Table 2.
Effect of Everyday Discrimination on Stress and Health Outcomes from Multivariable Regression Models
| Outcome | Multiple Imputation (N=535,m=4) | Complete Case Analysis | Model | |||||
|---|---|---|---|---|---|---|---|---|
| Effect of Discrimination | 95% CI | p-value | Effect of Discrimination | 95% CI | p-value | N (%) | ||
| Perceived Stress | 2.45 | (1.66, 3.24) | <0.0001 | 2.62 | (2.47, 2.77) | <0.0001 | 371 (69%) | Linear |
| Chronic Pain | 1.52 | (1.15, 2.02) | 0.0034 | 1.86 | (1.29, 2.67) | <0.0001 | 360 (67%) | Logistic |
| Anxiety/Depression | 1.88 | (1.23, 2.87) | 0.0038 | 1.35 | (0.75, 2.44) | 0.3138 | 354 (66%) | Logistic |
| Diabetes | 1.13 | (0.79, 1.62) | 0.4904 | 1.17 | (0.75, 1.84) | 0.4921 | 359 (67%) | Logistic |
| Hypercholesterolemia | 1.21 | (0.89, 1.63) | 0.2224 | 1.14 | (0.80, 1.63) | 0.4757 | 352 (66%) | Logistic |
| Hypertension | 1.31 | (0.93, 1.85) | 0.1248 | 1.29 | (0.86, 1.95) | 0.2234 | 354 (66%) | Logistic |
We display the regression coefficient for the linear model, the relative risk ratio (RRR) for the Poisson model, and the odds ratio (OR) for the logistic regression models. Multiple Imputed Estimates were derived using Rubin’s combining rules for the m=4 imputed datasets..