Nut and Peanut Butter Consumption and the Risk of Total Cancer: A Prospective Cohort Study

Growing scientific evidence has indicated that nut intake might have beneficial effects on cancer-related mortality and the risk of several cancer (sub)types, which might be related to their antioxidative and anti-inflammatory components (1). From a public health perspective, it is also relevant to know whether nut and peanut butter intake might contribute to a reduction of the total cancer burden. To our knowledge, this relation has only been investigated in one prospective cohort study, in which no associations were observed between nut intake and total cancer risk (2).

In this study, we investigated the associations of nut and peanut butter intake with the risk of total cancer in the prospective Netherlands Cohort Study (NLCS). These associations were also studied for smoking- and alcohol-related cancers, because the potential cancer preventive effects of nuts might be stronger for these carcinogen-related cancers.

At baseline, in September 1986, 58,279 men and 62,573 women aged 55 to 69 years were included in the NLCS (3). All participants filled in a self-administered baseline questionnaire on potential cancer risk factors. A case–cohort approach was applied, in which cases were derived from the entire cohort and person-time at risk from a subcohort consisting of 5,000 randomly sampled participants. Follow-up of cancer incidence was performed through annual record linkage with the Netherlands Cancer Registry and the Netherlands Pathology Registry (PALGA).

After 20.3 years of follow-up, we included 19,255 cancer cases (of which 765 in the subcohort) and 3,499 subcohort members after excluding participant with prevalent cancer (excluding skin cancer), incomplete or inconsistent dietary data, missing data on exposures and confounders, and cases with noninvasive tumors or without microscopic confirmation. Smoking-related cancers comprised cancers of the oral cavity (including lip), pharynx, esophagus, stomach, colorectum, liver, pancreas, nasal cavity, paranasal sinuses, larynx, trachea, lung, uterine cervix, ovary, kidney, ureter, urinary bladder, and myeloid leukemia (4). Alcohol-related cancers included cancers of the oral cavity (including lip), pharynx, esophageal squamous cell carcinoma, colorectum, liver, larynx, and breast (4). The smoking- and alcohol-unrelated cancer subgroups included all cancers not described above.

Habitual diet was assessed using a validated 150-item self-administered semiquantitative food frequency questionnaire (FFQ). Nut and peanut butter intake was measured as described previously (5).

Cox proportional hazards models were used to estimate multivariable-adjusted HRs and 95% confidence intervals (CI) for the associations between total nut intake and total cancer risk. Associations were also examined for tree nut, peanut, and peanut butter intake, and for smoking- and alcohol-(un)related cancers. SEs were calculated using a robust Huber–White sandwich estimator (6). Scaled Schoenfeld residuals, −ln(−ln) survival plots, and time-varying covariates confirmed that assuming proportional hazards was appropriate. Heterogeneity between the smoking- and alcohol-(un)related subgroups was tested using a competing risk procedure that applies a bootstrapping method to estimate SEs in case–cohort designs.

Mean total nut intake in the subcohort was 7.9 g/day in men and 4.5 g/day in women. Baseline characteristics are presented in Table 1. We observed no associations between total nut, tree nut, peanut, and peanut butter intake and total cancer risk in men or women (Table 2). Also no associations were found for smoking- and alcohol-(un)related cancers, except for a borderline significant positive association between 0.1–<5 g peanut butter intake/day and smoking-unrelated cancer in men [HR (95% CI), 1.17 (1.00–1.37)].

Heterogeneity tests between smoking-related and smoking-unrelated cancers were significant for all nut variables in men (P_{heterogeneity} < 0.001). This is probably the result of the large number of cases and the opposite directions of the nonsignificant associations in the smoking-related and smoking-unrelated subgroups in men.

Our results are in line with the only other cohort study that investigated the relation between nut intake and total cancer risk, in which also no significant associations were observed (2). Prospective evidence on total cancer is thus very limited. In a recent meta-analysis, a significantly reduced “overall” cancer risk was observed after pooling 33 prospective cohort studies that investigated the relations between nut intake and several cancer (sub)types (7). However, this pooled estimate is not informative, because it does not include all cancer types and might be affected by publication bias.

Here, we presented results for total cancer, but associations for several cancer (sub)types have been investigated previously in the NLCS, for example, ref. 5. Strengths of the NLCS include the prospective design, the inclusion of all cancer types (except skin cancer), the large number of cases, and the long and complete follow-up of 20.3 years. Unfortunately, we only have baseline measurements. Nevertheless, the FFQ showed adequate reproducibility (8) and the NLCS consists of an older population with relatively stable dietary habits. Furthermore, we cannot completely exclude residual confounding by (un)measured factors.

In conclusion, the results of this prospective cohort study do no support the hypothesis that nut intake is associated with a reduced total cancer risk in men or women.

No potential conflicts of interest were disclosed.

L. Nieuwenhuis: Formal analysis, investigation, writing–original draft, writing–review and editing. P.A. van den Brandt: Conceptualization, resources, supervision, funding acquisition, investigation, methodology, project administration, writing–review and editing.

This study was funded by the Dutch Cancer Society (grant no. um 2015-7860 to P.A. van den Brandt). The authors would like to thank the participants of the Netherlands Cohort Study (NLCS), the Netherlands Cancer Registry, and the Dutch Pathology Registry. Furthermore, NLCS staff members are acknowledged for their valuable assistance and advice.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.