Table 3. Multivariable Logistic regression for the frequency of WD use and physical activity levels among older adults in the US.

The primary predictor for each model is the frequency of wearable device use (never user, infrequent user, or frequent user; never user is the reference category). Models were *adjusted for sex, age category, educational level, insurance status, marital status, race, income level, health care provider status, confidence in one's own ability to take care of one's own health, self-health, and presence of comorbidities.

WD: wearable device; PA: physical activity; OR: odds ratio; CI: confidence intervals

Variables	Unadjusted OR OR (95% C.I.)	p-value	Adjusted OR* OR (95% C.I.)	p-value
Outcome: weekly PA level ("reference group" is "never user")
Infrequent user	1.32 (0.76, 2.31)	0.320	0.92 (0.49, 1.71)	0.786
Frequent user	2.87 (2.09, 3.98)	<0.001	2.05 (1.35, 3.09)	0.001
Outcome: strength training (reference category for primary predictor is ‘never user’)
Infrequent user	1.39 (0.89, 2.17)	0.144	1.21 (0.64, 2.48)	0.556
Frequent user	2.45 (1.78, 3.38)	<0.001	1.75 (1.25, 2.45)	0.001
Outcome: sedentary Level (reference category for primary predictor is ‘never user’)
Infrequent user	1.26 (0.76, 2.07)	0.363	1.28 (0.66, 1.71)	0.467
Frequent user	0.64 (0.47, 0.88)	0.006	0.83 (0.54, 1.27)	0.390