Background Higher diet quality shows associations with decreased risk of all-cause, cardiovascular disease, and cancer mortality in adults. To understand whether these associations also apply in younger age groups, we can study proximal metabolic risk factors: abdominal obesity, insulin resistance, hypertension and dyslipidemia. In this study we investigate how associations between diet quality and metabolic risk vary with age.

Methods We use data (n=2105) from the UK-representative National Diet and Nutrition Survey (2008–2016), across three age groups: adolescents (age 11–18), young adults (age 19–35), older adults (age 36–60). Four-day food diaries were processed to give an energy-adjusted diet quality index, based on the Dietary Approaches to Stop Hypertension (DASH) diet. Measures of plasma vitamin C, beta-carotene and lutein were combined to form a fruit and vegetable (F&V) biomarker score. Data on the five components of metabolic syndrome (waist circumference, blood triglycerides, blood high-density lipoprotein (HDL) cholesterol, blood pressure, fasting plasma glucose) were standardized by age, sex and ethnicity and combined to give a metabolic risk z-score. We assessed associations of (1) standardized DASH index and (2) standardized F&V biomarker score with metabolic risk z-score, across all ages, adjusted for potential confounders. We tested for interaction of the exposure with the three age groups, to understand moderation of effect estimates by age.

Results Adolescents and young adults showed lower self-reported diet quality (p<0.001), and lower F&V biomarker scores (p<0.05) compared to older adults. Across the whole analysis sample, both standardized DASH index (beta=-0.15, CI -0.22, -0.08) and standardized F&V biomarker score (beta=-0.33, CI -0.39, -0.27) were associated with metabolic risk z-score. Both DASH index and F&V biomarker score showed significant interactions with age group, with smaller associations with metabolic risk seen among adolescents and young adults compared to older adults (p<0.05). Associations between F&V biomarker score and metabolic risk remained significant across all age groups (adolescent: beta=-0.17, CI -0.26, -0.07, young adult: beta=-0.26, CI -0.36, -0.17, older adult beta=-0.39, CI -0.47, -0.32) while associations between DASH index and metabolic risk were attenuated below significance in adolescent and young adult groups (adolescent: beta=-0.00, CI -0.07, 0.08, young adult: beta=-0.07, CI -0.19, 0.04).

Conclusion Higher diet quality was associated with decreased metabolic risk, with stronger and more persistent associations seen using nutritional biomarkers, compared to self-reported dietary data. Across both diet measures, we find weaker cross-sectional associations between diet quality and metabolic risk in young people compared to older populations.