Table 3.
Linear regression analyses identifying correlates of PHQ-4 scores and COVID-related stress
| Variables | PHQ-4 | COVID-related stress | ||||
|---|---|---|---|---|---|---|
| B | SE | p | B | SE | p | |
| Sociodemographics | ||||||
| Age | –0.01 | .02 | .810 | 0.01 | .00 | .877 |
| Female (ref: male) a | 0.70 | .14 | <.001 | 0.15 | .03 | <.001 |
| Sexual orientation (ref: heterosexual) | 1.15 | .16 | <.001 | 0.08 | .04 | .023 |
| Race (ref: White) | ||||||
| Black | –0.83 | .31 | .008 | –0.28 | .07 | <.001 |
| Asian | –0.85 | .21 | <.001 | –0.01 | .05 | .816 |
| Other | 0.08 | .22 | .712 | –0.02 | .05 | .744 |
| Hispanic (ref: non-Hispanic) | –0.17 | .22 | .431 | 0.06 | .05 | .226 |
| Predictors of interest | ||||||
| MSA (ref: Oklahoma City) b | ||||||
| Atlanta | –0.07 | .27 | .792 | 0.04 | .06 | .547 |
| Boston | –0.13 | .27 | .636 | 0.08 | .06 | .165 |
| Minneapolis | 0.13 | .28 | .626 | 0.15 | .06 | .016 |
| San Diego | –0.29 | .28 | .290 | 0.12 | .06 | .046 |
| Seattle | 0.21 | .28 | .453 | 0.11 | .06 | .072 |
| Social distancing adherence | 0.26 | .08 | .001 | 0.04 | .02 | .030 |
| Sanitary adherence | –0.17 | .09 | .046 | 0.08 | .02 | <.001 |
| Employment status (ref: work outside of home) | ||||||
| Unemployed | 0.51 | .22 | .020 | 0.04 | .05 | .360 |
| Work from home | 0.10 | .30 | .752 | 0.02 | .07 | .746 |
| Switched to work from home | –0.23 | .18 | .209 | 0.05 | .04 | .193 |
| Laid off | 0.83 | .24 | <.001 | 0.21 | .05 | <.001 |
| New employment | 0.21 | .34 | .540 | 0.13 | .08 | .080 |
| Other | 0.36 | .35 | .313 | 0.04 | .08 | .638 |
| Household composition * | ||||||
| Partner/spouse | 0.29 | .17 | .083 | 0.04 | .04 | .308 |
| Children | –0.42 | .22 | .062 | 0.05 | .05 | .286 |
| Others (other than partner/children) | 0.37 | .17 | .033 | 0.04 | .04 | .261 |
| Political orientation (ref: no lean) | ||||||
| Democrat | 0.47 | .20 | .021 | 0.12 | .05 | .007 |
| Republican | –0.78 | .26 | .003 | –0.05 | .06 | .400 |
| Adjusted R2 | .102 | .072 | ||||
Notes * Dummy coded, so each is compared to “no” for that variable. a 69 participants responded as “other” and were thus excluded from analyses. b 198 participants resided in “other” MSA at W5 and were excluded from primary analyses