Major depression increases cardiovascular risk despite lower cholesterol levels. Little is known about effects of antidepressants on metabolic risk factors. We studied lipoprotein composition, insulin sensitivity (quantitative insulin sensitivity check index), and saliva cortisol in 78 depressed patients before and after 35 days of amitriptyline or paroxetin treatment. Data were analyzed by principal component factor analyses and analysis of variance (ANOVA). At baseline, quantitative insulin sensitivity check index was inversely correlated with cortisol (r = -0.46; P = 0.005) in normal weight patients, with body mass index in overweight patients (r = -0.50; P < 0.001). In overweight patients, hypercortisolemia correlated inversely with total and low density lipoprotein cholesterol (eg, cortisol at 4:00 PM and low density lipoprotein cholesterol: r = -0.49, P = 0.002). After treatment, quantitative insulin sensitivity check index was unchanged. Triglycerides increased in responders to amitriptyline only (P < 0.05). Parameters of cholesterol metabolism improved slightly without differences between treatment groups (eg, high density lipoprotein: pre 43.5 +/- 12.0; post 47.6 +/- 13.0 mg/dL; P = 0.01; low density lipoprotein triglycerides, a measure of low density lipoprotein atherogenicity: pre 458 +/- 120; post 415 +/- 130 mg/g; P < 0,01). The inverse correlation of cortisol and cholesterol, at least in the obese subgroup, proposes a mechanism for the known association of depression with low cholesterol. As determinants of plasma lipids in major depression, we identified body mass index, insulin sensitivity, and cortisol. Although uncontrolled, our data suggest that treatment of depression exerts a mainly beneficial effect on lipid regulation.