Krauss · The American journal of cardiology 1995 · narrative review · n=?

Dense low density lipoproteins and coronary artery disease.

Cited 160 times in the scientific literature.

Level 5 - mechanism / opinion, no new human data

Narrative review synthesizing observational case-control studies, trial data, and mechanistic research.

PubMed 7863975 · doi:10.1016/0002-9149(95)80012-h · record verified 2026-08-30

What was done

This narrative review synthesizes evidence from case-control studies, angiographic intervention trials, and in vitro mechanistic studies evaluating the relationship between small, dense low-density lipoprotein (LDL) particles and coronary artery disease (CAD).

What was found

Case-control studies show that 40% to 50% of patients with myocardial infarction and angiographically confirmed CAD possess the small, dense LDL phenotype, conferring a 2- to 3-fold increase in disease risk. This trait correlates with elevated plasma triglycerides and apolipoprotein B-100, reduced high-density lipoprotein cholesterol and apoA-I, and associations with insulin resistance and familial combined hyperlipidemia. One cited trial reported that equivalent LDL cholesterol lowering reduced coronary angiographic progression in patients with predominantly dense LDL, but showed no angiographic benefit in those with buoyant LDL. In vitro findings demonstrated that small, dense LDL has increased susceptibility to oxidation and greater binding affinity for arterial wall proteoglycans.

Why it matters

The small, dense LDL phenotype helps explain familial CAD risk in a large proportion of patients and may identify individuals who derive disproportionate benefit from LDL-lowering therapies.

Limits

As a narrative review, it lacks systematic search criteria, quality appraisal, and meta-analytic pooling. Due to strong collinearity with high triglycerides, low HDL, and insulin resistance, isolating the independent causal contribution of small, dense LDL from other components of the metabolic phenotype remains challenging. Numerical sample sizes and confidence intervals are not provided in the abstract.

Cited by