Bonora · Stroke 1997 · Prospective population-based cohort study · n=888

Relationship between insulin and carotid atherosclerosis in the general population. The Bruneck Study.

Cited 44 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective population-based cohort study (mislabeled as a clinical trial in indexing metadata)

PubMed 9183341 · doi:10.1161/01.str.28.6.1147 · record verified 2026-08-27

What was done

In an age- and sex-stratified random population sample of adults aged 40 to 79 years (the Bruneck Study), baseline serum insulin was measured at fasting (n = 888) and 2 hours after an oral glucose load (n = 811; individuals with known diabetes were excluded). Atherosclerosis in the common and internal carotid arteries was assessed by duplex sonography at baseline (1990) and at 5-year follow-up (1995). Progression was defined as an atherosclerosis score increase of >27% or the appearance of new plaques. Participants were stratified into quintiles of baseline insulin, and multivariable logistic regression evaluated associations adjusting for age, sex, BMI, glucose tolerance, triglycerides, apolipoproteins A1 and B, fibrinogen, blood pressure, behavioral factors, and socioeconomic status.

What was found

Logistic regression revealed a significant U-shaped relationship where both low and high insulin levels were associated with carotid atherosclerosis. This U-shaped association persisted after multivariable adjustment, applied to both fasting and 2-hour postload insulin, and was stronger in the 5-year prospective analysis than in the cross-sectional analysis. The abstract reports no exact odds ratios, confidence intervals, or p-values.

Why it matters

These findings suggest that both hypoinsulinemia and hyperinsulinemia are independent risk markers for carotid atherosclerosis progression in the general population, showing a non-linear relationship with vascular risk.

Limits

Individuals with known diabetes were excluded, limiting generalizability to that group. The abstract omits exact point estimates, confidence intervals, and event rates. As an observational cohort study, residual confounding cannot be ruled out.

Cited by