{"title":"Plant-based dietary patterns and frailty in older adults: a multicohort longitudinal study with within-between decomposition and network analysis","authors":"","doi":"10.1093/ageing/afag304","DOIUrl":null,"url":null,"abstract":"<span class=\"paragraphSection\"><div class=\"boxTitle\">Objective</div>Plant-based dietary patterns may influence healthy ageing, but their longitudinal and multidimensional associations with frailty remain unclear. We examined these associations across three primary ageing cohorts and explored links between dietary components and frailty domains.<div class=\"boxTitle\">Methods</div>The primary sample included 8628 older adults from Chinese Longitudinal Healthy Longevity Survey (CLHLS), Health and Retirement Study (HRS) and English Longitudinal Study of Ageing (ELSA). Overall, healthful and unhealthful plant-based diet indices (PDIs, hPDIs and uPDIs) were derived from cohort-specific food-group data. A supplementary analysis of 10 821 SHARE participants used a two-component plant–animal contrast. Frailty was assessed using a multidomain deficit-accumulation index. Generalised linear mixed models examined diet–frailty associations, and Mundlak decomposition separated between-person differences from within-person changes where repeated diet data were available. Spline and network analyses examined non-linearity and component-level associations.<div class=\"boxTitle\">Results</div>Higher PDI was associated with lower frailty risk across all three primary cohorts (ORs = 0.71–0.80), whereas higher uPDI was associated with higher risk (ORs = 1.24–1.41). hPDI was inversely associated only in ELSA and was null in CLHLS and HRS. In CLHLS, between-person dietary differences showed stronger associations than within-person changes. Spline analysis provided limited evidence of non-linearity. Network analyses highlighted nuts, vegetables and plant oils as showing predominantly inverse associations with several frailty domains. SHARE analysis showed an inverse between-person association, no within-person association, with the plant-based component also showing predominantly inverse links with frailty domains.<div class=\"boxTitle\">Conclusion</div>Higher overall plant-based diet scores were associated with lower frailty risk, whereas unhealthful plant-based diets were associated with higher risk, underscoring the importance of diet quality.</span>","PeriodicalId":7682,"journal":{"name":"Age and ageing","volume":"3 1","pages":""},"PeriodicalIF":7.4000,"publicationDate":"2026-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Age and ageing","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1093/ageing/afag304","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GERIATRICS & GERONTOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Objective
Plant-based dietary patterns may influence healthy ageing, but their longitudinal and multidimensional associations with frailty remain unclear. We examined these associations across three primary ageing cohorts and explored links between dietary components and frailty domains.
Methods
The primary sample included 8628 older adults from Chinese Longitudinal Healthy Longevity Survey (CLHLS), Health and Retirement Study (HRS) and English Longitudinal Study of Ageing (ELSA). Overall, healthful and unhealthful plant-based diet indices (PDIs, hPDIs and uPDIs) were derived from cohort-specific food-group data. A supplementary analysis of 10 821 SHARE participants used a two-component plant–animal contrast. Frailty was assessed using a multidomain deficit-accumulation index. Generalised linear mixed models examined diet–frailty associations, and Mundlak decomposition separated between-person differences from within-person changes where repeated diet data were available. Spline and network analyses examined non-linearity and component-level associations.
Results
Higher PDI was associated with lower frailty risk across all three primary cohorts (ORs = 0.71–0.80), whereas higher uPDI was associated with higher risk (ORs = 1.24–1.41). hPDI was inversely associated only in ELSA and was null in CLHLS and HRS. In CLHLS, between-person dietary differences showed stronger associations than within-person changes. Spline analysis provided limited evidence of non-linearity. Network analyses highlighted nuts, vegetables and plant oils as showing predominantly inverse associations with several frailty domains. SHARE analysis showed an inverse between-person association, no within-person association, with the plant-based component also showing predominantly inverse links with frailty domains.
Conclusion
Higher overall plant-based diet scores were associated with lower frailty risk, whereas unhealthful plant-based diets were associated with higher risk, underscoring the importance of diet quality.
期刊介绍:
Age and Ageing is an international journal publishing refereed original articles and commissioned reviews on geriatric medicine and gerontology. Its range includes research on ageing and clinical, epidemiological, and psychological aspects of later life.