PLASMA-LIPIDS AND DESOGESTREL AND ETHINYL ESTRADIOL - A METAANALYSIS

Objective: To review the literature to determine the magnitude of the effect of 150 mu g desogestrel-30 mu g ethinyl E(2), an oral contracep tive (OC) formulation, on plasma lipid concentrations in healthy women using meta-analysis techniques. Data sources: All English-language pu blished reports (1981 to 1991) on lipid parameters in women taking 150 mu g desogestrel and 30 mu g ethinyl E(2) for up to 6 months obtained via an Embase database search and via a subsequent review of the refe rence lists. Methods of study selection: Of 98 articles, 18 met eligib ility criteria and were included in the meta-analysis. Data extraction and synthesis: Data on total cholesterol, high-density lipoprotein ch olesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and tr iglycerides were extracted. The change in each parameter from baseline to cycle 6 was estimated as a weighted mean of changes reported in ea ch study; a standard error of the estimate was calculated. This proced ure was validated by goodness-of-fit tests. Results: The following sta tistically significant changes from baseline to cycle 6 were estimated (mean +/- SE): HDL-C: 0.15 +/- 0.02 mmol/L (5.80 +/- 0.62 mg/dL) trig lycerides: 0.28 +/- 0.03 mmol/L (24.49 +/- 2.59 mg/dL); and LDL: -0.12 +/- 0.04 mmol/L (-4.53 +/- 1.55 mg/dL). There was a nonsignificant tr end toward an increase in total cholesterol. Conclusion: When given in combination with 30 mu g ethinyl E(2), desogestrel increased HDL-C an d triglycerides and decreased LDL-C. The positive impact on HDL-C and LDL-C suggests that a potential cardioprotective benefit (rather than an atherosclerosis risk) may occur with prolonged use of such an OC, b ut this hypothesis will be difficult to prove.