Ridker · The New England journal of medicine 1997 · nested case-control study within a randomized controlled trial · n=1086

Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men.

Cited 5540 times in the scientific literature.

Level 2 - randomized trial

Nested case-control study analyzing biomarker interaction within a randomized controlled trial.

PubMed 9077376 · doi:10.1056/NEJM199704033361401 · record verified 2026-08-28

What was done

A nested case-control study was conducted within the Physicians' Health Study randomized controlled trial. Baseline plasma C-reactive protein (CRP) was measured in 543 apparently healthy men who subsequently developed myocardial infarction, stroke, or venous thrombosis over >8 years of follow-up, and compared with 543 matched men who remained free of vascular disease. Participants had been randomly assigned to aspirin or placebo at baseline.

What was found

Baseline CRP was higher among men who developed myocardial infarction (1.51 vs. 1.13 mg/L, P < 0.001) or ischemic stroke (1.38 vs. 1.13 mg/L, P = 0.02) compared to controls, but not venous thrombosis (1.26 vs. 1.13 mg/L, P = 0.34). Men in the highest CRP quartile had higher risk of myocardial infarction (RR 2.9, P < 0.001) and ischemic stroke (RR 1.9, P = 0.02) versus the lowest quartile. Aspirin use significantly reduced myocardial infarction risk among men in the highest CRP quartile (55.7% reduction, P = 0.02), with only a small, non-significant reduction in the lowest quartile (13.9%, P = 0.77).

Why it matters

This study showed that baseline systemic inflammation marked by CRP predicts future arterial cardiovascular events and suggests that the preventive benefit of aspirin against myocardial infarction is greatest in individuals with high inflammation.

Limits

The study was conducted exclusively in male physicians, limiting generalizability to women and other populations. Biomarker levels were measured from baseline plasma samples, and subgroup analysis by CRP quartile reduces statistical power for secondary outcomes.

Cited by