METS-IR Linked to Cardiovascular Risk Across Chinese and U.S. Cohorts

A study of middle-aged and older adults in China and the United States found that METS-IR was consistently associated with cardiovascular disease, while the Conicity index contributed mainly through classifications that captured differences between metabolic status and central body geometry. Researchers analyzed 5,837 adults in the China Health and Retirement Longitudinal Study (CHARLS) and 10,903 participants in the 1999–2018 U.S. National Health and Nutrition Examination Survey (NHANES). The results do not show that either measure improves cardiovascular risk prediction or has established clinical usefulness, the authors said.

The two datasets provided complementary evidence but did not offer directly interchangeable estimates. CHARLS had a prospective design and assessed self-reported incident cardiovascular disease recorded at the 2013 or 2015 follow-up. It included 761 cardiovascular events among 5,837 participants aged 45 or older. NHANES supplied an external, survey-weighted cross-sectional comparison involving 10,903 adults aged 45 or older and 1,881 prevalent cardiovascular disease cases. Modified Poisson models estimated risk ratios in CHARLS and prevalence ratios in NHANES. The distinction between incident and prevalent disease, together with the different survey designs and analytical approaches, limits direct comparison of the risk ratios and prevalence ratios.

The clearest prospective association involved METS-IR. In CHARLS, each one-standard-deviation increase in METS-IR was associated with a 12.5% higher risk of incident cardiovascular disease, corresponding to a risk ratio of 1.125 and a 95% confidence interval of 1.053–1.202. The continuous association for the Conicity index was compatible with the null, with a risk ratio of 1.016 and a 95% confidence interval of 0.950–1.085. In other words, the prospective analysis did not provide clear support for a relationship between increasing Conicity values and cardiovascular risk when the index was evaluated continuously. METS-IR, by contrast, showed the more consistent cardiovascular association across the study’s two analytical components.

The investigators also evaluated the measures continuously and divided them into prespecified lower and higher categories, identified as quartiles Q1–Q2 and Q3–Q4. Combining the two categorical dimensions produced four joint phenotypes labeled G00, G10, G01 and G11. In CHARLS, all three non-reference groups were associated with higher adjusted risk of incident cardiovascular disease relative to G00. The adjusted risk ratios were 1.377, with a 95% confidence interval of 1.115–1.702, for G10; 1.244, with a 95% confidence interval of 1.005–1.540, for G01; and 1.339, with a 95% confidence interval of 1.112–1.612, for G11. The G11-to-G10 comparison was near the null, and the analyses did not support either multiplicative or additive interaction between METS-IR and the Conicity index. Thus, the joint-category estimates suggested cardiovascular associations beyond the reference group, but they did not establish a stronger statistical effect from combining high values of the two measures.

NHANES produced a broadly similar but less complete cross-sectional pattern. The G01 estimate was directionally similar to the CHARLS result but imprecise. G11 was associated with higher prevalence of cardiovascular disease, with a survey-weighted prevalence ratio of 1.255 and a 95% confidence interval of 1.087–1.448. The researchers additionally compared Conicity-based categories with waist-to-height ratio and body roundness index benchmarks. Paired G01 contrasts between the Conicity index and those alternative classifications were supported in CHARLS, whereas the corresponding NHANES estimates were imprecise. Overall, the authors characterized the Conicity index as contributing mainly at the phenotype level by identifying metabolic–body-geometry discordance, rather than through a stronger continuous association or evidence of statistical interaction with METS-IR. The study remains observational: CHARLS relied on self-reported incident outcomes, NHANES examined prevalent disease cross-sectionally, and neither analysis demonstrated incremental predictive value. Prospective validation is therefore needed before the joint phenotypes can be considered useful for clinical assessment.

Subject of Research: Associations of METS-IR and Conicity joint phenotypes with cardiovascular disease in adults aged 45 and older

Article Title: Metabolic dysfunction and central body geometry in cardiovascular disease: joint METS-IR–Conicity phenotypes in CHARLS and NHANES

Article References: Li, H., Zhao, C., & Wu, X. (2026). Metabolic dysfunction and central body geometry in cardiovascular disease: joint METS-IR–Conicity phenotypes in CHARLS and NHANES. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02665-5

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02665-5

Keywords: METS-IR, Conicity index, cardiovascular disease, CHARLS, NHANES, metabolic dysfunction, body geometry, risk ratios, prevalence ratios, prospective cohort, Metabolic, dysfunction