Healthful and Unhealthful Plant-Based Diets and Risk of Breast Cancer in U.S. Women: Results from the Nurses' Health Studies

Abstract

Background: Plant-based diets have been associated with lower risk of various diseases, including type 2 diabetes, cardiovascular disease, and other cardiometabolic risk factors. However, the association between plant-based diet quality and breast cancer remains unclear.

Methods: We prospectively followed 76,690 women from the Nurses' Health Study (NHS, 1984-2016) and 93,295 women from the NHSII (1991-2017). Adherence to an overall plant-based diet index (PDI), a healthful PDI (hPDI), and an unhealthful PDI (uPDI) was assessed using previously developed indices. Cox proportional hazards models were used to estimate HR and 95% confidence intervals (CI) for incident invasive breast cancer.

Results: Over 4,841,083 person-years of follow-up, we documented 12,482 incident invasive breast cancer cases. Women with greater adherence to PDI and hPDI were at modestly lower risk of breast cancer [(HR_{Q5 vs. Q1}, 0.89; 95% CI, 0.84-0.95); (HR_{Q5 vs. Q1}, 0.89; 95% CI, 0.83-0.94)]. We observed significant heterogeneity by estrogen receptor (ER) status, with the strongest inverse association between hPDI and breast cancer observed with ER-negative tumors [HR_{Q5 vs. Q1}, 0.77; 95% CI, 0.65-0.90; P _trend < 0.01]. We also found an inverse association between extreme quintiles of healthy plant foods and ER-negative breast cancer [HR, 0.74; 95% CI, 0.61-0.88; P _trend < 0.01].

Conclusions: This study provides evidence that adherence to a healthful plant-based diet may reduce the risk of breast cancer, especially those that are more likely to be aggressive tumors.

Impact: This is the first prospective study investigating the relation between healthful and unhealthful plant-based dietary indices and risk of total and subtype-specific breast cancer.

Figure 1.

Multivariable spline analysis of the association between adherence to a healthful and unhealthful plant-based dietary indexes and risk of incident breast cancer in the NHS (1984-2016) and NHSII (1991-2017). ER= estrogen receptor; Model stratified by cohort, age in months and 2-year period at risk, adjusted for race (Non-Hispanic Caucasian, African, Asian, Hispanic Caucasian), cumulatively updated total caloric intake (kcal/day, quintiles), age at menarche (<12, 12, 13, 14, >14 years), age at menopause (premenopausal, <45, 45-49, 50-52, 53+), postmenopausal hormone use (never user, past user, current user– estrogen only for <5 years, current user – estrogen only for ≥5 years, current estrogen + progestin user for < 5 years, current estrogen + progestin user for ≥5 years, current user of other types), parity and age at first birth (nulliparous, 1 child before age 25, 1 child at ≥25 years of age, 2+ children before age 25, 2+children ≥25 years of age), breastfeeding history (never, breastfed for ≤ 6 months, breastfed for > 6 months), family history of breast cancer (yes or no), history of benign breast disease (yes or no), body mass index ate age 18 years (<20.0, 20.0-21.9, 22.0-23.9, 24.0-26.9, ≥27.0), oral contraceptives use (never/ever), height (<1.60, 1.60-1.64, 1.65-1.69, 1.70-1.74, 1.75+ m), cumulatively updated alcohol intake (0, <5, 5-9, 10-14, 15+ g/day), physical activity (linear METs-h/week) and neighborhood-based socioeconomic status indicator (continuous). For ER negative breast cancer cases, the p values for test of curvature for healthful plant-based diet index (hPDI)= 0.95 and for unhealthful plant-based diet index (uPDI)=0.29. For ER positive cases, the p values for test of curvature for hPDI=0.35 and uPDI=0.15.

Figure 2.

Pooled hazard ratios of estrogen receptor negative breast cancer per 10-units increment in the three dietary indices (PDI, hPDI and uPDI) across subgroups (physical activity, current BMI and menopausal status). The hazard ratios (HRs) and p values for women were obtained after combining all 2 cohorts (NHS; NHSII). Stratified by age in months, cohort and calendar year, adjusted for race (Non-Hispanic Caucasian, African, Asian, Hispanic Caucasian), age at menarche (<12, 12, 13, 14, >14 years), age at menopause (premenopausal, <45, 45-49, 50-52, 53+), postmenopausal hormone use (never user, past user, current user– estrogen only for <5 years, current user – estrogen only for ≥5 years, current estrogen + progestin user for < 5 years, current estrogen + progestin user for ≥5 years, current user of other types), oral contraceptive use history (never, ever), parity and age at first birth (nulliparous, 1 child before age 25, 1 child at ≥25 years of age, 2+ children before age 25, 2+ children ≥25 years of age), breastfeeding history (never, breastfed for ≤ 6 months, breastfed for > 6 months), family history of breast cancer (yes or no), history of benign breast disease (yes or no), height (<1.60, 1.60-1.64, 1.65-1.69, 1.70-1.74, 1.75 + m), cumulatively updated alcohol intake (0, <5, 5-9, 10-14, 15+ g/day), cumulatively updated total caloric intake (kcal/day, quintiles), physical activity (linear MET-hours/week), body mass index at age 18 years (<20.0, 20.0-21.9, 22.0-23.9, 24.0-26.9, ≥27.0) and socioeconomic status (continuous).For the analysis of BMI, we further adjusted for current body mass index (linear, kg/m2).To test whether the PDI, hPDI, uPDI and breast cancer association differed by current BMI, physical activity, or menopausal status we added interaction terms and used the Wald test. BMI= body mass index; CI= confidence interval; MET= metabolic equivalent task.