Serum Lipoproteins Improve After Successful Pharmacologic Antidepressant Treatment: A Randomized Open-Label Prospective Trial

Objective: Despite reports of lower plasma cholesterol in depressed patients, major depressive disorder has been shown to increase cardiovascular risk. Our objective was to study the composition of lipoproteins in depressed patients and controls and to examine the effects of pharmacologic treatment and treatment response on lipoprotein composition.

Method: Lipoprotein composition was analyzed in 65 adult inpatients at a university psychiatric hospital in Germany with DSM-IV major depressive disorder and 33 healthy controls (recruited via newspaper and radio ads) matched for age and sex. After the cross-sectional study phase, the patients were randomized in an open-label prospective trial to treatment with either mirtazapine or venlafaxine. Lipoproteins were reanalyzed after 4 weeks of treatment. Main outcome measures were total cholesterol, the low-density lipoprotein (LDL) to high-density lipoprotein (HDL) cholesterol ratio, and the LDL triglycerides to apolipoprotein B ratio. Secondary outcome measures were total triglycerides, HDL and LDL cholesterol levels, and apolipoproteins A1 and B levels. Comparisons were made between the 2 drug groups and between remitters and nonremitters as measured by the 21-item Hamilton Depression Rating Scale. The study was conducted from April 2003 through December 2007.

Results: Total cholesterol at baseline was lower in patients than in controls (mean ± SD = 4.99 ± 0.98 mmol/L vs 5.63 ± 1.01 mmol/L; P = .003), with significantly lower HDL cholesterol (P < .001) and LDL cholesterol (P = .03) in patients. However, the ratio of LDL triglycerides to apolipoprotein B, an index of size and atherogenic potential of LDL particles, was higher in depressed subjects (mean ± SD = 0.46 ± 0.14 mmol/g vs 0.38 ± 0.09 mmol/g; P = .002). Irrespective of treatment allocation, we found significant improvement of cardiovascular risk parameters in remitters but found deterioration in nonresponders. The LDL cholesterol mean change from baseline (remitters vs partial responders vs nonresponders) was −0.06 mmol/L versus +0.39 mmol/L versus +0.56 mmol/L (P = .014); the mean change in LDL/HDL cholesterol ratio was −0.50 versus +0.14 versus +0.80 (P = .002); and the mean change in the LDL triglycerides per apolipoprotein B ratio was −0.01 versus −0.01 versus +0.08 (P = .045). No drug-specific changes in lipid concentrations during treatment were observed except for total cholesterol (venlafaxine group mean = −0.02 mmol/L and mirtazapine group mean = +0.37 mmol/L; P = .033).

Conclusions: In depressed patients, lipoprotein structure is changed toward LDL particles with a higher atherogenic potential. Remission from depression is associated with an improvement of the LDL/HDL cholesterol ratio, shifting lipoproteins toward a less atherogenic composition. Our findings should be confirmed in a larger study, as they have relevance for both researchers and clinicians.

Trial Registration: German Clinical Trial Registry Identifier: DRKS00000008

J Clin Psychiatry 2011;72(7):885-891

Submitted: November 23, 2009; accepted October 11, 2010.

Online ahead of print: January 25, 2011 (doi:10.4088/JCP.09m05853blu).

Corresponding author: Daniel Kopf, MD, Bethanien Hospital, Center for Geriatric Medicine, University of Heidelberg, Rohrbacher Str 149, 68126 Heidelberg, Germany (dkopf@bethanien-heidelberg.de).