Are potatoes bad for health? Review of 'Potato intake and incidence of hypertension: results from three prospective US cohort studies'
In their recent report, Borgi and colleagues concluded that; ‘Higher intakes of boiled, baked, or mashed potatoes, and French fries were associated with an increased risk of hypertension in adult women and men.’ This conclusion was based on analyses of three cohort samples; two groups of women; the Nurses’ Health Study (NHS), the Nurses’ Health Study II (NHS II) and men in the Health Professionals Follow-up study (HPFS) from 1986 to 2010.
Cohort research designs can be effective for investigating the long-term health effects of nutritional patterns. One limitation of cohort studies is the risk of confounding. To reduce this possibility statistical analyses typically adjust for potential confounders.
The reported hazard ratios were adjusted for multiple covariates including; age, race/ethnicity, BMI, smoking status, physical activity, weight change, frequency questionnaire cycle, alcohol intake, family history of hypertension, total energy intake, animal flesh intake (combination of processed and unprocessed red meat, poultry, and seafood), whole grains, sugar sweetened drink intake, artificially sweetened diet drink intake, total fruit, total vegetables (see complete list in original paper). The multivariate adjusted hazard ratios for hypertension are shown in Table 1. Although the increased risk of hypertension associated with consumption of French Fries was consistent across all three samples, the consumption of baked boiled or mashed potatoes was only associated with hypertension in the two NHS samples of women. The association of potato chips with hypertension showed small and inconsistent trends across the three samples.
The inconsistent findings between the samples of men and women, indicates potential problems with the conclusions. A major limitation of the analyses was the lack of adjustment for known dietary patterns. Previous research using factor analysis with two of the three samples reported two general dietary patterns; a ‘prudent’ pattern based on high intakes of vegetables, fruit, whole grains fish and poultry; and a ‘western’ pattern based on refined grains, red and processed meat, French fries and sweets and desserts.[2 ,3]. The factors loadings for the ‘Western’ pattern are shown in Table 2. In both the women’s and men’s samples the ‘western’ pattern was associated with a significantly increased risk of coronary heart disease, for which hypertension is a contributing risk factor.
The data in Table 2 indicate that potatoes and French fries are a central component (loadings ≥0.40) of the western diet for women (NHS samples), but only French Fries for men (HPFS sample). Two known risk factors for hypertension (desserts and sweets, and processed meats) were not adjusted in the reported hazard ratios. Two covariates included in the adjustment were ‘sugar sweetened drinks’ and ‘animal flesh.’ These covariates would not remove confounding from the central components of the western dietary pattern shown in Table 2. A more plausible explanation for the hazard ratios reported is that the western dietary pattern is a risk factor for hypertension, not necessarily potatoes per se. The lack of relationship reported between mashed, boiled or baked potatoes and hypertension in the HPFS sample is consistent with the less central contribution of potatoes to the western diet in the male sample. In contrast, French fries were a core component of the western diet in both the male and female samples, consistent with the Hazard ratios shown in Table 1.
In summary, the hazard ratios reported in Table 1 did not adjust for the confounding of potato intake with core components of the western dietary pattern which are known high risk foods for hypertension. An alternative conclusion is consumption of up to one serving of baked boiled or mashed potatoes per day is unlikely to be a risk factor for hypertension. In contrast to the authors’ view, because of ‘residual confounding’ the study did not provide sufficient evidence to remove potatoes from food assistance programmes.
As has been noted previously, analyses focusing on individual foods in cohort research should take into account the association of those foods with dietary patterns to establish whether the effect of a specific food is independent from its association with dietary patterns. The use of factor analysis is an effective method for establishing patterns among food measures.
1. Borgi L, Rimm EB, Willett WC, et al. Potato intake and incidence of hypertension: results from three prospective US cohort studies. BMJ 2016;353:i2351 doi:10.1136/bmj.i2351
2. Fung TT, Willett WC, Stampfer MJ, et al. Dietary patterns and the risk of coronary heart disease in women. Arch Intern Med 2001;161(15):1857-62 doi:10.1001/archinte.161.15.1857
3. Hu FB, Rimm EB, Stampfer MJ, et al. Prospective study of major dietary patterns and risk of coronary heart disease in men. Am J Clin Nutr 2000;72(4):912-21
4. Jalal DI, Smits G, Johnson RJ, et al. Increased Fructose Associates with Elevated Blood Pressure. J Am Soc Nephrol 2010;21(9):1543-49 doi:10.1681/asn.2009111111
5. Micha R, Wallace SK, Mozaffarian D. Red and Processed Meat Consumption and Risk of Incident Coronary Heart Disease, Stroke, and Diabetes Mellitus: A Systematic Review and Meta-Analysis. Circulation 2010;121(21):2271-83 doi:10.1161/circulationaha.109.924977
6. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol 2002;13(1):3-9
Competing interests: No competing interests