Hong Kong Physiotherapy Journal

Validity and feasibility of using a seated push-up test among community-dwelling older adults 127 Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com Based on the concept of global physiological gender, height, bodyweight, vital signs, underlying by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. change throughout the body systems and the diseases, and walking device used (if any). Then, closed-association of the musculoskeletal system, the participants were assessed for their ability to the present researchers hypothesised that various perform SPUTs, and standard measures for the forms of SPUTs could be applied in older indivi- body composition, muscle strength, and mobility of duals with di®erent functional levels and that their the older individuals. Details of the tests are outcomes would re°ect many aspects necessary for explained below. the independence of these individuals, depending upon the particular characteristics of the tests. Seated push-up tests: Many forms of SPUTs are Therefore, this study assessed the discriminative clinically available, including untimed and timed and concurrent validity, as well as the feasibility, of SPUTs.9,10 With the aim to report the feasibility of various types of SPUTs—including the 1SPUT using SPUTs among older individuals, this study along with its ULL-SPUT, 5-time SPUT (5SPUT), applied simple SPUTs in increasingly demanding 10-time SPUT (10SPUT), and 1-min SPUT forms, including the 1SPUT along with its ULL- (1minSPUT)—as compared to the results of stan- SPUT, 5SPUT, 10SPUT, and 1minSPUT, dard measures for body composition, muscle according to the participant's ability to complete strength, and mobility among community-dwelling the tests (with no pressure to complete all SPUTs if older adults. they were unable). Before and after the tests, the participants engaged in a warm-up and stretching Methods session to reduce the risk of musculoskeletal injury, which can occur after completing such demanding Participants measures. The various SPUTs were completed as follows. This observational study was conducted among community-dwelling individuals aged 65 years and The 1SPUT was executed using push-up loading older with a body mass index (BMI) of between devices at the size of standard clinical push-up 18.5 kg/m2 and 29.9 kg/m2. The eligible partici- boards (18 cm in height) to quantify the ULL- pants needed to have the ability to stand up in- SPUT. The devices were developed from digital dependently, walk with or without a walking load cells (Model L6E3-C, 50 kg-3G, with the device, and understand the instructions for the standard calibration method based on UKAS LAB tests in this study. Individuals were excluded from 14: 2006; mini-patent application number the study if they had any signs or symptoms that 2103001612).11,12 After calibration, the tools were might a®ect their participation in the study, such accurate up to < 0:1 kg, with a measurement un- as uncontrolled medical conditions (e.g., hyper- certainty of Æ 0.1 kg. Participants were in a ring tension or heart disease); pain in the musculoskel- sitting position and placed their hands on the push- etal system that might a®ect outcomes of the up loading devices slightly anterior to their hips study, such as a rotator cu® injury; and a history of (Fig. 1). Then, they pushed both hands against the shoulder or upper limb problems that limits their devices, lifted the body from the °oor while slightly ability to perform SPUTs (i.e., a pain score of more bending the trunk forward and depressing both than 5 out of 10 on a visual analog scale). All scapulars, and gradually bent the elbows to sit participants signed written informed consent forms down on the °oor.9,11,12 The test was repeated over that were approved by the Institutional Ethics three trials, with a su±cient rest period between Committee for Human Research (HE 611600). The the trials. Outcomes of the test were recorded in estimated minimum sample size for this study was terms as either pass or fail; a pass was de¯ned as 82 participants, when R0 ¼ 0:0 and the lowest R1 the ability to lift the body from the °oor success- from a pilot study of 0.31 (n ¼ 40), with 90% fully in at least two of the three trials; if not, the power and an alpha value of 0.05.14 outcome was regarded as a fail.12 In addition, the average data of the maximum ULL-SPUT over the Research protocols three trials, which was automatically generated by the push-up loading devices, was recorded.11,12 The eligible participants were interviewed and assessed for their demographics, including age, Participants who failed a 1SPUT terminated the SPUTs, and were assessed using standard measures. Participants who passed continued the timed-based SPUTs, including the 5SPUT,

128 P. Poncumhak et al. Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com Fig. 1. Testing protocols of seated push-up tests. (A) Starting position with push-up loading devices. (B) Position while lifting by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. the body from the surface with push-up loading devices. (C) Starting position with push-up boards. (D) Position while lifting the body from the surface with push-up boards. 10SPUT, and 1minSPUT—using standard clinical (at least a minute). The number of participants push-up boards and a starting position similar to who could complete each method of SPUT was that used for the 1SPUT. For the 5SPUT and recorded, along with the adverse events (if any), 10SPUT, the participants were timed on their such as musculoskeletal pain, chest pain, or acci- ability to complete 5 and 10 SPUT repetitions in dental events, for the consideration on the feasi- the fastest and safe manner from the instruction bility of the SPUTs. \\start\" until the participant's buttock touched the °oor on the last repetition. The average time over Standard measures: The participants were the three trials was recorded. For the 1minSPUT, assessed using standard measures to indicate their the participants were assessed for the maximum body composition (i.e., lean body mass [LBM], number of SPUT repetitions they could do in one bone mineral content [BMC], and body fat mass), minute over one trial. During the test, they could muscle strength, and mobility (Table 1) by an take a period of rest as required and continue the experienced assessor in a random order. The par- test as soon as they could; otherwise, they termi- ticipants could take a period of rest between the nated the test if they were unable to continue. tests and the trials as required (or at least 30 s), in order to minimise the learning e®ects and fatigue The SPUTs were assessed on a hard and level that might occur due to the sequence of the tests. surface by an experienced rater (intraclass corre- They were fastened with a lightweight safety belt lation coe±cients [ICCs] = 0.932–1; p < 0:001). around their waist so that the assessor could pro- The participants could take a period of su±cient vide e±cient assistance if needed. Details of the rest between the trials and the tests as required standard measures are described in Table 1.15–34

Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com Validity and feasibility of using a seated push-up test among community-dwelling older adults 129 by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Notes: SPUT = seated push up test, ULL-SPUT = upper limb loading during a seated push up test Fig. 2. Participation °owchart.

130 P. Poncumhak et al. Table 1. Details of the standard measures for body compositions, muscle strength, and mobility. Outcome measure Aims of the assessments/ Details of the measurement psychometric properties Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com Body Dual-energy X-ray Body compositions (lean body Participants were in a supine position by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. compositions absorptiometry mass [LBM], bone mass content on a DXA table with the arms (DXA) [BMC], and body fat mass placed on their sides, the legs Muscle strength [FM]).16 Inter- and intra-tester extended, and the toes facing tests Handgrip test reliability (ICC = 0.97–0.98).17 upward according to the standard protocols recommended by GE- Mobility Back and leg strength Upper limb muscle strength, Healthcare.18 The data of body assessments test physical frailty, and disability compositions, including LBM, in older individuals. Excellent BMC, and body fat mass, were Five times sit-to-stand test-retest reliability (ICC = automatically generated by the test 0.912–0.954).19 machine in kilograms. Timed up-and-go test The back and leg extensor muscle Participants were in a sitting position strength. Excellent intra-tester and squeezed the handle as much 10-m walk test reliability (ICC = 0.97).21 as they could over three trials per hand with a su±cient rest interval Functional lower extremity (at least 30 s) between the trials, muscle strength and dynamic and the maximum force was balance control while changing recorded.20 postures.15,23 Test-retest reliability (ICC = 0.81; SEM = Participants stood on the base of a 0.9 s).23,24 MDC95 ¼ 2:5 s.24 back-leg-chest dynamometer with Concurrent validity with the the chain adjusted according to TUG test; r ¼ 0:64, their height and the protocols for p < 0:001.24 the trunk or leg extensor muscles explained previously. Then, they Mobility, dynamic balance pulled the chain with their control, and risk of falling in maximum force, holding it for 3 s. older individuals.26,27 Test- A rest period between the trials retest reliability (ICC = was allowed as needed (at least 0.97).28 Inter-tester reliability 30 s). The average force in the (ICC = 0.99).26 three trials for each method was recorded in kilograms.22 Overall quality of gait, community participation, Participants were timed their ability health condition, morbidity, to complete ¯ve chair-rise cycles at and mortality rates.29 Test- the fastest possible safe speed retest reliability (ICC = 0.92) without using their arms. The and MDC = 0.22 m/s.30 average time over the three trials was reported.25 Participants were timed their ability of standing up from a standard armrest chair, walking at the fastest and safe speed around a tra±c cone at 3 m from the chair, returning, walking back to sit back down on the chair. The average time over the three trials was reported.25 Participants were timed their ability of walking at a comfortable pace along the middle 4 m of the total 10-m walkway. The average time over the three trials was then converted to a walking speed.31,32

Validity and feasibility of using a seated push-up test among community-dwelling older adults 131 Table 1. (Continued ) Outcome measure Aims of the assessments/ Details of the measurement Two-minute step test psychometric properties Functional capacity and physical Participants raised their knee to a endurance.33 Test-retest mid-thigh level, that is, the mid distance between the iliac crest reliability (ICC = 0.90) and and the patella, marking the point on the wall. The total number of concurrent validity with 1-mile steps in place, that is, the number walking time (r ¼ 0:73).34 of times the right knee reached the target level in 2 min over one trial Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com was recorded.33 by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Notes: ICC: intraclass correlation coe±cient; SEM: standard error of measurement; MDC: minimal detectable change. Statistical analysis Results Descriptive statistics were used to describe the In total, 82 individuals, with an average age of 74 participants' characteristics and the ¯ndings of the years and a normal BMI, completed the study. study. The independent samples t-test and Mann– Most participants were well-functioning females, Whitney U-test were used to compare the data physically active, and able to perform daily activ- between the participants who passed and failed the ities independently, without mobility devices 1SPUT for the data with a normal and non-normal (n ¼ 72; 88%; Table 2). Of all, 57 participants were distribution, respectively (i.e., the discriminative assessed for their body composition because this validity). The Pearson correlation coe±cient was variable was additionally included after the initia- used to analyse the correlations of continuous data tion of the study. between the SPUTs and standard measures (i.e., the concurrent validity). The correlation level All 82 participants could complete a 1SPUT; 62 was interpreted as being very low or negligible participants (75.6%) passed the test and proceeded (r ¼ 0–0.30), low (r ¼ 0:30–0.50), moderate to be assessed with other forms of the SPUTs, (r ¼ 0:50–0.70), high or strong (r ¼ 0:70–0.90), or while the rest of them (n ¼ 20) failed in the test excellent (r ¼ 0:90–1.00).14 Therefore, the closer and continued with standard measures. Most par- the correlation coe±cient was to 1, regardless of ticipants who failed were female (n ¼ 18; 90%), the direction, the stronger was the existing asso- and approximately one-third of them (n ¼ 7; 35%) ciation, indicating a linear relationship between the used a single cane for daily movement. The average SPUT data and the standard measures. The level ULL-SPUT of the participants who passed was of statistical signi¯cance was set at p < 0:05. 85% of their bodyweight and for those who failed was 71% of their bodyweight (p < 0:001; Table 3). Table 2. Personal data of all participants and of those who passed and failed a 1-seated push up test (1SPUT). Variable Total (n=82) Fail (n ¼ 20) Pass (n ¼ 62) P -value Agea (years) 74:6 Æ 6:5 (73.1–76) 76:5 Æ 7:4 (73–79.9) 74 Æ 6:1 (72.4–75.5) 0.120 Genderb (female) 49 (61) 18 (90) 29 (46.7) 0.002 Ã Bodyweighta (kg) Body heighta (m) 55:6 Æ 9:9 (53.5–57.8) 55:9 Æ 11:9 (50.3–61.5) 55:6 Æ 9:3 (53.2–57.9) 0.893 Body mass indexa (kg/m2) 1:5 Æ 0:1 (1.5–1.6) 1:5 Æ 0:1 (1.5–1.6) 1:6 Æ 0:1 (1.5–1.6) 0.906 Daily walking deviceb (Cane) 23 Æ 3:2 (22.3–23.8) 3:9 Æ 3:7 (22.2–25.6) 22:8 Æ 3:1 (22–23.6) 0.185 10 (12.2) 7 (35) 3 (4.8) < 0:001Ã Notes: Participants who could lift the body up from the °oor successfully in at least two over the three trials were arranged into the \\pass\" group, if not, they were placed in the \\fail\" group. aData are presented as mean Æ SD (95% con¯dence intervals), and compared between the pass and fail groups using the independent samples t-test. bThe data are presented using number (%) and compared between the groups using the Chi square test. ÃIndicated signi¯cant di®erences between the groups.

Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Table 3. Data comparisons between participants who passed and failed a one-time seated push up test (1SPUT). 132 P. Poncumhak et al. Variable Total (n ¼ 82) Fail (n ¼ 20) Pass (n ¼ 62) P -value Upper limb loading during a seated push up test (ULL-SPUT) ULL-SPUT (BW%) 82:3 Æ 9:8 (80.1–84.5) 71:2 Æ 12:9 (65.2–77.2) 85:9 Æ 5:1 (84.6–87.2) < 0:001a Standardmeasures 60:5 Æ 5:1 (56.6–64.5) 66:0 Æ 7:6 (63.8–68.2) 0.016 a 3:2 Æ 0:3 (2.9–3.4) 3:5 Æ 0:7(3.3–3.7) 0.044 a Body compositionc Lean body mass (BW%) 65:1 Æ 7:5 (63.1–67.1) 30:5 Æ 8 (28.2–32.9) 0.033 a 3:4 Æ 0:7 (3.3–3.6) 36 Æ 6 (31.4–40.7) Bone mineral content (BW%) 31:4 Æ 8 (29.3–33.5) 21:6 Æ 5:3 (20.3–23) < 0:001a 16:4 Æ 3:7 (14.6–18.1) 28:1 Æ 15 (24.3–31.9) 0.037 b Fat mass (BW%) 19:6 Æ 10:6 (14.6–24.5) 33:3 Æ 18:5 (28.6–38) 0.02 b 20 Æ 12:3 (14.3–25.8) Muscle strength Hand grip test (kg) 20:3 Æ 5:4 (19.2–21.5) 11:9 Æ 2 (11.4–12.4) < 0:001a Back extensor (kg) 26 Æ 14:5 (22.8–29.2) 16:5 Æ 3:2 (15–18) 13:2 Æ 2:9 (12.5–14) < 0:001b Leg extensor (kg) 30:1 Æ 18 (26.1–34) 16:1 Æ 3:2 (14.6–17.6) < 0:001a 0:8 Æ 0:2 (0.7–0.9) 1 Æ 0:2 (1–1.1) Mobility Five times sit-to-stand test (s) 13 Æ 3:1 (12.4–13.7) 47:1 Æ 16:3 (39.5–54.7) 57:2 Æ 13:5 (53:7 Æ 60:6) 0.007 a Timed up and go test (s) 13:9 Æ 3:2 (13.2–14.6) 10-m walk test (m/s) 2-min step test (times) 1 Æ 0:2 (0.9–1) 54:7 Æ 14:8 (51.5–57.9) Notes: The data are presented using mean Æ standard deviation (95% con¯dence interval), BW = bodyweight, s = second, m = meter. Superscripts indicates the p-values from athe independent samples t-test, bthe Mann-Whitney U-test, cthere were 57 participants in this variable as it was additionally included after initiation of the study, with 48 participants passed and 9 participants failed a 1SPUT. Table 4. The correlation between outcomes of seated push up tests and standard measures, including body compositions, muscle strength and mobility measures of the participants. Body compositions (g) Muscle strength (kg) Mobility Lean body Bone mineral Body fat Back muscles Leg muscles Five times Timed up 10-m walk 2-min step Variable mass content mass Handgrip strength strength sit-to-stand test (s) and go test (s) test (m/s) test (times) ULL-SPUT a (kg) 0.785 ÃÃ 0.628 ÃÃ 0.515 ÃÃ 0.547 ÃÃ 0.456 ÃÃ 0.345 ÃÃ À0.416 ÃÃ À0.288 ÃÃ 0.332 ÃÃ 0.247 Ã p-value < 0:001 < 0:001 < 0:001 < 0:001 < 0:001 0:001 < 0:001 0:009 0:002 0:025 5SPUTb (s) À0.0281 À0.306 Ã À0.059 À0.182 À0.115 À0.059 À0.393 ÃÃ 0.173 0.364 ÃÃ 0.398 ÃÃ p-value 10SPUTb (s) 0.053 0:035 0.691 0.156 0.372 0.650 0:002 0.178 0:004 0:001 p-value À0.166 À0.164 À0.032 À0.049 À0.149 À0.126 À0.355 ÃÃ 0.207 0.361 ÃÃ 0.357 ÃÃ 1minSPUTb (times) p-value 0.261 0.264 0.828 0.703 0.247 0.329 0:005 0.107 0:004 0:004 0.325 Ã 0.244 À0.100 0.356 ÃÃ 0.425 ÃÃ 0.466 ÃÃ À0.526 ÃÃ À0.306 Ã 0.332 ÃÃ 0.421 ÃÃ 0:024 0:005 0:001 < 0:001 0.095 0.498 < 0:001 0:016 0:008 0:001 Notes: p-values were derived from Pearson correlation coe±cients. Bold characters indicate the data with signi¯cant correlation at Ãp < 0:05 and ÃÃp < 0:01. aThis variable was analysed in 57 participants for body compositions, and 82 participants for muscle strength and mobility measures. bThese variables were analysed in 48 participants for body compositions, and 62 participants for muscle strength and mobility measures. ULL-SPUT: Upper limb loading during a seated push up test; 5SPUT: ¯ve-time seated push-up test; 10SPUT: 10-time seated push-up test; 1minSPUT: 1-min seated push-up test.

Validity and feasibility of using a seated push-up test among community-dwelling older adults 133 Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com The body composition, muscle strength, and mo- test (at approximately 71% of their bodyweight; by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. bility also showed signi¯cant di®erences between p < 0:001; Table 3). Previous studies reported that the groups (p < 0:05; Table 3). The ¯ndings fur- the ability to increase the ULL-SPUT requires the ther indicated a moderate to strong correlation complex interaction of many upper limb and upper between the ULL-SPUT data, and all body com- trunk muscles, as well as perceptual information position (r ¼ 0:515–0.785; p < 0:001; Table 4), working cooperatively to generate muscle force and muscle strength, and mobility measures joint torque to lift the body upward by both (r ¼ 0:247–0.547; p < 0:05) of the participants arms.9–11 Such ability requires SMM, a major part (Table 4). of LBM, in order for the muscles involved in the task to convert chemical energy into mechanical All participants who passed the 1SPUT could energy for force and power generation.10,19 Mus- complete other forms of SPUTs (n ¼ 62). Their cular contraction also imposes mechanical loading average 5SPUT time was 9:2 Æ 4:1 s (95% con¯- onto the bones, along with cardiovascular stress, dence interval [CI]: 8.6–9.7 s), 10SPUT time was while body fat mass acts as resistance when com- 17:7 Æ 3:5 s (95% CI: 16.9–18.7 s), and 1minSPUT pleting the task.19,20 Therefore, the 1SPUT and repetitions was 28:3 Æ 8:0 times (95% CI: 26.4–30.3 ULL-SPUT outcomes signi¯cantly correlated to times). There were no adverse events related to the body composition of the arms. Then the age- SPUTs reported by any of the participants. The related physiological changes occurring throughout outcomes of the 1minSPUT showed a signi¯cant the body and in all body systems enabled outcomes low to moderate correlation to LBM and muscle of the tests involving upper limb and upper trunk strength tests and mobility measures (r ¼ À0:306– muscles to re°ect standard measures involving 0.526; p < 0:05; Table 4) but not to BMC or body other body parts.1,2 Consequently, the present fat mass (p > 0:05). By contrast, the 5SPUT and ¯ndings indicate a signi¯cant correlation between 10SPUT data showed a signi¯cant low correlation the ULL-SPUT and all body composition, muscle to only the ¯ve times sit-to-stand test (FTSST), strength, and mobility measures investigated in 10-m walk test (10MWT) and 2-min step test this study (Table 4). Furthermore, the participants (2MST) (r ¼ À0:355–0.398; p < 0:01; Table 4). who passed the 1SPUT had the outcomes of body composition, muscle strength, and mobility Discussion measures signi¯cantly better than those of the participants who failed the test (p < 0:05; Table 3). This study assessed the validity and feasibility for various types of SPUTs among community-dwell- Nonetheless, outcomes of the 1SPUT and ULL- ing older people. All participants could complete SPUT may face ceiling e®ects in older individuals the 1SPUT and ULL-SPUT, wherein a pass or fail with good ability (i.e., score limitation at the top of in a 1SPUT could clearly discriminate participants a scale, i.e., always getting a pass in the test or with di®erent body composition, muscle strength, nearly 100% of their bodyweight35). Thus the and mobility outcomes (p < 0:05; Table 3). In ad- 1SPUT and ULL-SPUT outcomes in these cases dition, the ULL-SPUT showed a signi¯cant corre- may not represent the changes occurring in these lation to all body composition, muscle strength, participants, even there is actual change in the and mobility measures (Table 4). Among the time- participants' body composition, muscle strength, based SPUTs, the 1minSPUT showed signi¯cant or mobility. In such cases, other timed-based correlation with LBM and all muscle strength and SPUTs (e.g., the 5SPUT, 10SPUT, and 1min- mobility measures, whilst the 5SPUT and 10SPUT SPUT)—which can be applied only in those who showed a signi¯cant low correlation with only some pass the 1SPUT—may be utilised to further chal- mobility measures (Table 4). lenge the ability of these individuals. However, the 5SPUT and 10SPUT, which can be completed Of all forms of SPUTs investigated in this study, within a short duration, may be unable to clearly the 1SPUT along with its ULL-SPUT are the least re°ect the variability in ability level among these demanding measures and, thus they could be participants. Therefore, the 5SPUT and 10SPUT completed by all participants with poor and good outcomes showed a low correlation to only some functional ability (Table 3). Participants who mobility measures, including the FTSST, 10MWT, passed a 1SPUT had the ULL-SPUT approxi- and 2MST (r ¼ À0:355–0.398; p < 0:01; Table 4). mately 85% of their bodyweight which was signif- On the contrary, outcomes of the most challenging icantly greater than that of those who failed the

Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com 134 P. Poncumhak et al. well-functioning older individuals. The 1SPUT and by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. ULL-SPUT can be applied among those with good form of SPUT investigated in this study—namely, or poor functional ability, and they may be com- the 1minSPUT—might be able to detect the vari- pleted using a digital bathroom scale placed on a ability of participants with good functional ability. hard and even surface. The 1minSPUT—which Therefore, 1minSPUT outcomes showed signi¯- may be assessed using an armchair or using small cant correlation with LBM, muscle strength and wooden boxes placed on a ¯rm surface—can be mobility of the participants (r ¼ À0:306–0.526; employed to further challenge older individuals p < 0:05; Table 4). who have good functional ability. Such measures may be applied to early detect the abnormality Previous studies have suggested that the relating to their muscle strength, mobility and slightest physical reduction could transform a body composition change that may occurred in person from independence into one with disability. older individuals, especially among those with Therefore, it is highly suggested to detect and lower limb limitations or in settings with limited monitor any abnormality early on, as well as to area and equipment. improve their physical conditions since they are still functioning independently, so that disability Con°icts of Interest can be prevented or delayed.36,37 The present ¯ndings suggest the feasibility and validity of the The authors have no con°icts of interest relevant SPUTs—in various forms—as another practical to this paper. upper limb measure to indicate body composition, muscle strength, and mobility, which are necessary Funding/Support for the independence and safety of community- dwelling older individuals. Such measures can be This study was supported by funding from the completed in a small area over a hard and smooth Royal Golden Jubilee (RGJ) Ph.D. programme, surface (e.g., over a bed); and thus they can be Thailand Research Fund (TRF, grant no. PHD/ applied in various clinical, community-, and home- 0174/2560). For the remaining authors none were based settings. declared. However, there are some noteworthy limitations Author Contributions to this study. First, the study was conducted cross- sectionally among mostly well-functioning older All authors were involved in concept, design, and participants who had a BMI of less than 30 kg/m2. planning of the study. In addition, all of them also Thus, the ¯ndings may not clearly indicate the took part in critical revision of the paper for in- ability of the SPUTs to monitor changes in the tellectual content and ¯nalised the manuscript. body composition, muscle strength, and mobility of PP, PC and RI additionally contributed in collec- older individuals over time or of those who are frail tion and assembling of the data. PP and SP also or obese. Second, the SPUT measurements were took part in drafting of the manuscript. SA, SP taken using clinical push-up boards of a standard and PA provided the study materials and technical size, which may in°uence the outcomes of the test support. SA additionally provided the funding for participants with di®erent heights. Third, only support and administration of this project. 57 participants (69.5%) were assessed for their body composition because this variable was addi- References tionally included after the initiation of the study. The lower number of participants as compared to 1. Amarya S, Singh K, Sabharwal M. Changes during that required in this study may a®ect clinical aging and their association with malnutrition. contribution for the ¯ndings of this variable, i.e., J Clin Gerontol Geriatr 2015;6:78–84. the correlation between SPUTs and body compo- sition. Thus, a further study addresses all these 2. Colon- Emeric CS, Whitson HE, Pavon J, Hoenig limitations, as well as other psychometric proper- H. Functional decline in older adults. Am Fam ties needed for clinical application are still required Physician 2013;88:388–94. to con¯rm the present ¯ndings and extend clinical bene¯t of SPUTs. 3. Frontera WR, Hughes VA, Fielding RA, Fiatarone MA, Evans WJ, Roubeno® R. Aging of skeletal In conclusion, particular forms of SPUTs are feasible and valid measures to re°ect body com- position, muscle strength, and mobility among

Validity and feasibility of using a seated push-up test among community-dwelling older adults 135 Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com muscle: A 12-yr longitudinal study. J Appl Physiol 16. Choi YJ. Dual-energy X-ray absorptiometry: by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. 2000;88:1321–26. Beyond bone mineral density determination. 4. Herman S, Kiely DK, Leveille S, O'Neill E, Cyberey Endocrinol Metab (Seoul) 2016;31:25–30. S, Bean JF. Upper and lower limb muscle power relationships in mobility-limited older adults. J 17. Bakkum AJ, Janssen TW, Rolf MP, et al. A reli- Gerontol A Biol Sci Med Sci 2005;60:476–80. able method for measuring proximal tibia and dis- 5. Putho® ML, Nielsen DH. Relationships among tal femur bone mineral density using dual-energy impairments in lower-extremity strength and X-ray absorptiometry. Med Eng Phys 2014;36:387– power, functional limitations, and disability in 90. older adults. Phys Ther 2007;87:1334–47. 6. Shinohara M, Latash ML, Zatsiorsky VM. Age 18. Blake GM, Fogelman I. The role of DXA bone e®ects on force produced by intrinsic and extrinsic density scans in the diagnosis and treatment of hand muscles and ¯nger interaction during MVC osteoporosis. Postgrad Med 2007;83:509–17. tasks. J Appl Physiol 2003;95:1361–69. 7. Kwon I, Kim J-S, Shin C-H, Park Y, Kim J-H. 19. Bohannon RW, Schaubert KL. Test-retest reli- Associations between skeletal muscle mass, grip ability of grip-strength measures obtained over a strength, and physical and cognitive functions in 12-week interval from community-dwelling elders. elderly women: e®ect of exercise with resistive J Hand Ther 2005;18:426–27. theraband. J Exerc Nutrition Biochem 2019; 23:50–5. 20. Roberts HC, Denison HJ, Martin HJ, et al. A re- 8. Garcia-Pena C, Garcia-Fabela LC, Gutierrez- view of the measurement of grip strength in clinical Robledo LM, Garcia-Gonzalez JJ, Arango-Lopera and epidemiological studies: towards a standar- VE, Perez-Zepeda MU. Handgrip strength predicts dised approach. Age Ageing 2011;40:423–29. functional decline at discharge in hospitalized male elderly: a hospital cohort study. PLoS One 2013;8: 21. Bethards S, Everitt-Smith S, Roberts H, Scarbor- e69849. ough G, Tate S, Bandy WD. Intrarater test-retest 9. Short FX, Winnick JP. Test items and standards reliability of an instrument used to measure back related to muscle strength and endurance on the and leg strength. Isokinet Exerc Sci 1995;5:31–5. Brockport Physical Fitness Test. Adapt Phys Activ Q 2005;22:371–400. 22. Ten Hoora GA, Muscha K, Meijerb K, Plasquia G. 10. Yoshimura Y, Ise M. Analysis of the push-up Test-retest reproducibility and validity of the back- movement base on action potentials of upper ex- leg-chest strength measurements. Isokinet Exerc tremity muscles and ground reaction force between Sci 2016;24:209–16. the palms and a force plate. Kawasaki J Med Welf 2005;11:1–7. 23. Bohannon RW. Test-retest reliability of the ¯ve- 11. Wiyanad A, Amatachaya P, Sooknuan T, et al. repetition sit-to-stand test: a systematic review of The use of simple muscle strength tests to re°ect the literature involving adults. J Strength Cond body compositions among individuals with spinal Res 2011;25:3205–7. cord injury. Spinal Cord 2021;60:99–105. 12. Poncumhak P, Wiyanad A, Siriyakul C, Kosura N, 24. Goldberg A, Chavis M, Watkins J, Wilson T. The Amatachaya P, Amatachaya S. Validity and fea- ¯ve-times-sit-to-stand test: validity, reliability and sibility of a seated push-up test to indicate skeletal detectable change in older females. Aging Clin Exp muscle mass in well-functioning older adults, Phy- Res 2012;24:339–44. siother Theory Pract 2022;1–8.doi:https://doi.org/ 10.1080/09593985.2021.2023931. 25. Thaweewannakij T, Wilaichit S, Chuchot R, et al. 13. Dehghan M, Merchant AT. Is bioelectrical imped- Reference values of physical performance in Thai ance accurate for use in large epidemiological elderly people who are functioning well and dwell- studies? Nutr J 2008;9:26. ing in the community. Phys Ther 2013;93:1312–20. 14. Mohamad AB, Nurakmal B. Sample size guideline for correlation analysis. World J Soc Sci Res 26. Podsiadlo D, Richardson S. The timed \"Up & Go\": 2016;3:37–46. A test of basic functional mobility for frail elderly 15. Whitney SL, Wrisley DM, Marchetti GF, Gee MA, persons. J Am Geriatr Soc 1991;39:142–48. Redfern MS, Furman JM. Clinical measurement of sit-to-stand performance in people with balance 27. Shumway-Cook A, Brauer S, Woollacott M. Pre- disorders: validity of data for the ¯ve-times-sit-to- dicting the probability for falls in community- stand test. Phys Ther 2005;85:1034–45. dwelling older adults using the timed up & go test. Phys Ther 2000;80:896–903. 28. Ste®en TM, Hacker TA, Mollinger L. Age- and gender-related test performance in community- dwelling elderly people: Six-minute walk test, Berg balance scale, timed Up & go test, and gait speeds. Phys Ther 2002;82:128–37. 29. Studenski S, Perera S, Patel K, et al. Gait speed and survival in older adults. J Am Med Assoc 2011;305:50–58. 30. Lang JT, Kassan TO, Devaney LL, Colon-Semenza C, Joseph MF. Test-retest reliability and minimal

Hong Kong Physiother. J. 2022.42:125-136. Downloaded from www.worldscientific.com 136 P. Poncumhak et al. 34. Rikli RE, Jones CJ. Development and validation of by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. a functional ¯tness test for community-residing detectable change for the 10-meter walk test in older adults. J Aging Phys Act 1999;7:129–61. older adults with Parkinson's disease. J Geriatr Phys Ther 2016;39:165–70. 35. Wang L, Zhang Z, McArdle JJ, Salthouse TA. In- 31. Amatachaya S, Kwanmongkolthong M, Thong- vestigating ceiling e®ects in longitudinal data jumroon A, et al. In°uence of timing protocols and analysis. Multivariate Behav Res 2009;43(3):476– distance covered on the outcomes of the 10-meter 496. walk test. Physiother Theory Pract 2020;36:1348– 53. 36. Batsis JA, Daniel K, Eckstrom E, et al. Promoting 32. Middleton A, Fritz SL, Lusardi M. Walking speed: healthy aging during COVID-19. J Am Geriatr Soc the functional vital sign. J Aging Phys Act 2021;69(3):572–580. 2015;23:314–22. 33. Haas F, Sweeney G, Pierre A, Plusch T, Whiteson 37. Netz Y, Ayalon M, Dunsky A, Alexander N. The J. Validation of a 2 minute step test for assessing multiple-sit-to-stand' ¯eld test for older adults: functional improvement. Open J Ther Rehabil what does it measure? Gerontology 2004;50:121– 2017;5:71–81. 126.

Research Paper Hong Kong Physiotherapy Journal Vol. 42, No. 2 (2022) 137–149 DOI: 10.1142/S1013702522500135 Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com Hong Kong Physiotherapy Journal by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. https://www.worldscientific.com/worldscinet/hkpj Chinese (Mandarin) translation of the incremental shuttle walk test and its validity and reliability: A cross-sectional study Wei Qin Ang1, Hong Ting Tan1, Si Min Goh1, Samantha W. Seng1, Katherin S. Huang1,2, Melissa Y. Chan1 and Meredith T. Yeung1,* 1Health and Social Sciences Cluster Singapore Institute of Technology, Singapore 2Department of Physiotherapy Khoo Teck Puat Hospital, Singapore *[email protected] Received 30 July 2021; Accepted 19 July 2022; Published 31 August 2022 Background/Purpose: To date, there are no published validated Chinese versions of the incremental shuttle walk test (ISWT) instructions despite its wide clinical applications. Translation of the Chinese ISWT instruction is done in an ad-hoc manner within the Chinese-speaking populations, a®ecting the test's reli- ability and validity since translation can di®er signi¯cantly between individuals. This warrants the need for psychometric testing of such translation. Objectives: To develop a Chinese (Mandarin) version of the ISWT instructions (ISWT-CHN) that is conceptually equivalent to the original English version (ISWT-ENG) and establish its reliability and validity. Methods: Forward and backward translations from the ISWT-ENG were done to generate the ISWT-CHN. Face and content validity was determined during the translation process. Intra-rater and inter-rater reli- ability of the ISWT-CHN, construct and criterion validity were established by analysing the ISWT and the gold standard cardiopulmonary exercise test results. Results: The Item-Content validity index (I-CVI), Scale-level-Content validity index (S-CVI), and content validity ratio (CVR) of the ISWT-CHN were 1.0. Intra-class Correlation Coe±cient (ICC) for inter-rater reliability between two raters were excellent (ICC ¼ 0:99, 95% CI 0.97–1.0, p < 0:001; SEM ¼ 0:85 m, MDC ¼ 2:35 m). The intra-rater reliability of both Raters A (ICC ¼ 0:92, 95% CI 0.53–0.98, p ¼ 0:003; *Corresponding author. Copyright@2022, Hong Kong Physiotherapy Association. This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 (CC BY- NC-ND) License which permits use, distribution and reproduction, provided that the original work is properly cited, the use is non-commercial and no modi¯cations or adaptations are made. 137

Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com 138 W. Q. Ang et al. by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. SEM ¼ 35 m, MDC ¼ 97 m) and B (ICC ¼ 0:90, 95% CI 0.76–0.96, p < 0:001; SEM ¼ 32 m, MDC ¼ 88 m) were good. In a sample of 32 healthy participants, both ISWT-CHN and ISWT-ENG instruction results showed low-positive correlations with the VO2max determined from the cardiopulmonary exercise test (r = 0.439, p < 0:001; r ¼ 0:448, p < 0:001). There is a very high correlation between ISWT-ENG and ISWT-CHN results with no statistically signi¯cant di®erences (r ¼ 0:967, p < 0:001). The construct and criterion validity of the ISWT-CHN were established. Conclusion: This study developed the ISWT-CHN and showed that it is a valid and reliable measure conceptually comparable to the ISWT-ENG. It will bene¯t the determination of functional exercise capacity in Chinese-speaking populations. Key messages . This study is aimed to develop a Chinese (Mandarin) version of the ISWT instructions. . The ISWT Chinese translation is valid and reliable that is conceptually comparable to the original English instruction. . The translated ISWT-Chinese instruction will enable the use of ISWT among the Chinese-speaking populations. Keywords: Chinese translations; cross-sectional studies; incremental shuttle walk test; reproducibility of results. Introduction specialised equipment or extensive training to conduct the tests. The externally-paced and in- The Incremental Shuttle Walk Test (ISWT) is a cremental characteristics of the ISWT are also commonly-used maximal ¯eld test that measures similar to laboratory CPETs,1 making the ISWT the total distance covered by the number of com- an excellent alternative to assess exercise capacity pleted shuttles (ISWD).1 The ISWT was ¯rst used in clinical settings. to determine the functional exercise capacity in people with Chronic Obstructive Pulmonary Dis- The European Respiratory Society (ERS) and ease (COPD),1 and was proven to be reliable and the American Thoracic Society (ATS) have valid.2–5 The application of ISWT was then ex- established standardised English instructions for panded and proven to be reliable and valid in the ISWT.17 However, its use with non-English populations with other health conditions such as speaking populations in regions and countries such cardiovascular diseases,6,7 lung cancer,8 and pe- as Hong Kong, Taiwan, Malaysia, China, and ripheral arterial diseases.9 Functional exercise ca- Singapore may be limited, given the di±culty in pacity refers to the maximum amount of aerobic understanding the instructions. The Chinese work an individual can sustain, de¯ned by the (Mandarin) language is the second-most spoken maximal oxygen uptake (VO2 max). This variable is language worldwide,18 and approximately 75% of highly associated with a patient's ability to per- Singaporeans are Chinese, with 48% of them have form activities of daily living and the results are Chinese as their primary language.19 This high- often used for exercise prescription among phy- lights the need for Chinese instructions. However, siotherapists.10 The ISWT is used to assess physi- there is no published validated Chinese version of cal ¯tness and design personalised walking the ISWT instructions to date. As a result, clin- programmes,11–13 as well as assess treatment out- icians have to translate the ISWT instructions on comes and predict rehospitalisation, morbidity, an ad-hoc basis, which a®ects the reliability and and mortality rates.14,15 validity of the test since translation can di®er sig- ni¯cantly between individuals, warranting the Traditionally, Cardiopulmonary Exercise Tests need for additional psychometric testing on an (CPETs) are the gold standard for measuring an established measurement after cross-cultural individual's exercise capacity.16 However, they re- translation and adaptation.20 This is further quire sophisticated equipment and trained sta® to attested by the results from the European Social conduct the test and analyse its results. In con- Survey, which showed di®erences in measurement trast, the ISWT is simple and does not require

Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com quality across countries. Their study attempted to Cross-cultural adaptation of ISWT in Chinese 139 by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. explain such di®erences and found that mis- translations and/or keywords lost in translations The development of the ISWT-CHN adopted the could a®ect the investigation results.21 Therefore, cross-cultural adaptation process recommended by this study aimed to develop a Chinese (Mandarin) the World Health Organisation (WHO).22 Figure 1 version of the ISWT instructions (ISWT-CHN) describes the translation and cross-cultural adap- that is conceptually equivalent to the original En- tation process of the ISWT-CHN. Two bilingual glish version (ISWT-ENG) and subsequently translators performed the forward translations in- established its reliability and validity. dependently. These two versions (Versions A and B) were combined to form the interim Methods instructions (Version C), and it was then compared and revised by two investigators and the two for- Design ward translators in agreement to form the revised translated version (Version D). Twenty volunteers This was a translation and cross-sectional validity who were bilingual in English and Chinese but had study assessing the reliability and validity of the no prior experience with the ISWT performed the ISWT after the adaptation to the Chinese (Man- backward translation of the revised-Chinese ver- darin) language. This study took place between sion (Version D) to English (Versions E1–E20). June 2020 and April 2021. Ethical approval was Lastly, the panel of researchers, which consisted of obtained from the Institutional Review Board of four investigators and two translators, reviewed the university (Number: 2020022), and all subjects the 20 backward translated versions for face and provided written informed consent before partici- content validity and modi¯ed them to produce the pating in the study. ¯nal ISWT-CHN instructions (Version F). Development of ISWT-CHN Participants, investigators and centres Before conceptualising this study, permission was ¯rst sought from the test originator.1 The participants were recruited via convenience sampling, with the following inclusion criteria: (1) Between 21 and 65 years old; (2) ambulant and not Fig. 1. Flow diagram of the translation and cross-cultural adaptation process.

Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com 140 W. Q. Ang et al. SpO2, HR, and BP had to return to baseline before by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. the subsequent trial. using any walking aids; (3) colloquially pro¯cient in Chinese and English; (4) has adequate mental Bruce protocol treadmill test capacity and able to follow instructions; and (5) was not involved with the backward translation in The CPET was conducted on a treadmill (h/p/ the development of the ISWT-CHN. Participants cosmos quasarr med, Germany) with metabolic gas were excluded from the study if they had any analysis (COSMED Quark CPET Gas Analyser, cardiovascular, respiratory, neuromuscular and/or Germany). The CPET was conducted using the musculoskeletal disorder(s) that could hinder their standard Bruce treadmill protocol.35 Participants ability to exercise. The Physical Activity Readiness were instructed to continue the test until maximal Questionnaire for Everyone (PAR-Qþ)23 was used exhaustion while trained investigators monitored for screening. Prospective participants were ex- the necessary parameters [HR and rhythm via a 12- cluded if they were presented with any contra- lead electrocardiogram, BP and RPE (0–10)]. The indications for CPET24 or spirometry results that maximal oxygen consumption (VO2max) recorded suggested any potential air°ow limitation, de¯ned by the cardiopulmonary diagnostic software as a forced expiratory volume in one second (OMNIA 1.6.3, COSMED, Rome, Italy) was used (FEV1) to forced vital capacity (FVC) ratio of less for data analysis. The CPET results served as the than 0.70.25,26 According to other cross-cultural benchmark for criterion validity testing between adaptation studies27–31 and the minimum sample ISWT-ENG and ISWT-CHN. The ISWT and size suggested by statisticians,32 30 participants CPET testings' were conducted on separate days. would provide adequate study power to observe the validity and reliability. All exercise testing, Reliability and validity of ISWT- namely ISWT and CPET, were conducted by two CHN investigators (Rater A and Rater B), undergradu- ate physiotherapy students with pro¯ciency in A pilot trial was ¯rst conducted on six healthy conducting the tests, under close supervision from participants who ful¯lled the same aforementioned the principal investigator. All testing procedures inclusion and exclusion criteria with the ISWT- and data collection took place in the exercise lab- CHN instructions (Version F) to ¯nalise the oratory of the university. translation and establish the inter-reliability of the ISWT-CHN (Raters A and B). Both investigators The ISWT protocol contributed to the collection of the demographic data. Each participant completed a total of ¯ve With adherence to the Technical Standard,17 the trials on the same day. Considering the learning ISWT was conducted with an open 10 m course e®ect on the ISWT,36,37 a practice trial was in- marked by two cones that were placed 0.5 m in- corporated before the participants were rando- wards from either end. Participants were required mised into A-B-B-A (n ¼ 3) or B-A-A-B (n ¼ 3) to listen to a pre-recorded standardised ISWT sequence via a computer-generated list for optimal instruction in either English or Chinese (Manda- e±ciency and inter-rater reliability estimates.38 rin) before the test started and had to keep up with The better of the two ISWD from each rater was the walking speed dictated by the pre-recorded used to establish the inter-rater reliability of the audio signals while walking up and down the ISWT-CHN. No further changes to the ISWT- course during the test. The test was terminated if CHN instructions were made after the pilot trial, the participant was limited by (1) fatigue; (2) and Version F was concluded as the ¯nal ISWT- dyspnoea; or (3) inability to maintain the required CHN instructions. Subsequently, the intra-rater speed and failed to complete two consecutive reliability was established with an additional 14 shuttles. The distance covered was calculated from healthy participants. the total number of completed shuttles (ISWD). Parameters such as oxygen saturation (SpO2), In the validity study, participants performed heart rate (HR), blood pressure (BP), and the three trials of ISWT on the same day and the modi¯ed Borg's scale (0–10) for Dyspnoea and CPET on the following day. The block randomi- Rate of Perceived Exertion (RPE)33,34 were taken sation method with a block size of three was used. at various time points of the test. There was a The block size of three was the minimum minimum 30 min break between each test, and the

Cross-cultural adaptation of ISWT in Chinese 141 Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com Notes: ECE ¼ English–Chinese–English, EEC ¼ English–English–Chinese, CEC ¼ Chinese–English–Chinese, CCE ¼ Chinese–Chinese–English, by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. ISWT-CHN ¼ Chinese (Mandarin) version of the Incremental Shuttle Walk Test, ISWT-ENG ¼ original version of the Incremental Shuttle Walk Test. Fig. 2. Design and °ow of participants through the validity study. permutation for two variables, yet it was able to Data analysis reduce bias and achieve balance in the allocation of participants compared to simple randomisation, Statistical analysis was performed with IBM particularly applicable with a small sample size.39 SPSSr Statistics for Windows Version 26.0 (IBM The recruited participants were randomised using Corporation, Armonk, New York, USA). The level a computer-generated list into four groups with of signi¯cance was set at p < 0.05. Demographic di®erent test sequences for the validity testing of and anthropometric data of participants were ex- the ISWT-CHN. The groups were as follows: The amined for normal distribution using the Shapiro– English–Chinese–English (ECE; n ¼ 8), English– Wilk test. ANOVA was used to analyse normally English–Chinese (EEC; n ¼ 8), Chinese–English– distributed continuous variables between groups Chinese (CEC; n ¼ 8) or Chinese–Chinese–English (Height, Weight, FEV1/FVC ratio, Chinese and (CCE; n ¼ 8) groups (Fig. 2). The better result of English ISWD, Absolute VO2 max) and the Krus- the two trials in the same language and the result kal–Wallis test was used to analyse non-normal of the remaining trial in the other language within continuous variables (Age). The Fisher's exact test the same test sequence, were chosen for analysis to was used to analyse sex di®erences between groups. account for the learning e®ect.36,37 For example, the better result of the two English ISWD and the The inter-rater and intra-rater reliability were Chinese ISWD from the ECE group were used for analysed via the interclass correlation (ICC) with subsequent analysis. a two-way mixed-e®ects model and a Bland– Altman plot.40 The ICC was based on a 95% Con¯dence Interval (95% CI). Absolute reliability

142 W. Q. Ang et al. was calculated with the standard error of mea- panellists.43 Each sentence's essentiality was examined on a 3-point Likert scale (not essen- surement (SEM), minimal detectable change tial, useful but not essential, essential). For a panel size of six, CVR ¼ 1:0 (p ¼ 0:05)46 is (MDC), and 95% limits of agreement (LOA). SEM required to be statistically signi¯cant. wpaðs1 calculated based on the formula SD Construct and criterion validities were analysed À ICCÞ, while MDC at 95% level of con¯dpence using Pearson's correlation coe±cient r. Construct validity was determined from the correlation be- was calculated based on the formula SEM Â 2 Â tween the distances obtained from ISWT-ENG and ISWT-CHN, while criterion validity was deter- 1.96. The 95% LOA provides an interval within mined from the correlations between the ISWDs of the ISWT-CHN, ISWT-ENG and VO2 max of the which 95% of di®erences between the two mea- CPET. surements are expected to lie. The alternate form of reliability between the two measurements can be Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com ensured if the 95% LOA is narrow and at least 95% by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. of the points lie within the limits.41,42 Face validity was determined during the trans- lation process. Content validity was determined by the CVI and CVR. (1) CVI is the most widely reported method for Results determining content validity in instrument development that examines its relevance and Flow of participants clarity, and it can be calculated using the Item- CVI (I-CVI) and the Scale-level-CVI Six participants were recruited during the pilot (S-CVI).43 This study used a 4-point Likert trial to establish the inter-rater reliability, and an scale (not at all, needs some revision, needs additional 14 participants were included to estab- minor revision, complete), where the I-CVI was lish the intra-rater reliability. All screened parti- calculated by the total ratings scored by all the cipants ful¯lled the inclusion criteria, and none panel members (n ¼ 6; four investigators and were excluded according to the exclusion criteria. two translators) and divided by the total Table 1 presents the overall characteristics of the number of panel members. Where I-CVI is participants in the reliability study. Subsequently, greater than 0.79, the item is relevant; between 40 participants were assessed for eligibility during 0.70 and 0.79, the item requires revisions; and the validity study, and eight participants were when it is less than 0.70, the item is elimi- excluded according to the exclusion criteria. The nated.43,44 Similarly, S-CVI is determined by remaining 32 participants, 20 males (63%) and 12 the number of items in the instrument that females (37%) satis¯ed the inclusion and exclusion received a \\highly complete\" grade. The Uni- criteria and participated in the validity study. The versal Agreement (UA) among the panel demographic and clinical data (age, sex, height, members (S-CVI/UA) and the Average CVI weight, spirometry readings, ISWD, and VO2 max) (S-CVI/Ave) are two ways of determining S-CVI, the latter being a less conservative Table 1. Demographics of partici- method.43 S-CVI/UA is calculated by the sum pants in the reliability study (n ¼ 20). of all items with I-CVI equal to 1 divided by the total number of items, whereas S-CVI/Ave Participants is equal to the sum of all the I-CVIs divided by the number of items. Content validity is ex- Age (years) 26.7 (SD 5.1) cellent when the S-CVI/UA is more than 0.8 Sex and the S-CVI/Ave is more than 0.9.44 This 11 study used the S-CVI/UA method. Male 9 Female 169 (SD 7.6) (2) CVR measures the essentiality of an item45 Height (cm) 67 (SD 14.9) CVR ranges from −1 to 1, with a higher score Weight (kg) 86 (SD 5.0) indicating greater agreement among panel FEV1/FVC ratio (%) members. CVR ¼ (Ne – N/2)/(N/2), where Ne is the number of panellists who rated an item Notes: SD: Standard deviation; cm: as \\essential\" and N is the total number of centimetres; kg: kilograms; FEV1/FVC ratio (%): forced expiratory volume in one second (FEV1) to forced vital capacity (FVC) ratio in percentage.

Cross-cultural adaptation of ISWT in Chinese 143 Table 2. Demographics and clinical data of participants in the validity study (n ¼ 32). ECE EEC CEC CCE p-value (n ¼ 8) (n ¼ 8) (n ¼ 8) (n ¼ 8) Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com Age (years) 24 (SD 2.2) 27 (SD 3.7) 27 (SD 6.5) 24 (SD 3.1) 0.23 by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Sex 0.96 Male (n = 20) 6 5 5 4 Female (n = 12) 2 3 3 4 0.53 Height (cm) 168 (SD 5.4) 165 (SD 6.0) 168 (SD 8.1) 170 (SD 7.3) 0.18 Weight (kg) 62 (SD 7.3) 60 (SD 9.2) 72 (SD 16.0) 62 (SD 12.0) 0.15 FEV1/FVC ratio (%) 90 (SD 3.3) 85 (SD 6.0) 85 (SD 3.0) 88 (SD 6.5) 0.52 Chinese ISWD (m) 699 (SD 155.0) 801 (SD 181.0) 738 (SD 150.0) 803 (SD 163.0) 0.47 English ISWD (m) 716 (SD 136.0) 801 (SD 174.0) 713 (SD 160.0) 806 (SD 157.0) 0.52 Absolute VO2 max (mL/min/kg) 47 (SD 8.7) 40 (SD 8.1) 41 (SD 14.0) 47 (SD 16.0) Notes: ECE ¼ English–Chinese–English, EEC ¼ English–English–Chinese, CEC ¼ Chinese–English–Chinese, CCE ¼ Chinese–Chinese–English, n ¼ number of participants, ISWD ¼ Incremental Shuttle Walk Distance, FEV1 ¼ forced expiratory volume in one second, FVC ¼ forced vital capacity, and VO2 max ¼ maximal oxygen consumption. ÃSigni¯cance of p values < 0:05. of the participants in the validity study are sum- reliability testing were 35 m for Rater A and 32 m marised in Table 2. for Rater B. The MDC values for the intra-rater reliability testing were 97 m for Rater A and 88 m Reliability of ISWT-CHN for Rater B. The SEM values for the reliability testing could be considered small as they were Table 3 illustrates the inter- and intra-rater reli- below 10% of the highest achievable distance of ability. The ISWT-CHN displayed excellent inter- 1020 m.47 rater reliability (ICC ¼ 0:99, 95% CI 0.97–1.0, p < 0:001) between Raters A and B, with the SEM Validity of ISWT-CHN and MDC being 0.85 metres (m) and 2.35 m, re- spectively. Figure 3 shows the Bland–Altman plot The rigorous translation process established the for the inter-rater reliability, and the dots on the face validity of the ISWT-CHN according to the plot represents each of the six participants for the translation and cross-cultural adaptation guide- pilot study. From the plot, all the points were lines by the WHO.22 The panel of six investigators within the 95% LOA (þ95% LOA ¼ 37 m, À95% synonymously agreed that the face meaning of the LOA ¼ À37 m). Good intra-rater reliability (Rater 20 copies of backward translations collected was A: ICC = 0.92, 95% CI 0.53–0.98, p = 0.003; Rater equivalent to the original English instructions. B: ICC ¼ 0:90, 95% CI 0.76–0.96, p < 0:001) was Subsequently, content validity was established observed. The SEM values for the intra-rater via quantitative and qualitative assessments. Table 3. Reliability of the ISWT-CHN. ICC 95% CI p-values SEM (m) MDC (m) Inter-rater reliability (n ¼ 6) 0.99 0.97-1.00 < 0.001Ã 0.85 2.35 Intra-rater reliability (n ¼ 20) 35.0 97.0 32.0 88.0 Rater A 0.92 0.53-0.98 ¼ 0:003Ã Rater B 0.90 0.76-0.96 < 0.001Ã Notes: ISWT-CHN ¼ Chinese (Mandarin) version of the Incremental Shuttle Walk Test, ICC ¼ Intra-class Correlation Coe±cient, CI ¼ Con¯dence Interval, SEM ¼ Standard Error of Measurement, and MDC ¼ Minimal Detectable Change. ÃSigni¯cance of p values < 0:05.

144 W. Q. Ang et al. Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com Notes: The di®erences in incremental shuttle walk distances between raters are plotted against the mean scores. The straight line represents the by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. mean di®erence between the two raters; dashed lines represent the 95% limits of agreement. ISWD ¼ Incremental Shuttle Walk Distances and LOA ¼ limits of agreement. Fig. 3. Bland–Altman plot comparing the agreement between two raters. Six investigators provided feedback and ratings on Discussion the essentiality, relevance, and clarity of each sentence of the translated instruction. There was This study aimed to cross-culturally adapt the unanimous agreement that the conceptual mean- English ISWT instructions to Chinese (Mandarin) ing of the IWST-CHN was consistent with the and evaluate the translation's reliability and va- original version. The I-CVI, S-CVI/Ave and CVR lidity. This study revealed that ISWT-CHN is a of the ISWT-CHN were 1.0. reliable and valid outcome measure. Intra-rater reliability of both Raters A and B were good, with The ISWT-CHN had a mean distance of 760 m an ICC of 0.92, p ¼ 0:003 and 0.90, p < 0:001, [standard deviation (SD) 161] and ISWT-ENG had respectively. Inter-rater reliability was excellent a mean distance of 759 m (SD 156). The Pearson's with an ICC of 0.99, p < 0:001, and the Bland– coe±cient of ISWT-CHN versus CPET (r ¼ 0:439, Altman plot revealed all points within Æ1:96 SD of p < 0:001) as compared to ISWT-ENG versus the mean di®erence implying that the two raters CPET (r ¼ 0:448, p < 0:001). This similarly weak agreed. Absolute reliability measurements also positive correlation with the VO2 max established showed that the ISWT-CHN results were precise the criterion validity of the ISWT-CHN. Addi- and had low margins of error between raters, with tionally, the ISWT-CHN had satisfactory con- an SEM of 0.85 m and an MDC of 2.35 m. This struct validity, as shown by the very high positive agrees with the results from other studies that correlation between the ISWD of the ISWT-CHN showed good to excellent reliability, for instance, and ISWT-ENG (r ¼ 0:967, p < 0:001). Figure 4 in the COPD population (ICC ¼ 0:80–0.93)2 shows the scatterplots that depict the relationships and patients with peripheral obstructive arterial between VO2 max (mL/min/kg), ISWT-CHN disease (ICC ¼ 0:95).9 distance (m) and ISWT-ENG distance (m).

Cross-cultural adaptation of ISWT in Chinese 145 Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com (a) (b) by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. (c) Fig. 4. Scatterplots of VO2 max, ISWT-CHN distance and ISWT-ENG distance. (a) VO2 max and ISWT-CHN distance, (b) VO2 max and ISWT-ENG distance, (c) ISWT-CHN distance and ISWT-ENG distance. ISWT-CHN ¼ Chinese (Mandarin) version of the Incremental Shuttle Walk Test, ISWT-ENG ¼ original version of the Incremental Shuttle Walk Test, VO2 max ¼ maximal oxygen consumption, r ¼ Pearson's correlation, and p ¼ probability value. The validity of the ISWT-CHN was established. ISWT-CHN and ISWT-ENG demonstrated the Face validity was determined during the transla- equivalence of the instructions in both the lan- guages. Therefore, the establishment of criterion tion process. Content validity was ascertained by validity was inferred. the determination of CVI, in the form of I-CVI and This study revealed a weaker correlation be- S-CVI, as well as CVR. Construct and criterion tween ISWT results in both languages and CPET validity of ISWT-CHN were also established. The VO2 max results (r ¼ 0:439–0.448) as compared to other studies. The Pearson's correlation coe±cient construct validity of the ISWT-CHN was estab- r ranged from 0.72 to 0.85 in the COPD popula- lished through the excellent correlation with the tion,3–5 from 0.73 to 0.83 in patients with cardiac ISWT-ENG (r ¼ 0:967, p < 0:001). The ISWT- diseases,6,7 and was 0.67 in lung cancer patients.8 A ENG and ISWT-CHN results were compared with possible reason for this is because the original CPET VO2 max results, the gold standard to mea- ISWT1 was developed for patients with COPD, sure exercise capacity.16 The criterion validity of and the protocol does not allow running to keep up the ISWT-CHN was indicated by the largely with the set speeds. This, however, might not be sensitive to predict functional exercise capacity in similar signi¯cant positive correlation of the the healthy population as they are less likely to ISWT-CHN and ISWT-ENG with the VO2 maxðr ¼ reach 85% of the predicted maximal HR or become 0:439; p < 0:001; r ¼ 0:448; p < 0:001Þ, which sug- too breathless to continue just by walking. More- gested strong agreement between both versions of over, when observing the best ISWT trial results ISWT. The similar degree of weak positive corre- lation between the VO2 max and ISWT in both languages and the strong correlation between the

Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com 146 W. Q. Ang et al. with data skewed to the left. This may be due to by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. the study's recruitment process, which required the regardless of the language used in this study, none participants to be on social media to obtain infor- of the participants terminated the test due to ex- mation about the study. This method may have cessive dyspnoea and/or leg fatigue.1 In fact, most limited the study's reach to only those pro¯cient in participants (78%, n ¼ 25) terminated the test due technology navigating the social media platforms. to the inability to maintain the required speed.1 Furthermore, a portion of interested participants Hence, in most participants, functional exercise who were in older age groups chose not to partic- capacity could not be accurately predicted as the ipate in the study after understanding what CPET test was terminated before adequate cardiovascu- was, and declined with the main reason being lar stress. The remaining 22% of participants \\afraid of over-exerting\". Although interested and (n ¼ 7) completed all 12 levels at least once out of willing to perform CPET, some older adults were the three ISWT trials, and 86% (n ¼ 6) of them excluded from the study as they had chronic were males. Functional exercise capacity could not medical conditions and were deemed unsafe for be accurately predicted in this subgroup of parti- exercise until certi¯ed by quali¯ed personnel, as cipants due to ceiling e®ects. To overcome these stated in the PAR-Qþ. Additionally, it is worth limitations when using ISWT in a healthy popu- noting that the study was conducted amid the lation, Probst et al.13 suggested allowing the par- COVID-19 global pandemic, where multiple mea- ticipants to run if needed and increasing the sures have been put in place, such as movement number of stages beyond 12 levels by increasing restrictions and proper social distancing to curb the speed by 0.17 metres per second (m/s) each transmissions. Thus, participant recruitment was minute. However, this modi¯ed protocol37 is not mainly limited to the younger population, and the commonly used in clinical settings. sample size fell short of the 30 participants required to provide adequate study power. Notwithstand- The ISWT is widely used in research and reha- ing, the reliability results are promising despite the bilitation as a measure of exercise capacity. It is small sample size. It is also unclear if there exists a often chosen over other measures of exercise ca- di®erence in colloquial usage and understanding of pacity, such as the 6-min walk test (6MWT), as the language between age groups which warrants this test necessitates at least 30 m of walking further exploration. space48 which may not be available at clinical or research facilities. Furthermore, given that the Nonetheless, this study is the ¯rst to develop a 6MWT is associated with several limitations re- Chinese version of the ISWT instructions with lated to its self-paced nature, the ISWT is superior established reliability and validity. The study when determining exercise intensity because of its results were comparable to other published data, progressive nature.48,49 The ISWT is relevant and and it showed signi¯cant results to con¯rm that particularly useful when prescribing exercise in- the ISWT-CHN was conceptually equivalent to its tensity using the percentage of peak performance48 original English instructions. Thus, this study can and thus may be more bene¯cial in the rehabili- serve as a framework for future studies with larger tation or general health promotion setting when sample sizes and participants of varied back- tailoring individualised exercise intervention.11–13 grounds despite its limitations. The newly translated Chinese instructions of the ISWT may o®er a reliable and validated method of Conclusion ¯eld-based exercise testing, bene¯ting Chinese- speaking populations around the globe. This study translated and cross-culturally adapted the original English instructions of the ISWT into There are several limitations of this study. the Chinese (Mandarin) language and subse- Firstly, the study used convenience sampling and quently established its reliability and validity. The recruited only a small sample size, which might ISWT-CHN was shown to be a reliable and valid have introduced bias where speci¯c subgroups of measurement of exercise capacity with psycho- populations might have been over- or under- metric properties similar to the original English represented. For instance, our participants were version.1 This development is helpful to the Chi- relatively young, with a mean age of 26.7 (SD 5.1) nese-speaking population as clinicians can now and 25 (SD 4.2) years in the reliability and validity studies, respectively, although the study's age in- clusion criteria were set between 21 and 65 years old. This also resulted in a non-normal distribution

Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com conduct the ISWT using standardised Chinese Cross-cultural adaptation of ISWT in Chinese 147 by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. instructions instead of performing ad-hoc transla- tion. However, given the limited sample size, future study, drafted and revised the paper critically for research on the ISWT-CHN with a larger sample important intellectual content. Meredith T. Yeung size and participants of di®erent backgrounds is conceptualised and designed the study, analysed recommended. and interpreted the data, revised the paper criti- cally for important intellectual content and ap- Acknowledgment proved the ¯nal paper. Wei Qin Ang and Hong Ting Tan are the co-¯rst authors who contributed We are grateful to the participants who took their and credited equally to this work. time o® to participate in this study. References Con°ict of Interest 1. Singh SJ, Morgan MD, Scott S, Walters D, The author(s) declared no potential con°icts of Hardman AE. Development of a shuttle walking interest concerning this paper's research, author- test of disability in patients with chronic airways ship, and/or publication. obstruction. Thorax 1992;47:1019–24. Funding/Support 2. Campo LA, Chilingaryan G, Berg K, Paradis B, Mazer B. Validity and reliability of the modi¯ed The author(s) disclosed receipt of the following shuttle walk test in patients with chronic obstruc- ¯nancial support for the research, authorship, and/ tive pulmonary disease. Arch Phys Med Rehabil or publication of this paper. This work was sup- 2006;87:918–22. ported by the Singapore Institute of Technology, Grant No. R-MOE-A404-H009. 3. Onorati P, Antonucci R, Valli G, et al. Non-inva- sive evaluation of gas exchange during a shuttle Ethics Approval walkingest vs. a 6-min walking test to assess exer- cise tolerance in COPD patients. Eur J Appl The Singapore Institute of Technology-Institu- Physiol 2003;89:331–6. tional Review Board (SIT-IRB Project Number: 2020022) approved this study. All participants 4. Satake M, Shioya T, Takahashi H, Kawatani M. gave written informed consent before data collec- Ventilatory responses to six-minute walk test, in- tion began. cremental shuttle walking test, and cycle ergometer test in patients with chronic obstructive pulmonary Competing Interests disease. Biomed Res 2003;24: 309–16. The authors declare that they have no competing 5. Turner SE, Eastwood PR, Cecins NM, Hillman interests. DR, Jenkins SC. Physiologic responses to incre- mental and self-paced exercise in COPD: A com- Author Contributions parison of three tests. Chest 2004;126:766–73. Wei Qin Ang and Hong Ting Tan collected, ana- 6. Green DJ, Watts K, Rankin S, Wong P, O'Driscoll lysed and interpreted the data, prepared and re- JG. A comparison of the shuttle and 6 minute vised the paper critically for important intellectual walking tests with measured peak oxygen con- content. Si Min Goh collected, analysed, and sumption in patients with heart failure. J Sci Med interpreted the data. Samantha W. Seng collected Sport 2001;4:292–300. and interpreted the data. Katherin S. Huang and Melissa Y. Chan conceptualised and designed the 7. Lewis ME, Newall C, Townend JN, Hill SL, Bonser RS. Incremental shuttle walk test in the assessment of patients for heart transplantation. Heart 2001;86:183–7. 8. Win T, Jackson A, Groves AM, Sharples LD, Charman SC, Laroche CM. Comparison of shuttle walk with measured peak oxygen consumption in patients with operable lung cancer. Thorax 2006;61:57–60. 9. da Cunha-Filho IT, Pereira DA, de Carvalho AM, Campedeli L, Soares M, de Sousa Freitas J. The reliability of walking tests in people with claudi- cation. Am J Phys Med Rehabil 2007;86:574–82. 10. Arena R, Myers J, Williams MA, et al. Assessment of functional capacity in clinical and research set- tings. Circulation 2007;116:329–43.

Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com 148 W. Q. Ang et al. 26. Miller MR, Hankinson J, Brusasco V, et al. Stan- by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. dardisation of spirometry. Eur Respir J 2005;26: 11. Harrison SL, Greening NJ, Houchen-Wollo® L, 319–38. et al. Age-speci¯c normal values for the incremental shuttle walk test in a healthy British population. 27. Macfarlane DJ, Lee CC, Ho EY, Chan KL, Chan J Cardiopulm Rehabil Prev 2013;33:309–13. DT. Reliability and validity of the Chinese version of IPAQ (short, last 7 days). J Sci Med Sport 12. Itaki M, Kozu R, Tanaka K, et al. Reference 2007;10: 45–51. equation for the incremental shuttle walk test in Japanese adults. Respir Investig 2018;56:497–02. 28. Poon DM, Chow LC, Au DK, Hui Y, Leung MC. Translation of the dizziness handicap inventory 13. Probst VS, Hernandes NA, Teixeira DC, et al. into Chinese, validation of it, and evaluation of the Reference values for the incremental shuttle walk- quality of life of patients with chronic dizziness. ing test. Respir Med 2012;106:243–8. Ann Otol Rhinol Laryngol 2004;113:1006–11. 14. Emtner MI, Arnardottir HR, Hallin R, Lindberg E, 29. Dhippayom JP, Trevittaya P, Cheng ASK. Cross- Janson C. Walking distance is a predictor of cultural adaptation, validity, and reliability of the exacerbations in patients with chronic obstructive patient-rated Michigan hand outcomes question- pulmonary disease. Respir Med 2007;101:1037–40. naire for Thai patients. Occup Ther Int 2018;2018:8319875. 15. Nutt CL, Russell JC. Use of the pre-operative shuttle walk test to predict morbidity and mortal- 30. Passos L, Tavares J, Figueiredo D. Cross-cultural ity after elective major colorectal surgery. Anaes- adaptation and analysis of the psychometric prop- thesia 2012;67:839–49. erties of the Portuguese version of the Edinburgh feeding evaluation in Dementia scale. Port J Public 16. Ferrazza AM, Martolini D, Valli G, Palange P. Health 2020;38:142–50. Cardiopulmonary exercise testing in the functional and prognostic evaluation of patients with pulmo- 31. Leung RW, Leung M-L, Chung P-K. Validity and nary diseases. Respiration 2009;77:3–17. reliability of a Cantonese-translated rating of per- ceived exertion scale among Hong Kong adults. 17. Holland AE, Spruit MA, Singh SJ. How to carry out Percept Mot Ski 2004;98:725–35. a ¯eld walking test in chronic respiratory disease. Breathe (She±eld, England) 2015; 11:128–39. 32. Onwuegbuzie AJ, Collins KM. A typology of mixed methods sampling designs in social science re- 18. Statista. The most spoken languages worldwide search. Qual Rep 2007;12:281–16. in 2021. https://www.statista.com/statistics/ 266808/the-most-spoken-languages-worldwide/: 33. Borg GA. Perceived exertion. Exerc Sport Sci Rev Statista, 2021. 1974;2:131–53. 19. DoS. Singapore Census of Population 2010, Sta- 34. Borg GA. Psychophysical bases of perceived exer- tistical Release 1. Demographic Characteristics, tion. Med Sci Sports Exerc. 1982;14:377–81. Education, Language and Religion. Department of Statistics, 2011. 35. Bruce RA, Blackmon JR, Jones JW, Strait G. Exercising testing in adult normal subjects and 20. Peña ED. Lost in translation: Methodological con- cardiac patients. Pediatrics 1963;32:742–56. siderations in cross-cultural research. Child Dev 2007;78:1255–64. 36. Dyer F, Marriner P, Cheema K, Bott J. Is a prac- tice incremental shuttle walk test really necessary? 21. Oberski D, Saris WE and Hagenaars J. Why are Chronic Respir Dis 2011;8:201–5. there di®erences in measurement quality across countries. In: Loosveldt G, Swyngedouw M and 37. Bradley J, Howard J, Wallace E, Elborn S. Validity Billiet J, (eds.). Measuring Meaningful Data in of a modi¯ed shuttle test in adult cystic ¯brosis. Social Research, 1st ed. Leuven: Acco, 2007. Thorax 1999;54:437–9. 22. WHO. Process of translation and adaptation of 38. Hallgren KA. Computing inter-rater reliability for instruments. Wold Health Organization, 2020. observational data: An overview and tutorial. Tutor Quant Meth Psychol. 2012;8:23–34. 23. Warburton DE, Jamnik VK, Bredin SS, Gledhill N. The physical activity readiness questionnaire for 39. E¯rd J. Blocked randomization with randomly se- everyone (PAR-Q+) and electronic physical activ- lected block sizes. Int J Environ Res Public Health. ity readiness medical examination (ePARmed-X 2010;8:15–20. +). Health Fit J Canada 2011;4:3–17. 40. Bland JM, Altman DG. Measurement error. Bmj 24. ATS/ACCP Statement on cardiopulmonary exer- 1996;313:744. cise testing. Am J Respir Crit Care Med 2003;167:211–77. 41. Atkinson G, Nevill AM. Statistical methods for assessing measurement error (reliability) in vari- 25. Rabe KF, Hurd S, Anzueto A, et al. Global strat- ables relevant to sports medicine. Sports Med. egy for the diagnosis, management, and prevention 1998;26:217–38. of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med 42. Hopkins WG. Measures of reliability in sports 2007;176:532–55. medicine and science. Sports Med 2000;30:1–15.

Hong Kong Physiother. J. 2022.42:137-149. Downloaded from www.worldscientific.com 43. Zamanzadeh V, Ghahramanian A, Rassouli M, Cross-cultural adaptation of ISWT in Chinese 149 by Horizon College Physiotherapy on 01/26/23. Re-use and distribution is strictly not permitted, except for Open Access articles. Abbaszadeh A, Alavi-Majd H, Nikanfar A-R. Design and implementation content validity study: methods of calculation. Meas Eval Couns Dev Development of an instrument for measuring pa- 2013;47:79–86. tient-centered communication. J Caring Sci 47. Prince B, Makrides L, Richman J. Research 2015;4:165–78. methodology and applied statistics. Part 2: The literature search. Physiother Can 1980;32:201–6. 44. Polit DF, Beck CT. The content validity index: Are 48. Holland AE, Spruit MA, Troosters T, et al. An you sure you know what's being reported? critique o±cial European Respiratory Society/American and recommendations. Res Nursing Health Thoracic Society technical standard: Field walking 2006;29:489–97. tests in chronic respiratory disease. Eur Respir J 2014;44:1428–46. 45. Yamada J, Stevens B, Sidani S, Watt-Watson J, 49. Wise RA, Brown CD. Minimal clinically important De Silva N. Content validity of a process evaluation di®erences in the six-minute walk test and the in- checklist to measure intervention implementation cremental shuttle walking test. J Chron Obstruct ¯delity of the EPIC intervention. Worldviews Pulmon Dis 2005;2:125–9. Evid-Based Nurs 2010;7: 158–64. 46. Ayre C, Scally AJ. Critical values for Lawshe's content validity ratio: Revisiting the original