Table I.
Conditional Models Examining Family Differences in Social Determinants of Health and COVID-19 Impacts on Social Determinants of Health
| Study groups |
|||||||||
|---|---|---|---|---|---|---|---|---|---|
| NHW parents of healthy childrenb |
NHW parents of children w/asthma |
BIPOC parents of healthy children |
BIPOC parents of children w/asthma |
||||||
| Social determinants of healtha | Mean (SE)/N (%) | Referent group | Mean (SE)/N (%) | OR (CI)/B using robust SEsc | Mean (SE)/N (%) | OR (CI)/B using robust SEsc | Mean (SE)/N (%) | OR (CI)/B using robust SEsc | Wald test/F |
| Healthcare access | |||||||||
| Didn’t get medical care in past year due to cost | 8 (11.8%) | — | 5 (8.1%) | 0.49 (0.14–1.76) | 7 (7.0%) | 0.52 (0.17–1.58) | 12 (13.2%) | 0.86 (0.30–2.48) | 2.13, p = .55 |
| Food insecurity in past year | |||||||||
| Food didn’t last, and couldn’t afford more (range: 0–2) | 0.26 (11) | — | 0.34 (0.10) | B = 0.08, p = .38, η2 = 0.003 | 0.34 (0.10) | B = 0.08, p= .28, η2 = 0.004 | 0.51 (0.10) | B = 0.24, p = .02, η2 = 0.02 | — |
| Couldn’t afford to eat balanced meals (range: 0–2) | 0.50 (0.11) | — | 0.53 (0.10) | B = 0. 03, p= .75, η2 = 0.00 | 0.53 (0.10) | B = 0.03, p = .69, η2 = 0.001 | 0.68 (0.10) | B = 0.18, p = .07, η2 = 0.01 | — |
| Adults reduced/skipped meals because there wasn’t enough money for food | 6 (8.8%) | — | 7 (11.3%) | 1.36 (0.39–.4.70) | 8 (8.0%) | 1.13 (0.35–3.60) | 14 (15.4%) | 1.93 (0.63–5.92) | 1.73, p = .63 |
| Discrimination | |||||||||
| Ever a time you would have gotten medical care if you had belonged to a different race | 3 (4.4%) | — | 1 (1.6%) | 0.29 (0.03–3.01) | 19 (19.0%) | 5.47 (1.49–20.11) | 13 (14.3%) | 3.51 (0.88–13.97) | 12.63, p = .01 |
| Sum Frequency Score on Everyday Discrimination Scale (range: 0–50) | 13.13 (1.70) | — | 14.54 (1.62) | B = 1.41, p = .36, η2 = 0.003 | 17.41 (1.63) | B = 4.28, p ≤ .001, η2 = 0.05 | 18.12 (1.52) | B = 5.00, p ≤ . 001, η2 = 0.05 | — |
| Changes in social determinants of health during COVID-19a | |||||||||
| Caregiver(s) lost job during pandemic | 9 (13.2) | — | 8 (12.9%) | 0.99 (0.30–3.02) | 14 (14.0%) | 1.26 (0.47–3.35) | 17 (18.7%) | 1.69 (0.62–4.59) | 1.66, p = .65 |
| Having difficulty paying bills/buying food during pandemic | 6 (8.8%) | — | 9 (14.5%) | 1.87 (0.56–6.26) | 16 (16.0%) | 2.13 (0.72–6.31) | 28 (30.8%) | 5.27 (1.80–15.40) | 12.05, p = .01 |
| Adults having to work longer hours | 5 (7.4%) | — | 8 (12.9%) | 2.88 (0.79–10.53) | 12 (12.0%) | 2.14 (0.63–7.26) | 14 (15.4%) | 3.35(0.98–11.46) | 4.01, p = .26 |
| Adult filed for unemployment | 9 (13.2%) | — | 6 (9.7%) | 0.72 (0.22–2.36) | 11 (11.0%) | 0.90 (0.33–2.49) | 9 (9.9%) | 0.76 (0.25–2.30) | 0.41, p = .94 |
| Applied for public assistance/food stamps | 5 (7.4%) | — | 3 (4.8%) | 0.93 (0.19–4.67) | 7 (7.0%) | 1.40 (0.37–5.25) | 8 (8.8%) | 1.73 (0.45–6.65) | 1.04, p = .79 |
| Degree COVID-19 has changed household income (range: 0–3) | 0.90 (0.19) | — | 1.35 (0.18) | B = 0.45, p = .02, η2 = 0.02 | 1.32 (0.19) | B = 0.43, p = .01, η2 = 0.03 | 1.31 (0.17) | B = 0.42, p = .01, η2 = 0.02 | F = 2.96, p = .03, η2 = 0.03 |
| Degree COVID-19 has changed access to food (range: 0–3) | 0.75 (0.15) | — | 0.91 (0.14) | B = 0.16, p = .28, η2 = 0.004 | 0.90 (0.14) | B = 0.15, p = .18, η2 = 0.01 | 0.94 (0.13) | B = 0.19, p = .14, η2 = 0.01 | F = 0.81, p = .45, η2 = 0.01 |
| Degree COVID-19 reduced access to medical care (range: 0–4) | 1.17 (0.21) | — | 1.69 (0.20) | B = 0.52, p = 0.03, η2 = 0.02 | 1.25 (0.20) | B = 0.09, p = .63, η2 = 0.001 | 1.56 (0.21) | B = 0.40, p = .045, η2 = 0.02 | F = 3.18, p = .02, η2 = 0.03 |
| Degree COVID-19 reduced access to mental healthcare (range: 0–4) | 0.84 (18) | — | 1.10 (0.17) | B = 0.26, p = .19, η2 = 0.01 | 0.96 (0.18) | B = 0.13, p = .39, η2 = 0.003 | 0.97 (0.16) | B = 0.13, p = .42, η2 =.0002 | — |
Note. NHW = Non-Hispanic White. Estimated marginal means are presented. The cell N’s and percentages are based on the full sample (N = 321); sample sizes for specific analyses were lower to missing data on covariates.
a
Analyses controlled for variables found to be different (p < .1) between groups (i.e., parent age, parent education, and parent asthma status) and if the child had been tested for COVID-19 and if parent had been diagnosed with COVID-19.
b
Referent Group is Non-Hispanic, White Parents of Healthy Children.
c
If significant heteroscedasticity was detected, parameter estimates with robust SEs were reported.