Coronary heart disease (CHD) is the single leading cause of death in the industrialized world. In 2005, CHD claimed 450,000 lives in the U.S., accounting for 18% of all reported deaths.1 Evidence strongly suggests a direct relationship between high levels of low density lipoprotein (LDL) and increased risk of CHD. More importantly, therapies aimed at lowering LDL cholesterol, particularly statins, have demonstrated to reduce CHD morbidity and mortality by as much as 35%.2
In an effort to further reduce the deaths attributed to CHD, researchers and clinicians have turned to other components of the lipid profile. Studies such as the Multiple Risk Factor Intervention Trial, Framingham Heart Study and Coronary Primary Prevention Trial identified high density lipoprotein (HDL) as being inversely and independently associated with the risk of developing CHD. These studies pointed out that for each 1 mg/dL increment in HDL, there was an associated decrease in CHD risk of 2% in men and 3% in women.3 Therefore, therapies aimed at increasing HDL cholesterol have been recommended by both experts and guidelines. Unfortunately, clinical trial data have failed to show that the increases in HDL, rather than the reductions in LDL, are responsible for the decreases in CHD morbidity and mortality.
Briel and colleagues4 recently published the results of a systematic review and meta-regression analysis that evaluated the effect of lipid modifying interventions on CHD risk. To be included in the analysis, trials had to follow patients for more than 6 months, have a randomized control design, and report HDL cholesterol and mortality or myocardial infarctions (MI) separately for treatment groups. Over 10,000 articles were identified through an extensive Medline, Embase, Central, CINAHL and AMED search. However, only 108 met the inclusion criteria and were included in the analysis.
The clinical endpoints were total death, CHD death, and CHD events. Treatment effects across studies were pooled together. Inverse variance weighted meta-regression analysis was used to determine if there was an association between changes in HDL and LDL levels among the intervention and control groups and clinical endpoints. To take into account the non-lipid effects of certain agents, a multivariable meta-regression analysis that included LDL, HDL, and drug class was performed. In order to focus on a more homogeneous sample of trials, a pre-specified analysis of trials that used interventions known to raise HDL (rather than reduce TG) was performed. Therefore, trials that used n-3 fatty acids, low fat diets, or probucol were excluded. Trials with agents associated with harmful effects such as torcetrapib and hormones were also excluded.
Researchers found that, for all meta-regression models, a reduction in LDL cholesterol was associated with a reduction in the risk for CHD events, CHD death and total death. Although the univariable models for HDL pointed to a potential benefit in terms of reducing CHD events and CHD death, they were not statistically significant. Also, after adjusting for changes in LDL cholesterol (bivariable and multivariable models), increasing HDL ...