Abstract
BACKGROUND: Obesity indices that combine anthropometric and lipid measurements, such as the atherogenic index of plasma (AIP), cardiometabolic index (CMI), lipid accumulation product (LAP), metabolic score for insulin resistance (Mets-IR), triglyceride-glucose (TyG) index, and visceral adiposity index (VAI), have been recommended as surrogates for assessing cardiometabolic risk. This study aimed to examine the association between these indices and type 2 diabetes mellitus (T2DM) incidence risk using large routine health checkup data. METHODS: This retrospective cohort study involved 195,989 participants aged 35-74 years who attended the Specific Health Checkups in 2015 and were observed until 2021 in Fukushima, Japan. We used multivariable Cox regression analysis to examine the associations between the indices and T2DM risk and evaluated the diagnostic utility of the indices. RESULTS: The accumulated T2DM incidence rate per 100,000 person-years over a mean follow-up of 4.61 years was 706.6 in males and 441.2 in females. Multiple adjusted hazard ratios were significantly higher in the highest quartile of indices (1.33-4.22 fold higher) compared to the lowest quartile. The increased risk in the higher quartiles was more pronounced in participants aged < 50 years than in those aged ≥ 50 years for both sexes. The areas under the receiver operating characteristic curves (AUC) of AIP, CMI, LAP, and VAI were between 0.821 and 0.844, whereas those of Mets-IR and TyG were between 0.756 and 0.780. For both sexes, AUCs for AIP, CMI, and LAP showed no significant differences (P values > 0.05), but other pairwise indices had significant differences (P values < 0.001). AUCs were higher in participants with a body mass index (BMI) of < 25 kg/m(2) or aged < 50 years than in those with a BMI ≥ 25 kg/m(2) or aged ≥ 50 years, respectively. Net reclassification improvements of models for CMI and LAP were similar to AIP, while those for Mets-IR, TyG, and VAI were significantly lower than AIP. CONCLUSIONS: Obesity indices are positively associated with and applicable for predicting new-onset T2DM in Japanese adults, particularly AIP, CMI, and LAP.