Fig 1.
Conceptual model relating cognitive factors from the Health Belief Model and resilience to pandemic non-work, non-household contact rates.
The light blue box shows the cognitive factors from the Health Belief Model, ‘Health Beliefs’, and resilience constructs that we evaluated and the variables we used to assess them. Contextual factors were assessed as potentially confounding the association between cognitive factors/resilience and contacts, and modifying factors were considered to modify the association.
Table 1.
Participant characteristics by study cohort, with source state population profiles provided for comparison.
Table 2.
Results of negative binomial models testing the effect of Health Belief Model (HBM) and resilience constructs on non-work, non-household contact rates during April and May 2020.
Fig 2.
Health Belief Model and resilience constructs stratified by race/ethnicity for the family cohort (C1).
(A) shows all constructs. (B) is limited to those variables with rate ratio (RR) estimates <3.0 for closer examination. RRs were estimated from negative binomial models adjusted for age (linear and quadratic terms), gender, income, state of residence, week surveyed, and number of COVID-19-related symptoms experienced in the past 30 days. Bars indicate 95% confidence intervals.
Fig 3.
Health Belief Model and resilience constructs stratified by socioeconomic status variables.
(A) shows rate ratio (RR) estimates in the family cohort (C1) from negative binomial models by income adjusted for age (linear and quadratic terms), gender, race/ethnicity, state of residence, week surveyed, and number of COVID-19-related symptoms experienced in the past 30 days. (B) shows RR estimates in the older adult cohort (C2) from negative binomial models by education level adjusted for age, gender, race/ethnicity, Parkinson’s disease, having or living with someone with comorbidities, and the EQ-5D measure of health status. Bars indicate 95% confidence intervals.