Differential associations of diet with hepatic and muscle insulin resistance: insights from an dietary pattern analysis in the PERSON study

Abstract Purpose The relationship between dietary patterns (DPs) and type 2 diabetes is well established, but the potential role of tissue-specific insulin resistance (IR) in this association remains unclear. This study aimed to derive DPs using reduced rank regression (RRR), incorporating hepatic IR index (HIRI) and muscle insulin sensitivity index (MISI) as response variables. We also examined whether these patterns are associated with insulin sensitivity and pancreatic β-cell function. Methods We conducted a cross-sectional analysis of 700 adults with overweight or obesity participating in the screening phase of the PERSON study. Dietary intakes were assessed using a food frequency questionnaire. RRR was used to derive DPs based on HIRI and MISI. Associations with HOMA-IR, HOMA-β, Matsuda index and Disposition index were tested using multiple regression models adjusted for socio-demographic and lifestyle factors. Results One DP was retained, explaining 13.7% of the variation in HIRI, 2.8% in MISI, and 8.2% of the combined variation. This DP was characterized by high intakes of unprocessed red meat, processed meat, fresh cream and whipped cream, and low intakes of fruits, vegetables, and tea. It was significantly associated with higher HOMA-IR (β-coefficient ± SE: 0.04 ± 0.02) and HOMA-β (0.05 ± 0.01), and lower Matsuda index (− 0.08 ± 0.02). Conclusion The identified DP was more strongly associated with hepatic than muscle IR. This finding highlights differential associations between diet and tissue-specific IR, and supports the relevance of considering tissue-specific insulin resistance phenotypes when investigating the relationship between diet, insulin resistance and type 2 diabetes risk. Trial registration The trial was registered at https://clinicaltrials.gov/study/NCT03708419 (identifier NCT03708419).