Association of Healthful Plant-based Diet Adherence With Risk of Mortality and Major Chronic Diseases Among Adults in the UK

Abstract

Importance: Plant-based diets have gained popularity for both environmental and health reasons, but a comprehensive assessment of their quality in relation to risk of mortality and major chronic diseases is lacking.

Objective: To examine whether healthful vs unhealthful plant-based dietary patterns are associated with mortality and major chronic diseases among UK adults.

Design, setting, and participants: This prospective cohort study used data from adults in the UK Biobank, a large-scale population-based study. Participants were recruited between 2006 and 2010 and followed up using record linkage data until 2021; follow-up for different outcomes ranged between 10.6 and 12.2 years. Data analysis was conducted from November 2021 to October 2022.

Exposures: Adherence to a healthful vs unhealthful plant-based diet index (hPDI vs uPDI) derived from 24-hour dietary assessments.

Main outcomes and measures: The main outcomes were hazard ratios (HRs) and 95% CIs of mortality (overall and cause specific), cardiovascular disease (CVD [total, myocardial infarction, ischemic stroke, and hemorrhagic stroke]), cancer (total, breast, prostate, and colorectal), and fracture (total, vertebrae, and hip) across quartiles of hPDI and uPDI adherence.

Results: This study included 126 394 UK Biobank participants. They had a mean (SD) age of 56.1 (7.8) years; 70 618 (55.9%) were women. The majority of participants (115 371 [91.3%]) were White. Greater adherence to the hPDI was associated with lower risks of total mortality, cancer, and CVD, with HRs (95% CIs) of 0.84 (0.78-0.91), 0.93 (0.88-0.99), and 0.92 (0.86-0.99), respectively, for participants in the highest hPDI quartile compared with the lowest. The hPDI was also associated with lower risks of myocardial infarction and ischemic stroke, with HRs (95% CIs) of 0.86 (0.78-0.95) and 0.84 (0.71-0.99), respectively. By contrast, higher uPDI scores were associated with higher risks of mortality, CVD, and cancer. The associations observed did not show heterogeneity across strata of sex, smoking status, body mass index, or socioeconomic status or with polygenic risk scores (specifically with regard to CVD end points).

Conclusions and relevance: The findings of this cohort study of middle-aged UK adults suggest that a diet characterized by high-quality plant-based foods and lower intakes of animal products may be beneficial for health, irrespective of established chronic disease risk factors and genetic predisposition.

Figure 1.. Multivariable-Adjusted Hazard Ratios and 95% CIs for All-Cause Mortality (n = 126 217), Cancer (n = 117 569), Cardiovascular Disease (CVD; n = 123 134), and Fracture (n = 112 208) Across Sex-Specific Healthful vs Unhealthful Plant-based Diet Index Quartiles

All models used age as the underlying time variable and were adjusted for sex, body mass index, race and ethnicity, physical activity level, smoking status, alcohol intake, education level, energy intake, polypharmacy index, multimorbidity index, and aspirin use, stratified by region. For all-cause mortality analyses, models were further adjusted for prevalent CVD and prevalent cancer. For any cancer analyses, models were further adjusted for menopausal status and use of menopause hormone therapy. For any CVD analyses, models were further adjusted for polygenic risk score. For any fracture analyses, models were further adjusted for vitamin or mineral supplement use and polygenic risk score (osteoporosis). hPDI indicates healthful plant-based diet index; HR, hazard ratio; uPDI, unhealthful plant-based diet index.

Figure 2.. Multivariable-Adjusted Hazard Ratios and 95% CIs for Postmenopausal Breast Cancer, Colorectal Cancer, and Prostate Cancer Across Sex-Specific Healthful vs Unhealthful Plant-based Diet Index Quartiles (n = 117 569)

All models used age as the underlying time variable and were adjusted for sex (excluding breast and prostate cancer), body mass index, race and ethnicity, physical activity level, smoking status, alcohol intake, education level, energy intake, polypharmacy index, multimorbidity index, and aspirin use, stratified by region. For breast cancer analyses, models were restricted to postmenopausal case patients with breast cancer and were further adjusted for use of menopause hormone therapy, use of oral contraception, polygenic risk score, age at menarche, and age at first live birth. For colorectal cancer analyses, models were further adjusted for menopausal status, polygenic risk score, and menopause hormone therapy. For prostate cancer analyses, models were further adjusted for polygenic risk score. hPDI indicates healthful plant-based diet index; HR, hazard ratio; uPDI, unhealthful plant-based diet index.

Figure 3.. Multivariable-Adjusted Hazard Ratios and 95% CIs for Myocardial Infarction, Ischemic Stroke, and Hemorrhagic Stroke Across Sex-Specific Healthful vs Unhealthful Plant-based Diet Index Quartiles (n = 123 134)

All models used age as the underlying time variable and were adjusted for sex, body mass index, race and ethnicity, physical activity level, smoking status, alcohol intake, education level, energy intake, polypharmacy index, multimorbidity index, and aspirin use, stratified by region. For myocardial infarction analyses, models were further adjusted for polygenic risk score (coronary artery disease). For ischemic stroke analyses, models were further adjusted for polygenic risk score (ischemic stroke). For hemorrhagic stroke analyses, models were further adjusted for polygenic risk score (cardiovascular disease). hPDI indicates healthful plant-based diet index; HR, hazard ratio; uPDI, unhealthful plant-based diet index.

Figure 4.. Multivariable-Adjusted Hazard Ratios and 95% CIs for Cardiovascular Disease (CVD) Across Strata of Genetic CVD, Ischemic Stroke, or Coronary Artery Disease Risk, With Healthful Plant-based Diet Index (hPDI) Score Modeled as a Continuous Trend (10-Point Increments)

Analyses used age as the underlying time variable and were adjusted for sex, body mass index, race and ethnicity, physical activity level, smoking status, alcohol intake, education level, energy intake, polypharmacy index, multimorbidity index, and aspirin use, stratified by region. Heterogeneity was tested by comparing 2 models: one model without an interaction term between the subgroup of interest and hPDI compared with another model with an interaction term between the subgroup of interest and hPDI. The likelihood ratio test was used to produce P interaction values. HR indicates hazard ratio; PRS, polygenic risk score.