· JAMA 2009 · individual participant data meta-analysis of prospective cohort studies · n=302430 participants across 68 studies

Major lipids, apolipoproteins, and risk of vascular disease.

Cited 2789 times in the scientific literature.

Level 1 - systematic review of randomized trials

Individual participant data meta-analysis of prospective cohort studies

PubMed 19903920 · doi:10.1001/jama.2009.1619 · record verified 2026-08-27

What was done

Individual-participant data meta-analysis from 68 long-term prospective cohort studies (mostly in Europe and North America) including 302,430 individuals without baseline vascular disease followed for 2.79 million person-years. Investigators assessed associations between baseline lipid and apolipoprotein levels and vascular outcomes (8,857 nonfatal myocardial infarctions, 3,928 coronary heart disease [CHD] deaths, 2,534 ischemic strokes, 513 hemorrhagic strokes, and 2,536 unclassified strokes), adjusting for conventional cardiovascular risk factors and within-person variation.

What was found

Adjusted hazard ratios (HRs) per 1-standard deviation higher baseline level for CHD were: - Triglyceride (0.52 log(e)): 0.99 (95% CI, 0.94-1.05) - High-density lipoprotein cholesterol (HDL-C, 15 mg/dL): 0.78 (95% CI, 0.74-0.82) - Non-HDL-C (43 mg/dL): 1.50 (95% CI, 1.39-1.61) A combination of 80 mg/dL lower non-HDL-C and 15 mg/dL higher HDL-C was associated with an HR of 0.35 (95% CI, 0.30-0.42) for CHD. For ischemic stroke, adjusted HRs were 1.02 (95% CI, 0.94-1.11) with triglyceride, 0.93 (95% CI, 0.84-1.02) with HDL-C, and 1.12 (95% CI, 1.04-1.20) with non-HDL-C. In subsets with available measurements, the non-HDL-C/HDL-C ratio (HR 1.50, 95% CI 1.38-1.62) performed similarly to the apo B/apo AI ratio (HR 1.49, 95% CI 1.39-1.60), and non-HDL-C (HR 1.42, 95% CI 1.06-1.91) performed similarly to directly measured LDL-C (HR 1.38, 95% CI 1.09-1.73). Associations were at least as strong in non-fasting participants as in fasting participants.

Why it matters

Vascular disease risk assessment can be simplified by measuring non-HDL-C and HDL-C without requiring fasting or measuring apolipoproteins.

Limits

Participants were drawn mostly from Europe and North America, potentially limiting generalizability to other populations. Observational design cannot entirely rule out residual confounding or determine causality on its own.

Cited by