Table 2 Zero-inflated negative binomial regressions predicting alcohol use (N = 2299).
Variables | OR/IRR | 95% CI for IRR/OR | P value | |
---|---|---|---|---|
Lower | Upper | |||
Logistic portion of model (for predicting excess zeros, OR) | ||||
CES-D-10 score | 0.916 | 0.870 | 0.964 | < .001*** |
Cognitive score | 1.005 | 0.951 | 1.060 | .868 |
Age | 0.992 | 0.974 | 1.010 | .384 |
Education Level | 0.934 | 0.886 | 0.985 | .012 |
Gender (Men versus Women) | 2.296 | 1.642 | 3.211 | < .001*** |
Religion (Christians versus Muslims) | 2.256 | 1.620 | 3.142 | < .001*** |
Count portion of model (negative binomial regression with log link, IRR) | ||||
CES-D-10 score | 0.978 | 0.960 | 0.996 | .016 |
Cognitive score | 0.953 | 0.919 | 0.988 | .010 |
Age | 0.983 | 0.971 | 0.995 | .007 |
Education Level | 1.045 | 1.007 | 1.083 | .018 |
Gender (Men versus Women) | 0.474 | 0.375 | 0.599 | < .001*** |
Religion (Christians versus Muslims) | 0.643 | 0.517 | 0.799 | < .001*** |