INDEPENDENT THINKING. WHOLE-FOOD LIVING.NO ANIMALS / NO OIL
THENANODIET
← Cardiology / Health Desk

Nuts and cardiovascular outcomes: coronary disease is not stroke

A prospective-study meta-analysis found different levels of evidence for coronary disease and stroke.

Whole-food ingredients prepared without added oil

Illustrative food image, not a photograph of the study intervention.

Beyond a cholesterol trial

Aune and colleagues brought together 20 prospective studies reported in 29 publications, searching evidence published through July 19, 2016. Their outcomes included coronary heart disease, stroke, cardiovascular disease and mortality. This differs from a controlled feeding trial measuring LDL after a few weeks: participants were observed rather than assigned to nut intake. The analysis used random-effects models to combine relative risks and confidence intervals. It therefore asks whether higher habitual nut consumption accompanies different subsequent outcomes, not whether randomly prescribing nuts produces those outcomes. Keeping that distinction visible is essential when translating its results into a heart-health headline.

Aune D, Keum N, Giovannucci E, Fadnes LT et al.. BMC medicine, 2016. PMID 27916000 · DOI

Coronary disease had the clearer association

For each 28 g per day increase in nut intake, the pooled relative risk for coronary heart disease was 0.71, with a 95% confidence interval of 0.63 to 0.80, based on 11 studies. The corresponding estimate for cardiovascular disease was 0.79, with an interval of 0.70 to 0.88, from 12 studies. These estimates describe lower relative risks associated with higher intake. They do not give an individual's absolute chance of disease, and the 28 g increment is an analytical comparison, not a dose whose effect was established by random assignment in these studies.

Aune D, Keum N, Giovannucci E, Fadnes LT et al.. BMC medicine, 2016. PMID 27916000 · DOI

Stroke requires a different sentence

The stroke estimate was 0.93 per 28 g daily increase, with a 95% confidence interval of 0.83 to 1.05, based on 11 studies. Because that interval includes 1.0, the pooled result did not establish a statistically clear inverse association for stroke. It would be misleading to attach the coronary heart disease estimate to stroke or to describe all cardiovascular endpoints as equally convincing. The contrast is useful precisely because both outcomes came from the same review. A favorable overall cardiovascular association can coexist with an uncertain result for a specific component outcome in prospective evidence.

Aune D, Keum N, Giovannucci E, Fadnes LT et al.. BMC medicine, 2016. PMID 27916000 · DOI

Mortality and between-study variation

All-cause mortality had a pooled relative risk of 0.78 per 28 g daily increase in nuts, with a 95% confidence interval of 0.72 to 0.84 across 15 studies. Between-study heterogeneity was substantial enough to note: I-squared was 66% for all-cause mortality and 60% for cardiovascular disease. The authors reported similar results for tree nuts and peanuts. These are still observational associations, not proof that nuts themselves caused the differences. Indeed, the abstract explicitly made its estimate of deaths attributable to low intake conditional on the associations being causal. That condition should not disappear in a practical summary.

Aune D, Keum N, Giovannucci E, Fadnes LT et al.. BMC medicine, 2016. PMID 27916000 · DOI