23. Re-use and distribution is strictly not permitted, except for Open Access articles. treatment or other conservative treatments.11 Low-back pain (LBP) continues to be one of the main problems for which su®erers seek treatment Therefore, the purpose of this study is to com- in primary care,1 and presents a large challenge to pare the e®ect between the isometric and isotonic the healthcare system despite improving scienti¯c exercises on pain, functional disability, and endurance technology, medical insight, and suggested man- on students and employees of PAU in Bethlehem, agement strategies.2 Therefore, according to the su®ering with LBP. Palestinian Ministry of Health, the prevalence of LBP in the Palestinian persons reached to Methods 59.4%. It was a comparative study conducted in the phys- Non-speci¯c LBP is de¯ned as LBP not attrib- iotherapy department of Allied Medical Science utable or recognizable, known speci¯c pathology Faculty. Ethical permission was obtained from (e.g., infection, tumour, osteoporosis, lumbar spine the Faculty Ethical Committee, Allied Medical fracture, structural deformity, in°ammatory dis- Science Faculty, Palestine Ahliya University, order, radicular syndrome, or cauda equine syn- Palestine. The sample study was calculated by the drome). Non-speci¯c low back is often associated following equation: n ¼ ðZ=2 þ ZÞ2 Ã 2 Ã 2=d2, with a history of lifting or twisting while holding where Z=2 is the critical value of the normal dis- heavy object, operating a machine that vibrates, tribution at =2, Z is the critical value of the prolonged sitting, fall, coughing, sneezing, and normal distribution at , 2 is the population variance, straining.3,4 Non-speci¯c LBP develops into and d is the di®erence you would like to detect. chronic low back pain (CLBP), which is de¯ned as mild to severe pain in the lower back that has A total of 36 patients (18 males and 18 female) lasted for more than three months, morning sti®- were selected in orthopaedic clinic according to ness, sleep interruptions due to pain tiredness and/ inclusion and exclusion criteria. To address the or irritability, depression, and inability to sit or issue of LBP, a special announcement was made by stand for long periods of time.1,5 the Palestine University's physical therapy de- partment. It invited patients to visit, its specialist In the treatment of CLBP, the focus has been on for a free examination. Written informed consent analgesia and limited activity. However, this ap- was obtained and the study procedures were proach is not ideal and can lead to harmful e®ects.6 explained. A detailed musculoskeletal evaluation Currently, clinical recommendations are focused on was done to screen the patients. Participants were the use of a biopsychosocial framework to guide the equally divided into two groups, enrolled in the treatment of patients. This method supports the study and randomly assigned to one of the two use of education and daily activities, but it does not groups. Baseline treatment was given to both the recommend routine use of complementary tests.7,8 groups which consisted of transcutaneous electrical nerve stimulation (TENS) and IR. Group A was Majiwala B and his colleagues reported the ef- given isometric exercise (N ¼ 18, M ¼ 9, F ¼ 9) fect of isometric and isotonic exercise training on and group B was given isotonic exercise (N ¼ 18, core muscle in patients with non-speci¯c LBP. As a M ¼ 9, F ¼ 9). All patients were divided ran- result, both isometric and isotonic exercises are domly. Three patients in group A and three equally e®ective in reducing pain, increase endur- patients in group B dropped out for personal rea- ance, and improve functional disability in patients sons. Sessions were performed for 1 h, three times with non-speci¯c LBP.9 weekly, for four weeks. Van Tulder et al. reported that exercise for the Inclusion criteria were as follows: (1) Both male treatment of LBP was e®ective in accelerating and female, (2) Age group 22–50 years, (3) Patients improvements in daily life activities and return to willing to participate in exercise program, and (4) work.10 In a meta-analysis, the patients with LBP History of LBP for three months. treated with exercise therapy showed a signi¯cant improvement in terms of pain and functional