Background: Whether the plasma triglyceride level is a risk factor for coronary heart disease has been controversial, and evaluation of the triglyceride level as a risk factor is fraught with methodologic difficulties.
Methods: We studied the association between plasma triglyceride levels and the 12-year incidence of death from coronary heart disease in 10 North American populations participating in the Lipid Research Clinics Follow-up Study, while adjusting for the potential confounding effects of other risk factors for cardiovascular disease, including the level of high-density lipoprotein (HDL) cholesterol. All analyses were sex-specific, and separate analyses were performed in high and low strata of HDL cholesterol, low-density lipoprotein (LDL) cholesterol, fasting plasma glucose, and age.
Results: The rates of coronary death in both men and women increased with the triglyceride level. In Cox proportional-hazards models adjusted for age, in which the natural log of the triglyceride levels was used to give a normal distribution, the relative risk per natural-log unit of triglyceride (e.g., a triglyceride level of 150 mg per deciliter vs. a level of 55 mg per deciliter) was 1.54 (95 percent confidence interval, 1.19 to 1.98; P < 0.001) in men and 1.88 (95 percent confidence interval, 1.19 to 2.98; P < 0.007) in women. After an adjustment for potential covariates, however, these relative risks were not statistically significant. Analyses based on lipoprotein cholesterol levels revealed a positive association between the triglyceride level and coronary mortality in the lower stratum of both HDL and LDL cholesterol, but not in the higher stratum. Conversely, the HDL cholesterol level was unrelated to coronary mortality in the lower stratum of LDL cholesterol, but was strongly inversely associated with coronary death in the higher stratum of LDL cholesterol. The relative risk of coronary death associated with triglyceride level was higher at younger ages. The associations between the triglyceride level and coronary mortality in the lower HDL cholesterol, LDL cholesterol, and age strata were small and were further reduced by an adjustment for the fasting plasma glucose level.
Conclusions: Overall, the plasma triglyceride level showed no independent association with coronary mortality. However, in subgroups of subjects with lower HDL and LDL cholesterol levels and in younger subjects, defined a priori, an association between the triglyceride level and coronary mortality was observed, although this association was small and was not statistically significant after an adjustment for the plasma glucose level.