Clinical Question

Is nut consumption associated with lower mortality?

Bottom Line

In a study of health professionals, nut consumption was associated with reduced all-cause mortality. This was primarily because of a reduction in heart disease associated with any nut consumption and a reduction in cancer associated with tree nut consumption. (Level of Evidence = 2a)

Synopsis

Nuts are a component of the Mediterranean diet, which has been shown in observational studies and clinical trials to be associated with reduced cardiovascular mortality. The authors of the current study used data from two large prospective cohort studies: the Health Professionals Follow-Up Study (male health professionals) and the Nurses' Health Study (female nurses). Anyone with cancer or cardiovascular disease at baseline was excluded, as was anyone who did not provide data on physical activity, nut consumption, or height and weight. That left 76,464 women and 42,498 men for the study. Nut consumption was quantified using a food questionnaire administered every two to four years; in later years, the questionnaire distinguished peanuts from other nuts. A serving was defined as 28 g. Only dietary surveys given before a diagnosis of cancer or cardiovascular disease were included, because these diagnoses may have prompted a change in diet. All-cause mortality was determined using the National Death Index, and the cause of death was determined by a physician masked to nut consumption who reviewed death certificates and medical records.

Persons who ate a lot of nuts tended to be healthier in many ways than those who did not: they were thinner, more physically active, and were much less likely to smoke; they were also more likely to take a multi-vitamin, eat fruits and vegetables, and drink alcohol. A multivariate analysis was done to attempt to adjust for known confounders (the above, plus hypertension, hyperlipidemia, or diabetes mellitus; aspirin use and red meat intake; and family history of cardiovascular disease or cancer), but the possibility of residual confounding by other variables remains. Another limitation is that the health professionals may have overestimated their own nut consumption, leading to response bias. The adjusted analysis showed a dose-response effect of nut consumption, with the multivariate adjusted hazard ratio for all-cause mortality decreasing from 1.0 for those who never ate nuts to 0.89 for those who ate them once per week, 0.85 for five or six times per week, and 0.80 for seven or more times per week. These reductions in mortality were statistically significant, as was the trend, and the authors argue that the study was large enough that there would have to be a lot of unknown confounding variables to cancel out this effect.

Study design: Cohort (prospective)

Funding source: Industry plus government

Setting: Population-based

Reference: BaoYHanJHuFBet alAssociation of nut consumption with total and cause-specific mortality. N Engl J Med.2013; 369( 21): 2001– 2011.