Journal of Rehabilitation Medicine 51-8 | Page 73

Relationships among measures of physical fitness in adults with heart failure every other second, timed by a metronome, in accordance with Gaffney et al. (29). The contralateral foot was held slightly above the floor, and the number of maximal heel rises completed for each leg was recorded. The bilateral isometric shoulder abduction was performed by having the patient sit on a stool with his or her back touching the wall and a 1-kg dumbbell in each hand while both arms were elevated to 90°. This position was held as long as possible. The patient could be asked to correct the position once during the test, but the test was discontinued if this instruction need to be given a second time. The time (in s) for which the 90° angle of abduction was held was recorded. The unilateral isotonic shoulder flexion was performed by having the patient sit on a stool with his or her back touching the wall and a weight (3 kg for males and 2 kg for females) held in the hand of the arm to be tested. The patient was asked to elevate the arm from a 0° to 90° flexion as many times as possible at a speed of 20 lifts per min timed by metronome. The patient could be asked to correct the flexion once; however, the test was discontinued on the second occurrence of an incorrect flexion. The number of flexions completed for each arm was recorded. The test-retest reliability of MFT for all 3 evaluations have been shown to be very high (r s  = 0.90–0.99) (12). Functional capacity The DASI was used to measure functional capacity. The DASI is a brief 12-item scale that determines functional capacity by assessing patients’ ability to participate in a spectrum of daily activities (18). Patients report their ability to perform personal care, ambulation, housework, yard work, sexual relations, and recreational activities. The possible responses for each item are “yes” or “no”, and each “yes”’ corresponds to a weighted score in terms of the metabolic equivalent (MET) associated with the proposed activity. The total DASI score ranges between 0 and 58.2, with higher scores reflecting better functional capacity. DASI validity was shown with significant correlation with the peak oxygen consumption (r = 0.580) (18), the N-terminal pro- brain natriuretic peptide levels (r = 0.670) (17) and NYHA class (r = –0.653) (30). Moreover, the functional capacity assessed by the DASI has a strong prognostic value in risk stratification for long-term adverse clinical events and mortality at 5 years in cardiac patients (31, 32), and the measure is an independent predictor of both death and myocardial infarction (33). DASI re- liability was shown to be adequate in several studies (Cronbach’s alpha ranging from 0.86 to 0.93) and also in the present study (Cronbach’s alpha = 0.80). Before its use, the DASI underwent to translation and back-translation and cross-cultural adaptation according to international guidelines (34). Sociodemographic data The patients reported their age, sex, education, and marital status. Data on HF aetiology, heart rhythm, EF, and NYHA functional class collected from the patients’ medical records. Regarding the inclusion and exclusion criteria, information on upper or lower limbs impairment was obtained by collecting information from patients’ medical files (e.g. clinical history, reports), and by asking patients for confirmation before the exercise tests. Statistical analysis Statistical analysis was performed in 5 steps. First, descriptive analyses were used to describe the sociodemographic and 609 clinical characteristics, as well as the physical fitness data, of the participants (6MWT, MFT and DASI scores). Normal dist- ribution of the data was analysed by the Kolmogorov-Smirnov test and by visual inspection of quantile-quantile plot (q-q). Secondly, an exploratory factor analysis was performed to detect the dimensionality of DASI. Principal component analysis with promax rotation was used, considering factor loading > 0.30, eigenvalue more than one and the scree plot of eigenvalues. Thirdly, 1-way analysis of variance was used to assess dif- ferences in age, NYHA classification and EF in the patients who scored above or below the median of each physical fitness test. The Kolmogorov-Smirnov test was used to define the differences in sex distribution (male vs female) of the patients who scored above or below the median. Fourthly, to determine the relationships among the 6MWT, MFT and DASI scores, Spearman’s ranked correlation coefficient was performed. Cor- relations among demographic and clinical variables, such as age, sex and NYHA class, were also assessed. The possible influence of movement impairment on physical fitness assessment was explored by using Spearman’s rho to determine the relationships among the number of impaired limbs and the 6MWT, MFT and DASI scores. A Bonferroni correction on correlation was conducted to protect from Type I error. Interpretation of the size of the correlation was performed according to Hinkle et al. (35) as follows: 0.00–0.30 “little if any correlation”, 0.30–0.50 “low correlation”, 0.50–0.70 “moderate correlation”, 0.70–0.90 “high correlation”, and 0.90–1 “very high correlation”. Fifthly, to detect the structure in the relationships among exercise ca- pacity, muscle function (considering all 5 tests) and functional capacity, principal component analysis was conducted using a promax rotation. Principal component analysis was evaluated considering factor loading > 0.30, eigenvalue more than 1 and scree plot of eigenvalues. All of the data were analysed using IBM SPSS Statistics for Windows version 22.0 (IBM Corp., Armonk, NY, USA). A p-value below 0.05 was considered sta- tistically significant. Sample size was performed using G*Power 3.1.9.2. With an estimated medium effect size of 0.30, a power of 80% and α level of 5%, a sample of 84 participants was considered adequate. The adequacy of the correlation matrix for factor analysis was investigated with the Bartlett’s test of sphericity (which should have a p-value < 0.05) and the Kaiser- Meyer Olkin (KMO) test (which should have a value > 0.70). RESULTS Sociodemographic and clinical characteristics of the sample A total of 328 patients were screened, and, of these, 96 patients with HF agreed to participate in the study (mean age 72 standard deviation (SD) 10 years). The patients who declined to participate were older (76 vs 72 years; p < 0.001) and in a higher NYHA class (χ 2  = 14.619; df = 3; p = 0.002) than those enrolled in the study. Most of the participants were men (73%), classified as NYHA II and NYHA III (97% in total), with a mean EF of 44% (SD 10) (Table I). Of the patients with HF participating in the study, 28% had reduced ability because of impairment of at least one limb, but all of them were able to swing their arms 10 times in a row. The prevalence of limb impairments J Rehabil Med 51, 2019