Table 3.

Association of pre-lockdown characteristics and total physical activity ^a during the lockdown or pre-post improvement or maintenance of sufficient physical activity ^a.

Prelockdown Characteristics (from the Last Available Assessment)	Linear Regression			Logistic Regression
	BPAAT during-Lockdown Total Score a			Improve or Maintain Sufficient PA during vs. Prelockdown b
	B	(95% CI)	p	OR	(95% CI)	p
Age	0.01	−0.04; 0.05	0.752	1.03	0.95; 1.12	0.494
Female	0.30	−0.31; 0.91	0.336	2.61	0.85; 8.04	0.094
Education	0.03	−0.13, 0.18	0.705	0.87	0.65; 1.17	0.370
Depressive symptoms (pre-lockdown) c	−1.15	−1.89; −0.41	0.003	0.12	0.02; 0.55	0.006
Social contact different than family (during the lockdown)	0.99	0.41; 1.57	0.001	5.07	1.60; 16.08	0.006
Self-reported fatigue	−1.25	−1.87; −0.63	<0.001	0.11	0.03; 0.44	0.002
Reading to stay active (during the lockdown)	0.74	0.08; 1.39	0.028	6.29	1.66; 23.90	0.007
Lives alone (pre-lockdown)	−1.30	−2.14; −0.46	0.003	-		-
Lives alone (during the lockdown)	−0.78	−1.74; 0.07	0.073	-	-	-
Diagnosis of cognitive impairment (pre-lockdown) d	-	-	-	0.29	0.08; 1.06	0.061

Stepwise multivariable linear and logistic regression were performed as appropriate. Age, sex and education level were set as lockterm in both cases. Variables with empty cells were not included in the final model. PA: Physical Activity. ^a Brief Physical Activity Assessment score, range from 0–8. ^b Change in physical activity level: Improve insufficient or maintain sufficient physical activity level vs reduction or maintain insufficient physical activity level. ^c Geriatric Depression Scale Yesavage: range from 0–15 points (>5 points: depression). ^d Previous diagnosis of cognitive impairment or dementia or last Minicog© assessment <3. Minicog© range 0–5 (<3 positive screening for cognitive impairment).