Patient Characteristics:
HDFIT enrolled 195 patients from 13 of 14 sites throughout Brazil during August 2016 to October 2017 (Figure 2). On average, patients were 53 years old, 71% male, 59% white race, 35% had diabetes, 17% had coronary artery disease, 8% had congestive heart failure; 89% used an arteriovenous fistula/graft and 55% had ESKD attributable to diabetes/hypertension. Family income was 2-to-10 times the Brazilian minimum wage in 73% and below that range in 18%. Fifty-six percent completed high school or higher. Seventy-two percent resided ≥7 kilometers from the dialysis clinic; 43% drove by car and 33% by public transportation to the HD clinic. Mean albumin was 4.0±0.4 g/dL, phosphate 5.3±1.4 mg/dL, iPTH 350.5±290.0 pg/mL, hemoglobin 11.1±1.6 g/dL, and Kt/V 1.5±0.4 (Table 1).
Profiles of Physical Activity:
Among 195 HD patients with PA monitored, 176 had valid data. On average, patients took 4654±3468 steps, their energy expenditure rate was 1.1±0.2 MET (kcal/kg/hr), and they performed 23.9±32.6 minutes of MVPA, 361.2±134.5 minutes of light PA, and 488.0±148.0 minutes of sedentary time per 24-hour period (Table 2). ANOVA of PA on HD days and the first, and second non-HD days revealed differences between groups for steps (p<0.001), MET (p=0.043), MVPA (p<0.001), light PA (p<0.0001) and sedentary time (p<0.001) per 24 hours.
On the first non-HD days, patients performed 1389 more steps (p=0.032), 7.2 more minutes of MVPA (p=0.021), 107.9 more minutes of light PA (p<0.0001), and 69.8 fewer minutes of sedentary time (p<0.0001) per 24-hours compared to HD days. MET levels were the same on the first non-HD days versus HD days (Table 2).
On the second non-HD day, patients performed 1007 more steps (p<0.001), 0.02 lower MET (p=0.027), 92.2 more minutes of light PA (p<0.0001), and 35.8 fewer minutes of sedentary time (p=0.026) per 24-hours compared to HD days. Minutes of MVPA were similar for the second non-HD day versus HD days (Table 2).
Granular Profiles of Physical Activity:
Assessment of granular PA levels revealed, as expected, that sitting/lying during HD was associated with the most remarkable differences on HD days versus concurrent/parallel times on non-HD days. There were small increases in PA during the 2-hour post-HD period versus the same times on non-HD days (Table 3).
At concurrent times to HD on the subsequent first non-HD day, patients performed an average of 1,308 more steps (p<0.001), 0.1 higher MET (p<0.001), 7.2 more minutes of MVPA (p<0.001), 88.1 more minutes of light PA (p<0.001), and 79.2 fewer minutes of sedentary time; on the second non-HD day patients performed 1,128 more steps (p<0.001), 0.1 higher MET (p<0.001), 5.5 more minutes of MVPA (p<0.001), 82.4 more minutes of light PA (p<0.001), and 64.8 fewer minutes of sedentary time compared to HD days (Table 3). During HD patients exhibited MET levels indicative of sitting still/resting (MET=1.0) and had 178.0 minutes of sedentary time, however they did perform low levels of some types of PA (97 steps, 0.3 minutes of MVPA, 13.3 minutes of light PA) that are likely attributable to slight movements during HD and could also be due to discrepancies in self-reported dialysis times captured from patient diaries.
Patients performed slightly more PA during the 2-hour post-HD period on HD days versus concurrent times on subsequent non-HD days. During the 0-to-30, 31-to-60, and 61-to-90-minute post-HD slices on HD days, patients tended to take more steps, perform more MVPA/light PA, and spend less time in a sedentary state compared to concurrent times on both the first and second non-HD days. During the 91-to-120-minute slice on HD days, patients tended to perform slightly more steps than concurrent times on the first non-HD days (Table 3). There was a small signal for patients performing lower PA during the >2-to-≤6.5-hour post-HD period on HD days versus the first non-HD days, where patients performed 13.9 fewer minutes of light PA (p=0.009).
Rates of Granular Physical Activity:
Normalized rates of steps and minutes of MVPA performed per hour were calculated to visualize relative distinctions in prespecified periods (Figures 3 and 4). Compared to the treatment time on HD days, patients performed rates of 327 and 282 more steps/hour and 1.8 and 1.4 more minutes of MVPA/hour during the concurrent times on the subsequent first and second non-HD days respectively (all p<0.001).
Patients generally performed higher rates of steps and minutes of MVPA in the 2-hour post-HD period on HD days (Figures 3 and 4). In the 0-to-30-minute post-HD period on HD days, patients performed rates of 161 and 248 more steps/hour compared to the first and second non-HD days respectively (both p<0.01); patients performed 1.4 more minutes of MVPA/hour on HD days versus the second-non-HD day (p<0.01). During the 31-to-60, 61-to-90, and 91-to-120-minute slices of PA following HD, patients performed significantly more steps than concurrent times on both the first and second non-HD day (Figure 3).
Granular Profiles of Physical Activity by Transportation Type:
An exploratory sub-group assessment of granular PA levels for step counts by categories of transportation type on HD days found patients using public transportation (category including public transportation or walking) performed on average a sum of 604 more steps during the 2-hour post-HD period versus patients using car transportation (category including family car, taxi, or ambulance) (Table 4). Patients using public transportation had on average 3.1 fewer minutes of sedentary time during only the predefined period >1.5-to-≤2 hours after HD compared to patients using car transportation. No other differences were observed by transportation type for step counts and sedentary time in other prespecified timepoints in the 24 hours during and after HD.
Profiles of Physical Activity by Dialysis Shift and Day of the Week:
Patients with HD sessions starting before 1500 hours (first and second HD shift) performed 2157 fewer steps (p<0.0001), 18.6 fewer minutes of MVPA (p<0.0001), and had 65.8 more minutes of sedentary time (p<0.0001) per 24-hours on HD days versus patients with HD sessions starting after 1500 hours (third HD shift) (Table 5).
An exploratory sub-group analysis of step counts on HD days by shift and age category showed patients starting HD before 1500 hours who were <65 years old performed 1668 more steps per 24 hours after HD compared to patients ≥65 years old (p<0.0001), yet there were no significant distinctions in sedentary time (Table 6). There were no significant differences between step counts and sedentary time by age categories in patients with HD starting after 1500 hours.
Another sub-group analysis of PA per 24 hours on HD days by shift and family income level found no differences in step counts among patients starting HD before 1500 hours with a family income above and below 2 minimum wages. However, patients starting HD before 1500 hours with a family income <2 minimum wages had 42.2 fewer minutes of sedentary time per 24 hours after HD as compared to patients with a family income >2 minimum wages (p=0.029) (Table 6). Among patients with HD starting after 1500 hours, those with a family income <2 minimum wages took 3872 more steps (p<0.0001) and had 62.6 fewer minutes of sedentary time (p=0.030) per 24 hours after HD versus patients with a family income >2 minimum wages.
Granular assessment of PA identified patients with HD starting before 1500 hours performed 116, 1741, and 871 fewer steps during the 0-to-30-minute, >11.0-to-≤15.5-hour, and >15.5-to-≤20.0-hour post-HD periods on HD days respectively versus patients with HD starting after 1500 hours (all p<0.05) (Table 7). Despite this, patients with HD starting before 1500 hours performed 68 and 680 more steps during the 91-to-120-minute and >2.0-to-≤6.5-hour post-HD periods versus patients with HD starting after 1500 hours (both p<0.05). Similar patterns of PA were observed for MVPA. Minutes of sedentary time during the periods 2-hours post-HD did not differ between patients starting HD before versus after 1500 hours. Patients starting before 1500 hours had more minutes of sedentary time in the >2.0-to-≤11.0 hours periods after dialysis HD (both p<0.0001) and less minutes of sedentary time in the >11.0-to-≤20.0 hour periods post-HD (both p<0.05) compared to those starting HD after 1500 hours.
PA levels on HD days were similar on the first (Monday/Tuesday), second (Wednesday/Thursday) or third (Friday/Saturday) HD session of the week and did not differ between groups in ANOVA and t-test comparisons (Table 8).
Granular steps did not differ by day of the week. Patients performed 0.4 more minutes of MVPA during the 91-to-120-minute post-HD period on the first versus the third HD session of the week (p=0.042); no other differences were observed (Table 9).