Archives of Physiotherapy

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 8 of 19 patient-reported questionnaire for measuring the impact to close their eyes and feel themselves moving their right of MS on individual lives, with 20 items associated with a or left leg to a board placed in front of their feet. physical subscale and 9 items with a psychological sub- scale [49]. Responses use a 5-point Likert scale, ranging Repeatability and reliability of gait analysis instruments from (1) “not at all” to (5) “extremely”. The MSIS-29 Participants were asked to synchronise their steps with showed the strongest psychometric properties compared instrumental music in regular metre at 110 bpm. The to other QoL scales [50, 51]. The items ask about the choice of the music beat frequency of 110 bpm was impact of MS on day-to-day living over the past two based on the gait literature in people with MS [59–61]. weeks, higher scores indicating a greater impact of MS on People with mild to moderate MS showed a mean (SD) daily function. According to a Rasch analysis, the two cadence of 109.1 ± 23.3 steps/min [59]; patients with MS scales are different and should not be combined to a total and higher disability levels walked mean 98.97 ± 19.95 score [52]; in the current study, the analysis was steps/min [60]. The cadence in pwMS with an EDSS of performed accordingly. up to 5.7 was between 100.0 ± 23.3 steps/min and 112.1 ± 11.3 steps/min [61]. The music was played with an Motor imagery ability Apple i-phone and X-Mi X Mini II Capsule- Evidence has recommended to comprehensively assess the Loudspeakers in a calm hallway free from obstacles. Gait MI ability using at least two different approaches because synchronisation analysis with musical pulse requires some people may have problems generating vivid images knowing step lengths and step times, therefore, 2- and intense sensations and/or subjectively assessing their dimensional (2D) video recording in the frontal plane MI ability, and others with the duration of their mental was performed [62, 63]. The videos were taken with a imagery, in relation to real movements [53, 54]. In other Panasonic HC-WX979 4 K camcorder with a frame rate words, both questionnaire and mental chronometry (that of 50 fields per second, which was mounted on a 1.3 m is, temporal congruence) tests are required to assess the high Haehnel 9,994,180 Triad 40 Lite Tripod placed 5 m MI capability and both were used in this study. from the participant [62, 64]. Participants walked be- tween two marked lines, 1 m apart, on a 30 m hallway, The Kinaesthetic and Visual Imagery Questionnaire so that they could accelerate and decelerate their speed, (KVIQ-10) is the short version of the KVIQ-20, which is and adjust their gait to the music beat. Thus, only the a MI ability questionnaire developed for people with central 4.5 m were video-recorded while the participant physical disabilities [55, 56]. Thus, the questionnaire is walked 4 to 6 times back and forth, depending on their researcher-administered, and the movements are easily step length (Fig. 2). Participants were allowed to wear performed while participants are seated [55, 56]. The shoes or walk barefoot and the type of footwear or its KVIQ-10 consists of a visual (V) and kinaesthetic (K) absence was to be kept consistent during all trials. This subscale and assesses the clarity of the image (V) and procedure enabled the acquisition of 25–35 steps per the intensity of the sensations (K) on a five-point ordinal participant. These video footages were used to measure scale [55]. The five visual categories range from (1) “no the average step length and step time per participant image” to (5) “image as clear as seeing”; the five (test measures). The video-footage was analysed using kinaesthetic categories range from (1) “no sensation” to CCC Utilius Fairplay 5 Software. This software allowed (5) “as intense as executing the action”. The maximum calibration of the field of view, in order to reduce paral- KVIQ-10 total scale is 50 points, with 25 points in each lax error [64]. Step length was defined as the distance subscale, and higher scores representing higher levels of between the initial contact of one foot with the floor, MI ability. The KVIQ-10 has been attributed excellent followed by the opposite foot, and was measured in me- validity and reliability in people with stroke [55] and MS tres [64]. Step time was defined as the period between [57]. For this study, the validated German version of the the initial contact of one foot and the other foot, and KVIQ-10 was used, the KVIQ-G-10 [56]. The minimum was measured in seconds [64]. In other words, the expected level of the MI ability at baseline and post- crucial moment was the heel contact with the ground, intervention, as assessed by the KVIQ-G-10, was median or, in the case of severe spasticity, the toe contact. Excel- 3 out of 5 points on both subscales [58]. lent image exposure was achieved, due to the camcorder 4 K technology and a well-lit hallway, therefore, no The Time-Dependent MI (TDMI) screening test is a reflective markers were used. mental chronometry test which measures the number of imagined stepping movements in seated participants Prior to the study, the accuracy of the video analysis over 3 time periods (15, 25 and 45 s) [53]. Good to ex- software calibration grid was evaluated, by using a 7 × 1 cellent reliability of the TDMI was shown in people with metres large grid mat with 0.2 × 0.2 m large grid fields stroke with intra-class correlation coefficients (ICCs) be- which accurately covered the marked walkway (Fig. 3). tween 0.87 and 0.93 [12].To encourage kinaesthetic MI Accuracy was confirmed, as the grid fields of the grid mat from a first-person perspective, participants were asked

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 9 of 19 Fig. 2 Representative image of the 2D video-based gait assessment and the video calibration grid were congruent. Moreover, reported for reliability data, and raw count (frequency, the measurement technique described above was esta- percentage) for nominal data (recruitment rates, reten- blished in 13 healthy team colleagues. After that, this feasi- tion rate, missing data, falls and adverse events). Median bility study assessed the repeatability and reliability of the MI vividness scores were calculated by dividing the me- gait analysis system. In addition to the test measures dian KVIQ-G-10 scores by the number of items, that is, described above, the step lengths and step times were mea- 5 for the visual/kinaesthetic subscales, and 10 for the sured 4 times on the video footage by the same rater, using total score. Spearman’s correlation coefficients (median Utilius Fairplay 5 software. From each of these measures with interquartile ranges; all baseline pairwise corre- the mean step length and step time was calculated, so that lations) were calculated for the numbers of imagined 4 retest measures were available. stepping movements over 15, 25 and 45 s (TDMI). Dif- ferences between the three groups at the baseline were Statistical analysis examined using Fisher’s Exact test (nominal data: gender; All statistics were performed using IBM SPSS software, walking aid use and fatigue yes/no) and Kruskal Wallis release 24.0 (IBM Corporation, Armonk, NY, USA) and test (continuous data: age; walking speed and walking dis- GraphPad Prism 6 (GraphPad, San Diego, California, tance; ordinal data: fatigue, QoL, MI ability). Corrected USA). Intention-to-treat analysis was performed for all Fisher’s exact contingency table analysis was calculated cases, which were analysed by the originally assigned using an online calculator (http://www.physics.csbsju.edu/ groups. Descriptive statistics were reported for all out- stats/exact_NROW_NCOLUMN_form.html). comes. Medians (interquartile ranges) were reported for continuous and ordinal data: fatigue, QoL, KVIQ-G-10 The recruitment rate (%) was determined by dividing MI ability (all: 5 categories), TDMI MI ability, walking the number of participants who consented by the num- speed and walking distance; medians (range) were ber of patients eligible, multiplied by 100. The retention reported for age, compliance (7 categories: 0–6 times rate was estimated: (N who completed the study/N total per week), and disability (EDSS). Means (95% CI) were sample)*100, where N is the number of participants. The adherence rate was reported as the percentage of the Fig. 3 Representative image of the accuracy evaluation of the video analysis software calibration

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 10 of 19 scheduled (cued) MI practice (6×/week), actually per- month, which is a recruitment rate of 8.6% (95% CI formed by the participants over the 4 week study period 5.2, 13.8%). This recruitment rate exceeded the [65]. The eligibility, recruitment and adherence rates target recruitment rate of 5.7%. were calculated with their 95% CI according to the b) All 15 participants completed the study and there Wilson ‘score’ method cited by Newcombe [66]; when were no missing data, both corresponding to a 100% the proportion was close to 0 or 1, a Poisson approxima- retention rate (95% CI 76.4, 100%). This retention tion as described by Brown was used [67]. rate surpassed the target retention rate of 80%. c) With reference to a maximum practice frequency of The ICC was estimated as a measure of intra-rater reli- 6 times per week, participants reported to have ability. The ICCs were calculated with their 95% CI using practised median 5 (range 4, 6) times per week. This a two-way mixed model based on absolute agreement with adherence rate of 83% (95% CI 0.42, 0.99) was single measure analysis. Point estimates of the ICC were greater than the target adherence rate of 67%. rated as excellent (0.9–1), good (0.73–0.9), moderate (0.4– 0.74) and poor (0–0.39) [68]. The standard error of meas- Safety, adverse events and acceptability urement (SEM) was calculated to assess the random error, No safety-related events such as falls occurred in this or variability, of the step length and step time measure- study. No adverse events related to this home-based ments. It is the difference between an observed score and study were reported. Phone calls were considered its ‘true’ score [69]. The SEM was calculated using the for- supportive by participants. One participant in the MI mula: SEM = SD*√(1-ICC), where SD is the sample stand- group reported minor concentration problems during ard deviation [70, 71]. The SEM was multiplied by 1.96 in the MI which resolved with practice. Overall, partici- order to compute the error that is expected in 95% of the pants in the MI group appeared to be satisfied with measurements [69]. This value was then multiplied by the interventions. All participants in the MVMI and √(2) to compute the MDC95, i.e. the smallest amount of MMI groups reported that they liked the music styles, change between two repeated measurements, based on a melodies, and tempo changes of the music pieces. 95% confidence interval, which is beyond the measure- Moreover, 9 out of 15 participants reported as an ad- ment error and hence is a genuine change [72]. junctive comment that they found safe and conveni- ent practicing the intervention at home, in a sitting Results position. To summarise, the interventions were found to be acceptable or even pleasurable. Baseline characteristics Participants´ baseline data are shown in Table 2 and Secondary outcomes Additional file 3. The female-to-male ratio was 6.5:1, as Walking speed and walking distance only two males participated in this study. Participants in Change in walking performance between baseline and the MI group were younger and walked faster and a lon- post intervention for participants in all groups is ger distance. In the MI group, one participant used two presented in Table 3 and Fig. 4. Improvements in walking sticks during all walking tests. Participants in the walking speed and walking distance were observed in MVMI group were more disabled, showed higher fatigue all groups with the greatest walking distance improve- values, lower QoL and lower kinaesthetic MI ability. ments in the cued MI groups. Only one participant These differences between the groups did not reach statis- in the MVMI group reached a clinically significant tically significance. Apart from slightly lower values in the improvement in walking speed of ≥20%. Three partic- MVMI group, with a 25th percentile of 2.7, participants ipants in the MVMI group and two participants in seemed to have been able to practice MI, as indicated by both MMI and MI groups showed a clinically mean- median KVIQ-G-10 vividness scores of 4.0 (interquartile ingful improvement in walking distance of ≥20% from range 3.2, 4.6) out of 5.0. Additionally, there was a moder- baseline to post intervention. ate to strong positive mean Spearman’s correlation of r = 0.78 (interquartile range 0.77, 0.84) for the numbers of imagined stepping movements during 15, 25 and 45 s for both lower limbs. Primary outcomes Fatigue Feasibility As can be seen in Table 3 and Fig. 5 and was consistent with our previous results, improvement in fatigue was a) 174 out of 339 people with MS were eligible for the observed in all groups. At baseline, 6 out of 15 partici- study, corresponding to an eligibility rate of 51.3% pants in all groups had fatigue, with ≥38 points on the (95% CI 46.0, 56.6%) (see Fig. 1). Of these 174 MFIS total score, which was reduced to 2 participants in participants, 15 consented to the study within one total from the MVMI group.

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 11 of 19 Table 2 Baseline characteristics Parameter Group 1 Group 2 Group 3 P-value Music and verbally cued motor Music cued motor imagery (MMI) Motor imagery (MI) imagery (MVMI) Gendera (F:M) N =5 N =5 N =5 0.999d N = 4:1 N = 5:0 N = 4:1 Age (years)b 52.0 (41.0, 69.0) 54.0 (34.0, 72.0) 37.0 (27.0, 74) 0.500e EDSSb 4.5 (2.0–4.5) 2.5 (2.5–4.5) 2.5 (1.5–4.5) 0.387e Walking aid use during testinga No/unilateral/bilateral aid N = 5/0/0 N = 5/0/0 N = 4/0/1 0.999d T25FWc (seconds) 5.6 (5.2, 6.2) 7.0 (5.3, 7.0) 4.9 (4.6, 5.4) 0.264e 6MWTc (metres) 367.5 (348.0, 435.2) 380.6 (337.2, 392.0) 460.5 (442.9, 476.8) 0.249e MFIS physical subc 21.0 (2.0, 25.0) 12.0 (10.5, 25.0) 16.0 (9.0, 21.0) 0.983e MFIS cognitive subc 17.0 (11.5, 22.5) 3.0 (0.0, 16.5) 11.0 (11.0, 14.0) 0.098e MFIS psychosoc subc 2.0 (1.0, 4.0) 2.0 (0.0, 5.0) 3.0 (1.5, 5.0) 0.638e MFIS total scorec 40.0 (17.5, 51.0) 28.0 (13.0, 39.0) 30.0 (23.5, 38.0) 0.564e MFIS total score ≥ 38 N = 3/2 N = 2/3 N = 1/4 0.800d MSIS-29 physical subc 45.0 (17.5, 49.4) 20.0 (17.5, 39.4) 16.2 (10.6, 35.6) 0.296e MSIS-29 psychological subc 27.8 (18.0, 33.3) 11.1 (17.5, 39.4) 11.1 (8.3, 25.0) 0.137e KVIQ-G-10 visual subc 21.0 (17.5, 23.5) 20.0 (17.5, 24.5) 21.0 (17.5, 24.0) 0.992e Median visual scorec 4.2 (3.5, 4.7) 4.0 (3.5, 4.9) 4.2 (3.5, 4.8) 0.992e KVIQ-G-10 kinaesthetic subc 15.0 (13.5, 18.5) 24.0 (15.5, 24.5) 23.0 (18.0, 23.5) 0.189e Median kinaesthetic scorec 3.0 (2.7, 3.7) 4.8 (3.1, 4.9) 4.6 (3.6, 4.7) 0.189e KVIQ-G-10 totalc 40.0 (31.0, 40.0) 43.0 (33.5, 49.0) 44.0 (36.0, 47.0) 0.328e Median total scorec 4.0 (3.1, 4.0) 4.3 (3.3, 4.9) 4.4 (3.6, 4.7) 0.328e TDMI 25 s rightc 15.0 (12.0, 22.0) 14.0 (12.0, 16.0) 16.0 (11.5, 23.0) 0.688e TDMI 15 s leftc 10.0 (9.0, 11.5) 9.0 (7.5, 11.0) 13.0 (6.5, 14.5) 0.572e TDMI 45 s rightc 27.0 (19.5, 36.0) 23.0 (19.0, 31.0) 33.0 (20.0, 38.5) 0.664e TDMI 15 s left2c 10.0 (8.5, 12.0) 8.0 (8.0, 9.5) 12.0 (7.5, 13.0) 0.231e TDMI 25 s leftc 16.0 (14.5, 19.5) 16.0 (14.5, 18.0) 20.0 (12.0, 21.0) 0.883e TDMI Spearman’s rc 0.72 (0.47, 0.82) 0.59 (0.46, 0.72) 0.87 (0.80, 0.97) Abbreviations: N: Counted number of participants; F:M: Females: Males; EDSS: Expanded Disability Status Scale; MFIS: Modified Fatigue Impact Scale; MFIS physical sub: MFIS physical subscale; MFIS cognitive sub: MFIS cognitive subscale; MFIS psychosoc sub: MFIS psychosocial subscale; MSIS-29 physical sub: Multiple Sclerosis Impact Scale-29 physical subscale; MSIS-29 psychological sub: Multiple Sclerosis Impact Scale-29 psychological subscale; KVIQ-G-10: Kinaesthetic and visual imagery questionnaire, German short version; KVIQ-G-10 visual sub: KVIQ-G-10 visual subscale; KVIQ-G-10 kinaesthetic sub: KVIQ-G-10 kinaesthetic subscale; KVIQ-G-10 total: KVIQ-G-10 total score; r: Spearman’s correlation coefficient (all pairwise correlations; all significant ≤0.05); TDMI: Time-dependent motor imagery screening test; T25FW: Timed 25-Foot Walk; 6MWT: 6-Minute Walk Test aCounted number of participants bMedian (range) cMedian (interquartile range) dAnalysed with corrected Fisher’s Exact contingency table analysis eAnalysed with Kruskal Wallis test Quality of life by median total KVIQ-G-10 scores, reached similar As shown in Table 3 and Fig. 6, QoL improved in all values in all groups, with 4.6 (interquartile range 4.0 to groups or remained at least stable. Throughout the 4.8) points in the MVMI group, when compared to 4.8 study, the lowest QoL values were observed in the (interquartile range 4.2 to 4.8) points in the MMI group MVMI group. and 4.2 (interquartile range 4.0 to 4.6) points in the MI group. Better MI abilities were also suggested by Motor imagery ability an overall median KVIQ-G-10 score of 4.0 (interquar- Post-intervention, as compared to baseline, participants tile range 4.1 to 4.8) and strongly correlated TDMI showed higher MI abilities. The MI ability, as assessed measures of median r = 0.89 (interquartile range 0.84

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 12 of 19 Table 3 Walking, fatigue and quality of life post-intervention data for each study group Parameter MVMI group MMI group MI group T25FW (seconds) −0.1 (−0.7, −0.1) −0.4 (−0.9, −0.3) −0.4 (− 0.4, − 0.3) Change from baselinea 5.1 (5.0, 5.5) 6.0 (4.9, 6.0) 5.0 (4.2, 5.1) Post-interventiona Participants with clinically significant N = 1/5 N = 0/5 N = 0/5 improvement (≥20%) 6MWT (metres) 85.5 (59.4, 97.1) 65.1 (39.5, 74.8) 33.6 (11.6, 77.7) Change from baselinea 453.2 (450.5, 470.0) 418.0 (412.0, 469.0) 476.5 (465.0, 569.3) Post-interventiona Participants with clinically significant N = 3/5 N = 2/5 N = 2/5 improvement (≥20%) MFIS physical subscale −3.0 (−7.5, 4.0) −2.0 (−13.5, −0.5) −3.0 (−6.5, −1.5) Change from baselinea 14.0 (9.0, 19.5) 9.0 (1.0, 21.5) 12.0 (7.5, 15.0) Post interventiona MFIS cognitive subscale −4.0 (−5.5, 0.5) −3 (−9.5, 0.0) −2.0 (−4.5, −0.5) Change from baselinea 15.0 (10.0, 18.0) 0.0 (0.0, 7.0) 10.0 (6.5, 13.0) Post interventiona MFIS psychosocial subscale 0.0 (−1.0, 0.0) 0.0 (−3.5, 0.0) −2.0 (−2.5, −0.5) Change from baselinea 2.0 (0.5, 3.5) 0.0 (0.0, 2.5) 1.0 (0.0, 3.5) Post interventiona MFIS total score −9.0 (−13.0, 4.5) −6.0 (−23.5, −3.0) −9.0 (− 10.5, −4.5) Change from baselinea 26.0 (22.0, 41.0) 12.0 (2.5, 28.0) 23.0 (15.0, 30.5) Post interventiona MFIS total score ≥ 38 (post-intervention) N = 2/5 N = 0/5 N = 0/5 MSIS-29 physical subscale −1.2 (− 11.9, 3.7) −7.5 (− 13.7, − 2.5) − 2.5 (−8.1, 6.2) Change from baselinea 27.5 (18.7, 48.1) 16.2 (8.1, 30.6) 21.2 (10.6, 30.0) Post interventiona MSIS-29 psychological subscale 0.0 (−11.1, 9.7) −5.5 (−6.9, 2.8) −5.5 (−9.7, 1.4) Change from baselinea 30.6 (11.1, 37.5) 5.6 (0.0, 23.6) 8.3 (4.2, 19.4) Post interventiona Abbreviations: T25FW Timed 25-Foot Walk, 6MWT 6-Minute Walk Test, MFIS Modified Fatigue Impact Scale, MFIS physical MFIS physical subscale, MFIS cognitive MFIS cognitive subscale, MFIS psychosocial MFIS psychosocial subscale, MFIS total MFIS total score, MSIS-29 = MS Impact Scale-29; N: Counted number of participants With walking speed (T25FW), improvement is indicated by a minus and worsening by a plus; with walking distance (6MWT), improvement is indicated by a plus and worsening by a minus aMedian (interquartile range) to 0.91). Strong correlations of the TDMI measures and for step time 0.038 s. This means that only when an were observed in all groups, as presented in Fig. 7. individual’s change from baseline to post-intervention or follow-up in step length and step time exceeds 3.7 cm or Reliability of gait analysis instruments 0.038 s one can be 95% confident that the individual was Results from the reliability analysis of the gait analysis actually changed. system are shown in Table 4. Excellent intra-rater reli- ability between measures was found, as evidenced by Discussion ICCs around 0.98 for step length and of 0.88 for step As far as we know, this study was the first study to time. The SEM for repeated measures was 0.013 m and explore the feasibility of the methods to be used for a for step time 0.014 s. In other words, the SEM was small full-scale RCT that will investigate the mechanisms and to moderate for step length measures and small for step effects of cued and non-cued MI on walking, fatigue and time measures. The MDC95 for step length was 0.037 m QoL in people with MS.

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 13 of 19 Fig. 4 Walking performance pre- and post-intervention. a Walking Speed. b Walking distance. Medians are shown by lines in the centre of the box-plots; the interquartile ranges are indicated by the boxes and ranges by the whiskers. Abbreviations: MI group = non-cued motor imagery group; MMI group = music-cued motor imagery group; MVMI group = music- and verbally cued motor imagery group; T25FW = Timed 25-Foot Walk; 6MWT = 6-Minute Walk Test. The grey boxes indicate the baseline data and the white boxes present the post-intervention data Feasibility We expect a similar adherence rate in the main trial, The observed recruitment, retention and adherence rates mainly because the interventions were acceptable, if not exceeded the pre-specified target rates so that a larger pleasurable to participants. Nevertheless, we recognise that study appears feasible. The phone support could have even with careful monitoring via phone-calls, participants added to adherence rates of median 5 (range 4–6) out of 6. could have stated their adherence rates slightly incorrectly. Fig. 5 Total fatigue pre- and post-intervention. The red line represents Safety, adverse events and acceptability the cut-off point for fatigue as defined at ≥38 points on the MFIS [48]. Participants spontaneously reported that they considered Medians are shown by lines in the centre of the box-plots; the interventions safe and convenient because they were the interquartile ranges are indicated by the boxes and ranges by the allowed to practise the MI at home and in a sitting po- whiskers. Abbreviations: MFIS = Modified Fatigue Impact Scale; MI group sition. As there were no falls or adverse events, it seems = non-cued motor imagery group; MMI group = music-cued motor safe to continue with the study procedures. 10 out of 10 imagery group; MVMI group = music- and verbally cued motor imagery participants in both music-cued MI groups reported that group. The grey boxes indicate the baseline data and the white boxes they liked the music styles and they regarded the inter- present the post-intervention data vention acceptable, with 8 out of 10 participants report- ing the music intervention as pleasurable. Based on the high acceptability of the music styles, the same music will be used in the main study. All participants in the non-cued MI group were satisfied with the intervention, and two participants viewed the intervention as pleasur- able as they appreciated the focus on their body aware- ness without any distraction. These findings are not unexpected as the melody and beat of music have been shown to impact on emotional function, the urge to move and the perception of fatigue [73]. In addition, we selected appropriate music pieces following participant feedback from our previous study, so as to use a higher amount of music pieces with a faster beat. At first glance, pure MI practise might be less stimulating than music-cued MI, however, the repeated attention to- wards the MI could positively impact on movement coordination. We narratively assessed the acceptability of the inter- ventions in this study as we consider high acceptability

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 14 of 19 Fig. 6 Quality of life pre- and post-intervention. a Physical quality of life. b Psychological quality of life. Medians are shown by lines in the centre of the box-plots; the interquartile ranges are indicated by the boxes and ranges by the whiskers. Abbreviations: MI group = non-cued motor imagery group; MMI group = music-cued motor imagery group; MSIS-29 = Multiple Sclerosis Impact Scale-29 MVMI group = music- and verbally cued motor imagery group an extremely important fact; as otherwise, no-one would like to practise nearly every day. In addition, the motivation and commitment during the home-based practice would be lower, which might be accompanied by tiredness, particularly when sitting with closed eyes. Walking The preliminary analysis from this study showed an im- provement in walking speed and walking distance after both types of cued MI and non-cued MI. Consistent with our previous results, the greatest improvement in walking distance was seen after music- and verbally- cued MI. We suggest that the regular rhythmic cueing, both by music and verbally, facilitated the brain’s temporal mechanisms similar to that found during cued gait training [22]. Rhythmic-cued MI could then have induced a synchronisation of imagined walking with the beat [21, 22]. Additionally, repeated rehearsal of imag- ined walking may have facilitated motor learning in all groups [54]. Moreover, as demonstrated in the sports domain and physiotherapy practice, motor tasks need to be specifically trained, in order to produce relevant im- provements [74]. The same may apply to MI training, given the comparable characteristics of executed and imagined movements and an overlapping brain area Fig. 7 MI ability pre- and post-intervention. a MI vividness. b Mental Table 4 Repeatability and reliability of gait measurement chronometry during MI. Abbreviations: KVIQ-G-10 = Kinaesthetic and instruments Visual Imagery Questionnaire-10, German short version: Medians are shown by lines in the centre of the box-plots; the interquartile Gait parameters ICC (95% CI) SEM MDC95 ranges are indicated by the boxes and ranges by the whiskers; MI Step length 0.978 (0.973, 0.982) 0.013 m 0.037 m group = non-cued motor imagery group; MMI group = music-cued motor imagery group; MVMI group = music- and verbally cued motor Step time 0.880 (0.855, 0.902) 0.014 s 0.038 s imagery group. The grey boxes indicate the baseline data and the white boxes present the post-intervention data Abbreviations: SEM standard error of measurement, ICC (95% CI) intraclass- correlation coefficient (95% confidence interval), MDC95 minimum detectable change with a 95% confidence interval

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 15 of 19 activation during real movement and MI [75]; hence in group, so that their scores were comparable to the MMI the current project, a specific training of imagined walk- and MI groups. These improvements could be associated ing was used. Finally, based on the walking improve- with motor learning; motor performance curves typically ments we speculate that cued and non-cued MI start with a slowly ascending slope, followed by a sharp rehearsal induced neural plasticity, which could be an increase in performance improvement over a period of underlying mechanism [76], however, this was outside practice [77]. In agreement with another study from the scope of this study. Heremans et al. (2012), who found that external cueing significantly improved the MI quality in people with MS, Fatigue we suggest that the music- and verbal cueing enhanced Post intervention, our preliminary results showed a mild their kinaesthetic MI capability [17]. Starting from improvement in fatigue in all groups when compared to higher baseline MI capabilities in the MMI group, a baseline. These findings are in line with results from smaller improvement was seen after the intervention. A Catalan et al. (2011) who found significantly improved larger study is required to explore underlying fatigue after non-cued MI practice in people with MS mechanisms. also at 6 month follow-up [20]. These results also agree with our previous study in people with mild to moderate On the TDMI screening test, the numbers of imagined MS where we observed an improvement in fatigue after stepping movements of the left and right lower extrem- rhythmic-cued MI [19]. ities were strongly related to the three different time pe- riods of 15, 25 and 45 s, as evidenced by Spearman’s Quality of life correlations of median 0.78 (interquartile range 0.77, To evaluate the impact of the two types of music-cued 0.84) to the three different time periods of 15, 25 and MI and non-cued MI on the quality of the participants’ 45 s. These results are suggestive of high temporal con- day-to-day lives, this study also obtained preliminary gruence of imagined stepping movements [12]. Our re- information on MS-related QoL. Improvements in phys- sults appear to contradict findings from previous studies ical QoL were observed in all groups; surprisingly and that used different mental chronometry tests, related to inconsistently with our previous findings, psychological the upper extremities [13–15]. However, these authors QoL improved after music-cued and non-cued MI, but linked impaired MI in this population to cognitive remained stable after music- and verbally-cued MI. dysfunction [13, 15] and depression [14]. Therefore, we These results seem to contradict the acceptability re- did not include persons with cognitive impairment and/ ports of participants, and probably could be associated or depression in our study. However, we did not use with the lowest baseline QoL in participants of the formal testing for cognitive function and mood, by MVMI group and their higher disability, as evidenced by which we could have obtained accurate results. Import- median EDSS scores of 4.5, when compared to 2.5 in the antly, we do not intend to withhold the treatment from other groups; however, these between-group differences people with MS and cognitive impairment or depression were not statistically significant. in the main study, but it does not seem useful to recruit patients into an MI intervention group when the litera- Motor imagery ability ture has already demonstrated their diminished ability to In accordance with the literature about the multifaceted perform MI. construct of MI, we used two different approaches to as- sess the ability to imagine movements in participants, an Reliability and repeatability MI questionnaire and a mental chronometry test [9, 12]. Evaluation of the 2D quantitative gait analysis system MI ability testing showed that all participants seemed to showed excellent reliability and repeatability. This is a have been able to practice MI as indicated by median substantial requirement to assess the degree of gait KVIQ-G-10 values of 4.0 (interquartile range 3.2, 4.6) synchronisation with music beat in the main study. Our out of 5.0. These scores imply adequate MI vividness results are in line with work from Harris-Hayes et al. [55, 58] and are consistent with a study from Heremans (2014) who examined the reliability and validity of a 2D et al. (2012) in 30 people with MS who observed very video based quantitative gait analysis system, and found similar KVIQ-10 scores [13]. However, in the absence of substantial to excellent reliability between three raters, statistically significant group differences, participants in in addition to excellent validity [62]. Similarly, Norris the MVMI group had lower baseline kinaesthetic MI and Olson (2011) analysed concurrent validity and scores than the other groups. The discrepancy in results reliability of 2D video-based motion analysis of the knee could be associated with their higher disability and fa- and hip joint movement [63]. Using an appropriate mo- tigue. Post-intervention, the greatest improvement in the tion analysis software and goniometer measures for joint kinaesthetic MI ability was observed in the MVMI angles, they found excellent concurrent validity. Their reliability measures showed excellent intra- and inter-

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 16 of 19 rater as well as intra-rater reliability. Recently, Paul et al. no statistically significant imbalance between groups. The (2016) assessed the validity and reliability of a 2D mo- female-to-male ratio indicated a probable selection bias, tion analysis instrument. They compared an established most likely associated with the small sample size. Thirteen and sophisticated 3D motion analysis system against a females and two males were included in the study, which 2D motion analysis system including a commercially represents a female-to-male ratio of 6.5:1 and does not available camera, both with the corresponding video correspond to previously reported data from the UK with analysis software [78]. The authors found excellent 2.4:1 [82] and Austria with 2.7:1 [83]. agreement between the 2D and 3D systems during bal- ance testing, and excellent reliability of the 2D system. The lack of blinding is a relevant limitation of this study. To minimise this limitation, a script was used by the In this study, we did not use reflective markers to better physiotherapist who gave the instructions and assessed identify specific body parts, because we were mainly inter- participants, and support was provided consistently. Apart ested in measuring initial contact or heel-strike points of from the fact that blinding of the participants would not the gait cycle. Thanks to modern camcorder 4 K technol- have been feasible because they would have realised their ogy and a well-lit hallway, excellent image exposure was group allocation, so far, none of the interventions is achieved, so reflective markers were redundant. Our sug- known to be superior to the others. Both walking tests gestions were supported by recent work from Castelli et al. were performed according to internationally recognised [79] who compared a 2D motion analysis in the sagittal guidelines and instructions. Nonetheless, it is possible plane with and without markers [79]. They observed high that, even if not intentionally, the assessor might have correlations between kinematic measures across different exerted an influence through her knowledge of group gait speeds for all joints between the two techniques. allocation. Therefore, it is relevant to implement blinding of the assessors in an eventual subsequent study. The results from our reliability study showed that the gait analysis system was a reliable instrument with low Consequences for the main study measurement errors related to step time data (SEM The results from this study suggest that a full-scale RCT 0.014 s) and low to moderate measurement errors re- is feasible. Relevant requirements elaborated for the lated to the step length data (SEM 0.013 m). Accord- main study are described as follows. ingly, the MDC95 for the step time measures was 0.038 s and 0.037 m for the step length measures. Comparable Randomisation research showed similar measurement errors for step Considering the threat of selection bias, stratified ran- time (SEM 0.001–0.009 s) and step length (SEM 0.03– domisation will be applied by an independent researcher 0.07 m) measures while using a 2D video-based gait ana- to balance groups. The stratification will be according to lysis in healthy people [80]. In agreement with this, a relevant prognostic factors for a change in walking, further study, which investigated the SEM and MDC namely, age (< 40, ≥40), gender (female, male) and dis- using an electronic walkway, found measurement errors ability (EDSS: 1.5–3.0, 3.5–4.5) [84]. In our previous for step time of 0.007 s in younger individuals and of study, we already used such a randomisation procedure 0.015 s in older people, respectively; the SEM for step on a similar population [19]. length was 0.006 m and 0.017 m in the same population. The same study found a MDC of 0.019 s and 0.042 s for Allocation concealment step time and of 0.016 m and 0.047 m for step length for To prevent allocation bias, allocation concealment younger and older people, respectively [81]. Consistent will be performed in the main study. As with this with this, the above cited study from Paul et al. found feasibility study, each participant will be assigned a measurement errors of 0.024–0.064 s for step time and unique identification number (ID). A computer gener- of 0.001–0.015 m for step length, obtained from a 2D ated randomisation list based on the predefined strata gait analysis system versus a 3D system [78]. Our results will be generated by an independent researcher at the suggest that the music-cueing impacted on the partici- Medical University of Innsbruck. Sealed opaque enve- pants’ gait as this was the only study which used music- lopes including group allocation numbers 1, 2 and 3 cueing, but this could not be substantiated. Overall, the will be created by the same researcher, which will be findings from the reliability study showed that the gait stored in a sequentially numbered order based on the analysis system can be used with confidence to measure randomisation list. These envelopes will be allocated gait synchronisation with music beat in the main study. to each participant in the order in which they are recruited. Participants will be asked to unseal the Study limitations envelopes themselves and not to discuss their group Due to the small sample size, the results should be consi- allocation until study completion. dered preliminary and not generalisable to a larger popu- lation of individuals with MS. At the baseline, there was

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 17 of 19 Blinding Additional files Involvement of three intervention groups in this feasibil- ity study made it unlikely that the allocation to a certain Additional file 1: CONSORT checklist for pilot and feasibility studies. group influenced participant motivation, behaviour or (DOC 229 kb) adherence. Compliance reports were similar in all groups, and neither the participants nor the researchers Additional file 2: Questions to be asked during phone calls. could know in advance which intervention, if any, might (DOCX 14 kb) induce greater improvements in performance. Once they will be allocated to a treatment, participants will be Additional file 3: Baseline characteristics for 15 participants. aware of the type of intervention. Concerning the asses- (XLSX 548 kb) sors, it is relevant to implement blinding in an eventual subsequent study. Abbreviations 2D: 2-dimensional; 6MWT: 6-Minute Walk Test; BPM: beats per minute; Assessments CD: compact disc; CI: confidence interval; CV MAD: coefficient of mean During the participant selection procedure, a minimal deviation about the median; CV: coefficient of variation; EDSS: Expanded screening of hearing will be considered since a hearing Disability Status Scale; ICC: intra-class correlation coefficient; KVIQ-G- impairment could limit the participation in a (cued) MI 10: Kinaesthetic and Visual Imagery Questionnaire, German version, short intervention. For the evaluation of participants’ cognitive form; KVIQ-G-10: Kinaesthetic and Visual Imagery Questionnaire, short form; capacity and mood/depression, it seems useful to MDC95: minimum detectable change with a 95% confidence interval; consider standardised measurement tools such as the MFIS: Modified Fatigue Impact Scale; MS: multiple sclerosis; MSIS-29: Multiple Mini-Mental State Examination [85] and the Beck Sclerosis Impact Scale-29; SD: standard deviation; SDC: smallest detectable Depression Inventory [86]. change; SEM: standard error of measurement; T25FW: Timed 25-Foot Walk; TDMI: Time-Dependent Motor Imagery screening test Reliability study A further development of the reliability study will be Acknowledgements considered for a future study, using an extended sample Funding for the material costs of this study was received from the Austrian size and employing different assessors. Such a study Multiple Sclerosis Research Society (no funding reference number; grant to BS). could be used to confirm intra-rater reliability values The authors would like to sincerely thank the funding body for their support. and to examine test-retest reliability. Further, the authors wish to warmly thank the participants of this study. They also thank the physiotherapists at the Innsbruck MS Clinic, particularly chief Conclusion therapist Gudrun Sylvest Schoenherr, MSc that they were able to use their Results from our feasibility study demonstrated that a facilities, the staff of the Innsbruck MS Clinic for handing over the recruitment full-scale RCT is feasible to investigate the mecha- flyers to the patients and Dr. Markus Reindl for helpful advice. nisms of differently cued and non-cued MI interven- tions and their effect on walking, fatigue and QoL in Funding people with MS. The gait analysis instruments were Funding for the material costs of this study was received from the Austrian found to be reliable and a future study could validate Multiple Sclerosis Research Society (no funding reference number). The these results. Stratified randomisation using a funders advertised the study on their webpage, but they had no role in the computer-generated randomisation sequence and allo- study design, data collection and analysis, interpretation of results, decision cation concealment should be used in the main study. to publish or preparation of the manuscript. It is essential to implement blinding in a further study. Additional assessments will be considered to Availability of data and materials screen their hearing and evaluate their cognitive The datasets supporting the conclusions of this article are included within the function and potential depression. article and its additional file. All data described in the manuscript, including all relevant raw data, are freely available to reviewers and any scientist wishing to The preliminary improvements in walking speed, walk- use them for non-commercial purposes, without breaching participant ing distance, fatigue, QoL and MI ability in the three confidentiality. Please contact the corresponding author for data requests. groups are promising however need to be confirmed. After a familiarisation with cued and non-cued MI, the Authors’ contributions participants showed adequate MI ability, which could BS, RK, AG and TB conceived and designed the study. BS was responsible for have influenced these improvements. Results from the acquisition of funding, data collection and manuscript writing. BS, RK and AG ongoing main study shall be used to provide specific rec- were involved in the statistical analysis and interpretation. TB and BS ommendations for clinically working physiotherapists on contributed venue and material. All authors read, revised critically for the use of (cued) MI in people with MS. important intellectual content, and approved the final manuscript. Ethics approval and consent to participate This study was approved by the Tier 2 College Research Ethics Committee (CREC) of the University of Brighton on 17 December 2015 (no reference number) and the Ethics Committee of the Medical University of Innsbruck on 26 February 2016 (reference number AN2014–0052 334/4.14358/5.13 (3743a)). All participants provided written informed consent to the participant consent form and were reimbursed for travel expenses only. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. As this is part of a PhD study, all intellectual property related to this study belongs to the University of Brighton, according to the University of Brighton’s Doctoral College (2012) research code of practice.

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 18 of 19 Publisher’s Note 21. Thaut MH. Rhythm, music and the brain. Scientific foundations and clinical applications. Oxon: Routledge; 2005. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 22. Thaut MH, Hoemberg V, editors. Oxford handbook of neurologic music therapy. Oxford: Oxford University Press; 2014. Author details 1School of Health Sciences, University of Brighton, Robert Dodd Building, 49 23. Eldridge SM, Chan CL, Campbell MJ, Bond CM, Hopewell S, Thabane L, Darley Road, Eastbourne BN20 7UR, UK. 2Clinical Department of Neurology, Lancaster GA. CONSORT 2010 statement: extension to randomised pilot and Medical University of Innsbruck, Anichstrasse 35, 6020 Innsbruck, Austria. feasibility trials. Pilot Feasibility Stud. 2016;2:64. Received: 29 June 2017 Accepted: 31 January 2018 24. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, Fujihara K, Havrdova E, Hutchinson M, Kappos L, et al. Diagnostic criteria for multiple References sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol. 2011;69:292–302. 1. Compston A, Confavreux C, Lassmann H, McDonald I, Miller D, Noseworthy 25. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded J, Smith K, Wekerle H. McAlpine’s multiple sclerosis. 4th edn. London: disability status scale (EDSS). Neurology. 1983;33:1444–52. Elsevier; 2006. 2. Krupp L. Fatigue is intrinsic to multiple sclerosis (MS) and is the most 26. Wondrusch C, Schuster-Amft C. A standardized motor imagery introduction commonly reported symptom of the disease. Mult Scler. 2006;12:367–8. program (MIIP) for neuro-rehabilitation: development and evaluation. Front 3. Induruwa I, Constantinescu CS, Gran B. Fatigue in multiple sclerosis -a brief Hum Neurosci. 2013;7:477. review. J Neurol Sci. 2012;323:9–15. 4. Phan-Ba R, Calay P, Grodent P, Delrue G, Lommers E, Delvaux V, Moonen G, 27. Schuster C, Hilfiker R, Amft O, Scheidhauer A, Andrews B, Butler J, Kischka U, Ettlin Belachew S. Motor fatigue measurement by distance-induced slow down of T. Best practice for motor imagery: a systematic literature review on motor walking speed in multiple sclerosis. PLoS One. 2012;7:e34744. imagery training elements in five different disciplines. BMC Med. 2011;9:75. 5. Leone C, Severijns D, Dolezalova V, Baert I, Dalgas U, Romberg A, Bethoux F, Gebara B, Santoyo Medina C, Maamagi H, et al. Prevalence of walking- 28. Holmes PS, Collins DJ. The PETTLEP approach to motor imagery: a related motor fatigue in persons with multiple sclerosis: decline in walking functional equivalence model for sport psychologists. J Appl Sport Psych. distance induced by the 6-minute walk test. Neurorehabil Neural Repair. 2001;13:60–83. 2016;30:373–83. 6. Malouin F, Richards CL. Mental practice for relearning locomotor skills. Phys 29. Jeannerod M. The cognitive neuroscience of action. Oxford. Blackwell. 1997; Ther. 2010;90:240–51. 30. Callow N, Hardy L. The relationship between the use of kinaesthetic imagery 7. Jeannerod M. Mental imagery in the motor context. Neuropsychologia. 1995;33:1419–32. and different visual imagery perspectives. J Sports Sci. 2004;22:167–77. 8. Decety J. Do imagined and executed actions share the same neural 31. Guillot A, Collet C, Dittmar A. Relationship between visual and kinesthetic substrate? Brain Res Cogn Brain Res. 1996;3:87–93. 9. Di Rienzo F, Collet C, Hoyek N, Guillot A. Impact of neurologic deficits on imagery, field dependence-independence, and complex motor skills. J motor imagery: a systematic review of clinical evaluations. Neuropsychol Psychophysiol. 2004;18:190–8. Rev. 2014;24:116–47. 32. Hoffmann TC, Glasziou PP, Boutron I, Milne R, Perera R, Moher D, Altman 10. Guillot A, Hoyek N, Louis M, Collet C. Understanding the timing of motor DG, Barbour V, Macdonald H, Johnston M, et al. Better reporting of imagery: recent findings and future directions. Int Rev Sport Exerc Psychol. interventions: template for intervention description and replication (TIDieR) 2012;5:3–22. checklist and guide. BMJ. 2014;348:g1687. 11. Decety J, Jeannerod M, Prablanc C. The timing of mentally represented 33. Kaufman M, Moyer D, Norton J. The significant change for the timed 25-foot actions. Behav Brain Res. 1989;34:35–42. walk in the multiple sclerosis functional composite. Mult Scler. 2000;6:286–90. 12. Malouin F, Richards CL, Durand A, Doyon J. Reliability of mental 34. Bosma LV, Sonder JM, Kragt JJ, Polman CH, Uitdehaag BM. Detecting clinically- chronometry for assessing motor imagery ability after stroke. Arch Phys relevant changes in progressive multiple sclerosis. Mult Scler. 2015;21:171–9. Med Rehabil. 2008;89:311–9. 35. Cutter GR, Baier ML, Rudick RA, Cookfair DL, Fischer JS, Petkau J, Syndulko K, 13. Heremans E, D'Hooge AM, De Bondt S, Helsen W, Feys P. The relation Weinshenker BG, Antel JP, Confavreux C, et al. Development of a multiple between cognitive and motor dysfunction and motor imagery ability in sclerosis functional composite as a clinical trial outcome measure. Brain. patients with multiple sclerosis. Mult Scler. 2012;18:1303–9. 1999;122(Pt 5):871–82. 14. Tabrizi YM, Mazhari S, Nazari MA, Zangiabadi N, Sheibani V. Abnormalities of 36. Hobart J, Blight AR, Goodman A, Lynn F, Putzki N. Timed 25-foot walk: motor imagery and relationship with depressive symptoms in mildly disabling direct evidence that improving 20% or greater is clinically meaningful in relapsing-remitting multiple sclerosis. J Neurol Phys Ther. 2014;38:111–8. MS. Neurology. 2013;80:1509–17. 15. Tacchino A, Bove M, Pedulla L, Battaglia MA, Papaxanthis C, Brichetto G. 37. Goldman MD, Marrie RA, Cohen JA. Evaluation of the six-minute walk in Imagined actions in multiple sclerosis patients: evidence of decline in motor multiple sclerosis subjects and healthy controls. Mult Scler. 2008;14:383–90. cognitive prediction. Exp Brain Res. 2013;229:561–70. 38. Learmonth YC, Dlugonski DD, Pilutti LA, Sandroff BM, Motl RW. The 16. Azin M, Zangiabadi N, Moghadas Tabrizi Y, Iranmanesh F, Baneshi MR. reliability, precision and clinically meaningful change of walking Deficiency in mental rotation of upper and lower-limbs in patients with assessments in multiple sclerosis. Mult Scler. 2013;19:1784–91. multiple sclerosis and its relation with cognitive functions. Acta Med Iran. 39. Stratford PW, Binkley JM, Riddle DL. Health status measures: strategies and 2016;54:510–7. analytic methods for assessing change scores. Phys Ther. 1996;76:1109–23. 17. Heremans E, Nieuwboer A, Spildooren J, De Bondt S, D'Hooge AM, Helsen 40. Feys P, Baert I, Dalgas U, Smedal T. Need for differentiation of real and W, Feys P. Cued motor imagery in patients with multiple sclerosis. clinical important change in research on responsiveness of walking Neuroscience. 2012;206:115–21. outcome measures in multiple sclerosis. Mult Scler. 2014;20:761. 18. Kim JS, Oh DW, Kim SY, Choi JD. Visual and kinesthetic locomotor imagery 41. Applebee A, Goodman AD, Mayadev AS, Bethoux F, Goldman MD, Klingler M, training integrated with auditory step rhythm for walking performance of Blight AR, Carrazana EJ. Effects of Dalfampridine extended-release tablets on 6- patients with chronic stroke. Clin Rehabil. 2011;25:134–45. minute walk distance in patients with multiple sclerosis: a post hoc analysis of 19. Seebacher B, Kuisma R, Glynn A, Berger T. The effect of rhythmic-cued a double-blind, placebo-controlled trial. Clin Ther. 2015;37:2780–7. motor imagery on walking, fatigue and quality of life in people with 42. Chetta A, Zanini A, Pisi G, Aiello M, Tzani P, Neri M, Olivieri D. Reference multiple sclerosis: a randomised controlled trial. Mult Scler. 2017;23:286–96. values for the 6-min walk test in healthy subjects 20-50 years old. Respir 20. Catalan M, De Michiel A, Bratina A, Mezzarobba S, Pellegrini L, Marcovich R, Med. 2006;100:1573–8. Tamiozzo F, Servillo G, Zugna L, Bosco A, et al. Treatment of fatigue in 43. Wetzel JL, Fry DK, Pfalzer LA. Six-minute walk test for persons with mild or multiple sclerosis patients: a neurocognitive approach. Rehabil Res Pract. moderate disability from multiple sclerosis: performance and explanatory 2011;2011:670537. factors. Physiother Can. 2011;63:166–80. 44. Fisk JD, Ritvo PG, Ross L, Haase DA, Marrie TJ, Schlech WF. Measuring the functional impact of fatigue: initial validation of the fatigue impact scale. Clin Infect Dis. 1994;18(Suppl 1):S79–83. 45. Fischer JS, LaRocca NG, Miller DM, Ritvo PG, Andrews H, Paty D. Recent developments in the assessment of quality of life in multiple sclerosis (MS). Mult Scler. 1999;5:251–9. 46. Learmonth YC, Dlugonski D, Pilutti LA, Sandroff BM, Klaren R, Motl RW. Psychometric properties of the fatigue severity scale and the modified fatigue impact scale. J Neurol Sci. 2013;331:102–7.

Seebacher et al. Archives of Physiotherapy (2018) 8:6 Page 19 of 19 47. Rietberg MB, Van Wegen EE, Kwakkel G. Measuring fatigue in patients with 72. Donoghue D, Physiotherapy Research and Older people (PROP) group, Stokes multiple sclerosis: reproducibility, responsiveness and concurrent validity of EK. How much change is true change? The minimum detectable change of three Dutch self-report questionnaires. Disabil Rehabil. 2010;32:1870–6. the berg balance scale in elderly people. J Rehabil Med. 2009;41:343–6. 48. Flachenecker P, Kumpfel T, Kallmann B, Gottschalk M, Grauer O, Rieckmann 73. Karageorghis CI, Priest DL. Music in the exercise domain: a review and P, Trenkwalder C, Toyka KV. Fatigue in multiple sclerosis: a comparison of synthesis (part 1). Int Rev Sport Exerc Psychol. 2012;5:44–66. different rating scales and correlation to clinical parameters. Mult Scler. 2002;8:523–6. 74. Reiman MP, Lorenz DS. Integration of strength and conditioning principles into a rehabilitation program. Int J Sports Phys Ther. 2011;6:241–53. 49. Hobart J, Lamping D, Fitzpatrick R, Riazi A, Thompson A. The multiple sclerosis impact scale (MSIS-29): a new patient-based outcome measure. 75. Guillot A, Di Rienzo F, Collet C. The neurofunctional architecture of motor Brain. 2001;124:962–73. imagery. In Book The neurofunctional architecture of motor imagery. pp. 678. City: InTechOpen; 2014. p. 678. 50. Riazi A. Patient-reported outcome measures in multiple sclerosis. Int MS J. 2006;13:92–9. 76. Rocca MA, Gatti R, Agosta F, Broglia P, Rossi P, Riboldi E, Corti M, Comi G, Filippi M. Influence of task complexity during coordinated hand and foot 51. Hobart JC, Riazi A, Lamping DL, Fitzpatrick R, Thompson AJ. How responsive movements in MS patients with and without fatigue. A kinematic and is the multiple sclerosis impact scale (MSIS-29)? A comparison with some functional MRI study. J Neurol. 2009;256:470–82. other self report scales. J Neurol Neurosurg Psychiatry. 2005;76:1539–43. 77. Schmidt, R.A., and T.D. Lee. 2011. Motor control and learning. 5th edn. 52. Ramp M, Khan F, Misajon RA, Pallant JF. Rasch analysis of the multiple Champaign: Human Kinetics. sclerosis impact scale MSIS-29. Health Qual Life Outcomes. 2009;7:58. 78. Paul SS, Lester ME, Foreman KB, Dibble LE. Validity and reliability of two- 53. Malouin F, Richards CL, Durand A, Doyon J. Clinical assessment of motor dimensional motion analysis for quantifying postural deficits in adults with and imagery after stroke. Neurorehabil Neural Repair. 2008;22:330–40. without neurological impairment. Anat Rec (Hoboken). 2016;299:1165–73. 54. Guillot A, Collet C. The neurophysiological foundations of mental and motor 79. Castelli A, Paolini G, Cereatti A, Della Croce U. A 2D markerless gait analysis imagery. New York: Oxford University Press; 2010. methodology: validation on healthy subjects. Comput Math Methods Med. 2015;2015:186780. 55. Malouin F, Richards CL, Jackson PL, Lafleur MF, Durand A, Doyon J. The kinesthetic and visual imagery questionnaire (KVIQ) for assessing motor 80. Padulo J, Annino G, D'Ottavio S, Vernillo G, Smith L, Migliaccio GM, Tihanyi imagery in persons with physical disabilities: a reliability and construct J. Footstep analysis at different slopes and speeds in elite race walking. J validity study. J Neurol Phys Ther. 2007;31:20–9. Strength Cond Res. 2013;27:125–9. 56. Schuster C, Lussi A, Wirth B, Ettlin T. Two assessments to evaluate imagery 81. Almarwani M, Perera S, VanSwearingen JM, Sparto PJ, Brach JS. The test- ability: translation, test-retest reliability and concurrent validity of the retest reliability and minimal detectable change of spatial and temporal gait German KVIQ and Imaprax. BMC Med Res Methodol. 2012;12:127. variability during usual over-ground walking for younger and older adults. Gait Posture. 2016;44:94–9. 57. Tabrizi YM, Zangiabadi N, Mazhari S, Zolala F. The reliability and validity study of the kinesthetic and visual imagery questionnaire in individuals with 82. Mackenzie IS, Morant SV, Bloomfield GA, MacDonald TM, O'Riordan J. multiple sclerosis. Braz J Phys Ther. 2013;17:588–92. Incidence and prevalence of multiple sclerosis in the UK 1990-2010: a descriptive study in the general practice research database. J Neurol 58. Saimpont A, Malouin F, Tousignant B, Jackson PL. Assessing motor imagery Neurosurg Psychiatry. 2014;85:76–84. ability in younger and older adults by combining measures of vividness, controllability and timing of motor imagery. Brain Res. 2015;1597:196–209. 83. Trojano M, Lucchese G, Graziano G, Taylor BV, Simpson S, Jr., Lepore V, Grand'maison F, Duquette P, Izquierdo G, Grammond P, et al. Geographical 59. Wajda DA, Sandroff BM, Pula JH, Motl RW, Sosnoff JJ. Effects of walking variations in sex ratio trends over time in multiple sclerosis. PLoS One 2012, direction and cognitive challenges on gait in persons with multiple 7:e48078. sclerosis. Mult Scler Int. 2013;2013:859323. 84. Baert I, Freeman J, Smedal T, Dalgas U, Romberg A, Kalron A, Conyers H, 60. Sandroff BM, Klaren RE, Pilutti LA, Motl RW. Oxygen cost of walking in Elorriaga I, Gebara B, Gumse J, et al. Responsiveness and clinically persons with multiple sclerosis: disability matters, but why? Mult Scler Int. meaningful improvement, according to disability level, of five walking 2014;2014:162765. measures after rehabilitation in multiple sclerosis: a European multicenter study. Neurorehabil Neural Repair. 2014;28:621–31. 61. Kalron A, Givon U. Gait characteristics according to pyramidal, sensory and cerebellar EDSS subcategories in people with multiple sclerosis. J Neurol. 85. Folstein MF, Folstein SE, McHugh PR. \"mini-mental state\". A practical 2016;263:1796–801. method for grading the cognitive state of patients for the clinician. J Psych Res. 1975;12:189–98. 62. Harris-Hayes M, Steger-May K, Koh C, Royer NK, Graci V, Salsich GB. Classification of lower extremity movement patterns based on visual 86. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for assessment: reliability and correlation with 2-dimensional video analysis. J measuring depression. Arch Gen Psychiatry. 1961;4:561–71. Athl Train. 2014;49:304–10. Submit your next manuscript to BioMed Central 63. Norris BS, Olson SL. Concurrent validity and reliability of two-dimensional and we will help you at every step: video analysis of hip and knee joint motion during mechanical lifting. Physiother Theory Pract. 2011;27:521–30. • We accept pre-submission inquiries 64. Baker R. Measuring walking. A handbook of clinical gait analysis. London: • Our selector tool helps you to find the most relevant journal Mac Keith Press; 2013. • We provide round the clock customer support 65. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353: 487–97. • Convenient online submission 66. Newcombe RG. Two-sided confidence intervals for the single proportion: • Thorough peer review comparison of seven methods. Stat Med. 1998;17:857–72. • Inclusion in PubMed and all major indexing services 67. Brown L, Cai T, DasGupta A. Interval estimation for a binomial proportion. Stat Sci. 2001;16:101–33. • Maximum visibility for your research 68. Fleiss J. The design and analysis of clinical experiments. New York: John Submit your manuscript at Wiley & Sons; 1986. www.biomedcentral.com/submit 69. Harvill LM. Standard Error of Measurement. Educational Measurement: Issues and Practice. 1991;10:33–41. 70. Tighe J, McManus IC, Dewhurst NG, Chis L, Mucklow J. The standard error of measurement is a more appropriate measure of quality for postgraduate medical assessments than is reliability: an analysis of MRCP(UK) examinations. BMC Med Educ. 2010;10:40. 71. Atkinson G, Nevill AM. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med. 1998;26:217–38.

Ristori et al. Archives of Physiotherapy (2018) 8:7 https://doi.org/10.1186/s40945-018-0050-3 REVIEW Open Access Towards an integrated clinical framework for patient with shoulder pain Diego Ristori1,7, Simone Miele2,7, Giacomo Rossettini3,7, Erica Monaldi4,7, Diego Arceri5,7 and Marco Testa6,7* Abstract Background: Shoulder pain (SP) represents a common musculoskeletal condition that requires physical therapy care. Along the years, the usual evaluation strategies based on clinical tests and diagnostic imaging has been challenged. Clinical tests appear unable to clearly identify the structures that generated pain and interpretation of diagnostic imaging is still controversial. The current patho-anatomical diagnostic categories have demonstrated poor reliability and seem inadequate for the SP treatment. Objectives: The present paper aims to (1) describe the different proposals of clinical approach to SP currently available in the literature; to (2) integrate these proposals in a single framework in order to help the management of SP. Conclusion: The proposed clinical framework, based on a bio-psychosocial vision of health, integrates symptoms characteristics, pain mechanisms and expectations, preferences and psychosocial factors of patients that may guide physiotherapist to make a diagnostic triage and to choose the right treatment for the individual patient. Keywords: Shoulder pain, Diagnosis, Rehabilitation treatment, Clinical framework Background Schellingerhout [6] defined the shoulder classification Shoulder pain (SP) is a common musculoskeletal condi- process as “a Babylonian confusion of tongues and seem tion that can influence negatively the function of the en- to be of little benefit for those with SP”. This conclusion tire upper limb [1]. The prevalence of SP ranged is in line with Buchbinder [9]: analyzing 5 classification between 7 and 26% within the general population, in- systems based on patho-anatomical way of soft tissue creasing with age [2]. Most of the patients affected by SP disorders, she argued that they may not be acceptable describe the symptoms as “troublesome pain” [3]. When for lack of validity and reliability of the inclusion criteria these symptoms become persistent and recurrent, the that create an overlapping of categories. demand for medical consulting increases [4]. As consequence, we have thought it would be clinic- A large number of diagnostic categories have been de- ally useful to overcome the diagnostic difficulties by pro- veloped: they are based on patho-anatomical classifica- posing a new pragmatic and symptoms-based model, tions, such as tendinopathies, bursitis, labral tears, coherent with a bio-psychosocial approach and closer to tendon tears, impingement, etc. [5]. However, there is patient’s needs. Moving from this vision, this debate considerable uncertainty regarding these diagnostic cri- aims to: 1) describe the reasons for this diagnostic in- teria [6], and the basis for them has been repeatedly consistency; 2) present the different alternative proposals challenged [7, 8]. Clinically, it may not be possible to existing in the literature; 3) integrate the different pro- distinguish between these patho-anatomical diagnostic posals in a single framework, in order to provide physio- categories with certainty [9]. therapists with a helpful tool to deal with SP patients. * Correspondence: [email protected] Is the usual diagnostic process of SP valid and 6Via Magliotto, 2 17100, Savona, Italy helpful? 7Department of Neuroscience, Rehabilitation, Ophtalmology, Genetics, In clinical practice, the assessment of a patient with SP is based upon an in-depth conversation (relevant history tak- Maternal and Child Health, University of Genova, Campus of Savona, Savona, ing, understanding the patient’s complaints, and defining his/her psychosocial status) and a clinical assessment, Italy Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 2 of 11 which, in some cases, may be supported by imaging (e.g. concluding that there was a need of a new system of as- magnetic resonance imaging [MRI] or ultrasound [US]). sessment in order to classify patients with SP [20]. This approach is designed to enable a clinical diagnosis use- ful to guide the subsequent physiotherapy treatment [10]. Importance of diagnostic imaging in SP In clinical practice, diagnostic imaging (e.g. MRI, mag- Importance of clinical physical tests in SP netic resonance arthrography [MRA], US and radio- Physical tests are tools commonly used in clinical practice, graphs [X-RAY]) is considered to play an essential role created to help the physiotherapist to identify which ana- during the assessment of patients with musculoskeletal tomical structures are involved with the patient’s symp- disorders [21]. They are used both in specialist consult- toms. They are non-invasive, quick, convenient, and ation (e.g. MRI and X-RAY to quantify the lesions and provide immediate results [11]. However, their interpret- to support surgical planning) and in general practitioner ation may differ with the examiners’ clinical expertise [11]. (GP) consultation of primary care (e.g.US) [21]. Morpho- logical and degenerative alterations are commonly con- Anatomical basis sidered relevant and together with patient’s history and Green et al. state that only few studies give information examinations findings could support the choice of treat- concerning the anatomical basis of the proposed tests ment [22]. [12]. Only four tests among those included in their re- view present a clear anatomical base. For these reasons, Diagnostic accuracy the author suggests a lack of assumptions in order to Lenza et al. [21], stated that MRI, MRA and US, are use- know what is happening in the shoulder during these as- ful tools to identify massive rotator cuff tears in a popu- sessment procedures [12]. lation of patients included in a waiting list for surgery. Diagnostic performance of imaging decreases in line The Hawkins-Kennedy represents a well-fitting ex- with the reduction of size the lesion. Moreover, the avail- ample of the confusion surrounding the anatomical con- able studies generally present weak methodological qual- struct of these tests. It has been developed to identify ity and heterogeneity of the included populations. the presence of sub-acromial impingement [13]. During Indeed, the diagnostic accuracy of these tools dramatic- years, many hypotheses have been suggested: compres- ally decreases when applied in populations with poorly sion of supraspinatus tendon under the coraco-acromial defined clinical features of association between struc- ligament [13], compression of the structures of the tural lesions and symptoms [23]. sub-acromial space between the head of the humerus and the acromion [14], contact between acromion and Clinical usefulness and relevance coraco-acromial ligament [15], compression of the long In 2013 a Cochrane editorial debated the diagnostic ac- biceps head tendon [16]. curacy of imaging [24]; it was argued that the presence of asymptomatic rotator cuff tears [25] represents the Clinical usefulness “elephant in the room” responsible to challenge the rele- Others researchers have questioned the clinical useful- vance of diagnostic imaging. Some observational studies ness of the physical tests in SP. Hegedus et al. have pub- confirm this perspective: Girish stated that up to 2/3 of lished three literature reviews discussing this topic [17– people with a rotator cuff lesion are asymptomatic [26] 19]. They concluded that clinicians cannot confirm a and rotator cuff tears are common in symptomatic and diagnosis of the different shoulder problems neither with asymptomatic populations [27]. individual tests nor with cluster tests [18, 19]. They de- fined impingement as an “all-encompassing term” often Usually, in patients with SP, there is uncertainty con- meaningless with respect to the treatment [17]. There- cerning the cause of pain and which risk factors are rele- fore, the clinical history collected from the patient and vant to the onset of symptoms. Some authors suggest expert clinical reasoning seems to be crucial in the diag- that the possibility of symptoms increases with the size nostic process [19]. Hanchard et al. investigated physical of rotator cuff tear [28], while other authors proposed test for impingement and associated lesions. Authors that the development of symptoms is mostly correlated concluded that the body of evidence is extremely hetero- to other non-structural factors, such as gender, age and geneous both in terms of performance (e.g. reliability, psychosocial factors [29, 30]. Thus, clinical interpret- specificity and sensitivity) and relative interpretation ation of diagnostic imaging in patients with SP remains thus making impossible to perform a synthesis of avail- controversial, suggesting that the biomechanical classifi- able data and to draw conclusions about their clinical cation system is unsuitable (Fig. 1). application [11]. Furthermore, the reliability of these procedures was also found to be poor, with the authors

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 3 of 11 Fig. 1 Inconsistency of diagnostic labels in SP. The weak correlation still present some limitations that reduce their ability to between structural factors and shoulder pain, together with the interpret comprehensively all the features of SP. The limited diagnostic value of bio-imaging and clinical tests, caused a STAR-shoulder classification and SSMP do not assess lack of uniformity in diagnostic labelling the contribution of central sensitization (CS) [31, 32]. Klintberg et al., propose a pure mechanist approach Are integrative procedures of assessment based on a diagnostic algorithm for the assessment of available? movement patterns [33]. Moreover, excluding SSMP The lack of reliability of clinical tests and limited useful- [34], the reliability of the existing proposals is still ness of diagnostic imaging led some authors to suggest lacking. the integration of different assessment strategies, more pragmatic and focused on the results of functional as- Moving from the existing proposals, we have tried to sessment [31–33]. The existing proposals include: the overcome the limits of exclusively mechanistic classifica- shoulder symptom modification procedure (SSMP) [31], tion and to create a framework for assessment and treat- the staged approach for rehabilitation classification: ment of SP that integrates and includes them in a shoulder disorders (STAR-shoulder) [32] and the Klint- bio-psychosocial perspective (Fig. 2). Anamnesis, phys- berg proposal [33] that synthesize a consensus statement ical assessment, triage and treatment are the four clinical of several shoulder rehabilitation experts (see Table 1). procedures mainly affected by the implementation of our clinical perspective. Anamnesis The anamnesis is a milestone in the assessment of pa- tients with musculoskeletal dysfunction [35]. Different anamnestic elements must be collected (e.g. characteris- tics of symptoms, mechanisms of pain, expectations, preferences and psychosocial factors of patients), weighted and included in the clinical reasoning process to guide the subsequent physical examination [11, 21]. Is it the time to move towards an integrated Characteristics of symptoms clinical framework for the assessment and Specific information regarding impairments and symp- treatment of SP? toms of SP must be investigated during the colloquium These current approaches have evolved as a conse- with the patient, including; onset, quality, 24 h behavior, quence of the uncertainty in biomedical model, but they localization, alleviating and aggravating factors [35]. Table 1 Characteristics of existingproposals of assessment strategies Existing proposals of assessment strategies SSMP Star-shoulder Klintberg’s clinical algorithm The SSMP is a series of clinical The authors created a model The algorithm encompasses procedures aimed to reduce the providing a sub-classification the functional assessment patient’s symptoms. A procedure of patients on the basis of of a range of motion (ROM) able to eliminate/reduce the patho-anatomical features, and the evaluation of symptoms is adopted as a tissue irritability and individual presence/absence of abnormal treatment technique. If following impairments. Three steps are scapulohumeral the application of the SSMP, proposed: 1) screening, 2) motion pattern in order to symptoms have not completely patho-anatomical diagnosis identify patients with limited disappeared an exercise program (e.g. sub-acromial syndrome, passive ROM or with reduced is required; the SSMP is typically frozen shoulder, glenohumeral muscle performance that can embedded within a graduated instability) and 3) a be treated with specific exercises shoulder exercise program. rehabilitative step, based on or manual therapy. The algorithm Lewis suggests to apply the different the level of irritability. helps clinicians to choose techniques of the SSMP after the the adequate therapeutic approach. conduction of a preliminary assessment Moreover, it allows flexibility during (composed by detailed history, screening the assessment process. for potential red-flag, functional/disability Algorithm-based re-assessment questionnaires administration, evaluation of the patients allows monitoring of impairments and if necessary whether the proposed exercises orthopaedic tests and imaging). are correctly targeted towards the prevalent impairment or is necessary to test other clinical adjunctive problems.

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 4 of 11 Fig. 2 The integrated clinical model for the assessment and treatment of SP. By history taking, the physiotherapist investigates pain characteristics, its prevalent mechanisms and patient’s beliefs and expectations. Integrating this information with the results of the physical assessment, the physiotherapist classifies the shoulder pain condition with three diagnostic labels: Red Flags and Specific SP which require a referral to a specialist consultation and Non-specific SP which falls within the competence of the physiotherapist Moreover, there is a need to consider symptoms in other non-coherence of any referred pain and the widespread body segments reported by patients correlated with the irritability. Moreover, other elements are disturbed sleep, main problem (e.g. cervical or thoracic spine), the co- areas of numbness, misperception of the affected seg- morbidities, the previous consultancy with other health- ment, the feeling of swelling in absence of evident edema care professions (e.g. orthopaedic), the previous positive/ which may increase with closed eyes [39, 42]. negative experiences with a specific therapeutic ap- proach, the patient’s life context (e.g. family or work In addition to the anamnestic elements, physical signs problems) and the screening of red flags [35]. Physio- of CS must be investigated during physical examination therapists should also investigate the limitation of activ- (e.g. swelling, weakness or stiffness of the affected seg- ities and restriction of participation associated with SP ment, lack of correspondence between specific move- impairments with specific questionnaires (e.g. the dis- ments and pain) [42]. During the assessment process, abilities of the arm, shoulder and hand questionnaire the identification of pain mechanism on the basis of pa- [DASH], shoulder pain and disability index [SPADI], the tient’s dysfunction could help physiotherapists to better pain self-efficacy questionnaire [PSEQ]) [36–38]. manage the SP and to target more adequately the treat- ment [43]. Mechanisms of pain The features of symptom help physiotherapists to under- Unfortunately, there is only preliminary discriminative stand the underpinning mechanism behind patient’s pain validity of mechanism-based classification of musculo- presentation, such as nociceptive pain (NP) or CS. Po- skeletal pain [39]. The use of some self-reported tool tential indicators of NP are: the localization of pain in (e.g. central sensitization inventory [CSI]) could be use- the area of injury/dysfunction, the description as ful for physiotherapists to quantify symptoms severity of intermittent and sharp during movement or constant CS, thus guiding the clinical reasoning process [44]. dull or throbbing ache at rest, as well as a clear re- sponse consistent with aggravating and easing mech- Expectations, preferences and psychosocial factors of anical factors [39]. patients During the anamnesis, it is essential to investigate the Evidence suggests that CS phenomena are present in patient’s expectation, preferences and the presence of patients with SP [40, 41]. Potential indicators of CS are: psychosocial factors (yellow flags) in order to guide the the absence of correlation or inconsistency between pro- subsequent best treatment decisions and to reduce the vocative stimulus and response, the discontinuity of patient’s risk for developing long-term disability [45]. pain, its unpredictability and increase with non-specific movements, the variable anatomical distribution, the Recent evidence suggests that expectations and prefer- ences about the physiotherapy treatment play an import- ant role as influencers of musculoskeletal outcomes [46,

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 5 of 11 47] also in patients with SP [48]. It is useful to ask the non-specific pain. Firstly, the physiotherapist should ex- patients what he expects from physiotherapy to under- clude Red Flags, then distinguish patients, classifiable as stand if the achievement of the desired outcomes is pos- specific shoulder pain, with signs and symptoms of mus- sible or not. The physiotherapists should also investigate culoskeletal dysfunction for which is necessary the refer- the patient’s preferences towards a specific treatment ral to an orthopaedic evaluation before establishing a (e.g. manual therapy or exercises) [46, 47]. physiotherapy treatment [57, 58]. Finally, the physiother- apist can classify as non-specific shoulder pain [56] the Moreover, the predominance of specific psychological patients presenting clinical features that do not belong factors such as personal and environmental elements to the two categories described above. must be analysed. The physiotherapist should screen for older age (more of 50 years), higher perceived pain in- Red flags and specific shoulder pain tensity, longer duration of symptoms, previous injury, Red flags are sign and symptoms alerting the physiother- extensive sick leave, unemployment, co-morbidities, pre- apist on a possible presence of a non-musculoskeletal, vious SP, poor perceived general health, avoidance of ac- life-threatening pathology, fracture, infection, tumor and tivity for fear of pain and harm, perceived high job inflammatory rheumatic conditions [59]. However, phys- demands and low job satisfaction, higher body mass iotherapists must be careful in the evaluation of signs index, poor social support, personal problems (alcohol, and symptoms of patients [60]. The prevalence and inci- financial, marital) [49, 50]. dence of red flags in shoulder disorders are unknown [59, 61], thus limiting the identification of serious Physical assessment non-musculoskeletal pathology at the first consultation The evaluation of the quality of active and passive shoul- [62]. Specific shoulder pain indicates that symptoms der and cervical spine movements [51], the range of mo- could refer to a pathology that has a clear structural, tion (ROM) and the shoulder muscles strength are the patho-anatomic or pathophysiologic origin (e.g. symptom- priority of the assessment [33]. Physical assessment atic rotator cuff tears, superior labral tear from anterior to should explore the provocative movements of the pa- posterior [SLAP] or instability). It requires referral to an tient’s pain. It can be an active or passive movement of orthopaedic specialist to clarify diagnostic aspects or sur- the shoulder [31] or a movement of the neck region (in gical needs [63, 64]. Signs and symptoms characteristic of this case a comprehensive clinical assessment of this re- these two categories are listed in Table 2. It is not neces- gion should be considered) [20]. sary that all symptoms have to be present at the same time to guide physiotherapists during their clinical reasoning Physical assessment is also aimed to confirm the pres- process [65–68]. When a conservative approach has been ence of CS signs suspected during anamnesis, such as chosen for specific shoulder pain, physiotherapists may swelling, weakness or stiffness of the affected segment, refer to specific options for treatment available in litera- lack of correspondence between specific movements and ture (e.g. for conservative treatment of patients with pain [39, 42], while the clinical utility of quantitative massive rotator cuff tears, we could propose stretching, sensory testing (QST) for the detection of CS is ques- proprioceptive and active exercises towards functional tionable and still a source of debate [52]. Triage and treatment Table 2 Red Flags and symptoms of specific shoulder pain From a clinical-pragmatic perspective, there is the need to modify the diagnostic labeling used for patients with Anamnestic and clinical Sign and symptoms of SP [6, 17, 20]. As proposed for other body regions (e.g. Features of red flags Specific shoulder pain lumbar spine, cervical spine), also in the shoulder com- plex, there is a growing awareness of the very limited Fever, shivering, changes in body Recent trauma of the shoulder ability to identify a specific structure responsible for the patient’s symptoms [53]. Therefore, a transition of the temperature overnight, diaphoresis, complex, high reactivity of SP assessment from a strictly mechanistic to a more bio-psychosocial oriented approach seems necessary. nausea, unexplainable sweating symptoms, pain during the night, The analysis of the patient’s history, beliefs, preferences and functional movements, have recently assumed a key overnight, vomiting, sphincteric limitation of flexion (< 90° both role [54]. complaints, diarrhoea, paleness, passive and active), apprehension, In our clinical framework, we propose a process of diagnostic triage that adopts a classification system simi- fatigue, lurching, fainting, fear of movement and/or lar to the one already adopted for other regions (e.g. lumbar spine [55, 56]): red flags, specific pain, exhaustion, excessive and weakness during humeral unexplainable weakness, not linked external rotation. to any physical effort, unexplainable loss of weight, skin rash, unexplainable multiple hematoma, lumps over the body, deformities, inability to lay supine in bed, marked muscle weakness, marked restriction of movement, limb atrophy, local pain and pain during load when age is less than 20 years old and more than 50.

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 6 of 11 movements, increasing progressively the position of exe- through exercise, thus playing an important role for this cution and the resistance) (see Table 4) [57, 58]. category of patients with SP [79]. Indeed exercises have both the capacity of re-conditioning the anatomical Non-specific shoulder pain structures, with an effect on NP mechanisms [57, 80, 81] Once the patient is categorized as non-specific shoulder and the capacity of modulating the patient’s pain with an pain [69], the physiotherapist should recognize what is action on CS mechanisms [82]. This effect of movement the prevalent mechanism of pain elaboration of the pa- and exercises have been demonstrated also in other body tient, and identify what are the functional movements regions [81, 83]. Various load strategies will be described that provoke symptoms. If the patient does not in the section below. recognize precisely the pain provocative shoulder move- ments, physiotherapists can use shoulder orthopaedic Algorithm of treatment To organize the treatment of tests to provoke pain [31]. Three strategies, overlaid and patients with non-specific shoulder pain provoked by fused in every intervention of our clinical practice, shoulder movements, is advisable to adopt the same should be adopted to treat non-specific shoulder pain pa- functional approach proposed by Klintberg et al., that tients: education, de-sensitization and load management. seems to be flexible and easy to perform [33] (Fig. 3). Overall, in the choice of treatment, the physiotherapist Once the painful movement is identified, the patient must integrate, as much as possible, expectations and rates his/her pain on a numeric rating scale and then the preferences of patients thus adopting any previous posi- physiotherapist attempts to modify it applying specific tive physiotherapy solutions and avoiding the past nega- procedures [31, 57, 84–89] (see Table 3). tive experiences [46, 47]. Different procedures are administered to the patient Education It is important to inform patients about their until he/she reports a satisfying improvement occurring clinical condition, avoiding an excessive biomedical ter- during or after the intervention [31]. Some authors minology (e.g. “shoulder impingement”), explaining the stated that inter and intra-treatment changes may be pain mechanisms underpinning their symptoms, their fa- predictive of improvement of the specific symptom as vorable prognosis, the strategies of treatment that are well as of the general condition of the patient: this intended to use proprioand the value of self-management phenomenon can be of support for the clinical and home exercise [48, 70, 71]. This education process decision-making [90, 91]. If the procedures of symptoms should be promoted throughout the whole treatment, thus modification result effective, manual treatment in associ- enhancing the patient’s engagement and empowerment ation with exercises (with a progressive amount of load [70, 72]. In presence of high predominance of yellow flags, based upon the clinical evolution of the patient) are the patient should be monitored and educated, thus modi- adopted [51, 57, 85]. The pain-free therapeutic window fying any dysfunctional beliefs and overestimated expecta- identified by the positive response of symptoms modifi- tions about SP and reconceptualising on a cognitive level cation procedures can be used to propose pain-free exer- any fear, harm and avoidance about shoulder activity [73]. cises (adopting positive procedures as exercises), regardless of the presence of CS component of the clin- De-sensitization Manual therapy is one of the possible ical scenario (see Table 4). interventions to reduce SP. Mechanical stimuli applied to the skin of the patients by manual therapy, determine Regarding load management, the physiotherapist several neurophysiological mechanisms (e.g. peripheral, should play with an accurate tuning of the posology, in spinal and supra-spinal) that improve pain. Because of terms of specific target and modalities of execution (see this variety of effects sources, we can assume that man- Table 4) [32, 33, 43, 57, 79, 92, 93]. ual therapy can be considered as a therapeutic interven- tion able to de-sensitize the neurologic system that Even if some authors choose the structure(s) to which supports pain perception [74–77]. Manual therapy could target the exercises on the base of the effect of the previ- play an important role to decrease fear of movement ously performed procedure (e.g. gleno-humeral muscles and catastrophization [75]. The physiotherapists should if humeral head and muscular conditioning procedures consider also the adoption of drugs (e.g. pain killers) or were positive; or thoracic spine muscles if modification exercises to reduce pain and to desensitize the patient of thoracic kyphosis resulted positive) [33, 57, 87, 93], [51, 57], thus reducing the possibility of CS [78]. according with Littlewood et al. [79] we suggest as pref- erable to rely on the dysfunctional pattern of movement Load management We considered as load every move- of the patient in order to restore the activities of daily ment that could increase the ability to perform a lim- life (e.g. respecting the tolerance of movement in ited/painful movement. The load is usually administered terms of pain/fatigue, if abduction is the painful/lim- ited movement, the physiotherapist will increase and modulate number of repetition and resistance in ab- duction exercises).

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 7 of 11 Fig. 3 Non-specific SP: the algorithm of treatment. De-sensitization procedures should be adopted first. If an improvement of pain and/or patient’s satisfaction is obtained, the treatment load should be increased by using the positive procedures and specific exercises. If this first approach does not reach its goal, then therapeutic strategies based on the prevalent pain mechanism should be implemented. Symptom-contingent strategy or manual techniques (in cases of joint stiffness) and time-contingent strategy have to be used in patients respectively with prevalent NP or CS mechanisms. In case of lack of improvement, the patient should be re-assessed or referred to the specialist If symptoms reduction procedures resulted negative  Patients with CS mechanisms: de-sensitization and without influence patient’s SP, the load will be managed load management are coupled with the therapeutic by exercise, according with prevalent mechanisms of approach. The clinical conditions with prevalent CS pain elaboration showed by the patient: features are managed by patient’s education and “time-contingent” exercise (exercises have to be per-  Patients with NP mechanisms: if the clinical formed for a certain time, agreed with the patient, condition is characterized by high level of reactivity, despite the presence of symptoms) in order to the therapeutic approach will be based on patient education [32], de-sensitization with rest, drugs Table 3 Examples of diagnostic/therapeutic procedures to (refer to medical management) or graded motor im- reduce the patient’s symptoms agery (GMI) [94]. In the more active treatment phase, once the reactivity is reduced, load man- Example of procedures for symptoms reduction agement is predominant and exercises are pro- posed with the adoption of a “symptoms- SSMP (thoracic kyphosis, humeral contingent” strategy (the presence of symptoms head procedures, scapular position) limits the performance of exercises), targeting the dysfunctional motor task (e.g. a program of exer- Mulligan’s techniques of cises that aim to load the impaired movement, mobilization with movement starting from pain-free, simple, with low resist- ance exercises toward more complex functional- Scapular assisted test and scapular tasks) (see Table 4) [32, 33, 57, 80, 93, 95]; repositioning test Manual or dry needling treatment of myofascial trigger points (mTrPs) Manual treatment of cervical and thoracic joints (mobilization/manipulation)

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 8 of 11 Table 4 Load strategies for specific and non-specific shoulder pain Load strategies Specific shoulder pain Non-specific shoulder pain Responsive Non responsive Non responsive (NP or CS) NP CS Aim: Increase movement/strength Aim: reduce pain, fear, increase Aim: reduce pain, increase Aim: reduce pain, fear, increase and flexibility. movement/strength. movement/strength. movement/strength. What to do: symptoms-contingent What to do: symptoms-contingent What to do: symptoms-contingent What to do: strategies; stretch, proprioceptive strategies; commute positive strategies; stretch and active exercises time-contingent strategies; graded and active exercises progressively procedures in exercises plus load progressively changing the position exposure/activity starting from the changing the position of execution the dysfunctional pattern of of execution and increasing number identification of painful tolerated and increasing number of patient’s movement increasing of repetition and the resistance exercise in terms of number of repetition and the resistance and modulating number according with patient’s pain. repetitions and pain granted during according with patient’s pain. of repetition and resistance the execution. according with patient’s pain. restructure the patient belief of association between our best efforts, several limitations affected the sug- pain-danger-harm (e.g. graded exposure/activity gested framework: 1) it is based on a discretionary ex- starting from the identification of painful tolerated pert opinion; 2) it is created without an international exercise in terms of number of repetitions and pain expert consensus methodology (e.g. Delphi study); 3) the granted during the execution) [42, 43, 70, 73, 82, 96– selection of relevant articles was based on narrative review 98]. Communication with patients (including informa- instead of a declared approach (e.g. systematic review); 4) tion, reassurance and education) could also help exer- its applicability, efficacy, validity and reliability has not cises and it plays an important role in the achievement been tested. Moreover, we actually could not classify dif- of this aim [99]. GMI and low intensity, aerobic/non- ferent profile of patients under the label of non-specific specific exercise also seems to be particularly useful in shoulder pain [108]. When it will be possible, as previ- this category of patients [100] (see Table 4). ously happened for low back pain, it could permit us to optimize diagnostic and therapeutic proposals. Concerning the prognosis, when a positive progressive improvement is obtained, an exercise training of Conclusions 12 weeks duration is recommended [33, 92]. Moreover, Existing literature underlines the limits of a strictly ana- the presence of a lower baseline pain and disability, a pa- tomical model for the evaluation of patients with SP. tient expectation of a ‘complete recovery’ as ‘a result of The integration of the alternative purposes in that clin- physiotherapy treatment’ in comparison to ‘slight im- ical framework could help to orientate physiotherapists provement’, a higher pain self-efficacy and lower pain se- towards a more bio-psychosocial and pragmatic ap- verity at rest enhances the likelihood to reduce pain and proach. In the future, the category of non-specific shoul- improve disability in SP [101]. Opposite, when capsular der pain and its peculiarities should be taken into stiffness is a predominant feature of the clinical scenario, account in diagnostic and prognostic research studies. a longer time is needed to fully restore functional move- ment [95]. The concomitance of higher level of depres- Abbreviations sion symptoms, catastrophizing thoughts, fear of CS: Central sensitization; CSI: Central sensitization inventory; DASH: The movement, fear of pain and anxiety were related to disabilities of the arm, shoulder and hand questionnaire; GMI: Graded motor higher disability, greater pain severity, lowest perceptions imagery; GP: General practitioner; MRA: Magnetic resonance arthrography; of clinical improvement and increased possibility of de- MRI: Magnetic resonance imaging; NP: Nociceptive pain; PSEQ: Pain self- veloping a pattern of CS in patients with SP [102–105]. efficacy questionnaire; QST: Quantitative sensory testing; ROM: Range of Finally, in the short/medium term, the expected results motion; SLAP: Superior labral tear from anterior to posterior; SP: Shoulder are not reached, the patient should be re-evaluated pain; SPADI: Shoulder pain and disability index; SSMP: Shoulder symptom re-exploring the different steps of the framework or re- modification procedure; STAR-shoulder: Staged approach for rehabilitation ferred to other specialists (e.g. orthopaedic) [33]. classification: shoulder disorders; US: Ultrasound; X-RAY: Radiographs Study limitations Acknowledgements In this paper, we adopted an evidence-based approach to The authors want to thank Samuele Graffiedi for his valuable advice during guide physiotherapists in the management of SP, thus the advancement of this manuscript. proposing a framework inspired by the bio-psychosocial model [106] and aligned to the International Classifica- Authors’ contributions tion of Functioning, Disability and Health [107]. Despite All authors conceived, designed, drafted and approved the final manuscript. Ethics approval and consent to participate Not applicable. Competing interests The authors declare that they have no competing interests.

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 9 of 11 Publisher’s Note 21. Lenza M, Buchbinder R, Takwoingi Y, Johnston RV, Hanchard NC, Faloppa F. Magnetic resonance imaging, magnetic resonance arthrography and Springer Nature remains neutral with regard to jurisdictional claims in ultrasonography for assessing rotator cuff tears in people with shoulder published maps and institutional affiliations. pain for whom surgery is being considered. Cochrane Database Syst Rev. 2013;(9):CD009020. Author details 1Via Veneto, 6, Subbiano, Arezzo, Italy. 2Via Paolo VI, Cologne, Brescia, Italy. 22. Kuhn JE, Dunn WR, Ma B, Wright RW, Jones G, Spencer EE, Wolf B, Safran M, 3Via de Gaspari, 9, Montecchio Maggiore, Vicenza, Italy. 4Via Italo Svevo, 2 Spindler KP, McCarty E, et al. Interobserver agreement in the classification of Codogno, Lodi, Italy. 5Via Eugenio Scalfaro, 17, Catanzaro, Italy. 6Via rotator cuff tears. Am J Sports Med. 2007;35:437–41. Magliotto, 2 17100, Savona, Italy. 7Department of Neuroscience, Rehabilitation, Ophtalmology, Genetics, Maternal and Child Health, University 23. Frost P, Andersen JH, Lundorf E. Is supraspinatus pathology as defined by of Genova, Campus of Savona, Savona, Italy. magnetic resonance imaging associated with clinical sign of shoulder impingement? J Shoulder Elb Surg. 1999;8:565–8. Received: 3 January 2018 Accepted: 23 May 2018 24. Handoll H, Hanchard N, Lenza M, Buchbinder R. Rotator cuff tears and References shoulder impingement: a tale of two diagnostic test accuracy reviews. 1. Kuijpers T, Van Tulder MW, Van der Heijden GJ, Bouter LM, Van der Windt Cochrane Database Syst Rev. 2013;(10):ED000068. DA. Costs of shoulder pain in primary care consulters: a prospective cohort 25. Rees JL. The pathogenesis and surgical treatment of tears of the rotator study in the Netherlands. BMC Musculoskelet Disord. 2006;7:83. cuff. J Bone Joint Surg Br. 2008;90:827–32. 2. Luime JJ, Koes BW, Hendriksen IJ, Burdorf A, Verhagen AP, Miedema HS, Verhaar JA. Prevalence and incidence of shoulder pain in the general 26. Girish G, Lobo LG, Jacobson JA, Morag Y, Miller B, Jamadar DA. Ultrasound population; a systematic review. Scand J Rheumatol. 2004;33:73–81. of the shoulder: asymptomatic findings in men. AJR Am J Roentgenol. 2011; 3. Parsons S, Breen A, Foster NE, Letley L, Pincus T, Vogel S, Underwood M. 197:W713–9. Prevalence and comparative troublesomeness by age of musculoskeletal pain in different body locations. Fam Pract. 2007;24:308–16. 27. Yamaguchi K, Ditsios K, Middleton WD, Hildebolt CF, Galatz LM, Teefey SA. 4. Breivik H, Collett B, Ventafridda V, Cohen R, Gallacher D. Survey of chronic The demographic and morphological features of rotator cuff disease. A pain in Europe: prevalence, impact on daily life, and treatment. Eur J Pain. comparison of asymptomatic and symptomatic shoulders. J Bone Joint Surg 2006;10:287–333. Am. 2006;88:1699–704. 5. Magee D. Orthopedic physical assessment. 6th ed. Edmonton: Elsevier; 2013. 6. Schellingerhout JM, Verhagen AP, Thomas S, Koes BW. Lack of uniformity in 28. Keener JD, Galatz LM, Teefey SA, Middleton WD, Steger-May K, Stobbs- diagnostic labeling of shoulder pain: time for a different approach. Man Cucchi G, Patton R, Yamaguchi K. A prospective evaluation of survivorship Ther. 2008;13:478–83. of asymptomatic degenerative rotator cuff tears. J Bone Joint Surg Am. 7. Littlewood C, Cools AMJ. Scapular dyskinesis and shoulder pain: the devil is 2015;97:89–98. in the detail. Br J Sports Med. 2018;52:72–3. 8. Lewis J. The end of an era? J Orthop Sports Phys Ther. 2018;48:127–9. 29. Chester R, Jerosch-Herold C, Lewis J, Shepstone L. Psychological factors 9. Buchbinder R, Goel V, Bombardier C, Hogg-Johnson S. Classification systems are associated with the outcome of physiotherapy for people with of soft tissue disorders of the neck and upper limb: do they satisfy shoulder pain: a multicentre longitudinal cohort study. Br J Sports Med. methodological guidelines? J Clin Epidemiol. 1996;49:141–9. 2018;52:269–75. 10. Hermans J, Luime JJ, Meuffels DE, Reijman M, Simel DL, Bierma-Zeinstra SM. Does this patient with shoulder pain have rotator cuff disease?: the rational 30. Wylie JD, Suter T, Potter MQ, Granger EK, Tashjian RZ. Mental health has a stronger clinical examination systematic review. JAMA. 2013;310:837–47. association with patient-reported shoulder pain and function than tear size in 11. Hanchard NC, Lenza M, Handoll HH, Takwoingi Y. Physical tests for shoulder patients with full-thickness rotator cuff tears. J Bone Joint Surg Am. 2016;98:251–6. impingements and local lesions of bursa, tendon or labrum that may accompany impingement. Cochrane Database Syst Rev. 2013;(4):CD007427. 31. Lewis JS. Rotator cuff tendinopathy/subacromial impingement syndrome: is 12. Green R, Shanley K, Taylor NF, Perrott M. The anatomical basis for clinical it time for a new method of assessment? Br J Sports Med. 2009;43:259–64. tests assessing musculoskeletal function of the shoulder. Phys Ther Rev. 2013;13:17–24. 32. McClure PW, Michener LA. Staged approach for rehabilitation classification: 13. Hawkins RJ, Kennedy JC. Impingement syndrome in athletes. Am J Sports shoulder disorders (STAR-shoulder). Phys Ther. 2015;95:791–800. Med. 1980;8:151–8. 14. Roberts CS, Davila JN, Hushek SG, Tillett ED, Corrigan TM. Magnetic 33. Klintberg IH, Cools AM, Holmgren TM, Holzhausen AC, Johansson K, resonance imaging analysis of the subacromial space in the impingement Maenhout AG, Moser JS, Spunton V, Ginn K. Consensus for physiotherapy sign positions. J Shoulder Elb Surg. 2002;11:595–9. for shoulder pain. Int Orthop. 2015;39:715–20. 15. Yamamoto N, Muraki T, Sperling JW, Steinmann SP, Itoi E, Cofield RH, An KN. Impingement mechanisms of the Neer and Hawkins signs. J Shoulder 34. Lewis JS, McCreesh K, Barratt E, Hegedus EJ, Sim J. Inter-rater reliability of Elb Surg. 2009;18:942–7. the shoulder symptom modification procedure in people with shoulder 16. Tucker S, Taylor NF, Green RA. Anatomical validity of the Hawkins-Kennedy pain. BMJ Open Sport Exerc Med. 2016;2:e000181. test–a pilot study. Man Ther. 2011;16:399–402. 17. Hegedus EJ, Cook C, Lewis J, Wright A, Park JY. Combining orthopedic 35. Jull G, Moore A, Falla D, Lewis J, McCarthy C, Sterling M. Grieve's modern special tests to improve diagnosis of shoulder pathology. Phys Ther Sport. musculoskeletal physiotherapy. 4th ed. Australia, United Kingdom, Germany: 2015;16:87–92. Elsevier; 2015. 18. Hegedus EJ, Goode A, Campbell S, Morin A, Tamaddoni M, Moorman CT, Cook C. Physical examination tests of the shoulder: a systematic review with 36. Bot SD, Terwee CB, Van der Windt DA, Bouter LM, Dekker J, De Vet HC. meta-analysis of individual tests. Br J Sports Med. 2008;42:80–92. Clinimetric evaluation of shoulder disability questionnaires: a systematic 19. Hegedus EJ, Goode AP, Cook CE, Michener L, Myer CA, Myer DM, Wright review of the literature. Ann Rheum Dis. 2004;63:335–41. AA. Which physical examination tests provide clinicians with the most value when examining the shoulder? Update of a systematic review with meta- 37. Martinez-Calderon J, Zamora-Campos C, Navarro-Ledesma S, Luque-Suarez analysis of individual tests. Br J Sports Med. 2012;46:964–78. A. The role of self-efficacy on the prognosis of chronic musculoskeletal pain: 20. May S, Chance-Larsen K, Littlewood C, Lomas D, Saad M. Reliability of a systematic review. J Pain. 2018;19:10–34. physical examination tests used in the assessment of patients with shoulder problems: a systematic review. Physiotherapy. 2010;96:179–90. 38. Nicholas MK. The pain self-efficacy questionnaire: taking pain into account. Eur J Pain. 2007;11:153–63. 39. Smart KM, Blake C, Staines A, Doody C. The discriminative validity of “nociceptive” “peripheral neuropathic” and “central sensitization” as mechanisms-based classifications of musculoskeletal pain. Clin J Pain. 2011;27:655–63. 40. Noten S, Struyf F, Lluch E, D'Hoore M, Van Looveren E, Meeus M. Central pain processing in patients with shoulder pain: a review of the literature. Pain Pract. 2017;17:267–80. 41. Sanchis MN, Lluch E, Nijs J, Struyf F, Kangasperko M. The role of central sensitization in shoulder pain: a systematic literature review. Semin Arthritis Rheum. 2015;44:710–6. 42. Nijs J, Torres-Cueco R, Van Wilgen CP, Girbes EL, Struyf F, Roussel N, Van Oosterwijck J, Daenen L, Kuppens K, Vanwerweeen L, et al. Applying modern pain neuroscience in clinical practice: criteria for the classification of central sensitization pain. Pain Physician. 2014;17:447–57. 43. Nijs J, Lluch Girbes E, Lundberg M, Malfliet A, Sterling M. Exercise therapy for chronic musculoskeletal pain: innovation by altering pain memories. Man Ther. 2015;20:216–20.

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 10 of 11 44. Scerbo T, Colasurdo J, Dunn S, Unger J, Nijs J, Cook C. Measurement 68. Chang D, Mohana-Borges A, Borso M, Chung CB. SLAP lesions: anatomy, properties of the central sensitization inventory: a systematic review. Pain clinical presentation, MR imaging diagnosis and characterization. Eur J Pract. 2018;18:544–54. Radiol. 2008;68:72–87. 45. Menendez ME, Baker DK, Oladeji LO, Fryberger CT, McGwin G, Ponce BA. 69. Peek AL, Miller C, Heneghan NR. Thoracic manual therapy in the Psychological distress is associated with greater perceived disability and pain in management of non-specific shoulder pain: a systematic review. J Man patients presenting to a shoulder clinic. J Bone Joint Surg Am. 2015;97:1999–2003. Manip Ther. 2015;23:176–87. 46. Rossettini G, Carlino E, Testa M. Clinical relevance of contextual factors as 70. Nijs J. Paul van Wilgen C, van Oosterwijck J, van Ittersum M, Meeus M. How triggers of placebo and nocebo effects in musculoskeletal pain. BMC to explain central sensitization to patients with ‘unexplained‘ chronic Musculoskelet Disord. 2018;19:27. musculoskeletal pain: practice guidelines. Man Ther. 2011;16:413–8. 47. Testa M, Rossettini G. Enhance placebo, avoid nocebo: how contextual 71. Bishop MD, Bialosky JE, Cleland JA. Patient expectations of benefit from factors affect physiotherapy outcomes. Man Ther. 2016;24:65–74. common interventions for low back pain and effects on outcome: secondary analysis of a clinical trial of manual therapy interventions. J Man 48. Cuff A, Littlewood C. Subacromial impingement syndrome - what does this Manip Ther. 2011;19:20–5. mean to and for the patient? A qualitative study. Musculoskelet Sci Pract. 2018;33:24–8. 72. Miciak M, Mayan M, Brown C, Joyce AS, Gross DP. The necessary conditions of engagement for the therapeutic relationship in physiotherapy: an 49. Bonde JP, Mikkelsen S, Andersen JH, Fallentin N, Baelum J, Svendsen SW, interpretive description study. Arch Physiother. 2018;8:3. Thomsen JF, Frost P, Thomsen G, Overgaard E, et al. Prognosis of shoulder tendonitis in repetitive work: a follow up study in a cohort of Danish 73. Vlaeyen J, Morley S, Linton S, Boersma K, De Jong J. Pain-related fear: industrial and service workers. Occup Environ Med. 2003;60:E8. exposure-based treatment for chronic pain. Washington, D.C.: IASP Press; 2012. 50. Hopman K, Krahe L, Lukersmith S, McColl A, Vine K. Clinical practice 74. Bialosky JE, Bishop MD, Price DD, Robinson ME, George SZ. The mechanisms guidelines for the Management of Rotator cuff syndrome in the workplace. of manual therapy in the treatment of musculoskeletal pain: a 2013. https://rcs.med.unsw.edu.au/sites/default/files/rcs/page/ comprehensive model. Man Ther. 2009;14:531–8. RotatorCuffSyndromeGuidelines.pdf. Accessed 21 May 2018. 75. Zusman M. The modernisation of manipulative therapy. Int J Clin Med. 51. Littlewood C, Ashton J, Chance-Larsen K, May S, Sturrock B. Exercise for rotator 2011;2(5):644–9. cuff tendinopathy: a systematic review. Physiotherapy. 2012;98:101–9. 76. Bishop MD, Torres-Cueco R, Gay CW, Lluch-Girbes E, Beneciuk JM, Bialosky 52. Rolke R, Baron R, Maier C, Tolle TR, Treede RD, Beyer A, Binder A, Birbaumer JE. What effect can manual therapy have on a patient's pain experience? N, Birklein F, Botefur IC, et al. Quantitative sensory testing in the German Pain Manag. 2015;5:455–64. research network on neuropathic pain (DFNS): standardized protocol and reference values. Pain. 2006;123:231–43. 77. Braun C, Bularczyk M, Heintsch J, Hanchard NCA. Manual therapy and exercises for shoulder impingement revisited. Phys Ther Rev. 2013;18:263–84. 53. Jull G. Discord between approaches to spinal and extremity disorders: is it logical? J Orthop Sports Phys Ther. 2016;46:938–41. 78. Borstad J, Woeste C. The role of sensitization in musculoskeletal shoulder pain. Braz J Phys Ther. 2015;19:251–7. 54. McCarthy CJ, Arnall FA, Strimpakos N, Freemont A, Oldham JA. The biopsychosocial classification of non-specific low back pain: a systematic 79. Littlewood C, Bateman M, Brown K, Bury J, Mawson S, May S, Walters SJ. A review. Phys Ther Rev. 2004;9:17–30. self-managed single exercise programme versus usual physiotherapy treatment for rotator cuff tendinopathy: a randomised controlled trial (the 55. Koes BW, Van Tulder M, Lin CW, Macedo LG, McAuley J, Maher C. An SELF study). Clin Rehabil. 2016;30:686–96. updated overview of clinical guidelines for the management of non-specific low back pain in primary care. Eur Spine J. 2010;19:2075–94. 80. Cook JL, Purdam CR. Is tendon pathology a continuum? A pathology model to explain the clinical presentation of load-induced tendinopathy. Br J 56. Pillastrini P, Gardenghi I, Bonetti F, Capra F, Guccione A, Mugnai R, Violante Sports Med. 2009;43:409–16. FS. An updated overview of clinical guidelines for chronic low back pain management in primary care. Joint Bone Spine. 2012;79:176–85. 81. Rio E, Moseley L, Purdam C, Samiric T, Kidgell D, Pearce AJ, Jaberzadeh S, Cook J. The pain of tendinopathy: physiological or pathophysiological? 57. Lewis J. Rotator cuff related shoulder pain: assessment, management and Sports Med. 2014;44:9–23. uncertainties. Man Ther. 2016;23:57–68. 82. Littlewood C, Malliaras P, Bateman M, Stace R, May S, Walters S. The central 58. Warby SA, Pizzari T, Ford JJ, Hahne AJ, Watson L. The effect of exercise- nervous system–an additional consideration in ‘rotator cuff tendinopathy‘ based management for multidirectional instability of the glenohumeral and a potential basis for understanding response to loaded therapeutic joint: a systematic review. J Shoulder Elb Surg. 2014;23:128–42. exercise. Man Ther. 2013;18:468–72. 59. Goodman C, Snyder T. Differential diagnosis for physical therapists, 83. Van Ark M, Cook JL, Docking SI, Zwerver J, Gaida JE, Van den Akker-Scheek I, screening for referral. 5th ed. United States: Elsevier; 2012. Rio E. Do isometric and isotonic exercise programs reduce pain in athletes with patellar tendinopathy in-season? A randomised clinical trial. J Sci Med 60. Goodman C. Screening for medical problems in patients with upper Sport. 2016;19:702–6. extremity signs and symptoms. J Hand Ther. 2010;23:105–25. 84. Bron C, De Gast A, Dommerholt J, Stegenga B, Wensing M, Oostendorp RA. 61. Obuchowski NA, Graham RJ, Baker ME, Powell KA. Ten criteria for effective Treatment of myofascial trigger points in patients with chronic shoulder screening: their application to multislice CT screening for pulmonary and pain: a randomized, controlled trial. BMC Med. 2011;9:8. colorectal cancers. AJR Am J Roentgenol. 2001;176:1357–62. 85. Hing W, Hall T, Rivett D, Vicenzino B, Mulligan B. The mulligan concept of 62. Sizer PS, Brismee JM, Cook C. Medical screening for red flags in the diagnosis manual therapy. 1st ed. Chatswood: Churchill Livingstone; 2014. and management of musculoskeletal spine pain. Pain Pract. 2007;7:53–71. 86. Wassinger CA, Rich D, Cameron N, Clark S, Davenport S, Lingelbach M, 63. Oh JH, Lee YH, Kim SH, Park JS, Seo HJ, Kim W, Park HB. Comparison of Smith A, Baxter GD, Davidson J. Cervical & thoracic manipulations: acute treatments for superior labrum-biceps complex lesions with concomitant effects upon pain pressure threshold and self-reported pain in rotator cuff repair: a prospective, randomized, comparative analysis of experimentally induced shoulder pain. Man Ther. 2016;21:227–32. debridement, biceps Tenotomy, and biceps Tenodesis. Arthroscopy. 2016; 32:958–67. 87. Kibler WB, Ludewig PM, McClure PW, Michener LA, Bak K, Sciascia AD. Clinical implications of scapular dyskinesis in shoulder injury: the 2013 consensus 64. Kim YS, Lee HJ, Kim JH, Noh DY. When should we repair partial-thickness statement from the ‘Scapular Summit‘. Br J Sports Med. 2013;47:877–85. rotator cuff tears? Outcome comparison between immediate surgical repair versus delayed repair after 6-month period of nonsurgical treatment. Am J 88. Ong J, Claydon LS. The effect of dry needling for myofascial trigger points Sports Med. 2018;46:1091–6. in the neck and shoulders: a systematic review and meta-analysis. J Bodyw Mov Ther. 2014;18:390–8. 65. Litaker D, Pioro M, El Bilbeisi H, Brems J. Returning to the bedside: using the history and physical examination to identify rotator cuff tears. J Am Geriatr 89. Minkalis AL, Vining RD, Long CR, Hawk C, De Luca K. A systematic review of thrust Soc. 2000;48:1633–7. manipulation for non-surgical shoulder conditions. Chiropr Man Therap. 2017;25:1. 66. Collin P, Matsumura N, Ladermann A, Denard PJ, Walch G. Relationship 90. Cook C, Lawrence J, Michalak K, Dhiraprasiddhi S, Donaldson M, Petersen S, between massive chronic rotator cuff tear pattern and loss of active Learman K. Is there preliminary value to a within- and/or between-session shoulder range of motion. J Shoulder Elb Surg. 2014;23:1195–202. change for determining short-term outcomes of manual therapy on mechanical neck pain? J Man Manip Ther. 2014;22:173–80. 67. Merolla G, Cerciello S, Chillemi C, Paladini P, De Santis E, Porcellini G. Multidirectional instability of the shoulder: biomechanics, clinical presentation, 91. Cook CE, Showalter C, Kabbaz V, O'Halloran B. Can a within/between- and treatment strategies. Eur J Orthop Surg Traumatol. 2015;25:975–85. session change in pain during reassessment predict outcome using a

Ristori et al. Archives of Physiotherapy (2018) 8:7 Page 11 of 11 manual therapy intervention in patients with mechanical low back pain? Man Ther. 2012;17:325–9. 92. Littlewood C, Malliaras P, Chance-Larsen K. Therapeutic exercise for rotator cuff tendinopathy: a systematic review of contextual factors and prescription parameters. Int J Rehabil Res. 2015;38:95–106. 93. Abdulla SY, Southerst D, Cote P, Shearer HM, Sutton D, Randhawa K, Varatharajan S, Wong JJ, Yu H, Marchand AA, et al. Is exercise effective for the management of subacromial impingement syndrome and other soft tissue injuries of the shoulder? A systematic review by the Ontario protocol for traffic injury management (OPTIMa) collaboration. Man Ther. 2015;20:646–56. 94. Hoyek N, Di Rienzo F, Collet C, Hoyek F, Guillot A. The therapeutic role of motor imagery on the functional rehabilitation of a stage II shoulder impingement syndrome. Disabil Rehabil. 2014;36:1113–9. 95. Page MJ, Green S, Kramer S, Johnston RV, McBain B, Chau M, Buchbinder R. Manual therapy and exercise for adhesive capsulitis (frozen shoulder). Cochrane Database Syst Rev. 2014;(8):CD011275. 96. Butler D, Moseley L. Explain pain. Adelaide: Noigroup Publications; 2013. 97. Nijs J, Meeus M, Cagnie B, Roussel NA, Dolphens M, Van Oosterwijck J, Danneels L. A modern neuroscience approach to chronic spinal pain: combining pain neuroscience education with cognition-targeted motor control training. Phys Ther. 2014;94:730–8. 98. Leeuw M, Goossens ME, Van Breukelen GJ, De Jong JR, Heuts PH, Smeets RJ, Koke AJ, Vlaeyen JW. Exposure in vivo versus operant graded activity in chronic low back pain patients: results of a randomized controlled trial. Pain. 2008;138:192–207. 99. Miller J, MacDermid JC, Richardson J, Walton DM, Gross A. Depicting individual responses to physical therapist led chronic pain self-management support with pain science education and exercise in primary health care: multiple case studies. Arch Physiother. 2017;7:4. 100. Daenen L, Varkey E, Kellmann M, Nijs J. Exercise, not to exercise, or how to exercise in patients with chronic pain? Applying science to practice. Clin J Pain. 2015;31:108–14. 101. O'Malley KJ, Roddey TS, Gartsman GM, Cook KF. Outcome expectancies, functional outcomes, and expectancy fulfillment for patients with shoulder problems. Med Care. 2004;42:139–46. 102. Clark J, Nijs J, Yeowell G, Goodwin PC. What are the predictors of altered central pain modulation in chronic musculoskeletal pain populations? A Systematic Review Pain Physician. 2017;20:487–500. 103. George SZ, Stryker SE. Fear-avoidance beliefs and clinical outcomes for patients seeking outpatient physical therapy for musculoskeletal pain conditions. J Orthop Sports Phys Ther. 2011;41:249–59. 104. Wolfensberger A, Vuistiner P, Konzelmann M, Plomb-Holmes C, Leger B, Luthi F. Clinician and patient-reported outcomes are associated with psychological factors in patients with chronic shoulder pain. Clin Orthop Relat Res. 2016;474:2030–9. 105. Das De S, Vranceanu AM, Ring DC. Contribution of kinesophobia and catastrophic thinking to upper-extremity-specific disability. J Bone Joint Surg Am. 2013;95:76–81. 106. Karjalainen K, Malmivaara A, Van Tulder M, Roine R, Jauhiainen M, Hurri H, Koes B. Multidisciplinary biopsychosocial rehabilitation for neck and shoulder pain among working age adults. Cochrane Database Syst Rev. 2003;(2):CD002194. 107. Roe Y, Soberg HL, Bautz-Holter E, Ostensjo S. A systematic review of measures of shoulder pain and functioning using the International classification of functioning, disability and health (ICF). BMC Musculoskelet Disord. 2013;14:73. 108. Rabey M, Beales D, Slater H, O'Sullivan P. Multidimensional pain profiles in four cases of chronic non-specific axial low back pain: an examination of the limitations of contemporary classification systems. Man Ther. 2015;20:138–47.

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 https://doi.org/10.1186/s40945-018-0049-9 RESEARCH ARTICLE Open Access Ethics reporting practices in randomized controlled trials of physical therapy interventions after stroke Francesco Ferrarello1* , Matteo Viligiardi2 and Mauro Di Bari3 Abstract Background: Adequate reporting of ethics-related research methods promotes convergence on best ethics practices. In physical therapy, studies on ethics reporting are limited to few aspects, and none focuses on stroke research. Our objectives were to investigate the reporting of multiple ethics-related features and its variation over time, and the characteristics of the studies associated with ethics reporting in Randomized Controlled Trials (RCTs) of physical therapy interventions after stroke. Methods: A random sample of RCTs published in the years 2004, 2009 and 2014, was extracted from the PubMed database, regardless of the publishing journal. For each trial we investigated year of publication, trial registration, sample size, stroke subtype, phase of the disease, setting, interventions and dosing, outcome measures, outcome of the study, PEDro score and 5-year impact factor of the publishing journal. Reporting of ethics-related issues was analyzed. Differences between groups were examined. Multiple regression was used to evaluate the relationship between ethics-related issues reporting and some studies’ characteristics. Results: Eighty studies were reviewed. Groups differed in the proportion of registered trials (p = .009), 5-year impact factor (p = .011), assessment of cognitive capacity (p = .049), declaration about conflict of interest (p < .001), and number of ethics-related issues reported (p = .009). The proportion of issues reported ranged from 92.5% (consent obtaining) to 0% (eventual follow up care). Post-hoc comparisons showed significantly greater reporting of ethics issues in trials published in the year 2014 compared to 2004 (p = .014, 95%CI = 0.40/4.26). Year of publication and PEDro score were significant predictors of adequate reporting. Conclusions: Authors, editors, and reviewers should be more rigorous and demanding about the reporting of ethic-related methods in randomized controlled trials of physical therapy interventions after stroke. Keywords: Ethics reporting, Physical therapy, Randomized controlled trials, Stroke Background Randomized controlled trial (RCT), systematic review The importance of ethics in clinical research has been and meta-analysis of RCTs are considered reliable re- well-established for many decades [1]. A growing atten- search designs, able to evaluate the effectiveness of an tion is given to determine the best ethical practices for intervention [1]. The design of a clinical trial implies ad- conducting observational and experimental studies [2]. herence to challenging and multifaceted ethical and Methodological quality, approval by a research ethics methodological standards that must integrate each other committee, and obtaining informed consent from partic- [4]; Ashton et al. in their taxonomy indeed identified five ipants are the main ethical issues in research with hu- major categories, with over 5900 possible standards [4]. man beings [3]. Differences among research topics and study designs may add further variability in ethical requirements * Correspondence: [email protected] [5–8]. Research conducted with little methodological 1Functional Rehabilitation, Azienda USL Toscana Centro, Via Cavour 118/120, rigor does not lead to knowledge or benefit, and ex- 59100 Prato, Italy poses participants to unnecessary burden or harm [9]. Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 2 of 10 Thus methodological and ethical requirements of a recommendations [21]. Based on a sample of RCTs of study have a symbiotic relationship [10] and, in order to physical therapy interventions after stroke, our objectives make readers able to fully evaluate clinical research, both were to investigate the reporting of multiple ethics-related should be appropriately reported [3]. Adequate reporting features and its variation over time, as well as the charac- of research findings may promote the implementation of teristics of the studies associated with ethics reporting. evidence-based clinical practice in many fields, including physical therapy [11]. However, reporting of scientific A random sample of RCTs indexed in PubMed was methods receives considerably greater attention, as com- extracted, regardless of the publication journal. To be pared with reporting of research ethics issues [12]. Many included, the studies had to be published in English in guidelines have been published to improve the reporting the years 2004, 2009 and 2014, involve a parallel-group of RCTs, and adherence to these guidelines is usually design, and evaluate an experimental physical therapy considered necessary for publication [12]. Recommenda- intervention administered to adult (age ≥ 18 years) indi- tions on how ethics issues should be reported in re- viduals with stroke. search studies are also available [13], although evidence suggests flaws in publication requirements and reporting The rationale for choosing the three years was the fol- of ethical protections [14]. Thus, descriptions provided lowing. The year 2014 had just ended when we drafted by journal articles contain little information about the protocol; it was chosen to give a contemporary view. research ethics methods [12]. To have a time perspective, we chose the years 2004 and 2009 as comparators, thus covering a period of ten Research reproducibility refers to the possibility to du- years. plicate the results of prior studies, and is based on a clear and comprehensive description of study design [15]. The Interventions were considered suitable if included in the concept of ethical reproducibility was recently introduced. classification proposed by De Jong et al. [22], or listed in the Ethical reproducibility prescribes thorough reporting and book “Guide to Physical Therapist Practice”[23], or reviewed critical evaluation of the ethics methods employed in in the Clinician’s Handbooks of the Evidence-Based Review biomedical research [16]. Although ethics committee of Stroke Rehabilitation-motor rehabilitation [24]. approval is a crucial aspect of ethics in research and is usually considered as a proxy measure for the fulfillment For the selection of the studies, the term “stroke” of all ethical requirements in research [14], it is not suffi- followed by “physical therapy” was entered in the search cient to judge the overall ethical quality of an RCT [17]. box, filtered by the type of study (RCT). Additional fil- Including ethics-related methods in research reports may ters related to the year of publication (2004, 2009 and address concerns raised by researchers, clinicians, and 2014) were then added one at a time to obtain a list of other stakeholders [2]. Moreover, ethical reproducibility studies for each of the years of interest. The resulting can promote benefits such as learning, inner reflection, search strategy was the following: ((“stroke”[MeSH increase of ethical responsibility, critical research assess- Terms] OR “stroke”[All Fields]) AND (“physical therapy ment, and use of better ethical practices [12]. modalities”[MeSH Terms] OR (“physical”[All Fields] AND “therapy”[All Fields] AND “modalities”[All Fields]) OR In the physical therapy field, interest on ethical issues “physical therapy modalities”[All Fields] OR (“physical”[All of the profession is remarkable, yet it mainly focuses on Fields] AND “therapy”[All Fields]) OR “physical thera- clinical practice, whereas only few papers have been py”[All Fields])) AND Randomized Controlled Trial[ptyp] published over the years addressing ethics in physical AND ((“2004/01/01”[PDAT]: “2004/12/31”[PDAT]) OR therapy research [18]. Studies on issues related to ethics (“2009/01/01”[PDAT]: “2009/12/31”[PDAT]) OR (“2014/ in RCTs reporting were limited to ethics committee ap- 01/01”[PDAT]: “2014/12/31”[PDAT])). proval, informed consent, and confidentiality [14, 19]. In each year, a random sample of the studies in the Although stroke represents a major topic in physical PubMed-generated list was extracted. A formal sample therapy research [20], none of the studies published on size calculation was not performed because of lack of ethical aspects of physical therapy research focused on preliminary data; indeed, previous studies differed from stroke. Therefore, because available knowledge does not the present one in eligibility criteria, sources and appear to be exhaustive and satisfactory, we designed methods of studies selection, and variables of interest. In the present study with the aim to investigate ethics the absence of references, to quantify the sample size we reporting characteristics in RCTs of physical therapy were inspired by Pinto et al. research on RCTs registra- interventions after stroke. tion,[25] and we established 20 % of the studies as the target sample. We performed calculations based on the Methods total number of articles present in the lists, before the The reporting of this study conforms to the Strengthening screening process. The random selection was performed the Reporting of Observational studies in Epidemiology in August 2015, by entering the sequence numbers of the lists acquired through search in an on-line randomization program [26]; the studies were then

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 3 of 10 assessed for eligibility following the randomized sequence,  incentives or compensation for participants until the desired sample size was reached. Full texts of the  details about incentives or compensation given to articles were retrieved following a flow-chart (Fig. 1). participants Reasons for exclusion were given for each study ex-  funders cluded. Two reviewers (MV; FF) independently evaluated  details about funders the studies selected for final inclusion; disagreement was  potential conflicts of interest resolved by consensus.  details about conflicts of interest  steps taken to assess if eligible individuals were able Relevant data were extracted using a standard data recording spreadsheet, including characteristics of the to provide informed consent (e.g. use of validated studies and research ethics-related issues. screening tools for cognitive ability)  measures taken to ensure the best interests of Data related to year of publication, clinical trial regis- participants with reduced competence tration, sample size, stroke subtype, phase of the disease,  steps taken to ensure that the sample size was setting, interventions administered, dosing of the inter- sufficient to achieve adequate statistical power ventions, outcome measures, and outcome of the study  appropriateness of comparators were extracted from each article included. Information  justification for the eventual use of placebo about 5-year impact factor of the publishing journal was  potential harm for participants acquired [27].  plans for collecting, assessing, reporting, and managing adverse events and other unintended In the absence of a standardized comprehensive list, effects of trial interventions or trial conduct issues related to ethics in RCTs reporting were chosen  appropriate follow up care considering what has been contemplated in previous  steps taken to prevent unauthorized access to publications [2, 14, 17, 28]. The following aspects were personal and clinical data (confidentiality) investigated in each article:  accordance with the Helsinki Declaration.  ethics review committee or institutional review We assumed the definition of appropriateness of com- board study approval parator proposed by Caprino and Russo [29]; in the pres- ence of multiple interventions groups, we considered the  details about the ethics committees  obtainment of informed consent  details about the consent process Fig. 1 Full text retrieving flow chart

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 4 of 10 most active comparator intervention. For each study in- based on the year of publication were examined using cluded, the number of ethics-related issues reported was the Kruskal-Wallis or the one-way ANOVA tests for calculated by summing up any reported or mentioned interval variables. Pearson χ2 test was used for nominal issue. The maximum achievable was 22 issues. The and ordinal variables, except when cell counts were extended list of questions is detailed in the Appendix. < 5, in which case Fisher exact test was used. Linear Data were extracted from an investigator (MV) and subse- trends (across years) were considered as appropriate. In quently double-checked (FF). Any disagreement was re- case of significant results, post-hoc tests were performed solved by consensus. The PEDro score of methodological to explore differences between any two pairs of years. quality was verified for each study [30]. This tool is reli- Multiple linear regression was performed to assess the able [31] and useful for assessing the quality of studies in ability of year of publication, trial registration, PEDro stroke rehabilitation [30]. Criteria for quality assessment score, and 5-year impact factor to predict the number of are represented by randomness and concealment of ethics-related issues reported in each article selected. allocation, baseline comparability between groups, blind- Preliminary analysis was performed to ensure there was ing of participants, therapists, and assessors, adequacy no violation of the assumption of normality, linearity, of follow-up assessments, intention-to-treat analysis, and multicollinearity. between-group comparisons, reporting of point esti- mates and variability. The PEDro score ranges from 0 Analyses were performed using IBM SPSS Statistics (poor quality) to 10 (excellent quality); a score of 6 is for Windows (version 20.0; IBM Corp, Armonk, NY). conventionally considered as a threshold to identify The significance level was set at a p value of <.05. high-quality studies [32]. Results Statistical analysis The search on PubMed generated lists of 57, 119, and Descriptive statistics was initially performed. Five-year 177 articles (total 353) for the years 2004, 2009, and impact factor equal to zero were attributed to the three 2014, respectively. The target sample size was set at 11 journals not found in the Journal Citation Reports data- studies for 2004, 24 studies for 2009, and 35 studies for base [27]. In one case where only the impact factor was 2014. After randomization, a total of 162 papers were available, the same value was assigned as 5-year impact retrieved; the selection process led to the inclusion in factor. Expectation-maximization imputation was used the study of 80 articles (Fig. 2). to address missing values (2.5 and 5% respectively) of the number of sessions per week and the minutes per Study characteristics session needed to calculate the dosing of the interven- Studies’ characteristics are presented in Table 1. Eighteen tion in hours. Shapiro-Wilk’s test was used to test the percent of the selected trials had been registered. The normality of distribution. Differences between groups proportion of registered trials increased substantially across the three years; post-hoc comparison showed that Fig. 2 Studies selection flow chart. /*Including protocols, observational studies, preliminary reporting (e.g. recruitment or sample characteristics), non-randomized trials, and within-subject or cross-over design studies. PT = Physical Therapy. RCT = Randomized Controlled Trial

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 5 of 10 Table 1 Characteristics of the studies Year 2004 2009 2014 p All studies (n = 13) (n = 28) (n = 39) .011b (n = 80) 5-year Impact factora 2.967 (2.447/4.625) 4.206 (2.455/4.626) 2.179 (1.647/3.503) .009d 2.784 (2.130/4.626) Trial registeredc 2 (7.1) 12 (30.8) .212b 14 (17.5) Sample sizea 0 (0.0) 39 (31.5/65.0) 34 (23.0/52.5) .418d 38 (23.5/56.5) Intervention groupsc 28 (20.0/47.0) 21 (75.0) 34 (87.2) .305d 66 (82.5) 7 (25.0) 5 (12.8) 14 (17.5) 2 groups 11 (84.6) .516d 8 (28.6) 5 (12.8) .807d 14 (17.5) ≥ 3 groups 2 (15.4) 0 (0.0) 1 (2.6) .139b 1 (1.2) Etiologyc 16 (57.1) 21 (53.8) .855d 44 (55.0) 4 (14.3) 12 (30.8) .614d 21 (26.2) Ischemic 1 (7.7) .705d 15 (53.6) 14 (35.9) .073b 33 (41.2) Hemorrhagic 0 (0.0) 12 (42.9) 23 (59.0) 43 (53.8) 1 (3.6) 2 (5.1) 4 (5.0) Mixed 7 (53.8) 5 (17.9) 9 (23.1) 16 (20.0) 15.5 (11.0/27.0) 24.0 (10.0/34.7) 18.0 (9.0/30.0) None declared 5 (38.5) Phase of the disease-onsetc 20 (71.4) 25 (64.1) 54 (67.5) 25 (89.3) 36 (92.3) 72 (90.0) < 6 months 4 (30,8) 1 (3.6) 3 (7.7) 5 (6.2) 6 (5.0/7.0) 6 (5.0/7.5) 6 (5.0/7.0) ≥ 6 months 8 (61.5) None declared 1 (7.7) Placebo-sham comparatorc 2 (15.4) Intervention dosing-hoursae 18.0 (6.0/22.5) At least one outcomec Favorable 9 (69.2) Nonsignificant 11 (84.6) Unfavorable 1 (7.7) PEDro scorea 6 (5.0/6.0) a median (1th quartile/3rd quartile) b Kruskal-Wallis test c absolute frequencies (percentages) d Fisher exact test e after Expectation-maximization imputation the probability to be not registered was greater for trials balance (13 studies), walking (9 studies), arm-hand activ- published in the years 2004 and 2009 compared to those ities (26 studies), and activities of daily living (24 studies) published in the year 2014 (Table 2). of the International Classification of Functioning domains. Median 5-year impact factor was 2.784 (1th quartile Ethics reporting 2.130, 3rd quartile 4.626). Five-year impact factor was Ethics committee or institutional review board study greater in 2009 than in the other two years, with no approval was reported in 65 (81.2%) studies, and 74 clear linear-by-linear change; post-hoc comparison (92.5%) articles mentioned that consent was obtained. indicated that it was significantly greater in RCTs Details about the ethics committee and the consent published in the year 2009 compared to 2014 process were available in half of the cases (51.2 and (Table 2). PEDro score ranged from 3 to 8 (median 52.5%, respectively) (Table 3). One paper failed to report 6); there were no significant differences between years both ethics committee approval and obtainment of of publication (Table 1). informed consent (Fig. 3). Neuromuscolar interventions were most frequently The Mini-Mental State Examination or similar test applied (53 studies), followed by musculoskeletal inter- aimed to assess cognitive capacity were administered in ventions (20 studies). Standard care was frequently used the inclusion stage in 41 (51.2%) publications, namely as side intervention, or comparator (32 studies). The in 3 (23.1%), 18 (64.3%), and 20 (51.3%) studies in year outcome measures used to investigate the efficacy of the 2004, 2009, and 2014, respectively; this difference interventions mainly belonged to muscle and movement across the three years was borderline significant but functions (21 studies), gait pattern functions (17 studies),

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 6 of 10 Table 2 Post-hoc test results published in the years 2004 and 2009 compared to those published in the year 2014 (Table 2). Variable 95% CI p Trial registrationa Statements about sample size estimation were identi- 1.000b fied in 30 (37.5%) articles, 27 of which performed a 2004 vs 2009 1.077 .972 to 1.193 .024b power calculation. We considered appropriate the com- .031b parator interventions in 68 (85.0%) cases; participants 2004 vs 2014 1.444 1.172 to 1.781 received the same amount of attention in 60 (75.0%) .480 trials, whereas in the remaining ones the attention was 2009 vs 2014 1.341 1.062 to 1.693 .219 minor, or the control inactive. Justification for the use of 5-year impact factorc .002 placebo was always reported. Out of the 16 RCTs reporting the use of placebo/sham controls, the experimental inter- 2004 vs 2009 −.264 −1.659 to .629 .020b vention was represented by electrical stimulation in 10, by .110b acupuncture in 2, by a combination of acupuncture and 2004 vs 2014 .745 −.528 to 1.399 .289d electrical stimulation in 2, by action observation train- ing in 1, and by respiratory muscles training in 1. 2009 vs 2014 1.100 .328 to 1.842 1.000b .001b Potential harm for participants was mentioned in 17 Assessment of reduced competencea .031d (21.2%) studies, and mainly included increased pain and/ or fatigue, risk of falls, and unstable cardiovascular 2004 vs 2009 2.154 1.207 to 3.844 .494 status. Plans for collecting, assessing, reporting, and .014 managing adverse events and other unintended effects 2004 vs 2014 1.579 1.018 to 2.448 .078 of trial interventions were reported respectively in 13, 11, 3, and 8 of these articles. The reporting of presence/ 2009 vs 2014 .733 .405 to 1.325 absence of harm or adverse events was observed in 31 (38.8%) cases (Table 3). None of the studies mentioned Potential conflicts of interesta eventual follow up care. 2004 vs 2009 1.077 .737 to 1.573 Although we found a clear reference to steps taken to protect anonymity in only one article, confidentiality was 2004 vs 2014 3.333 1.748 to 6.358 always preserved. Eighteen publications (30%) showed pictures depicting models or participants: with only two 2009 vs 2014 1.341 1.062 to 1.693 exceptions, faces were shown only in part (twice masking the eyes region) or were shielded. As previously noted, Number of ethics-related issues reportede one study reported a specific consent for publication of the participant’s photos. 2004 vs 2009 −0.96 −2.98 to 1.06 The number of ethics-related issues reported was sig- 2004 vs 2014 −2.33 −4.26 to −0.40 nificantly different across the three years considered (one-way ANOVA F(5.07) = 6.345, p = .009) (Table 3); 2009 vs 2014 −1.37 −2.86 to 0.12 Tukey post-hoc comparisons showed that this number was significantly higher in RCTs published in the year CI Confidence Interval 2014 compared to 2004, whereas there was no signifi- a Relative risk for the absence of the characteristic cant difference between the years 2004 and 2009, and b Fisher exact test the years 2009 and 2014 RCTs (Table 2). c Mann-Whitney test and Hodges-Lehmann estimator, data are Multivariable predictors of ethics reporting median differences Multiple regression showed that the number of d Pearson χ2 test ethics-related issues reported could be predicted as e Tukey’s test, data are mean differences 1.167 + .893* year of publication (coded as 1 = 2004, 2 = 2009, and 3 = 2014) + .799* PEDro score (F[4,75] = 11.103, had no linear-by-linear association (Table 3). In a p < .001, adjusted R2 of .338). Five-year impact factor post-hoc test a significantly greater probability that and clinical trial registration (coded as 0 = absent and such assessment would not be performed was observed 1 = present) were not significant predictors (Table 4). for studies published in the year 2004 compared to 2009 (Table 2), Only 3 papers published in 2009 Discussion specified how consent was acquired for vulnerable In our study, we observed a broad variability in the participants. Separate consents for videotaping or for reporting proportions of ethics-related components of publication of participant’s photos were reported, each in one study. Incentives or compensation for participants (money or gifts, or free transportation to the research center) were stated in 3 articles published in 2014. Funders of the studies and details related were reported in 56 (70.0%) cases. The proportion of RCTs that declared the pres- ence/absence of conflict of interest increased signifi- cantly across the three years, from 23.1% (3/13) in 2004, to 28.6% (8/28) in 2009, to 76.9% (30/39) in 2014 (Table 3). Authors declared conflict of interest only in 2 (2.5%) studies. Post-hoc comparisons showed that the probability that potential conflict of interest and re- lated details were not reported was greater for trials

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 7 of 10 Table 3 Ethics-related issues reporting Year 2004 (n = 13) 2009 (n = 28) 2014 (n = 39) p All studies (n = 80) 33 (84.6) .470b 65 (81.2) Ethic committee study approval 9 (69.2) 23 (82.1) 22 (56.4) .667c 41 (51.2) 37 (94.9) .443b 74 (92.5) Details about ethic committee 6 (46.2) 13 (46.4) 22 (56.4) .245b 42 (52.5) 20 (51.3) .049b 41 (51.2) Consent 11 (84.6) 26 (92.9) 3 (7.7) .278b 3 (3.8) 26 (66.7) .848b 56 (70.0) Details about the consent process 4 (30.8) 16 (57.1) 30 (76.9) <.001b 41 (51.2) 17 (43.6) .455c 30 (37.5) Assessment of reduced cognitive competence 3 (23.1) 18 (64.3) 15 (38.5) .478c 27 (33.8) 34 (87.2) .638b 68 (85.0) Incentives or compensation, and details 0 (0.0) 0 (0.0) 30 (76.9) .509b 60 (75.0) 10 (25.6) .470b 17 (21.2) Funders and details 10 (76.9) 20 (71.4) 20 (51.3) .081b 31 (38.8) 6 (15.4) .403b 9 (11.2) Potential conflicts of interest 3 (23.1) 8 (28.6) 9.9 (2.9) .009d 9.0 (2.6) Statement about sample size estimates 5 (38.5) 8 (28.6) Performing of power calculations 5 (38.5) 7 (25.0) Appropriateness of comparators 10 (76.9) 24 (85.7) Matching of comparators 11 (84.6) 19 (67.9) Potential harm for participants 1 (7.7) 6 (21.4) Reporting presence/absence of adverse events 4 (30.8) 7 (25.0) Accordance with the Helsinki declaration. 0 (0.0) 3 (10.7) Number of ethics-related issues reporteda 7.5 (1.6) 8.5 (2.3) Data are presented as absolute frequencies (percentages) except a mean (standard deviation) b Fisher exact test, c Pearson χ2 test, d one-way ANOVA research methods. Ethics committee approval and events were definitively underreported, and details about obtaining an informed consent, the main ethical issues, incentives or compensation for participants and steps were the most reported. Approximately half of the stud- taken to protect confidentiality were almost ignored. ies reported details such as name and location of the The year of publication and PEDro score were associated committee, or how consent was acquired, but only 3 with the completeness of the reporting of ethics-related papers specified how consent was obtained for vulner- issues. able participants. Information about funding was fre- quently reported (70% of studies), whereas that about Limitations conflict of interest was mentioned in about half of the Articles analyzed in the present study were limited in studies, with a significantly increasing trend over the terms of year of publication (2004, 2009, and 2014). In years. Statement about sample size estimates, potential the absence of references, we established a 20 % of the harm for participants, and presence/absence of adverse gained studies as the target sample. However, we made Fig. 3 Ethic Committee approval and Informed Consent reporting. / Data are presented as percentages. EC = ethics review committee study approval. IC = obtainment of informed consent

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 8 of 10 Table 4 Multiple regression analysis results B ± Std. Error 95% CI for B β t p .828 .410 (Costant) 1.167 ± 1.410 −1.641 to 3.975 2.424 .018 1.591 .116 Year of publication .893 ± .368 .159 to 1.627 .251 .983 .329 3.761 <.001 5-year Impact Factor .261 ± .164 −.066 to .589 .162 Clinical trial registration .708 ± .721 −.727 to 2.144 .102 PEDro score .799 ± .212 .376 to 1.222 .391 B regression coefficients, followed by the respective standard error, CI confidence interval; β standardized regression coefficient our calculation based on the total number of articles International Committee of Medical Journal Editors rec- present in the PubMed-generated lists, and before the ommendations [13]. The statement underlines that details screening process. Thus, the representativeness of our about the approval from an ethics committee, the obtain- random sample may be hypothesized higher than we ing of informed consent, and the research funding should expected. We performed our search on PubMed, which be reported while publishing an article. The presence of has been recognized as a comprehensive database index- conflict of interest is also highly emphasized, and we ing RCTs of physical therapy interventions [33]. To be notice in our sample a significant increase over time in indexed in PubMed, journals should demonstrate one the related reporting. The document does not clearly for- certain quality of the editorial work, including features mulate recommendations about the ethical issues that we such as statements indicating adherence to ethical guide- found not properly reported. Nevertheless, ethics aspects lines and evidence that authors have disclosed financial like sample size dimension and risk of harm for partici- conflicts of interest. This condition suggests that in our pants are relevant [10, 18, 35, 36]. Compensation to sample the reporting of ethics issues may be of better participants may impact the statistical inferences, and quality than in the general population. Moreover, since transparency on the topic is desirable when reporting we reviewed only English-language publications, we do clinical trials [37]. not know if similar ethics reporting characteristics could be observed in non-English language publications. These Despite the progress observed, and in accordance with factors limit the generalizability of the findings. other authors, [35] our findings suggest that the reporting of many ethics-aspects needs to be improved in RCTs of A wide range of ethical issues was considered in our physical therapy interventions after stroke. Reporting study. When we drafted the protocol, we tried to be com- guidelines should be updated [35]; however, adding brief prehensive. However, establishing which ethics-related and clear sentences to the text (e.g., “The occurrence of issues should be contained in a fully comprehensive list adverse events has been monitored”, “No compensation would require consensus from a broad multidisciplinary was offered to participants”, “Data were managed and expert team, which was beyond our intentions and cap- accessed only by authorized personnel”) could be a abilities. Extracting data on the reporting of the ethical as- starting point. pects contained in our list was challenging in the absence of clear statements; for example, we found it difficult to Conclusion extrapolate data on issues such as those related to the Though improved over time, ethics reporting practices in potential harm for participants. Thus, we acknowledge RCTs of physical therapy interventions after stroke should that objectivity might have not been maintained in these be ameliorated. With its limitations, our study shows circumstances. deficiencies of various degrees. Authors, editors, and re- viewers should be more rigorous and demanding about Interpretation the reporting of ethic-related methods regarding the Associations between methodological quality and ethics reproducibility of research. Almost all the ethics-related reporting practices have already been observed, [3] as issues in our list have been recognized as part of the mini- well as improvement of ethical approval and consent mum set of items to be addressed drawing a protocol for a reporting in RCTs [34]. Compared to a previous study RCT [28]. If protocols are drawn up in accordance with on physical therapy publications, we observed a greater the expected standard, and subsequent trials are reported rate of trials reporting both ethics committee approval faithfully, this should result in an increase in quantity and and consent obtainment (+ 11%), as well as a smaller quality of the ethics reporting practices in RCTs. Thus, rate of those who did not report either (− 16%) [14]. the understanding of ethical methods and the convergence on best practices will be promoted, [12, 28] and a virtuous The reporting of research ethics-related issues and circle originated. methods observed in our sample met, to some extent, the

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 9 of 10 Appendix Table 5 Questions a1.0 Was ethics review committee or institutional review board a6.0 Were information about sample size estimates reported? study approval reported? 1.1 Was the name of the committee reported? a6.1 Were power calculations performed? 1.2 Was the location of the committee reported? a7.0 Were comparator interventions appropriate? a1.4 Were details about the ethics committee reported? a7.1 Was the dosage of the intervention provided to groups comparable? a2.0 Was it reported if informed consent was obtained? 8.0 Was the eventual use of placebo justified? 2.1 Was the consent for participation written? a9.0 Was potential harm for participants mentioned? 2.2 Was the consent for participation oral? a9.1 Were plans for collecting adverse events and other unintended effects mentioned? 2.3 How was consent (and possibly assent) acquired for members a9.2 Were plans for assessing adverse events and other unintended effects of vulnerable populations? mentioned? a2.4 Were details about the consent process reported? a9.3 Were plans for reporting adverse events and other unintended effects mentioned? a3.0 Were incentives or compensation for participants reported? a9.4 Were plans for managing adverse events and other unintended effects mentioned? 3.1 What did they receive, money, a gift, free medical care or a9.5 Were presence/absence of harm or adverse events reported? treatment, free transportation or other services? 3.2 Were details about incentives or compensation given to a10.0 Was mentioned if appropriate follow up care was assured? participants reported? a4.0 Were funders reported? a11.0 Were steps taken to assess if participants had reduced competence (eg. use of validated screening tools for cognitive ability) reported? 4.1 How many organizations were involved? 11.1 Which measures were taken to protect participants with reduced competence best interests? 4.2 What were the funding sources? Governmental agencies, private foundations, or some other type? a12.0 Were steps taken to prevent unauthorized access to personal and clinical data (confidentiality) mentioned? 4.3 Were details about funders reported? a12.1 Was confidentiality preserved? a5.0 Were there statements about potential conflicts of interest? 5.1 Was any conflict declared? a13.0 Was reported if the study was conducted in accordance with the Helsinki Declaration? 5.2 Were details about conflicts of interest reported? aitem counted in the number of ethics-related issues reported Abbreviation Author details RCT: Randomized controlled trial 1Functional Rehabilitation, Azienda USL Toscana Centro, Via Cavour 118/120, 59100 Prato, Italy. 2Outpatient Rehabilitation, CRT Clinica di Riabilitazione Availability of data and materials Toscana Terranuova Bracciolini Spa, Via Gaetano Donizetti 2, 52028 The datasets analysed during the current study are available from the Terranuova Bracciolini, AR, Italy. 3Department of Experimental and Clinical corresponding author on reasonable request. Medicine, Research Unit of Medicine of Aging, University of Florence, and Azienda Ospedaliero–Universitaria Careggi, Viale Pieraccini 18, 50139 Authors’ contributions Florence, Italy. FF and MDB provided concept/idea/research design. FF, MV and MDB provided writing. MV and FF provided data collection. FF and MDB provided Received: 14 February 2018 Accepted: 15 May 2018 data analysis. FF and MV provided project management. MDB provided consultation (including review of manuscript before submission). All authors References read and approved the final manuscript. 1. Vergnes JN, Marchal-Sixou C, Nabet C, et al. Ethics in systematic reviews. J Ethics approval and consent to participate Med Ethics. 2010;36(12):771–4. Not applicable. 2. Jacobsen KH. Reporting of ethics-related methods in epidemiological Competing interests research. J Med Ethics. 2009;35(4):262–7. The authors declare that they have no competing interests. 3. Ruiz-Canela M, de Irala-Estevez J, Martínez-González MA, et al. Publisher’s Note Methodological quality and reporting of ethical requirements in clinical trials. J Med Ethics. 2001;27(3):172–6. Springer Nature remains neutral with regard to jurisdictional claims in 4. Ashton CM, Wray NP, Jarman AF, Kolman JM, Wenner DM, Brody BA. A published maps and institutional affiliations. taxonomy of multinational ethical and methodological standards for clinical trials of therapeutic interventions. J Med Ethics. 2011;37(6):368–73. 5. Roberts LW. Ethical dimensions of psychiatric research: a constructive, criterion-based approach to protocol preparation. The research

Ferrarello et al. Archives of Physiotherapy (2018) 8:8 Page 10 of 10 protocol ethics assessment tool (RePEAT). Biol Psychiatry. 1999;46(8): 33. Michaleff ZA, Costa LO, Moseley AM, et al. CENTRAL, PEDro, PubMed, and 1106–19. EMBASE are the most comprehensive database indexing randomized 6. Harriss DJ, Atkinson G. Ethical standards in sport and exercise science controlled trials of physical therapy interventions. Phys Ther. 2011;91(2):190–7. research: 2016 update. Int J Sports Med. 2015;36(14):1121–4. 7. Chiumento A, Rahman A, Frith L, Snider L, Tol WA. Ethical standards for 34. Schroter S, Plowman R, Hutchings A, et al. Reporting ethics committee mental health and psychosocial support research in emergencies: review of approval and patient consent by study design in five general medical literature and current debates. Glob Health. 2017;13(1):8. https://doi.org/10. journals. J Med Ethics. 2006;32(12):718–23. 1186/s12992-017-0231-y. 8. Junod V, Elger B. Retrospective research: what are the ethical and legal 35. Trung LQ, Morra ME, Truong ND, et al. A systematic review finds requirements? Swiss Med Wkly. 2010;140:w13041. https://doi.org/10.4414/ underreporting of ethics approval, informed consent, and incentives in smw.2010.13041. clinical trials. J Clin Epidemiol. 2017;91:80–6. 9. Emanuel EJ, Wendler D, Grady C. What makes clinical research ethical? JAMA. 2000;283(20):2701–11. 36. Cesana BM, Antonelli P. Sample size calculations in clinical research should 10. Sim J. Methodology and morality in physiotherapy research. Physiotherapy. also be based on ethical principles. Trials. 2016;17(1):149. https://doi.org/10. 1989;75(4):237–43. 1186/s13063-016-1277-5. 11. Maher CG, Sherrington C, Elkins M, et al. Challenges for evidence-based physical therapy: accessing and interpreting high-quality evidence on 37. Swanson DM, Betensky RA. Research participant compensation: a matter of therapy. Phys Ther. 2004;84(7):644–54. statistical inference as well as ethics. Contemp Clin Trials. 2015;45(Pt B):265–9. 12. Anderson JA, Eijkholt M, Illes J. Ethical reproducibility: towards transparent reporting in biomedical research. Nat Methods. 2013;10(9):843–5. 13. Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals [Internet]. ICMJE, The International Committee of Medical Journal Editors [accessed 2016 Dec 10]. Available from: http://www.icmje.org/icmje-recommendations.pdf 14. Henley LD, Frank DM. Reporting ethical protections in physical therapy research. Phys Ther. 2006;86(4):499–509. 15. Goodman SN, Fanelli D, Ioannidis JP. What does research reproducibility mean? Sci Transl Med. 2016;8(341):341ps12. https://doi.org/10.1126/ scitranslmed.aaf5027. 16. Eijkholt M, Anderson JA, Illes J. Picturing neuroscience research through a human rights lens: imaging first-episode schizophrenic treatment-naive individuals. Int J Law Psychiatry. 2012;35(2):146–52. 17. Weingarten MA, Paul M, Leibovici L. Assessing ethics of trials in systematic reviews. BMJ. 2004;328(7446):1013–4. 18. Sim J. Addressing conflicts in research ethics: consent and risk of harm. Physiother Res Int. 2010;15(2):80–7. 19. Sabapathy SS, Janakiraman K, Swarnalatha CC, et al. Reporting of ethical issues in indian physiotherapy journals. J Phys Ther. 2010;1(1):25–31. 20. Moral-Munoz JA, Arroyo-Morales M, Herrera-Viedma E, Cobo MJ. An overview of thematic evolution of physical therapy research area from 1951 to 2013. Front Res Metr Anal 2018;3: article 13. doi:https://doi.org/10.3389/ frma.2018.00013 21. von Elm E, Altman DG, Egger M, et al. STROBE initiative. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344–9. 22. DeJong G, Horn SA, Gassaway JA, et al. Toward a taxonomy of rehabilitation interventions: using an inductive approach to examine the ‘black box’ of rehabilitation. Arch Phys Med Rehabil. 2004;85(4):678–86. 23. American Physical Theraphy Association. Guide to physical theraphy practice. 2nd ed. Alexandra: APTA; 2001. 24. Clinician’s Handbook, Motor rehabilitation [internet]. Evidence-Based Review of Stroke Rehabilitation. Available from: http://www.ebrsr.com/clinician-handbook. [Accessed 2017 Jan 22] 25. Pinto RZ, Elkins MR, Moseley AM, et al. Many randomized trials of physical therapy interventions are not adequately registered: a survey of 200 published trials. Phys Ther. 2013;93(3):299–309. 26. Urbaniak GC, Plous S (2015). Research Randomizer (Version 4.0) [Computer software]. Retrieved on January 22, 2017, from: http://www.randomizer.org/ 27. 2014 Journal Citation Reports ® Science Edition (Clarivate Analytics, 2017). Available from: https://jcr.incites.thomsonreuters.com/JCRJournalHomeAction.action. [accessed 2017 Jan 5] 28. Chan AW, Tetzlaff JM, Altman DG, et al. SPIRIT 2013 statement: defining standard protocol items for clinical trials. Ann Intern Med. 2013;158(3):200–7. 29. Caprino L, Russo P. Developing a paradigm of drug innovation: an evaluation algorithm. Drug Discov Today. 2006;11(21–22):999–1006. 30. Armijo Olivo S, Macedo LG, Gadotti IC, et al. Scales to assess the quality of randomized controlled trials: a systematic review. Phys Ther. 2008;88(2):156–75. 31. Maher CG, Sherrington C, Herbert RD, et al. Reliability of the PEDro scale for rating quality of randomized controlled trials. Phys Ther. 2003;83(8):713–21. 32. Foley NC, Teasell RW, Bhogal SK, Speechley MR. Stroke rehabilitation evidence-based review: methodology. Top Stroke Rehabil. 2003;10(1):1–7.

Adjenti et al. Archives of Physiotherapy (2018) 8:9 https://doi.org/10.1186/s40945-018-0048-x RESEARCH ARTICLE Open Access An ultrasonographic analysis of the activation patterns of abdominal muscles in children with spastic type cerebral palsy and in typically developing individuals: a comparative study Saviour Kweku Adjenti1* , Graham Jacob Louw2, Jennifer Jelsma3 and Marianne Unger4 Abstract Background: Abdominal muscles have stiffer appearance in individuals with spastic type cerebral palsy (STCP) than in their typically developing (TD) peers. This apparent stiffness has been implicated in pelvic instability, mal-rotation, poor gait and locomotion. This study was aimed at investigating whether abdominal muscles activation patterns from rest to activity differ in the two groups. Method: From ultrasound images, abdominal muscles thickness during the resting and active stages was measured in 63 STCP and 82 TD children. The thickness at each stage and the change in thickness from rest to activity were compared between the two groups. Results: Rectus abdominis (RA) muscle was the thickest muscle at rest as well as in active stage in both groups. At rest, all muscles were significantly thicker in the STCP children (p < 0.001). From rest to active stages muscle thickness significantly increased (p < 0.001) in the TD group and significantly decreased (p < 0.001) in the STCP children, except for RA, which became thicker during activity in both groups. In active stages, no significant differences in the thickness in the four abdominal muscles were found between the STCP and the TD children. Conclusion: Apart from the RA muscle, the activation pattern of abdominal muscles in individuals with STCP differs from that of TD individuals. Further studies required for understanding the activation patterns of abdominal muscles prior to any physical fitness programmes aimed at improving the quality of life in individuals with STCP. Trial registration: HREC REF: 490/2011. Human Research Ethics Committee, Faculty of Health Sciences, University of Cape Town, South Africa. November 17, 2011. Keywords: Spastic type cerebral palsy (STCP), Abdominal muscles, Muscle thickness, Utra-sound imaging, Rehabilitation Background with spastic type cerebral palsy (STCP), poor postural The abdominal muscles play an important role in stabi- control is noted to be a primary manifestation of the lising the trunk and providing postural stability [1]. motor dysfunction [2]. The need to target the control of These muscles include the internal and external oblique the trunk in therapy as early as possible in children with muscles (IO and EO respectively), the transverse abdom- neurodevelopmental problems has been emphasised by inis (TA) and the rectus abdominis (RA). In children Burtner and co-workers [3]. These authors reported that skeletal muscles possess remarkable plasticity and can * Correspondence: [email protected]; [email protected] quickly gain or lose contractile material according to 1Department of Anatomy, School of Biomedical & Allied Health Sciences, changes in loading regimens. Therapists often focus, College of Health Sciences, Korle-Bu Campus, University of Ghana, P.O. Box either directly [4] or indirectly [5], on the abdominal KB 143, Korle-Bu, Accra, Ghana muscles for improving postural control and function. Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Adjenti et al. Archives of Physiotherapy (2018) 8:9 Page 2 of 8 Targeting the trunk is particularly common in those Methods children who display an anterior pelvic tilt [4, 6]. This The design of this study was descriptive and analytical. position places a prolonged stretch on the TrA and RA Ethical approval was obtained from the Human Re- muscles needed to maintain a neutral pelvis and subse- search Ethics Committee of the XXXXXXXXXXXXX quently causes inhibition of the stretch reflex [7] thereby (HREC REF: 490/2011). decreasing reactivity in these muscles. Despite the recog- nised clinical importance of these muscles, it is evident Participants from the literature that little is known about the struc- The STCP group was recruited from individuals attend- ture, function and neuronal activity in persons with ing special schools in XXXXXXXX while the TD group STCP [8]. comprises children attending mainstream schools in the vicinity of the special schools. Informed consent and or Although STCP is a non-progressive disorder, over time assent were obtained from individuals and or guardians secondary complications occur due to weakness and tone from these convenient sampling. A learner was excluded imbalance [9]. According to Hungerford and co-workers, if he or she had any surgical operation involving the an- the abnormal forces imposed by the muscles on the skel- terior abdominal wall in the last six months before the etal system result in biomechanical mal-alignment such as start of the study. the anterior pelvic tilt mentioned above [10]. Additionally, in individuals with STCP abnormal recruitment such as For the STCP group, the Gross Motor Function Classi- the top-down recruitment of the trunk muscles is com- fication Scale [GMFCS] [18] was also used by a neurode- mon [11]. This abnormal recruitment has been associated velopmental therapist to determine the level of function with co-contraction of the extremity muscles [7]. Other of the participants to be included. Only children in levels abnormalities noted include a method of fixation of the I-IV formed part of the inclusion criteria. Children at trunk in some STCP population as well as a negligible level V were excluded because they were unable to per- muscular activity in others [10, 12]. With regards to indi- form the test manoeuvres. Another exclusion criterion viduals with STCP, the muscle groups which appear to for individuals with STCP was an involvement with any contribute to this fixation include the flexors, adductors medical treatment that would have impacted on muscle and the internal rotators of the hip, which gives rise to the function (e.g., Botulinum toxin injection, casting, and typical postural and gait patterns seen – couch gait in surgical intervention such as dorsal rhizotomy and bac- diplegia and equines gait in hemiplegia [11]. lofen pump placement) less than six months before the study. The force-generating capacity of a skeletal muscle and consequently muscle strength is reported to be a com- Assessment posite function of different aspects of the muscle archi- In all participants, anthropometric parameters (i.e. height tecture, including thickness [13]. Ultrasound imaging is and weight were measured before ultrasonographic a non-invasive method of recording changes in muscle assessment. thickness during activation, which was first exploited in muscle activity of the myocardium [14]. Ultrasonography A SIEMENS® ACUSONIC X150 ultrasound imaging has since been used to quantify muscle thickness in indi- machine (Munich, Germany) was used to capture the viduals with STCP [15, 16].Ultrasound imaging method thickness of the four abdominal muscles, rectus abdom- has been reported to be fast, inexpensive and above all inis (RA), internal oblique (IO), external oblique (EO) reliable. The advantages of the use of ultrasonography and transverse abdominis (TrA), in both the resting and over electromyography technique have been docu- active stages. To test the muscles in the resting stage, mented by Ohata et al. [16, 17]. children were asked to lie supine on the plinth with no activity. For the active stage, children were asked to lie The present study aimed to contribute to the under- supine on the plinth and then asked to perform the fol- standing of the functioning of abdominal muscles as a lowing activities: (i) To fully abduct the shoulder joint group and/or separate muscles of the anterior abdominal (ii) to tuck in the chin and lift head and neck slightly wall in individuals with STCP. The specific objectives of towards the chest; and (iii) to flex the hip as far as pos- the study were to: (i) measure the thickness levels in sible. The performance of these activities was aimed at each of the four anterior abdominal muscles during the initiating a simultaneous contraction of the abdominal resting and active stages and (ii) compare the changes in muscles, which was then measured. The average of these thickness, herein referred to as activation pattern three manoeuvres was recorded as the active stage thick- between these two stages in the two groups. It is ness. The side of active upper or lower limb motion and expected that the muscles of TD individuals would dem- of abdominal muscle thickness measurement was the onstrate greater activation pattern evidenced by a larger affected side in hemiplegic children, the right side in change in thickness from the resting to active stages diplegic, quadriplegic and TD children. The principal than their age-matched peers with STCP.

Adjenti et al. Archives of Physiotherapy (2018) 8:9 Page 3 of 8 investigator handled the transducer head (ultra-sound groups. This test was performed by the principal investi- probe) while one of the research assistants, a neurodeve- gator (SKA) on different occasions. Measurements were lopmental therapist, issued the instructions to the taken at three different points for a particular muscle participants. per participant and the averages were recorded. In each participant, muscle thickness was measured as described Using the umbilicus as a landmark the ultrasound above, in both the resting and active stages by SKA. The probe was placed two to three centimetres from the same assessor repeated the measurements one week midline and then was panned around in a semi-circular later. The outcome of this intra-rater reliability test was fashion until the bulk of the image from the deepest expressed as reliability indexes (Intraclass Correlation lying abdominal muscle, TrA, was observed on the Coefficient, typical error and mean differences ± SD be- image screen. This position was marked on the skin with tween the measures taken by the two assessments), a marker pen in order to ensure that the probe was kept showing good to excellent correlation (ICC ≥ 0.80) see in this position for subsequent measurements. The scan- Tables 4 and 5 in Appendix). ning head of the probe was then oriented along the mid-sagittal axis of each of the rest of the three antero- Statistical analysis lateral abdominal muscle (EO, IO and TrA) in a some- STATISTICA software package, version 11 (2012) was what oblique fashion. The pressure of the transducer used to analyse the data. The BMI was calculated using was kept to a minimum by using a generous amount of the standard formula, mass (kg)/height (m) x height (m) the contact gel in order to obtain optimum values for [19]. Descriptive statistics were presented for the data muscle thickness. All sites along a muscle from which sets: height, weight, age and muscle thickness. Due to images were taken at rest were then repeated during the relatively large sample size, normality was assumed each child’s head and shoulder/leg lift movement (active and parametric tests were used for all analyses. The stage). Images were stored on a personal computer and Chi-Square test was used to compare the sex distribu- then analysed with ImageJ Microsoft version 1.46, 2011 tion between the STCP and TD groups. Independent edition (Richmond, Virginia, USA). t-test was used to statistically compare the means of the two groups in both resting and active stages. A two-way Muscle thickness (MT) was determined using an elec- ANOVA with repeated measures was used to determine tronic calliper on a frozen image. The length of a per- if there was a significant group-stage interaction that af- pendicular line drawn between the echoes parallel to the fected muscle thickness. The changes in muscle thick- fascicles from the deep up to the superficial aponeurosis ness from rest to the active stage in each group were (inter-fascial planes) was measured (Fig. 1). Since thick- compared using paired t-test. A 95% confidence interval ness varies along the length of a muscle, measurements was used to determine the precision of the estimates of were taken at three different points for a particular the differences in muscle thickness between the resting muscle according to the clarity of the image and the and active stages for both groups. Association between average was recorded for that individual. muscle thickness and age of all participants was assessed using the Pearson’s correlation coefficient. The level of Before the test, the intra-rater reliability of the ultra- significance for all statistical tests was set at 0.05. sound measurements was assessed on fifteen (15) randomly selected participants in the STCP and TD Fig. 1 Sonogram showing the three anterolateral muscles. SF = skin Results and superficial fascia, AC = abdominal cavity. The RA is out of view Over 200 participants (more than 100 in each group) met the inclusion criteria and were invited to participate. The parents of 145 (63 children with STCP and 82 TD children) gave consent and their children were recruited. The demographic data of all the participants are shown in Table 1. There were no significant differences in age (p = 0.102) and in gender (X2 = 0.139; p = 0.709) between the two groups (STCP group: mean age 11.2 ± 2.9 years, 55.6% males: TD group: mean age 11.3 ± 2.9 years, 52.4% males). The groups were also similar as regards height and weight. `However, the children with STCP were both shorter and heavier than the children in the TD group, and therefore they had a significantly greater BMI (p < 0.001).

Adjenti et al. Archives of Physiotherapy (2018) 8:9 Page 4 of 8 Table 1 Comparison of demographic data between the groups STCP TD t-value Df p-value Mean SD Mean SD Age (years) 11.89 2.92 11.05 2.92 1.65 143 0.102 Height (cm) 139.19 16.04 143.32 17.13 −1.48 143 0.142 Weight (kg) 39.68 10.28 38.75 12.38 0.48 143 0.629 BMI (kg.m−2) 20.14 2.16 18.37 2.62 4.34 143 < 0.001 The distribution of the various subtypes of the spastic Fig. 2 Scatterplot diagrams of age of participants and mean resting cerebral palsy is shown in Table 2. More than half (N = 34) muscle thickness for both groups combined (N = 145), showing of the STCP participants were able to move independently Pearson’s correlation (r) between muscles thickness and age. All without appliances (GMFCS level I). There were between correlation coefficients were significant at p < 0.001. EO = external eight and eleven children in each of the other levels. oblique; IO = external oblique; TrA = transversus abdominis; Forty-four of the participants had hemiplegia (Table 2). RA = rectus abdominis Muscle thickness at rest showed a significant positive and then decreased in thickness when the active ma- association with the age of participants (r = 0.766–0.864, noeuvres were performed. The RA muscle of the STCP p < 0.001) in both groups (Fig. 2). The RA muscle group resembled those of TD children in both thickness remained the thickest muscle across all ages in both and activation patterns. It may be that the abdominal groups, followed by the IO, EO and TrA muscles. muscles have to stabilise the multi-jointed spine and pelvis even in supine position, which requires both con- The average thickness of each muscle at rest and the centric and eccentric contraction, compared to the com- active stages in the two groups and the results of paired paratively simple task of controlling one or two joints in and unpaired t-tests are reported in Table 3. Data are the case of the lower limb muscles. Therefore, since in- also shown graphically in Figs. 3, 4, 5 and 6. At rest, all dividuals with STCP require more support of the trunk muscles were significantly thicker in the STCP than in than their TD counterparts, the trunk stabilising role the TD children (p < 0.001). From rest to active stages, could inevitably predispose the abdominal muscles in muscle thickness significantly increased (p < 0.001) in individuals with STCP to become relatively thicker at the TD group and significantly decreased (p < 0.001) in rest than for the TD group. Ohata and co-workers also the STCP children, except for RA, which became thicker found the abdominal muscles thickness in individuals during activity in both groups. The repeated measures with STCP to be remarkably high at rest [20]. Apart ANOVA confirmed the signicant (p < 0.001) group x from the work of Ohata and co-workers [20], no other stage interaction effect on thickness for all muscles (EO: F comparable results for the abdominal muscles in STCP (1,143) = 283.097; IO: 310.669; TrA: 601.925; RA: 25.278). children are found in the literature. In active stages, no significant difference in the thickness of the four abdominal muscles was found between the The size of the RA, the thickest muscle in both STCP and TD children, and the RA muscle was still the groups, may be explained in terms of postural roles. thickest muscle in both groups. Although recruitment patterns differ between individuals and are influenced by changes in body position, a thicker Discussion RA than the rest of the abdominal muscles could prob- The counterintuitive results that emerged from the study ably be a result of the use this muscle in all global move- were that apart from the RA, the muscles of the STCP ments of the trunk during activity of daily living [21]. It were thicker at rest, than the muscles of the TD children has been suggested that the abdominal muscles may be in constant state of contraction, either due to neuro- Table 2 Gross Motor Classification System Level per distribution logical damage leading to spasticity/hypertonia or to the of STCP (N = 63) need to stabilise the trunk in the presence of deficient postural responses [2]. A related study which focused on LEVEL Hemiplegia Diplegia Quadriplegia Total for level the neuromuscular activity of the abdominal muscles, I 29 5 0 34 II 9 20 11 III 3 41 8 IV 3 34 10 All Groups 44 14 5 63 Note the high distribution of the hemiplegic subtypes of STCP in this study Note also that only four disability levels (ambulatory individuals) were recruited

Adjenti et al. Archives of Physiotherapy (2018) 8:9 Page 5 of 8 Table 3 Comparison of average raw muscle thickness at rest and active stage in both STCP and TD groups Rest Activity Mean diff. T-statistic p-value 95% CI of diff. EO STCP 3.64 ± 0.50 3.36 ± 0.52 0.28 12.43 < 0.001 0.24 to 0.32 TD 3.08 ± 0.50 3.29 ± 0.50 −0.21 −11.21 < 0.001 −0.24 to − 0.18 Mean diff. 0.56 0.08 95% CI of diff. 0.47 to 0.65 0.00 to 0.16 p-value < 0.001 0.086 IO STCP 4.76 ± 0.69 4.43 ± 0.73 0.33 11.04 < 0.001 0.27 to 0.38 TD 4.25 ± 0.52 4.45 ± 0.52 −0.20 −15.35 < 0.001 −0.23 to −0.17 Mean diff. 0.51 −0.02 95% CI of diff. 0.40 to 0.62 −0.12 to 0.08 p-value < 0.001 0.104 TrA STCP 2.86 ± 0.49 2.56 ± 0.48 0.30 16.30 < 0.001 0.27 to 0.33 TD 2.10 ± 0.53 2.38 ± 0.49 −0.28 −18.49 < 0.001 −0.31 to − 0.25 Mean diff. 0.76 0.18 95% CI of diff. 0.66 to 0.86 0.10 to 0.26 p-value < 0.001 0.082 RA STCP 6.33 ± 0.91 6.70 ± 0.93 −0.37 −17.65 < 0.001 −0.41 to −0.33 TD 5.44 ± 0.59 5.97 ± 0.63 −0.53 −23.97 < 0.001 −0.57 to − 0.49 Mean diff. 0.89 0.73 95% CI of diff. 0.75 to 1.03 0.59 to 0.87 p-value < 0.001 0.130 EO external oblique muscle, IO internal oblique muscle, TrA transverse abdominis muscle, RA rectus abdominis muscle, STCP spastic type cerebral palsy, TD typically developing developing, diff. difference reported high EMG activation patterns during periods of in a state of contraction formerly at rest. However, with inactivity (resting stage) in a cohort of children and ado- the exception of RA, muscle thickness in the STCP lescents with STCP [2]. Such evidence would support group was less when active than during the resting stage, the result from the present study and suggests that the indicating that these other abdominal muscles might dif- levels of abdominal muscles thickness at rest may have fer structurally and functionally from those of TD indi- underlying physiological/neurological activity, which viduals. Although similar decrease in abdominal muscle needs further investigation. This physiological/neuro- thickness from resting to active stage was reported by logical activity would also partially explain why the mus- Ohata et al. [20], further investigations would be cles in the STCP group showed less change during neck required to fully understand the anatomy and physiology and lower/upper limb activity, as they might have been of these muscles in individuals with STCP. Fig. 3 Error bar plots showing the mean thickness for the external Fig. 4 Error bar plots showing the mean thickness for the internal oblique muscle (EO) during resting and active stages in both groups. oblique muscle (IO) during resting and active stages in both groups. STCP = spastic type cerebral palsy; TD = typically developing STCP = spastic type cerebral palsy; TD = typically developing

Adjenti et al. Archives of Physiotherapy (2018) 8:9 Page 6 of 8 Fig. 5 Error bar plots showing the mean thickness for the transverse ability to activate the contractile materials optimally as a abdominis muscle (TrA) during resting and active stages in both decrease in thickness from resting to active stages was ob- groups. STCP = spastic type cerebral palsy; TD = typically developing served. Alternatively, a varying amount of non-contractile materials in these muscles might have contributed to the In both groups, the level of thickness and activation differences in thickness between the two groups. An inclu- pattern of the RA muscle suggest that this muscle may sion of this aspect in further investigation would be useful be suitable to take on the stabilising and flexing role of for a better understanding. the other abdominal muscles on activity. In individuals with STCP, the activation pattern of the RA muscle This study has some limitations. First more than might result from the inhibition of the other abdominal two-third (2/3) of the children with STCP enrolled were muscles that decreased their level of contraction during hemiplegic and more than half were at the highest func- activity. From a functional perspective, weak and inad- tional level. These proportions limit the generalisation of equately contracting oblique muscles are often associ- our finding since they are higher than those in ated with a lack of trunk rotation and an accompanying epidemiological studies [23]. This discrepancy may be due altered gait, all features that characterise individuals with to the sampling from special schools in which only educ- STCP. It could therefore, be inferred from the results of able children are admitted and more severe disability levels, this study that the function of trunk rotation by the EO which are associated with severe mental involvement [24], and IO muscles in individuals with STCP could be sacri- are excluded. Moreover, the researcher who analysed the ficed above the need to stabilise the trunk. The latter images was not blinded and this may have introduced some role taken on almost exclusively by the RA muscle: a bias. trunk flexor with no rotatory moment on activity [22]. Finally, in the children with STCP, the chosen neutral The larger thickness found at rest in children with STCP (plinth) position might not reflect the true thickness of in comparison with their TD counterparts might potentially abdominal muscles at rest. Actually, in children with indicate muscle hypertrophy. In that case, however, except some neuromuscular deficit the abdominal muscles for the RA muscle, children with STCP seem lack the might have been conditioned to contract also in that position in an attempt to stabilise the spine and pelvis. Conclusion The resting stage thickness of the anterior abdominal wall muscles of individuals with STCP is greater than those of their TD counterparts. The change in muscle thickness of abdominal muscles from the resting to active stages in in- dividuals with STCP differs from that of TD children ex- cept for the RA muscle, thereby implying that the RA muscle is unaffected or less affected by the condition. This knowledge could be useful in the problem-solving ap- proaches with regard to the functional aspect of the mus- culoskeletal system in the provision of quality-of-life benefits for individuals with STCP. Further research is needed to examine the patterns of abdominal muscle ac- tivity through dynamic electromyography (EMG). Highlights of this study include Fig. 6 Error bar plots showing the mean thickness for the rectus  In children with STCP, the activation patterns of abdominis muscle (RA) during resting and active stages in both abdominal muscles, except for the RA differ from groups. STCP = spastic type cerebral palsy; TD = typically developing those found in TD children.  Apart from the RA, the rest of the abdominal muscles in children with STCP are thicker at rest than during the active stage which might indicate hypertonicity or increased need to stabilise the trunk.  The role of abdominal muscles in stabilising the trunk / pelvis in individuals with STCP requires further investigation.

Adjenti et al. Archives of Physiotherapy (2018) 8:9 Page 7 of 8 Appendix Availability of data and materials The datasets used and/or analysed during the current study are available Table 4 Intra-rater reliability of the ultrasound measurements from the corresponding author on reasonable request. (direct method) in the feasibility study for the STCP group (n = 15). Data were expressed as typical error and intra-class coefficients Authors’ contributions (ICC) with their 95% confidence intervals (CI) and mean differences The conception of this research idea was by JJ, a Physiotherapist. ± standard deviation (SD) Accessibility to participants and equipment was overseen by MU, also a Physiotherapist. Recruitment of the participants and collection of data were Muscle & Stage Typical Error 95% CI ICC 95% CI Mean diff ± SD carried out by SKA as his doctoral research. The analysis of the data was undertaken by JJ and SKA. The drafting and editing of the manuscript was EO R 0.65 (0.06–0.24) 0.81 (0.66–0.96) − 0.19 ± 0.04 supervised by GJL. All four authors shared equal responsibilities in the final preparation of the manuscript. All authors read and approved the final EO Ac 0.59 (0.07–0.18) 0.88 (0.76–0.97) −0.10 ± 0.04 manuscript. IO R 0.60 (0.06–0.20) 0.82 (0.86–0.98) 0.10 ± 0.08 Our study is an original research No special grants or financial awards were received for the execution of this IO Ac 0.62 (0.08–0.24) 0.80 (0.65–0.95) 0.11 ± 0.05 study. There are no conflicts of interest pertaining to this study and with regard to any of the authors. Strict institutional ethical protocols (University TrA R 0.64 (0.05–0.20) 0.82 (0.68–0.98) 0.17 ± 0.07 of Cape Town, Human Research Ethics Committee) were observed. All individuals and institutions that provided support for this study were fully TrA Ac 0.63 (0.06–0.22) 0.85 (0.70–0.94) 0.08 ± 0.06 acknowledged. Limitations to the generalisability of this study were outlined. This manuscript has not been submitted to any other journal for publication. RA R 0.60 (0.08–0.25) 0.86 (0.75–0.98) −0.11 ± 0.06 Ethics approval and consent to participate RA Ac 0.62 (0.07–0.20) 0.88 (0.76–0.97) −0.19 ± 0.03 Ethics approval and consent to participate in the study was obtained from the Human Research and Ethics Committee (HREC) of the Faculty of Health Key: EO R external oblique resting stage, EO Ac external oblique active stage, Sciences, University of Cape Town, Ref. no.: HREC: 490/2011. IO R internal oblique resting stage, IO Ac internal oblique active stage, TrA R transverse abdominis resting. TrA Ac transverse abdominis active stage, RA R Consent for publication rectus abdominis resting stage, RA Ac rectus abdominis active stage Consent for publication was obtained from the parents/guardian of the participants. Table 5 Intra-rater reliability of the ultrasound measurements (direct method) in the feasibility study for the TD group (n = 15). Competing interests Data were expressed as typical error and intra-class coefficients The authors declare that they have no competing interests. The authors (ICC) with their 95% confidence intervals (CI) and mean differences alone are responsible for the content and writing of this article. ± standard deviation (SD) Publisher’s Note Muscle & Stage Typical Error 95% CI ICC 95% CI Mean diff ± SD Springer Nature remains neutral with regard to jurisdictional claims in EO R 0.62 (0.06–0.22) 0.80 (0.66–0.96) −0.13 ± 0.08 published maps and institutional affiliations. EO Ac 0.60 (0.08–0.20) 0.82 (0.76–0.97) −0.02 ± 0.09 Author details 1Department of Anatomy, School of Biomedical & Allied Health Sciences, IO R 0.64 (0.05–0.22) 0.81 (0.86–0.98 −0.16 ± 0.06 College of Health Sciences, Korle-Bu Campus, University of Ghana, P.O. Box KB 143, Korle-Bu, Accra, Ghana. 2Division of Clinical Anatomy & Biological IO Ac 0.62 (0.06–0.21) 0.79 (0.65–0.95) −0.09 ± 0.08 Anthropology, Department of Human Biology, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa. 3Division of TrA R 0.64 (0.04–0.20) 0.84 (0.68–0.98) 0.09 ± 0.09 Physiotherapy, Department of Health & Rehabilitation Sciences, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa. 4Division TrA Ac 0.64 (0.06–0.22) 0.85 (0.72–0.94) 0.09 ± 0.09 of Physiotherapy, Faculty of Medicine & Health Sciences, Stellenbosch University, Stellenbosch, South Africa. RA R 0.66 (0.08–0.24) 0.85 (0.75–0.98) −0.10 ± 0.09 Received: 16 February 2016 Accepted: 8 May 2018 RA Ac 0.63 (0.06–0.20) 0.86 (0.76–0.96) −0.01 ± 0.03 References Key: EO R external oblique resting stage, EO Ac external oblique active stage, 1. Hodges PW, Eriksson AE, Shierley D, Gandevia SC. Intra-abdominal pressure IO R internal oblique resting stage, IO Ac internal oblique active stage, TrA R transverse abdominis resting. TrA Ac transverse abdominis active stage, RA R increases stiffness of the lumbar spine. J Biomech. 2005;38(9):1873–80. rectus abdominis resting stage, RA Ac rectus abdominis active stage 2. Woollacott M, Shunway-Cook A, Hutchinson S, Ciol M, Price R, Kartin D. Abbreviations Effect of balance training on muscle activity used in recovery of stability in BMI: Body mass index; CP: Cerebral palsy; EMG: Electromyograph; children with cerebral palsy: a pilot study. Dev Med Child Neurol. 2005;47: EO: External oblique muscle; GMFCS: Gross motor function classification 455–61. system; IO: Internal oblique muscle; RA: Rectus abdominis muscle; 3. Burtner PA, Qualls C, Woollacott MH. Muscle activation characteristics of sEMG: Surface electromyography; STCP: Spastic type cerebral palsy; stance balance control in children with spastic cerebral palsy. Gait Posture. TD: Typically developing; TrA: Transversus abdominis muscle 1998;8:163–74. 4. Unger M, Faure M, Frieg A. Strength training in adolescent learners with Acknowledgements cerebral palsy: a randomized controlled trial. Clin Rehabil. 2006;20:469–77. Special thanks to the staff and learners of all the schools from which participants 5. Prosser LA, Lee SCK, Barbe MF, VanSant AF, Lauer RT. Trunk and hip muscle were recruited for this study. The authors are also grateful to the technical staff of activity in early walkers with or without cerebral palsy – a frequency the Division of Clinical Anatomy of the Department of Human Biology, Faculty of analysis. J Electromyo Kinesiol. 2010;20:851–9. Health Sciences, University of Cape Town, for the transporting of the equipment to and from the data sampling sites. Finally we appreciate the financial inputs of both faculty and management of the postgraduate units of the Faculty of Health Sciences of the Universities of Cape Town and Ghana Medical School toward the doctoral training of the principal investigator.

Adjenti et al. Archives of Physiotherapy (2018) 8:9 Page 8 of 8 6. Roussouly P, Gollogly S, Berthonnaud E, Dimnet J. Classification of the normal variation in the sagittal alignment of human lumbar spine and pelvis in the standing position. Spine. 2005;30:346–53. 7. Urquhart DM, Hodges PW, Allen TJ, Story IH. Abdominal muscle recruitment during a range of voluntary exercises. Man Ther. 2005;10(2):144–53. 8. Rose J, McGill KC. Neuromuscular activation and motor-unit firing characteristics in cerebral palsy. Dev Med Child Neurol. 2005;47:329–36. 9. Ando N, Ueda S. Functional deterioration in adults with cerebral palsy. Clin Rehabil. 2000;14:300–6. 10. Hungerford B, Gilleard W, Hodges P. Evidence of altered lumbopelvic muscle recruitment in the presence of sacroiliac joint pain. Spine. 2003;28: 1593–600. 11. Stackhouse SK, Binder-Macleod SA, Lee SCK. Voluntary muscle activation, contractile properties, and fatigability in children with and without cerebral palsy. Muscle Nerve. 2005;31:594–601. 12. Ferreira PH, Ferreira ML, Hodges PW. Changes in recruitment of abdominal muscles in people with low back pain: ultrasound measurement of muscle activity. Spine. 2004;29:2560–6. 13. Shortland AP, Harris CA, Gough M, Robinson RO. Architecture of the medial gastrocnemius in children with spastic diplegia. Dev Med Child Neurol. 2002;44:158–63. 14. Heimdal A, Stoylen A, Torp H, Skaerpe T. Real-time strain rate imaging of the left ventricle by ultrasound. J Am Soc Echocardiogr. 1998;11:1013–9. 15. Hodges PW, Pengel LHM, Herbert RD, Gandevia SC. Measurement of muscle contraction with ultrasound imaging. Muscle Nerve. 2003;27:682–92. 16. Ohata K, TsuboyamaT, Ichihashi N, Minami S. Measurement of muscle thickness as quantitative muscle evaluation for adults with cerebral palsy. Phys Ther. 2009;86:1231–9. 17. Benard MR, Becher JG, Harlaar J, Huijing PA, Jaspers RT. Anatomical information is needed in ultrasound imaging of muscle to avoid potentially substantial errors in measurement of muscle geometry. Muscle Nerve. 2009; 39:652–65. 18. Palisano RJ, Rosenbaum P, Bartlett D, Livingston MH. Content validity of the expanded and revised gross motor function classification system. Dev Med Child Neurol. 2008;50(10):744–50. 19. Brener ND, McManus T, Galuska DA, Lowry R, Wechsler H. Reliability and validity of self-reported height and weight among high school students. J Adolesc Health. 2003;32:281–7. 20. Ohata K, Haruta T, Kato T, Nakamura T. Relation between muscle thickness, spasticity, and muscle limitation in children and adolescents with cerebral palsy. Dev Med Child Neurol. 2008;50:152–6. 21. Vasseljen O, Fladmark AM. Abdominal muscle contraction thickness and function after specific and general exercises: a randomised controlled trial in chronic low back pain patients. Man Ther. 2010;15(5):482–9. 22. Andersson EA, Grundstrom H, Thorstensson A. Diverging intramuscular activity patterns in back and abdominal muscles during trunk rotation. Spine. 2002;27(6):e152–60. 23. Gorter JW, Rosenbaum PL, Hanna SE, Palisano RJ, Barlett DJ, Russell DJ, Walter SD, Raina P, Galuppi BE, Wood E. Limb distribution, motor impairment and functional classification of cerebral palsy. Dev Med Child Neurol. 2004;46:461–7. 24. Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M, Damiano D, Dan B, Jacobson B. A report: the definition and classification of cerebral palsy April 2006. Dev Med Child Neurol. 2007;109(suppl):8–14.

Bowman et al. Archives of Physiotherapy (2018) 8:10 https://doi.org/10.1186/s40945-018-0051-2 RESEARCH ARTICLE Open Access Predictors of mobility domain of health- related quality of life after rehabilitation in Parkinson’s disease: a pilot study Thomas Bowman, Elisa Gervasoni, Riccardo Parelli, Johanna Jonsdottir, Maurizio Ferrarin, Davide Cattaneo* and Ilaria Carpinella Abstract Background: Parkinson’s disease impacts health-related quality of life (HRQoL), however no studies inquired on predictors of HRQoL changes after rehabilitation. This study assessed the relationship between mobility domain of HRQoL measured by Parkinson’s Disease Questionnaires-39 (PDQ-39) and clinical-demographic characteristics and developed a model predicting changes after rehabilitation. Methods: Subjects with Parkinson’s disease underwent rehabilitation treatment and completed the following predictors: 10-m walking test (10MWT), Timed Up and Go (TUG), Berg Balance scale (BBS), Activities-specific Balance Confidence scales (ABC), Freezing of Gait (FOGQ) and PDQ-39. Two general linear models were calculated to predict the relationship between HRQoL at baseline and to predict HRQoL changes after rehabilitation. Results: Forty-two subjects (age 74.9 ± 7.3 years, Hoehn&Yahr 2.8 ± 0.6) completed the baseline evaluation. The first model (multiple R2 = 0.59, F = 5.86, P < 0.001) showed that ABC (B = − 0.51, CI = − 0.86 to 0.15, R2 = 0.41, P = 0.005) and FOGQ (B = 2.38, CI = 1.03 to 3.73, R2 = 0.07, P = 0.001) were statistically significant predictors of mobility aspect of HRQoL at baseline. Thirty seven subjects completed the rehabilitation sessions, data were entered in the second model (multiple R2 = 0.40, F = 4.24, P < 0.004) showing that gender (B = − 5.12, CI = − 9.86 to − 0.39, R2 = 0.23, P = 0.034), Hoehn&Yahr (B = 10.93, CI = + 3.27 to + 18.61, R2 = 0.22, P = 0.006) and PDQ-39 mobility at baseline (B = − 0.38, CI = − 0.63 to − 0.14, R2 = 0.55, P = 0.002) were statistically significant predictors of changes of the mobility aspect of HRQoL. Conclusions: Balance confidence and Freezing of Gait are associated with the mobility aspect of HRQoL. Changes in mobility domain of HRQoL (as assessed by PDQ-39) are likely to be greater in males, in people at higher stages of the disease and in people with more severe limitation in mobility domain of HRQoL (as assessed by PDQ-39) before rehabilitation. Results might be different when considering different outcomes or different measures for the same outcome (performance mobility test instead of self-report questionnaires). Further investigations are needed to better understand other components of HRQoL in addition to mobility. Trial registration: NCT02713971 registered March 8, 2016. Keywords: Neurological disease, Parkinson, Neurorehabilitation, Prediction, Health-related quality of life, Participation * Correspondence: [email protected] IRCCS Fondazione Don Carlo Gnocchi, Via Capecelatro 66, 20148 Milan, Italy © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 2 of 9 Background knowledge about clinical and demographic characteris- Parkinson’s Disease (PD) is a common chronic neurode- tics associated to mobility domain of PDQ-39 will be generative disease affecting 1% of the population over useful for physiotherapist and clinicians to better under- 60 years of age with incidence and prevalence 1.5 to 2.0 stand the component of the HRQoL related to mobility times higher in men than in women [1]. PD affects phys- and the complexity of the whole construct. ical, mental and psychosocial health, impacting quality of life (QoL) [2, 3]. QoL is a multi-dimensional construct In addition, no previous studies inquired on variables defined as the individual self-perceived life function. [3, that could predict changes in HRQoL after rehabilita- 4] Most of the existing questionnaire used to describe tion. Since in the rehabilitation field it has become in- and assess subjects’ QoL actually measure Health-Re- creasingly important to understand what is associated to lated QoL (HRQoL) defined as the aspects of QoL most improvement in HRQoL, we hypothesize that clinical affected by ill health and self perceived health status. [5] and demographic characteristics can predict changes in Recent studies found that several health-related factors the mobility domain of PDQ-39 measured before and such as disease severity, disability, gait impairments, after rehabilitation to predict which subjects will im- complications arising from medication therapy, depres- prove in this domain after rehabilitation. For this reason, sive symptoms, psychosocial well-being and autonomic predictive models could be performed to improve the dysfunction are important contributors to HRQoL in design of interventions by providing a framework of in- people with PD [6, 7]. dividual characteristics and parameters prone to change thus improving estimation of patient care needs and im- Specific tools have been developed to measure HRQoL provement in therapeutic plans. Furthermore, predictive in people with PD and to assess emotional status, models could be used to provide sample size estimates cognitive functions and life’s health before and after re- for trials aimed at improving HRQoL in people with PD. habilitation [3]. The Parkinson’s Disease Questionnaire (PDQ-39) is a widely used tool to evaluate different do- In keeping with our research hypothesis, the aims mains related to HRQoL in people with PD, including of this pilot study were to: 1) assess the relationship mobility, activities of daily living, emotional well-being, between HRQoL mobility domain of PDQ-39 at base- stigma, social support, cognition, communication and line and clinical and demographic characteristics of a pain [8]. These domains of HRQoL have been found sample of people with PD, and 2) develop a model to related to subjects’ demographic and clinical characteris- predict changes in HRQoL mobility domain of PDQ- tics such as age, gender, disease severity, disease dur- 39 after rehabilitation. ation, motor and non-motor PD specific symptoms, and subsequent limitations in mobility and gait [9]. Methods Mobility and gait limitations are major issues for Design people with PD and these clinical characteristics affect This study combines a cross-sectional design to assess their daily life activities and participation in society. Sev- the relationship between HRQoL mobility domain of eral studies have been carried out to understand the re- PDQ-39 at baseline and clinical and demographic char- lationship between mobility, gait limitations and HRQoL acteristics of a sample of people with PD and a longitu- with mixed results [10, 11]. Other authors [3] showed dinal design (assessments pre and post treatment) to that clinical characteristics (dynamic balance and execu- develop a model to predict changes in HRQoL mobility tive function) result as significant predictors of HRQoL domain of PDQ-39 after rehabilitation. but this finding was in contrast by previous studies [6, 7] and the current literature is not exhaustive with respect Subjects to this topic. Further, the effect of other clinical charac- Forty-two people with PD were consecutively recruited teristics of mobility on HRQoL, such as self-perception for the study between 2013 and 2015. The eligible popu- of walking and balance abilities, that were not consid- lation included all PD (inpatient and outpatient) living ered in previous studies, need further investigation. Fi- in the catchment areas requiring rehabilitation. Inclusion nally we know that studies assessing HRQoL tend to use criteria were: subjects older than 18 years, Hoehn & Yahr questionnaires reducing this multi-dimensional con- stage between I and IV, ability to walk for 6 m with or struct as a summary index [5] but no studies have without walking aids, ability to maintain standing pos- investigated the relationship between the mobility com- ition for at least 10s but inability to stay on one leg ponents of HRQoL construct (measured by PDQ-39 stance more than 10s, having mini-mental state examin- mobility domain) and the clinical and demographic ation (MMSE) > 24 to be able to fulfill patients reported characteristics. outcome and stable drug therapy (to be monitored throughout the treatment period). People with PD who Since mobility domain of PDQ-39 is a fundamental had deep brain stimulation surgery, mini-mental state component of HRQoL construct we hypothesize that examination (MMSE) < 24 and changed drug therapy

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 3 of 9 before the end of the treatment were excluded.People impairment [18] and 8) The Parkinson’s Disease with PD underwent individual training consisting of 20 Questionnaire (PDQ-39), used to evaluate HRQoL in sessions of 45 min each, 3 times a week. They received people with PD. It consists of 39 items divided in 8 balance and gait exercises defined by clinical staff for domains, items 1 to 10 for mobility, items 11 to 16 each patient with or without biofeedback. Exercises were for activities of daily living, items 17 to 22 for emo- performed in different sensory conditions and/or includ- tional well-being, items 23 to 26 for stigma, items 27 ing a dual-task. Few minutes of muscle stretching, and to 29 for social support, items 30 to 33 for cognition, mobilization exercises were also provided. Each session items 34 to 36 for communication and items 37 to 39 was performed while subjects were in the “on” medica- for pain. Scores for each domain are expressed as a tion state, for further information on the study protocol percentage (100% indicating greater dissatisfaction/dis- see Carpinella et al [12] ruption within a domain). The total score is com- puted by summing the 8 domain scores divided by Subjects signed an informed consent form before the the total number of domains. In the present study, beginning of the study. The study was approved by the the mobility domain of the PDQ-39 was used to de- Ethical Committee of Don Gnocchi Foundation. scribe the mobility aspect of HRQoL [8, 19]. Clinical assessment Statistical analysis Subjects were assessed by a rater before and after the 20 Parametric descriptive statistics were used to describe rehabilitation sessions. Each assessment was performed demographic and clinical characteristics of the sample while subjects wore normal shoes and were in the “on” and to detect the presence of outliers. Data distribu- medication state. tion was checked for normality and data with a skew- ness score greater than 1 or − 1 were transformed, The clinical assessment included: 1) The 10-m BBS cubed scores and TUG logarithm scores were walking test (10MWT) used to assess gait speed. Sub- calculated. Paired T test was used to compare general jects walked with or without walking device at a pre- improvement of PDQ-39 mobility domain before and ferred walking speed along a 10-m walkway. The time after rehabilitation. required to cover the middle 6 m of the walkway was recorded [13]; 2) The Timed Up and Go (TUG) test, Aim 1) To assess the relationship between HRQoL used to assess mobility and dynamic balance, by mobility domain of PDQ-39 at baseline and clinical and measuring the time taken by the subject to rise from demographic characteristics the Hosmer and Lemeshow a chair, walk 3 m, turn around, walk back to the chair [20] two-step approach has been used to decrease re- and sit down [14]; 3) The Berg Balance Scale (BBS), dundancy and to reduce number of predictors due to used to measure subject’s balance during tasks involv- small sample size. In the first step univariate analyses ing sitting, standing and positional changes. The scale (Pearson correlation coefficient and independent sample consists of 14-items that rates function on as scale T test) between mobility domain of PDQ-39 at baseline from 0 (worst) to 4 (best). Maximum total score is 56 and clinical and demographic characteristics were calcu- [15]; 4) The Activities-specific Balance Confidence lated. In the second step (multivariate analysis) only var- (ABC) questionnaire, used to assess balance confi- iables associated at univariate analyses (P < 0.2 with dence and balance self-perception. Score range is PDQ-39 mobility domain at baseline) were entered in a from 0 to 100, where 100 means high self-perception general linear model as independent variables while in balance skills [16]; 5) The Freezing of Gait Ques- PDQ-39 mobility domain score at baseline was used as tionnaire (FOG-Q), used to assess freezing during dependent variable. walking and its severity. It consists of six items with a score ranging from 0 to 24, where 0 means no Aim 2) The same two-step statistical approach [20] freezing [17]; 6) The Unified Parkinson Disease Rating of the first aim has been used to develop a model to Scale-Motor Section (UPDRS III), used to assess dis- predict changes in HRQoL mobility domain of ease severity and disease-specific impairments, PDQ-39 after rehabilitation. In the first step univari- through 27 motor items about the clinical spectrum ate analysis (Pearson correlation coefficient and inde- of motor symptoms typical of PD, such as tremor, ri- pendent sample T test) was calculated between gidity, bradykinesia and impairment of axial motor PDQ-39 mobility domain change score (post re- function [6]; 7) The Mini Mental State Examination habilitation score - baseline score) and demographic (MMSE) is used to assess cognitive aspects of mental and clinical variables (including PDQ-39 mobility do- functions and consists of 11 simple questions or tasks main at baseline) to calculate associations between grouped into 7 cognitive domains. The score ranging HRQoL changes and clinical predictors. In the sec- from 0 to 30 and score of < 24 is the generally an ac- ond step (multivariate analysis) clinical predictors as- cepted cutoff indicating the presence of cognitive sociated at univariate analysis) were entered in the

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 4 of 9 second general linear model as independent predic- Table 2 Correlations (Pearson) between demographic and tors while PDQ-39 mobility domain change score clinical variables and PDQ-39 mobility domain at baseline was used as dependent variable. For both models we checked for collinearity (variance inflation factor < 6), PDQ-39 mobility P-value distribution of residuals and influential points and domain (baseline) the level of statistical significance was set at P < 0.05. Statistical analyses were performed using STATIS- GENDER 0.37 0.019* TICA 9.0. AGE (years) 0.12 0.452 TIME FROM ONSET (years) 0.15 0.357 H&Y 0.38 0.015* Results 10MWT(s) 0.46 0.003* Demographic and clinical characteristics of the recruited sample (28 males and 14 females) are shown in Table 1. TUG(s) 0.47 0.002* BBS −0.40 0.009* ABC − 0.63 < 0.001* Aim1-assessment of the relationship between HRQoL FOGQ 0.53 < 0.001* mobility domain of PDQ-39 at baseline and clinical and UPDRS part 3 0.28 0.073* demographic characteristics MMSE −0.06 0.729 Table 2 shows correlations and p-values between demo- graphic and clinical characteristics and PDQ-39 mobility PDQ39 mobility domain Parkinson’s Disease Questionnaire-mobility domain, domain at baseline. H&Y Hoehn and Yahr, 10MWT 10-m walking test, TUG Timed up and go, BBS Berg Balance Scale, ABC Activities Balance Confidence, FOGQ Freezing of gait Males had statistically significant lower scores on the questionnaire, UPDRS part 3 Unified Parkinson Disease Rating Scale-Motor PDQ-39 mobility domain compared to female (Male: Section part 3, MMSE Mini Mental State Examination. * P-value < 0.2 41.9 ± 24.7, Female 61.1 ± 21.7, P = 0.02). Correlations between PDQ-39 mobility domain at baseline and gen- and the ABC (B = − 0.51, CI = − 0.86 to 0.15, R2 = 0.41, P = der; H&Y, 10MWT, TUG, BBS, ABC, UPDRS part 3 and 0.005) and the FOGQ (B = 2.38, CI = 1.03 to 3.73, R2 = 0.07, FOGQ (P < 0.2) were entered in the multivariate model (Table 3) investigating effect of baseline values on mobil- P = 0.001) were significantly associated with the PDQ-39 ity aspect of HRQoL. No statistically significant correla- tions were found between PDQ-39 mobility domain at mobility domain at baseline. baseline and age, time from onset and MMSE (Table 2) therefore these variables were not entered in the model. Aim2-development of a model to predict changes in The general linear model fitted the data (multiple R2 = 0.59, adjusted R2 = 0.49, F = 5.86, P < 0.0001) HRQoL mobility domain of PDQ-39 after rehabilitation Table 4 shows correlations and p-values between demo- Table 1 Demographic and clinical characteristics of the sample graphic and clinical characteristics and PDQ-39 mobility domain change score. Measure Mean (SD) Max Min No statistically significant correlations were found be- AGE (years) 74.9 (7.3) 89.8 57.56 tween PDQ-39 mobility domain change score and age, time from onset, 10MWT, TUG, BBS, ABC and UPDRS TIME FROM ONSET (years) 9.2(5.0) 21.3 0.44 part 3 (Table 4). Statistically significant correlations (P < 0.2) were instead found between PDQ-39 mobility PDQ-39 31.7(14.1) 63.8 7.1 change score and gender, H&Y, FOGQ, MMSE and PDQ-39 mobility domain at baseline and these variables PDQ-39 mobility domain 48.3(25.2) 95 2.5 were entered in the multivariate model. H&Y 2.8(0.6) 42 Change score on the PDQ-39 mobility domain was − 6.2 (±14.6) points suggesting an overall improve- 10MWT (s) 13.3(7.4) 50.3 6.94 ment in HRQoL after rehabilitation (Paired T Test, T = 2.56, P = 0.015). TUG (s) 19.5(12.9) 72 7 Five subjects were missing at post assessment BBS 44.2(9.7) 56 17 resulting in data of thirty-seven subjects being in- cluded in the second multivariate model (Table 5) ABC 52.4(21.6) 97.5 13.13 predicting rehabilitation outcome of HRQoL in people with PD. FOGQ 12.6(4.4) 21 2 The general linear model fitted the data (multiple UPDRS part 3 19.8(7.3) 37 4 R2 = 0.40, adjusted R2 = 0.31 F = 4.24, P < 0.004) and Gender (B = − 5.12, CI = − 9.86 to − 0.39, R2 = 0.23, P = MMSE 27.6(1.9) 30 22.4 0.034), H&Y (B = 10.93, CI = + 3.27 to + 18.61, R2 = 0.22, P = 0.006) e PDQ-39 mobility domain at baseline (B = − 0.38, Data are represented as means, standard deviations (SD), max and min values CI = − 0.63 to − 0.14, R2 = 0.55, P = 0.002) were significant PDQ-39 Parkinson’s Disease Questionnaire, PDQ-39 mobility domain Parkinson’s Disease Questionnaire - mobility domain, H&Y Hoehn and Yahr, 10MWT 10-m walking test, TUG Timed up and go, BBS Berg Balance Scale, ABC Activities Balance Confidence, FOGQ Freezing of gait questionnaire, UPDRS part 3 Unified Parkinson Disease Rating Scale-Motor Section part 3, MMSE Mini Mental State Examination

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 5 of 9 Table 3 Model analysis between demographic and clinical variables and PDQ-39 mobility domain at baseline DEPENDENT VARIABLE multiple R2 adjusted R2 F P value INDEPENDENT VARIABLES Β CI (−95 to + 95%) R2 P value PDQ-39 mobility domain (baseline) 0.59 0.49 5.86 < 0.0001* Gender −2.05 −9.34 to + 5.24 0.30 0.571 H&Y 0.01 −15.55 to + 15.58 0.58 0.998 TUG 36.16 −29.78 to + 102.1 0.84 0.272 10MWT −9.50 −86.31 to + 67.30 0.81 0.802 BBS 0.00004 −0.002 to + 0.003 0.68 0.758 ABC −0.51 − 0.86 to + 0.15 0.41 0.005* FOGQ 2.38 + 1.03 to +3.73 0.07 0.001* UPDRS part 3 −0.26 −1.45 to + 0.91 0.52 0.645 PDQ-39 mobility domain Parkinson’s Disease Questionnaire mobility domain, H&Y Hoehn and Yahr, TUG Timed up and go, 10MWT 10-m walking test, BBS Berg Balance Scale, ABC Activities Balance Confidence, FOGQ Freezing of gait questionnaire, UPDRS part 3 Unified Parkinson Disease Rating Scale-Motor Section part 3. * P-value < 0.05 predictors of a change on the PDQ-39 mobility domain mobility related QoL. This information is useful to de- after rehabilitation. fine criteria to include people with PD in rehabilitation program having mobility related HRQoL as the main Discussion outcome, to set rehabilitation goals and to identify The aim of this study was to develop two models with causes of failure to recover. baseline demographic and clinical variables predicting, respectively, to mobility-relate HRQoL and its change Aim1-assessment of the relationship between HRQoL after rehabilitation in people with PD. The main results mobility domain of PDQ-39 at baseline and clinical and indicate that mobility aspect of HRQoL is mostly associ- demographic characteristics ated with subject’s perception of gait and balance disor- Univariate correlation analysis showed moderate correl- ders (measured by FOG-Q and ABC questionnaires ation between mobility related HRQoL and gait and bal- respectively). Conversely, improvement in mobility re- ance disorders. This was true both for self-administered lated HRQoL following rehabilitation was predicted by a tests inquiring on subject’s perception of their balance worse PDQ-39 mobility domain at baseline, higher confidence (ABC) and freezing of gait (FOG-Q), and disease severity, and being male suggesting that even rater-administered tests assessing static and dynamic subjects in their later stage of the disease can improve balance (BBS and TUG) and walking skills (10MWT). Conversely, mild correlations were found between mo- Table 4 Correlations (Pearson) between demographic and bility aspect of HRQoL and overall disability (H&Y and clinical variables and PDQ-39 mobility domain change score UPDRS III), cognitive function (MMSE) and demo- graphic characteristics (age and time from onset), sug- PDQ-39 mobility P-value gesting that limitations in walking and balance are more specifically associated to mobility related HRQoL [3]. domain (change score) These results are in line with previous studies [9, 21–25] investigating parameters affecting HRQoL in people with GENDER 0.21 0.20* PD. Some authors [9] found that postural instability and gait disorders predicted overall HRQoL, others [23] AGE (years) 0.14 0.41 pointed out that mobility disorders, mostly start hesita- tion, freezing, festination and difficulty in turning, are TIME FROM ONSET (years) −0.15 0.39 related to HRQoL. H&Y 0.22 0.20* When predictors were entered in the first general linear model balance self-perception (ABC) and freezing of gait 10MWT(s) 0.09 0.58 (FOG-Q) were better predictors of mobility aspect of HRQoL than rater-administered tests, even after control- TUG(s) −0.04 0.81 ling for all the other variables included in the model. Along those lines, previous studies [26–29] found that balance BBS −0.06 0.70 self-perception is associated with fear of falling in people with PD and is one of the main predictors of HRQoL. The ABC 0.09 0.58 relationship between balance self-perception, participation restrictions and HRQoL is also supported by studies FOGQ −0.27 0.10* UPDRS part 3 0.01 0.93 MMSE 0.23 0.17* PDQ39 mobility domain (baseline) −0.37 0.02* PDQ39 mobility domain Parkinson’s Disease Questionnaire-mobility domain, H&Y Hoehn and Yahr, 10MWT 10-m walking test, TUG Timed up and go, BBS Berg Balance Scale, ABC Activities Balance Confidence, FOGQ Freezing of gait questionnaire, UPDRS part 3 Unified Parkinson Disease Rating Scale-Motor Section part 3, MMSE Mini Mental State Examination. * P-value < 0.2

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 6 of 9 Table 5 Model analysis between demographic and clinical variables and PDQ-39 mobility domain change score DEPENDENT VARIABLE Multiple R2 Adjusted R2 F P value INDEPENDENT VARIABLES B CL (−95 to + 95%) R2 P value PDQ-39 mobility domain 0.40 0.31 4.24 0.004* Gender −5.12 −9.86 to −0.39 0.23 0.034* (change score) H&Y 10.93 +3.27 to + 18.61 0.22 0.006* MMSE 1.21 −1.17 to + 3.60 0.018 0.307 FOGQ −0.13 − 1.28 to + 1.03 0.36 0.824 PDQ-39 mobility domain (baseline) −0.38 −0.63 to − 0.14 0.55 0.002* PDQ-39 mobility domain Parkinson’s Disease Questionnaire mobility domain, H&Y Hoehn and Yahr, MMSE Mini Mental State Examination, FOGQ Freezing of gait questionnaire. * P-value < 0.05 suggesting that people with PD with low balance The results of the second general linear model taking self-perception as measured by the ABC are more likely to in to account univariate correlations showed that the use an assistive device to walk [27, 30] to improve the sense PDQ-39 mobility domain at baseline, severity of disease of safety increasing their mobility, independence and, con- (H&Y) and gender were significant predictors for sequently, HRQoL. changes in HRQoL. Besides low level of balance confidence, also freezing PDQ-39 mobility domain at baseline was found to be of gait was associated to lower mobility related to the best predictor. In particular those PD patients having HRQoL. This result is confirmed by previous studies worse HRQoL mobility domain at baseline tended to im- showing the impact of freezing on QoL in PD [31, 32]. prove more in HRQoL mobility after treatment. This Some authors [31, 33], showed that HRQoL decreases trend was confirmed by studies considering different proportionally with the severity of FOGQ scores and populations of subjects. For example, Asiri et al. [38] found that freezing has an independent, direct and sig- found that the most impaired post-stroke subjects nificant impact on HRQoL in people with PD even con- showed larger degree of improvement in gait speed after trolling for gait and mobility disorders. This can be due home-based training in subjects with lower HRQoL re- to the nature of freezing, consisting in an episodic event lated to mobility at baseline. Similarly, Altenburg et al. that causes a sudden and unpredictable inability to [39] found larger improvement after cardio-pulmonary maintain walking [32]. Often, people with PD are not rehabilitation in patients affected by chronic obstructive prepared for this event that can lead to perceived loss of pulmonary disease with low initial exercise capacity. It is control on their own body, compromising mobility and possible that PD subjects that were less affected had a leading to loss of independence and increased risk of floor effect on the PDQ-39 (12% of the sample) masking falling [31, 34, 35]. possible improvements in HRQoL. [40] On the other hand, worse baseline values might indicate a bigger po- Moreover, freezing of gait can have social conse- tential for improvement. We can also speculate that quences because frequent episodes in crowded situa- people with PD with low HRQoL related to mobility at tions, during social events or activities of everyday life baseline have entered a downward spiral of avoidance in become a source of stress, embarrassment and frustra- engaging activities of daily living, thus increasing partici- tion with consequences on emotional well-being [31, 36, pation restriction, deconditioning and demotivation. In 37]. Thus, it is possible that loss of control and motor this context rehabilitation may have increased ability in difficulties caused by freezing of gait, combined with participating in social events and motivation, maybe de- psychological distress, are reflected in a worse mobility creasing depression. Unfortunately, we did not take in related QoL. These findings underline the importance of account these psychological and non-motor symptoms efforts to alleviate freezing of gait and its related conse- that are considered as predictors of HRQoL outcome as quences, such as the negative impact on mobility aspect reported in recent studies [7, 41–43]. of HRQoL in people with PD. Disease severity was found to be the second best sta- Aim2-development of a model to predict changes in tistically significant predictor for rehabilitation outcome HRQoL mobility domain of PDQ-39 after rehabilitation in HRQoL. PD subjects with moderate disease severity Univariate correlation analysis showed that changes in (H&Y between 2 and 3), involving axial motor symp- mobility domain of HRQoL is mostly correlated to toms with balance and gait deficits, seemed to improve HRQoL mobility domain at baseline evaluation. Also, more their mobility aspect of HRQoL after rehabilita- the degree of disability (H&Y), cognitive function tion. Our results are in line with a systematic review [7] (MMSE) and demographic characteristics (gender) were suggesting that disease severity and motor features in- correlated with changes in mobility aspect of HRQoL. cluding gait impairments were the major predictors of poor HRQoL in people with PD in combination with

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 7 of 9 non-motor characteristics as depression and treatment- improvements after rehabilitation. Second, analysis of induced complications. Moreover, contemporary litera- follow-up data should be included to understand long ture [44] pointed out that factors as disease severity in- term predictors of changes in QoL related to mobility. fluences HRQoL but a better management strategy can Third, assessments were completed only during on slow down or lower their negative effects. A recent study medication state. Evaluation in off-medication state may by Rafferty et al., [45] demonstrated that long-term give more information about the relationship between HRQoL benefit following rehabilitation was greater in motor symptoms and mobility aspect of QoL. Fourth, people having moderate to advanced PD severity com- we did not take in account psychological and non-motor pared with those with mild PD severity. It is possible symptoms and a small percentage of people with PD that rehabilitation of the more impaired subjects leads to (12%) showed floor effects on the PDQ-39, leading to larger improvement in mobility and consequently in per- bias in data analysis. ception of mobility, thus increasing their confidence in performing activities considered too difficult before Even with these limitations our study shows that bal- rehabilitation leading to reduced disability and im- ance confidence and freezing of gait are associated with proved HRQoL. the mobility aspect of QoL. Changes in mobility (as assessed by PDQ-39) are likely to be greater in males, in Gender was the third significant predictor for rehabili- people at higher stages of the disease and in people with tation outcome in mobility aspect of HRQoL. Males had more severe mobility limitation (as assessed by PDQ-39) higher PDQ-39 mobility domain change scores following before rehabilitation. Results might be different when rehabilitation with a mean improvement of − 11.5 considering different outcomes or different measures for points, compared to − 0.4 points for female. As previ- the same outcome (performance mobility test instead of ously demonstrated, people with PD showed gender-re- self-report questionnaires). lated differences in disease experience and HRQoL perception factors that can have important clinical im- Considering that HRQoL is a multi-dimensional con- plications [46]. For example, being female has a negative struct, further research with larger sample size will be impact on drug and surgical outcome in PD treatments needed to find other predictors of the HRQoL domains [47–49], with females also showing poorer short and in addition to mobility domain. long-term motor outcome after subthalamic stimulation [47, 50]. Despite our results, contemporary literature Abbreviations stated that the effects of demographic characteristics 10MWT: 10-m walking test; ABC: Activities Balance Confidence; BBS: Berg (gender, age, level of education) on HRQoL in PD sub- Balance Scale; FOGQ: Freezing of gait questionnaire; H&Y: Hoehn and Yahr; jects are controversial [44] and considering our small HRQoL: Health Related Quality of like; MMSE: Mini Mental State Examination; sample size we should be careful to generalize our re- PDQ39: Parkinson’s Disease Questionnaire; QoL: Quality of life; TUG: Timed sults to the whole PD population. Although gender has up and go; UPDRS part 3: Unified Parkinson Disease Rating Scale-Motor Sec- been shown to influence brain anatomy, function, tion part 3 hormonal modulation, gene expression and levodopa bioavailability [49, 51, 52], further studies are needed Acknowledgements to better understand role of gender in rehabilitation Not applicable. outcome. Funding Keeping into account our results on demographic and The study was supported by the Italian Ministry of Health: grant no. GR-2009- clinical (motor) factors and the recent growing evidence 1604984. demonstrating the impact of non-motor characteristics on lives of people with PD [53] an efficient strategy to Availability of data and materials maintain and improve HRQoL in people with PD should The datasets used and/or analysed during the current study are available consist of a holistic, multidisciplinary, personalized and from the corresponding author on reasonable request. patient-centered approach with timely administration of palliative care and effectual involvement of caregivers Authors’ contributions and family members [44]. TB treated the subjects during rehabilitation phase, analyzed and interpreted the patient data and wrote the article, EG wrote the article, RP assessed the Limitation and conclusions subjects, JJ critical read the main document, MF critical read the main First and main limitation of this study is the sample size document, DC analyzed and interpreted the patient data and wrote the that was too small and, therefore, the results could not article, IC developed the feedback systems used during rehabilitation be generalized to all PD subjects and reduced the power phase, All authors read and approved the final manuscript. of the study. Further studies with larger sample size are needed to get firm conclusion on predictors of HRQoL Ethics approval and consent to participate The study was approved by the Ethical Committee of Don Gnocchi Foundation. Consent for publication Subjects signed an informed consent form before the beginning of the study. Competing interests The authors declare that they have no competing interests.

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 8 of 9 Publisher’s Note 22. Bloem BR, Grimbergen YA, Cramer M, Willemsen M, Zwinderman AH. Prospective assessment of falls in Parkinson's disease. J Neurol. 2001;248: Springer Nature remains neutral with regard to jurisdictional claims in published 950–8. maps and institutional affiliations. 23. Rahman S, Griffin HJ, Quinn NP, Jahanshahi M. Quality of life in Parkinson's Received: 21 February 2018 Accepted: 11 December 2018 disease: the relative importance of the symptoms. Mov Disord. 2008;23: 1428–34. References 1. Lee A, Gilbert RM. Epidemiology of Parkinson disease. Neurol Clin. 2016;34: 24. Gómez-Esteban JC, Zarranz JJ, Lezcano E, Tijero B, Luna A, Velasco F, Rouco I, Garamendi I. Influence of motor symptoms upon the quality of life of 955–65. patients with Parkinson's disease. Eur Neurol. 2007;57:161–5. 2. Opara JA, Brola W, Leonardi M, Błaszczyk B. Quality of life in Parkinson's 25. Lyons KE, Pahwa R, Troster AI, Koller WC. A comparison of Parkinson's disease. J Med Life. 2012;5:375–81. disease symptoms and self-reported functioning and well being. 3. Stegemöller EL, Nocera J, Malaty I, Shelley M, Okun MS, Hass CJ. NPF quality Parkinsonism Relat Disord. 1997;3:207–9. improvement initiative investigators timed up and go, cognitive, and 26. Bryant MS, Rintala DH, Hou JG, Protas EJ. Influence of fear of falling on gait quality-of-life correlates in Parkinson's disease. Arch Phys Med Rehabil. 2014; and balance in Parkinson's disease. Disabil Rehabil. 2014;36:744–8. 95:649–55. 4. Klepac N, Trkulja V, Relja M, Babić T. Is quality of life in non-demented 27. Bryant MS, Rintala DH, Graham JE, Hou JG, Protas EJ. Determinants of use of Parkinson's disease patients related to cognitive performance? A clinic- a walking de-vice in persons with Parkinson's disease. Arch Phys Med based cross-sectional study. Eur J Neurol. 2008;15:128–33. Rehabil. 2014;95:1940–5. 5. Karimi M, Brazier J. Health, health-related quality of life, and quality of life: what is the difference? PharmacoEconomics. 2016;34:645–9. 28. Bryant MS, Rintala DH, Hou JG, Protas EJ. Relationship of falls and fear of 6. Visser M, van Rooden SM, Verbaan D, Marinus J, Stiggelbout AM, van Hilten falling to activity limitations and physical inactivity in Parkinson's disease. J JJ. A comprehensive model of health-related quality of life in Parkinson's Aging Phys. 2015;23:187–93. disease. J Neurol. 2008;255:1580–7. 7. Soh SE, Morris ME, JL MG. Determinants of health-related quality of life 29. Grimbergen YA, Schrag A, Mazibrada G, Borm GF, Bloem BR. Impact of falls in Parkinson's disease: a systematic review. Parkinsonism Relat Disord. and fear of falling on health-related quality of life in patients with 2011;17:1–9. Parkinson's disease. J Parkinsons Dis. 2013;3:409–13. 8. Jenkinson C, Fitzpatrick R, Peto V, Greenhall R, Hyman N. The Parkinson's disease questionnaire (PDQ-39): development and validation of a 30. Constantinescu R, Leonard C, Deeley C, Kurlan R. Assistive devices for gait in Parkinson's disease summary index score. Age Ageing. 1997;26:353–7. Parkinson's disease. Parkinsonism Relat Disord. 2007;13:133–8. 9. Soh SE, McGinley J, Morris ME. Measuring quality of life in Parkinson's disease: selection of-an-appropriate health-related quality of life instrument. 31. Moore O, Peretz C, Giladi N. Freezing of gait affects quality of life of peoples Physiotherapy. 2011;97:83–9. with Parkinson's disease beyond its relationships with mobility and gait. 10. Christofoletti G, ME MN, Campbell MC, Duncan RP, Earhart GM. Investigation Mov Disord. 2007;22(15):2192–5. of factors impacting mobility and gait in Parkinson disease. Hum Mov Sci. 2016;49:308–14. 32. Walton CC, Shine JM, Hall JM, et al. The major impact of freezing of gait on 11. Amboni M, Barone P, Hausdorff JM. Cognitive contributions to gait and falls: quality of life in Parkinson's disease. J Neurol. 2015;262(1):108–15. evidence and implications. Mov Disord. 2013;28:1520–33. 12. Carpinella I, Cattaneo D, Bonora G, Bowman T, Martina L, Montesano A, 33. Perez-Lloret S, Negre-Pages L, Damier P, et al. Prevalence, determinants, and Ferrarin M. Wea-rable sensor-based biofeedback training for balance and effect on quality of life of freezing of gait in Parkinson disease. JAMA gait in Parkinson disease: a pilot ran-domized controlled trial. Arch Phys Neurol. 2014;71:884–90. Med Rehabil. 2017;98:622–30. 13. Amatachaya S, Naewla S, Srisim K, Arrayawichanon P, Siritaratiwat W. 34. Fahn S. The freezing phenomenon in parkinsonism. Adv Neurol. 1995;67:53–63. Concurrent validity of the 10-meter walk test as compared with the 6- 35. Bloem BR, Hausdorff JM, Visser JE, et al. Falls and freezing of gait in minute walk test in patients with spinal cord injury at various levels of ability. Spinal Cord. 2014;52:333–6. Parkinson’s disease. Mov Disord. 2004;19:871–84. 14. Huang SL, Hsieh CL, Wu RM, Tai CH, Lin CH, Lu WS. Minimal detectable 36. Giladi N, McMahon D, Przedborski S, et al. Motor blocks in Parkinson’s change of the timed “up & go” test and the dynamic gait index in people with Parkinson disease. Phys Ther. 2011;91:114–21. disease. Neurology. 1992;42:333–9. 15. Qutubuddin AA, Pegg PO, Cifu DX, Brown R, McNamee S, Carne W. 37. Okuma Y. Freezing of gait in Parkinson’s disease. J Neurol. 2006;253:27–32. Validating the berg balance scale for patients with Parkinson's disease: a key 38. Asiri FY, Marchetti GF, Ellis JL, et al. Predictors of functional and gait to rehabilitation evaluation. Arch Phys Med Rehabil. 2005;86:789–92. 16. Powell LE, Myers AM. The activities-specific balance confidence (ABC) scale. outcomes for persons poststroke undergoing home-based rehabilitation. J J Gerontol A Biol Sci Med Sci. 1995;50A:M28–34. Stroke Cerebrovasc Dis. 2014;23:1856–64. 17. Giladi N, Shabtai H, Simon ES, Biran S, Tal J, Korczyn AD. Construction of 39. Altenburg WA, de Greef MH, ten Hacken NH, et al. A better response in freezing of gait questionnaire for patients with parkinsonism. Parkinsonism exercise capacity after pulmonary rehabilitation in more severe COPD Relat Disord. 2000;6:165–70. patients. Respir Med. 2012;106(5):694–700. 18. Folstein MF, Folstein SE, McHugh PR. Mini-mental state: a practical method 40. Hagell P, Nygren C. The 39 item Parkinson's disease questionnaire (PDQ-39) for grading the cognitive state of patients for the clinician. J Psychiatr Res. revisited: implications for evidence based medicine. J Neurol Neurosurg 1975;12(3):189–98. Psychiatry. 2007;78(11):1191–8. 19. Peto V, Jenkinson C, Fitzpatrick R. PDQ-39: a review of the 41. Santos-García D, De la Fuente-Fernández R. Impact of non-motor symptoms development, validation and application of a Parkinson's disease quality on health-related and perceived quality of life in Parkinson's disease. J of life questionnaire and its associated measures. J Neurol. 1998;245 Neurol Sci. 2013;332:136–40. Suppl 1:S10–4. 42. Schrag A, Jahanshahi M, Quinn N. What contributes to quality of life in 20. Lemeshow S, Hosmer DW. A review of goodness of fit statistics for use in patients with Parkinson's disease? J Neurol Neurosurg Psychiatry. 2000;69: the development of logistic regression models. Am J Epidemiol. 1982;115: 308–12. 92–106. 43. Menon B, Nayar R, Kumar S, et al. Parkinson's disease, depression, and 21. Wielinski CL, Erickson-Davis C, Wichmann R, Walde-Douglas M, Parashos SA. quality-of-life. Indian J Psychol Med. 2015;37:144–8. Falls and injuries resulting from falls among patients with Parkinson's 44. Fereshtehnejad SM. Strategies to maintain quality of life among people disease and other parkinsonian syndromes. Mov Disord. 2005;20:410–5. with Parkinson's disease: what works? Neurodegener Dis Manag. 2016;6: 399–415. 45. Rafferty MR, Schmidt PN, Luo ST, et al. Regular exercise, quality of life, and mobility in Parkinson's disease: a longitudinal analysis of National Parkinson Foundation quality improvement initiative data. J Parkinsons Dis. 2016;7: 193–202. 46. Lubomski M, Louise Rushworth R, Lee W, et al. Sex differences in Parkinson's disease. J Clin Neurosci. 2014;21(9):1503–6. 47. Romito LM, Contarino FM, Albanese A. Transient gender-related effects in Parkinson's disease patients with subthalamic stimulation. J Neurol. 2010; 257:603–8. 48. Cantuti-Castelvetri I, Keller-McGandy C, Bouzou B, et al. Effects of gender on nigral gene expression and Parkinson disease. Neurobiol Dis. 2007;26:606–14.

Bowman et al. Archives of Physiotherapy (2018) 8:10 Page 9 of 9 49. Cahill L. Why sex matters for neuroscience. Nat Rev Neurosci. 2006;7:477–84 Review. 50. Accolla E, Caputo E, Cogiamanian F, et al. Gender differences in patients with Parkinson’s disease treated with subthalamic deep brain stimulation. Mov Disord. 2007;22:1150–6. 51. Shulman LM. Gender differences in Parkinson’s disease. Gend Med. 2007;4:8–18. 52. Kowal SL, Dall TM, Chakrabarti R, et al. The current and projected economic burden of Parkinson’s disease in the United States. Mov Disord. 2013;28:311–8. 53. Todorova A, Jenner P, Ray Chaudhuri K. Non-motor Parkinson’s: integral to motor Parkinson’s, yet often neglected. Pract Neurol. 2014;14:310–22.