Parameter   Inverse probability weighted (IPW) regression adjusted estimates Unweighted logistic regression coefficients
Effects of wealth Average effects of wealth (ATE)† (SE) Coefficient‡ (SE)
Middle-wealth category 0.005**§ (0.001) (counterfactual: if subjects were identical in all characteristics to those in ‘low-wealth’ group) 0.060* (0.027)
High-wealth category 0.001 (0.001) (counterfactual: if subjects were identical to those in ‘low-wealth’ group) -0.083** (0.025)
Potential outcome mean& 0.057***^ (0.001) (predicted value of prevalence for the low-wealth group)