Nuts and health: what large cohorts and PREDIMED actually show

People who eat nuts regularly have modestly lower rates of heart disease and early death than people who avoid them, a pattern that holds across two of the largest diet cohorts ever run [s1]. The strongest single trial to test nuts directly found that a Mediterranean diet supplemented with mixed nuts lowered major cardiovascular events, but almost everything else on nuts is observational — good enough to say the association is real and consistent, not good enough to prove nuts themselves are the cause [s1][s3].

What the cohorts show

The clearest read comes from the Nurses' Health Study and the Health Professionals Follow-up Study, which tracked 76,464 women from 1980 to 2010 and 42,498 men from 1986 to 2010, with nut intake reassessed every two to four years [s1]. Over 3,038,853 person-years, 16,200 women and 11,229 men died [s1]. Compared with people who never ate nuts, the risk of death fell step by step as intake rose: a hazard ratio of 0.93 for eating nuts less than once a week, 0.89 for once a week, 0.87 for two to four times a week, 0.85 for five or six times, and 0.80 for seven or more times a week (P<0.001 for trend) [s1]. Inverse associations also showed up for deaths from cancer, heart disease and respiratory disease [s1].

A 2016 dose-response meta-analysis pooling 20 studies put a number on the size of the effect [s2]. Per 28 grams a day — roughly a handful — nut intake was associated with a relative risk of 0.71 for coronary heart disease, 0.79 for cardiovascular disease overall, 0.85 for total cancer and 0.78 for all-cause mortality; the association with stroke, 0.93, was not statistically significant [s2]. Mortality from diabetes came in at 0.61 [s2]. The results were similar for tree nuts and peanuts, which matters because peanuts are cheaper and botanically a legume, not a tree nut [s2].

The same analysis reported still larger associations for less common causes of death — 0.48 for respiratory-disease mortality and 0.25 for deaths from infectious disease [s2]. Those numbers look dramatic, but they rest on only two or three studies each, so the confidence intervals are wide and the estimates fragile [s2]. The pattern to take from the meta-analysis is not any single cause-specific figure but the breadth: nut intake tracks with lower death rates across a range of unrelated conditions, which is as easily explained by the kind of people who eat nuts as by the nuts themselves.

The one trial that tested it

Observational findings like these are vulnerable to a familiar problem: people who eat nuts also tend to smoke less, exercise more and eat better overall, and no statistical adjustment fully removes that. The nearest thing to a direct test is the Spanish PREDIMED trial, which randomly assigned 7,447 people at high cardiovascular risk to a Mediterranean diet with extra-virgin olive oil, the same diet with mixed nuts, or a control diet advised to cut fat [s3].

That trial has a complicated history — it was withdrawn and re-analysed after protocol irregularities, which we cover in the PREDIMED retraction and reanalysis — but the revised results still favoured the food. Over a median 4.8 years, a first major cardiovascular event occurred in 3.4 percent of the nut group and 4.4 percent of the control group, a hazard ratio of 0.72 [s3]. The olive-oil group fared similarly, at 3.8 percent and a hazard ratio of 0.69 [s3]. The trial tested nuts inside a whole dietary pattern, so it credits the pattern rather than nuts in isolation.

What to make of it

Two cautions belong in any honest summary. First, the cohort mortality study was funded in part by the International Tree Nut Council Nutrition Research and Education Foundation alongside the National Institutes of Health — industry money that does not by itself invalidate the result but belongs on the record [s1]. The 2016 meta-analysis went further and estimated that, if the associations are causal, 4.4 million premature deaths a year across four world regions would be attributable to eating fewer than 20 grams of nuts a day — a projection that assumes causation the data cannot establish [s2].

Second, the direction and consistency are what carry weight here, not any single number. Across a randomised trial and two very large cohorts the signal points the same way, and the effect sizes are moderate rather than dramatic. For readers interested in where nuts sit within a broader eating pattern, they are a defining component of the Mediterranean diet and feature in the PREDIMED-Plus diabetes-prevention work. This is informational, not dietary advice.

Sources

• Association of Nut Consumption with Total and Cause-Specific Mortality — New England Journal of Medicine, 2013
• Nut consumption and risk of cardiovascular disease, total cancer, all-cause and cause-specific mortality — BMC Medicine, 2016
• Primary Prevention of Cardiovascular Disease with a Mediterranean Diet Supplemented with Extra-Virgin Olive Oil or Nuts — New England Journal of Medicine, 2018