Plasma triglyceride as a risk factor for cardiovascular disease.

OBJECTIVES: To determine whether cardiovascular disease can be predicted from plasma triglyceride levels independent of high density lipoprotein (HDL) cholesterol levels. DATA SOURCES AND STUDY SELECTION: Seventeen prospective population-based studies, ranging in publication date from 1965 to 1994, were selected based on uniform criteria. Among the studies, 16 were conducted in men, comprising data for more than 46,000 men, with an average follow-up period of more than eight years. The data from women came from five studies, including nearly 11,000 women with an 11-year average follow-up. DATA EXTRACTION: The meta-analysis protocol comprised the following steps: estimating the relative risk of triglyceride for each individual study; standardizing the relative risks to a 1 mmol/l increase; estimating the variance and weighting relative risks by the inverse of the variance, and estimating the summary relative risk by averaging the weighted relative risks. DATA SYNTHESIS: Among studies of men, relative risks ranged from 1.1 to 2.0, and the summary relative risk was 1.32 (P < 0.05). All five studies of women were statistically significant, with relative risks ranging from 1.7 to 2.0. The summary relative risk was 1.76, higher than that for men, indicating a 76% increase in risk for women. For the studies among men in which HDL cholesterol data were available, multivariate adjusted relative risk was attenuated to 1.14 but was still statistically significant. Although only two studies of women included HDL cholesterol adjustments, the summary relative risk for triglyceride remained statistically significant, with a value of 1.37.
CONCLUSIONS: When all available data are taken into account, increased triglyceride level is a risk factor for cardiovascular disease, independent of HDL cholesterol level. Clinical trials are urgently needed to determine whether lowering plasma triglyceride levels reduces subsequent risk of cardiovascular disease.