Dietary inflammatory index (DII) and attention deficit hyperactivity disorder: a cross-sectional study in US children and adolescents and Mendelian randomization study.

BACKGROUND: Diet may impact the risk of attention deficit hyperactivity disorder (ADHD) via altering inflammation. However, no investigations have been conducted to evaluate the relationship between ADHD and the diet's inflammatory potential. The study sought to investigate the relationship between ADHD and the dietary inflammatory index (DII). METHODS: We used data from the cross-sectional National Health and Nutrition Examination Survey 2001-2004 to investigate the relationship between DII and ADHD in 6481 children and adolescents aged 6-19 years. The relationship between ADHD and DII was investigated using weighted logistic regression and restricted cubic spline (RCS). We also explored the potential causal associations between ADHD and 187 types of food preferences using Mendelian randomization (MR). RESULTS: Despite thorough multivariate correction, a significant association between lower DII scores and an elevated probability of ADHD persisted (OR = 0.91, 95% CI 0.82-0.99). Nevertheless, when stratifying the DII into quartiles (T1-T4) for analysis, no statistically significant correlation was seen between DII and ADHD (P > 0.05). RCS analysis indicated a U-shaped correlation between DII and ADHD, with an inflection point at approximately 2.9. Mediation analysis indicated that the Frailty Index partly mediated the association between DII and ADHD, representing about 5.49% of the overall impact. The findings from the univariate MR analysis revealed that fruit liking is significantly linked to a reduced risk of ADHD. Conversely, a liking for red meat, such as lamb, as well as a fondness for flavorings like onions, is significantly associated with an increased risk of ADHD. Reverse two-sample MR analyses revealed that genetic predisposition to ADHD significantly altered 37 food preferences (FDR < 0.05). CONCLUSIONS: We observed a U-shaped association between the DII and ADHD risk. The Frailty Index may partially mediate (5.49%) the association between DII and ADHD. However, the cross-sectional design precludes causal inference. Fruit liking lowered ADHD risk, while red meat/onion liking increased risk. ADHD was associated with a generalized reduction in liking for a wide range of foods. Future studies should investigate the role of a moderate-inflammatory diet in ADHD prevention.