Table 2.
Association between education and high allostatic load presented as prevalence ratios (PRs) and associated 95% confidence intervals (CIs), among 4177 participants, using NHANES weighting an estimated 9,494,904 US Black women
| No. (weighted %)a | Prevalence ratios (95% confidence interval) | ||||
|---|---|---|---|---|---|
| Crude | Model 1b | Model 2c | Model 3d | ||
| Education level | |||||
| Less than high school (Referent) | 443 (25.3) | 1.000 (Referent) | 1.000 (Referent) | 1.000 (Referent) | 1.000 (Referent) |
| High school/GED | 393 (24.1) | 0.850 (0.849–0.850) | 0.946 (0.937–0.954) | 0.932 (0.925–0.940) | 0.926 (0.920–0.933) |
| Some college or associates degree | 546 (35.3) | 0.858 (0.857–0.859) | 0.959 (0.952–0.967) | 1.005 (0.993–1.017) | 1.027 (1.013–1.041) |
| College graduate or higher | 226 (15.3) | 0.746 (0.738–0.753) | 0.758 (0.750–0.767) | 0.835 (0.819–0.851) | 0.857 (0.839–0.876) |
Prevalence ratios for high allostatic load are estimated using modified Poisson regression with robust variance estimation and accounting for NHANES weighting. Confidence intervals estimated using delete-1 jackknife method accounting for complex statistical weighting, cluster, and strata
aNumber of participants with high allostatic load per stratum (weighted stratum proportion with high allostatic load)
bModel 1: Adjusted for age only
cModel 2: Additionally adjusted for total number of pregnancies, age at menarche, and poverty to income ratio,
dModel 3: Additionally adjusted for depressive disorder, smoker status, ever congestive heart failure, and ever heart attack