Table 3.
Factors associated with physical activity during lockdown.
| Factors | Insufficiently Active | Decrease in Physical Activity | |
|---|---|---|---|
| OR (95% CI) | OR (95% CI) | ||
| Country-level COVID-19 risk | 10-unit change of confirmed cases | 0.97 (0.90, 1.05) | 0.98 (0.92, 1.04) |
| 5-unit change of policy response score | 1.22 (1.03, 1.45) | 1.01 (0.88, 1.15) | |
| Mental health symptoms during lockdown | 10-unit change of anxiety score | 0.98 (0.92, 1.05) | 1.18 (1.11, 1.26) |
| 10-unit change of depression score | 1.44 (1.07, 1.22) | 1.15 (1.08, 1.22) | |
| Sex a | Female | 1.30 (1.19, 1.42) | 0.92 (0.85, 1.00) |
| Male | Reference | Reference | |
| Age group | 18–24 | Reference | Reference |
| 25–34 | 1.15 (1.01, 1.30) | 1.06 (0.93, 1.20) | |
| 35–44 | 1.16 (0.99, 1.36) | 1.19 (1.03, 1.39) | |
| 45–54 | 1.09 (0.92, 1.29) | 1.24 (1.05, 1.46) | |
| 55–64 | 0.98 (0.81, 1.18) | 1.37 (1.14, 1.64) | |
| 65 years or older | 1.50 (1.18, 1.91) | 1.10 (0.87, 1.40) | |
| Marital status a | Married | 1.15 (1.02, 1.31) | 0.98 (0.87, 1.11) |
| Single | Reference | Reference | |
| Education level | High school degree or lower | Reference | Reference |
| Associate degree | 1.01 (0.88, 1.15) | 1.06 (0.93, 1.21) | |
| Bachelor’s degree | 0.79 (0.71, 0.88) | 1.20(1.08, 1.33) | |
| Graduate degree | 0.79 (0.70, 0.89) | 1.30(1.16, 1.47) | |
| General health before COVID-19 | Excellent | 0.34 (0.28, 0.41) | 1.00 (0.83, 1.20) |
| Very Good | 0.47 (0.39, 0.56) | 1.18 (0.99, 1.40) | |
| Good | 0.72 (0.59, 0.86) | 1.17 (0.97, 1.40) | |
| Fair | Reference | Reference | |
| Living with person(s) ≥65 years old during lockdown | Yes | 0.96 (0.86, 1.06) | 0.93 (0.84, 1.03) |
| No | Reference | Reference | |
| Having a child in the household during lockdown | Yes | 1.17 (1.05, 1.31) | 0.89 (0.80, 0.99) |
| No | Reference | Reference | |
| Required to go out of the home to work during lockdown | Full-time | 1.16 (1.04, 1.30) | 0.91 (0.81, 1.01) |
| Part-time | 1.03 (0.93, 1.14) | 1.00 (0.90, 1.11) | |
| No | Reference | Reference | |
| ICC (null model) b | 0.075 | 0.045 | |
| AIC (full model) b | 14,798.77 | 15,369.79 | |
| BIC (full model) b | 14,808.72 | 15,379.73 | |
Note: Results were based on two separate hierarchical generalized linear models with a dichotomous outcome variable that assessed the factors associated with the outcomes of (1) insufficiently active, and (2) a decrease in physical activity during lockdown. Abbreviations: ICC: intralevel correlation coefficient; AIC: Akaike information criterion; BIC: Bayesian information criterion; CI: confidence intervals. a Estimates for participants that reported “Other” on variables of sex and marital status were not listed. b Results from null model tests were not listed, except for the ICC from null models. Null models were conducted without fixed factors but a random intercept of the country variable. For insufficiently active outcome, null model n = 11,746, ICC Country = 0.075, AIC = 15,473.24, and BIC = 15,474.04; for decrease in PA engagement outcome, null model n = 11,746, ICC Country = 0.045, AIC = 15,773.85, and BIC = 15,774.65.