Predominant Dietary Pattern Characteristics and Their Association with Obesity-Related Metabolic Phenotypes in Middle-Aged and Older Chinese Adults: Findings from a Nationwide Cross-Sectional Study

Background/Objectives: To identify major dietary patterns among adults aged ≥45 years and examine their associations with metabolic health-obesity phenotypes. Materials and Methods: We analyzed 32,091 adults aged ≥45 years from the 2015 China Adults National Chronic Disease and Nutrition Surveillance. Diet was assessed using three consecutive 24 h dietary recalls, with household condiment weighing used to improve estimation of cooking oil and salt intake. Dietary patterns were derived using principal component analysis and categorized into quartiles. Multinomial logistic regression adjusted for energy intake and key sociodemographic/lifestyle factors to estimate odds of metabolically unhealthy non-obesity (MUNO), metabolically healthy obesity (MHO) and metabolically unhealthy obesity (MUO) versus metabolically healthy non-obesity (MHNO). Results: Four patterns with geographic variation were identified: (1) the Rice-Vegetable-Pork pattern; (2) the Fruit-Egg-Dairy pattern; (3) the Red Meat-Offal-Snack pattern; and (4) the Soybeans-Tubers-Grains pattern. Compared with Q1, Q4 of Pattern 1 was associated with lower odds of MHO (OR = 0.42, 95% CI: 0.38-0.46) and MUO (OR = 0.40, 95% CI: 0.36-0.44). Pattern 2 and Pattern 3 were associated with higher odds of MHO (Q4 vs. Q1: OR = 1.42 and 1.21) and MUO (OR = 1.36 and 1.14, all p < 0.001). Pattern 4 was inversely associated with MHO (OR = 0.85, 95% CI: 0.79-0.92) but positively associated with MUNO (OR = 1.16, 95% CI: 1.08-1.24). Conclusions: The Rice-Vegetable-Pork pattern was associated with more favorable obesity-related metabolic phenotypes, whereas energy-dense, animal-derived patterns were associated with higher odds of obesity phenotypes; the Soybeans-Tubers-Grains pattern showed mixed associations.