Abstract
BACKGROUND AND AIMS: Psoriasis is a chronic inflammatory skin condition that adversely affects quality of life. Given limited evidence to inform dietary recommendations for psoriasis prevention, this study used data from the UK Biobank to examine associations between diet quality and incident psoriasis, and whether genetic susceptibility moderates this relationship.
METHODS: Participants without psoriasis at baseline who completed ≥2 Oxford WebQ 24-h dietary recalls were included. Psoriasis incidence was identified through self-report, primary care, and hospital records. Ten diet quality indices (DQIs) were computed: Low Carbohydrate Diet Score (LCDS), Alternative Mediterranean Diet (aMed) Score, Eatwell Score (EWS), Alternative Binary Eatwell Score (BEWS), Alternative Graded Eatwell Score (GEWS), Healthy Diet Index (HDI), Alternative Healthy Eating Index (aHEI), Dietary Approaches to Stop Hypertension Score (DASH), Dietary Inflammatory Index (DII), and Plant-Based Diet Index (PDI). A multivariable Cox proportional hazards regression model adjusted for sociodemographic, lifestyle, and clinical, estimated association between each DQI and psoriasis incidence.
RESULTS: Among 121,299 participants followed up for a mean 11.4 years, 822 developed psoriasis (incidence 0.68 %). In fully adjusted models, higher PDI scores (Q4: 56-77) were associated with a 19 % lower risk of psoriasis (HR: 0.806, [95 % CI: 0.651-0.997]) compared to lower PDI scores (Q1: 26-47). This was driven by lower intake of unhealthy plants (HR: 1.036, [95 % CI: 1.015-1.057], PFDR-trend = 0.021) and meat components (HR: 1.027, [95 % CI: 1.002-1.052]). While there was no strong evidence of interaction between PDI and genetic risk, the protective effect of PDI was stronger among individuals with low genetic risk (HR: 0.965, [95 % CI: 0.939-0.991], P = 0.009). Mediation analyses suggested partial effects via lower adiposity.
CONCLUSION: Individuals with the highest adherence to a plant-based dietary pattern were at lower risk of incident psoriasis in this large UK cohort, with no evidence that this association differed by genetic risk.