Empirically Derived Whole-Diet Patterns and Multidimensional Adiposity in Emerging Adulthood: A Cross-Sectional Study of Geographically Diverse Students at a UK University

Background: Emerging adulthood is a period in which dietary behaviours may become established before their longer-term effects on adiposity are fully expressed. Evidence linking empirically derived whole-diet patterns with multiple adiposity phenotypes remains limited in narrow-age university populations. This study examined associations between three previously derived dietary pattern scores and adiposity/body composition in geographically diverse young adults attending an English university. Methods: This historical cross-sectional analysis included 111 students aged 18-22 years assessed during 2016, within the 2015/2016 academic year, at De Montfort University, Leicester. Habitual diet was assessed using an adapted EPIC-Norfolk food frequency questionnaire. Three standardised dietary pattern scores-Plant-Rich, Animal-Source Foods and Energy-Dense/Discretionary-were evaluated in relation to BMI, body fat percentage and a bioelectrical impedance-derived visceral fat index. The prespecified primary multivariable models evaluated dietary pattern-adiposity associations conditional on age, sex, geographical background and log-transformed total energy intake. HC3 robust standard errors were used, with Benjamini-Hochberg false discovery rate correction across the nine prespecified primary associations. Results: Mean age was 20.4 ± 1.1 years and 70.3% of participants were female. Overall, 33.3% had BMI ≥ 25 kg/m², whereas visceral fat index values ranged from 1 to 9. None of the nine Model 1 dietary pattern-adiposity associations was nominally statistically significant before adjustment for total energy intake, and addition of energy intake did not uniformly attenuate the coefficients. None of the prespecified primary Model 2 associations survived false discovery rate correction. The most pronounced primary estimate was the inverse association between Animal-Source Foods and visceral fat index (standardised β = -0.172, 95% CI -0.358 to 0.014; p = 0.069; q = 0.481). Secondary, mutually adjusted, transformed-outcome, geographical background, physical activity-adjusted and influence analyses did not materially alter the overall interpretation. Conclusions: After conditioning on total energy intake, no robust independent association was detected between the three empirical dietary pattern scores and contemporary adiposity in this cohort. The findings therefore suggest limited detectable independent relevance of dietary pattern composition for contemporary adiposity, although smaller associations cannot be excluded with the available precision. The results are compatible with overall energy intake capturing adiposity-related variation overlapping with dietary pattern composition but do not establish that energy intake is a stronger causal determinant.