Adjusted hazard ratios (95% CI) for all-cause mortality, cardiovascular mortality and incident coronary heart disease predicted by anxious depression classes derived using latent class analysis (LCA) [4], and by cut-off scores on the Diabetes Anxiety Depression Scale (DADS) for different levels of anxious depression.

In each model, the psychiatric categories were added to the most parsimonious Cox proportional hazards model of independently non-psychiatric predictors of the clinical outcome described previously [4].