YouControl-AFib: Digital Health Intervention to Promote Physical Activity in Persons With Atrial Fibrillation

Author

Lu Mao

Published

October 17, 2026

1 Study and analysis overview

YouControl-AFib (YCA) was a single-arm prospective digital health intervention designed to support physical activity among persons with atrial fibrillation. The primary analysis evaluates change in 6-minute walk distance (6MWD) from baseline to post-intervention assessment. Group Medical Visits (GMV) provides a nonrandomized contextual comparator cohort.

The primary continuous outcome is the change in 6MWD, calculated as

\[ \Delta \mathrm{6MWD} = \mathrm{6MWD}_{\mathrm{post}} - \mathrm{6MWD}_{\mathrm{baseline}}, \]

in meters. A clinically meaningful improvement is defined as \(\Delta \mathrm{6MWD} \geq 30\) meters.

2 Analysis populations

Table 1: Analysis populations
Analysis_set N
YCA enrolled 51
YCA paired 6MWD (primary analysis) 48
YCA paired 6MWD without rhythm/medication changes (sensitivity) 39
GMV paired 6MWD 28

3 Statistical analysis

Baseline characteristics will be summarized with counts and percentages for categorical variables and mean (SD) or median (IQR) for continuous variables. The primary within-YCA analysis will estimate the mean paired change in 6MWD, with a 95% confidence interval and paired t-test. The proportion with a clinically meaningful improvement of at least 30 m will be reported with an exact 95% confidence interval.

The analysis excluding participants with a rhythm or relevant medication change will be presented as a sensitivity analysis. YCA-versus-GMV comparisons are observational and will be presented as contextual (rather than causal) comparisons. Predictor, adherence, residence subgroup, and composite-score analyses will be described as exploratory.

4 Results

4.1 Cohort characteristics

The YCA cohort included 51 enrolled participants, of whom 48 completed both 6MWD assessments. The following tables describe all enrolled participants. Because the urban/non-urban and YCA/GMV comparisons are exploratory and sample sizes are modest, the tables are descriptive; no baseline hypothesis tests are presented.

Table 2 compares baseline characteristics and 6MWD change between urban and non-urban YCA participants. Table 3 compares the YCA and GMV cohorts.

Table 2: Baseline characteristics by urban residence among all enrolled YouControl-AFib participants
Characteristic Non-urban
N = 281		 Urban
N = 231		 Overall
N = 511
Age, years 64 (7) 62 (12) 63 (10)
Gender


    Female 15 (54%) 11 (48%) 26 (51%)
    Male 13 (46%) 12 (52%) 25 (49%)
Body mass index, kg/m² 39.8 (5.0) 39.9 (6.2) 39.9 (5.5)
Hypertension 23 (82%) 19 (83%) 42 (82%)
Diabetes mellitus 8 (29%) 10 (43%) 18 (35%)
Heart failure history 10 (36%) 8 (35%) 18 (35%)
Prior ablations 0.75 (0.70) 0.70 (0.82) 0.73 (0.75)
Prior cardioversions 0.82 (1.19) 1.13 (1.69) 0.96 (1.43)
Rate-control agent at baseline 21 (75%) 17 (74%) 38 (75%)
Rhythm-control agent at baseline 11 (39%) 8 (35%) 19 (37%)
Rhythm at baseline


    A-paced 0 (0%) 1 (4.3%) 1 (2.0%)
    AFib 2 (7.1%) 1 (4.3%) 3 (5.9%)
    Sinus 26 (93%) 21 (91%) 47 (92%)
Any rhythm or medication change 3 (11%) 7 (30%) 10 (20%)
Left atrial enlargement


    Normal 8 (30%) 13 (57%) 21 (42%)
    Dilated 19 (70%) 10 (43%) 29 (58%)
Mitral regurgitation


    None/trace/trivial 17 (65%) 16 (73%) 33 (69%)
    Mild or greater 9 (35%) 6 (27%) 15 (31%)
LE8 BMI score 12 (21) 14 (22) 13 (21)
LE8 blood-pressure score 44 (31) 50 (27) 47 (29)
LE8 physical-activity score 49 (35) 53 (36) 51 (35)
LE8 blood-glucose score 58 (33) 50 (28) 55 (31)
LE8 cholesterol score 49 (33) 49 (32) 49 (33)
LE8 nicotine score 90 (13) 87 (21) 89 (17)
Change in 6-minute walk distance, m 39 (45) 17 (67) 29 (56)
1 Mean (SD); n (%)
LE8 PA was available for 50 of 51 participants; 6MWD change was available for 48 of 51.
Table 3: Baseline characteristics in enrolled YouControl-AFib and Group Medical Visit participants
Characteristic Group Medical Visits
N = 281		 YouControl-AFib
N = 511		 Overall
N = 791
Age, years 67 (10) 63 (10) 64 (10)
Gender


    Female 15 (54%) 26 (51%) 41 (52%)
    Male 13 (46%) 25 (49%) 38 (48%)
Body mass index, kg/m² 35 (7) 40 (6) 38 (6)
Hypertension 22 (79%) 42 (82%) 64 (81%)
Diabetes mellitus 4 (14%) 18 (35%) 22 (28%)
Heart failure history 11 (39%) 18 (35%) 29 (37%)
Prior ablations 0.64 (0.78) 0.73 (0.75) 0.70 (0.76)
Prior cardioversions 1.54 (1.71) 0.96 (1.43) 1.16 (1.55)
Rate-control agent at baseline 27 (96%) 38 (75%) 65 (82%)
Rhythm-control agent at baseline 10 (36%) 19 (37%) 29 (37%)
Rhythm at baseline


    A-paced 2 (7.1%) 1 (2.0%) 3 (3.8%)
    AFib 3 (11%) 3 (5.9%) 6 (7.6%)
    Sinus 23 (82%) 47 (92%) 70 (89%)
Any rhythm or medication change 5 (18%) 10 (20%) 15 (19%)
Left atrial enlargement


    Normal 11 (41%) 21 (42%) 32 (42%)
    Dilated 16 (59%) 29 (58%) 45 (58%)
Mitral regurgitation


    None/trace/trivial 19 (73%) 33 (69%) 52 (70%)
    Mild or greater 7 (27%) 15 (31%) 22 (30%)
LE8 BMI score 31 (31) 13 (21) 19 (26)
LE8 blood-pressure score 59 (30) 47 (29) 51 (30)
LE8 physical-activity score — 51 (35) 51 (35)
LE8 blood-glucose score 79 (26) 55 (31) 63 (31)
LE8 cholesterol score 74 (19) 49 (33) 57 (31)
LE8 nicotine score 89 (13) 89 (17) 89 (16)
Change in 6-minute walk distance, m 19 (49) 29 (56) 25 (54)
1 Mean (SD); n (%)
LE8 PA was available for 50 of 51 YCA participants and was not collected in GMV. Change in 6MWD was available for 48 YCA and 28 GMV participants.

4.2 Primary 6MWD analysis

The primary analysis included the 48 YCA participants with both baseline and post-intervention 6MWD assessments. The mean paired change in 6MWD was 29.2 m (95% CI: 12.8 to 45.5 m; paired t-test <0.001). A clinically meaningful improvement of at least 30 m was observed in 24 of 48 participants (50%).

The sensitivity analysis excluding participants with a rhythm or relevant medication change included 39 participants. Results are shown alongside the primary analysis in Table 4.

Table 5 summarizes the distribution of 6MWD changes in the primary paired-analysis population.

Table 4: Primary and sensitivity analyses of change in 6-minute walk distance among YouControl-AFib participants
Analysis set N Baseline 6MWD, mean (SD) Post-intervention 6MWD, mean (SD) Mean change, m (95% CI) Paired t-test p-value Improvement ≥30 m, n (%) Responder proportion, 95% CI
Primary analysis 48 423.7 (99.1) 452.9 (100.7) 29.2 (12.8, 45.5) <0.001 24 (50%) (35%, 65%)
Sensitivity analysis: no rhythm or medication change 39 421.2 (94.5) 450.0 (95.0) 28.8 (11.2, 46.5) 0.002 20 (51%) (35%, 68%)

Figure 1 displays individual and overall changes in 6-minute walk distance among YCA participants with paired assessments.

Figure 1: Six-minute walk distance at baseline and post-intervention among YouControl-AFib participants
Table 5: Categories of change in 6-minute walk distance among YouControl-AFib participants with paired assessments
Change in 6-minute walk distance n (%)
Decreased or no improvement (≤0 m) 14 (29%)
Modest improvement (>0 to <30 m) 10 (21%)
Clinically meaningful improvement (≥30 m) 24 (50%)

4.3 Patient-reported outcomes

Changes in patient-reported outcomes are summarized in Table 6. These analyses are exploratory; p-values are unadjusted for multiple comparisons.

For PHQ-9 and AFSS, negative changes indicate improvement. For AFKAT, AFEQT, and SF-12 scores, positive changes indicate improvement.

Table 6: Changes in patient-reported outcomes from baseline to post-intervention among YouControl-AFib participants
Outcome N Baseline, mean (SD) Post-intervention, mean (SD) Mean change, post minus baseline (95% CI) Paired t-test p-value
PHQ-9 score 46 4.8 (3.3) 3.9 (3.5) -0.9 (-1.7, -0.1) 0.022
AFSS score 46 6.9 (5.7) 6.1 (5.5) -0.9 (-2.2, 0.5) 0.208
AF knowledge assessment score (%) 49 91.3 (5.8) 92.2 (5.5) 0.9 (-0.7, 2.4) 0.267
AFEQT score 46 78.6 (15.6) 83.7 (14.6) 5.1 (1.1, 9.1) 0.013
SF-12 physical component score 44 40.4 (4.1) 41.8 (5.3) 1.5 (0.2, 2.8) 0.025
SF-12 mental component score 44 48.1 (5.6) 48.0 (5.1) -0.1 (-2.1, 1.8) 0.887
AFKAT scores are displayed as percentages. P-values are exploratory and unadjusted for multiple comparisons.

Figure 2 displays the distributions of the patient-reported outcomes at baseline and post-intervention. Lines connect paired observations from the same participant.

Figure 2: Patient-reported outcomes at baseline and post-intervention among YouControl-AFib participants

4.4 Differences by residence

Changes by residence and non-urban minus urban differences are shown in Table 7. Adjusted comparisons control for the baseline value of each outcome.

Table 7: Changes in 6MWD and patient-reported outcomes by residence.
Outcome Non-urban change (95% CI), N Urban change (95% CI), N Difference (95% CI) p-value Adjusted difference (95% CI) Adjusted p-value
6MWD (m) 38.8 (20.9, 56.7); N=27 16.8 (-13.8, 47.4); N=21 22.0 (-12.7, 56.7) 0.206 24.1 (-7.8, 56.0) 0.135
AFEQT 1.1 (-3.8, 6.0); N=25 9.8 (3.5, 16.2); N=21 -8.7 (-16.5, -0.9) 0.030 -4.8 (-12.1, 2.6) 0.196
SF-12 physical 2.0 (0.2, 3.7); N=25 0.8 (-1.3, 2.9); N=19 1.1 (-1.5, 3.8) 0.386 1.4 (-1.2, 4.0) 0.272

The non-urban–urban difference in 6MWD change was 22.0 (-12.7, 56.7) m (p = 0.206). The baseline-adjusted difference was 24.1 (-7.8, 56.0) m. The unadjusted AFEQT improvement was greater in urban participants, but the difference weakened after baseline adjustment. SF-12 physical score changes showed no clear difference by residence.

4.5 Contextual comparison with GMV

The GMV cohort is a nonrandomized contextual comparator. Accordingly, the following comparisons describe differences between the two cohorts and should not be interpreted as estimates of a causal intervention effect.

Table 8 summarizes 6MWD outcomes by cohort. Figure 3 displays paired baseline and post-intervention measurements in each cohort. The adjusted model controls for baseline 6MWD only.

Table 8: Six-minute walk distance outcomes in YouControl-AFib and Group Medical Visit cohorts
Cohort N Baseline 6MWD, mean (SD) Post-intervention 6MWD, mean (SD) Change in 6MWD, mean (SD) Improvement ≥30 m, n (%) Responder proportion, 95% CI
Group Medical Visits 28 382.9 (102.1) 402.0 (119.2) 19.1 (48.7) 12 (43%) (24%, 63%)
YouControl-AFib 48 423.7 (99.1) 452.9 (100.7) 29.2 (56.3) 24 (50%) (35%, 65%)
GMV is a nonrandomized contextual comparator.
Figure 3: Six-minute walk distance at baseline and post-intervention in YouControl-AFib and Group Medical Visit cohorts

In an unadjusted comparison, the mean difference in change in 6MWD between YCA and GMV was 10.1 m (95% CI: -14.4 to 34.6 m; 0.414).

After adjustment for baseline 6MWD, the estimated difference in post-intervention 6MWD for YCA versus GMV was 12.7 m (95% CI: -13.2 to 38.6 m; 0.331). These comparisons are contextual and not causal.

4.6 Exploratory predictors and composite score

The following analysis evaluates an exploratory composite score based on rural residence, LE8 blood-pressure score above the cohort median, and LE8 physical activity score above the cohort median. Because the components and cut points were selected using the present dataset, this analysis is hypothesis-generating and does not constitute validation of a prediction model.

One missing PA score is classified as not above the median in the composite score.

The blood-pressure and physical-activity score medians were 50 and 40, respectively. Table 9 summarizes 6MWD change and clinically meaningful response by composite score.

Table 9: Exploratory composite score and 6-minute walk distance outcomes among YouControl-AFib participants with paired assessments
Composite score N Change in 6MWD, mean (SD) Improvement ≥30 m, n (%) Responder proportion, 95% CI
0 6 -9.6 (28.5) 0 (0%) (0%, 46%)
1 16 19.9 (53.6) 6 (38%) (15%, 65%)
2 21 42.1 (61.3) 14 (67%) (43%, 85%)
3 5 50.8 (51.4) 4 (80%) (28%, 99%)
Composite score = rural residence + LE8 blood-pressure score above the paired-YCA median + LE8 physical-activity score above the paired-YCA median.

Figure 4 displays the relationship of the composite score with continuous 6MWD change and the proportion of participants achieving a clinically meaningful improvement.

Figure 4: Exploratory composite score in relation to 6-minute walk distance change and clinically meaningful improvement

4.7 Fitbit adherence and longitudinal activity

Weekly Fitbit step-count and activity-minute data were used to characterize intervention-period activity and post-intervention Fitbit continuation. Continued Fitbit use was defined as any weekly step count greater than zero after the participant’s intervention end date.

Table 10 summarizes post-intervention Fitbit continuation overall and its exploratory association with clinically meaningful improvement in 6MWD.

Table 10: Post-intervention Fitbit continuation and clinically meaningful improvement in 6-minute walk distance
Group N Post-trial Fitbit continuation, n (%) Improvement ≥30 m, n (%)
All enrolled YCA participants 51 47 (92%) —
Continued Fitbit use 45 — 23 (51%)
No recorded post-trial Fitbit use 3 — 1 (33%)
Fisher’s exact test p >0.999. Only three paired participants had no recorded post-trial use; interpret this comparison cautiously.

Activity is summarized at Weeks 1, 6, and 12. Week 1 represents early intervention activity, not a pre-intervention assessment.

Table 11: Fitbit step counts and activity minutes at selected study weeks by 6-minute walk distance response
Measure Study week 6MWD response group N Mean (SD)
Weekly activity minutes 1 No clinically meaningful improvement (<30 m) 24 258 (315)
Weekly activity minutes 1 Clinically meaningful improvement (≥30 m) 24 185 (202)
Weekly activity minutes 6 No clinically meaningful improvement (<30 m) 24 315 (348)
Weekly activity minutes 6 Clinically meaningful improvement (≥30 m) 24 220 (223)
Weekly activity minutes 12 No clinically meaningful improvement (<30 m) 24 273 (372)
Weekly activity minutes 12 Clinically meaningful improvement (≥30 m) 24 168 (130)
Weekly step count 1 No clinically meaningful improvement (<30 m) 24 42 394 (21 401)
Weekly step count 1 Clinically meaningful improvement (≥30 m) 24 35 478 (20 330)
Weekly step count 6 No clinically meaningful improvement (<30 m) 24 44 016 (23 690)
Weekly step count 6 Clinically meaningful improvement (≥30 m) 24 41 605 (17 536)
Weekly step count 12 No clinically meaningful improvement (<30 m) 24 41 565 (31 393)
Weekly step count 12 Clinically meaningful improvement (≥30 m) 24 42 659 (20 370)

Figure 5 displays the distributions of weekly step counts and activity minutes at Weeks 1, 6, and 12 by 6MWD response group.

Figure 5: Fitbit step counts and activity minutes at selected study weeks by 6-minute walk distance response

4.8 LE8 scores and early activity

LE8 BP and PA scores were examined separately. Continuous scores are reported per 10-point increase; the requested cutoff compares scores >30 with scores ≤30. Associations with 6MWD change adjust for baseline 6MWD. Activity-minute change is Week 12 minus Week 1, adjusted for Week 1 activity. All additional analyses are exploratory; p-values are unadjusted for multiple comparisons.

Table 12: Associations of LE8 BP and PA scores with walking response and Fitbit activity.
Scale Outcome N Measure Estimate (95% CI) p-value
LE8 BP
Per 10 points 6MWD change (m) 48 Mean difference 4.6 (-0.7, 9.9) 0.087
>30 vs ≤30 6MWD change (m) 48 Mean difference 21.6 (-10.3, 53.5) 0.179
Per 10 points 6MWD response ≥30 m 48 Odds ratio 1.231 (1.001, 1.514) 0.048
>30 vs ≤30 6MWD response ≥30 m 48 Odds ratio 2.318 (0.725, 7.413) 0.157
Per 10 points Week 1 steps 51 Mean difference -267.8 (-2304.8, 1769.1) 0.793
>30 vs ≤30 Week 1 steps 51 Mean difference -2760.9 (-14473.4, 8951.6) 0.638
Per 10 points Activity-minute change 51 Mean difference 4.9 (-12.4, 22.2) 0.569
>30 vs ≤30 Activity-minute change 51 Mean difference 7.0 (-93.0, 107.1) 0.888
LE8 PA
Per 10 points 6MWD change (m) 47 Mean difference 3.4 (-1.4, 8.1) 0.158
>30 vs ≤30 6MWD change (m) 47 Mean difference 14.9 (-19.9, 49.7) 0.393
Per 10 points 6MWD response ≥30 m 47 Odds ratio 1.194 (0.998, 1.428) 0.053
>30 vs ≤30 6MWD response ≥30 m 47 Odds ratio 2.082 (0.599, 7.232) 0.249
Per 10 points Week 1 steps 50 Mean difference 821.5 (-829.4, 2472.3) 0.322
>30 vs ≤30 Week 1 steps 50 Mean difference 4264.6 (-7676.9, 16206.2) 0.476
Per 10 points Activity-minute change 50 Mean difference -0.4 (-15.4, 14.5) 0.955
>30 vs ≤30 Activity-minute change 50 Mean difference 96.8 (-6.9, 200.5) 0.067
Models include one LE8 score at a time. 6MWD outcomes adjust for baseline 6MWD; activity-minute change adjusts for Week 1 activity. Week 1 step associations are unadjusted. Missing PA scores are excluded.

LE8 BP showed a borderline association with the ≥30 m response (OR 1.231 (1.001, 1.514) per 10 points; p = 0.048). The corresponding PA association was borderline (p = 0.053). Neither score showed clear associations with continuous 6MWD change, Week 1 steps, or activity-minute change. The >30 cutoffs did not strengthen the response associations.

Week 1 activity and baseline measurements are compared using Spearman correlations (Table 13).

Table 13: Correlations of early Fitbit activity with baseline walking distance and LE8 PA.
Pair N Spearman correlation (95% CI) p-value
Week 1 steps / baseline 6MWD 51 0.54 (0.30, 0.73) <0.001
Week 1 activity / baseline 6MWD 51 0.40 (0.07, 0.65) 0.004
LE8 PA / Week 1 steps 50 0.16 (-0.11, 0.42) 0.255
LE8 PA / Week 1 activity 50 0.14 (-0.14, 0.40) 0.349
LE8 PA / baseline 6MWD 50 0.03 (-0.26, 0.32) 0.839
95% CIs use 2,000 participant-level bootstrap samples. Activity refers to weekly activity minutes.
Table 14: Distribution of Week 1 Fitbit activity in YCA participants.
Measure N Median (IQR) Range
Weekly activity minutes 51 136 (40, 272) 0–1413
Weekly steps 51 40671 (21022, 55865) 0–74971

Week 1 steps and activity minutes correlated with baseline 6MWD (Spearman r = 0.54 and 0.40, respectively). LE8 PA correlations with Fitbit activity and baseline 6MWD were weak (r = 0.03–0.16).

These are weekly totals. Variation between participants does not by itself identify measurement error or inconsistency with self-reported activity.