Turning to the SEM, the fit between the data and the model in Figure 1 also was acceptable, χ2 (5, N = 208) = 6.33, p = .28, CFI = .99, RMSEA = .04. Standardized path estimates are reported in the figure. Results showed that subjective well-being at age 16 had a statistically significant positive association with alcohol use at age 18, controlling for earlier alcohol use and adverse consequences (and the covariates). Neither alcohol use nor adverse consequences at age 16 was significantly related to subsequent subjective well-being; however, alcohol use at age 18 had a statistically significant positive association with alcohol problems at age 21, independent of the influences of earlier adverse consequences and the covariates.
