Metabolically unhealthy obesity linked to diabetes and stroke in older Chinese adults

Older adults in China who carry excess weight are far from a uniform group, and the differences among them may matter more than obesity itself. A cohort study drawing on the China Health and Retirement Longitudinal Study, a nationally representative survey conducted between 2011 and 2020, mapped how several distinct obesity phenotypes are distributed among Chinese adults aged 60 years or older and assessed how each relates to four major late-life health outcomes: cardiovascular disease, stroke, diabetes, and depressive symptoms. The findings, published in BMC Geriatrics, suggest that the metabolic quality of obesity, and the loss of muscle strength that can accompany it, are associated with future disease risk in ways that a simple body mass index reading cannot capture.

The analysis included 5,971 participants, of whom 42.1 percent were classified as obese using body mass index, waist circumference, and additional criteria related to cardiometabolic health, inflammation, and muscle strength. Even at baseline, the obese group carried a heavier burden of disease: 20.2 percent had diabetes compared with 10.6 percent of the non-obese group, 19.8 percent had cardiovascular disease compared with 12.4 percent, and 4.3 percent had experienced a stroke compared with 2.3 percent. The researchers then subdivided obesity into six phenotypes: metabolically unhealthy obesity, metabolically healthy obesity, sarcopenic obesity, nonsarcopenic obesity, inflammatory obesity, and noninflammatory obesity.

The distribution was strikingly uneven. Metabolically unhealthy obesity dominated, accounting for 74.6 percent of obese participants, while sarcopenic obesity affected 47.1 percent and inflammatory obesity 19.6 percent. Metabolically unhealthy obesity was also the phenotype that overlapped most extensively with the others, meaning many older Chinese adults with obesity simultaneously met criteria for more than one profile.

Geography added another layer of variation. Obesity prevalence was higher in northern China, a descriptive pattern the authors characterize as hypothesis-generating rather than explanatory. Major cities such as Beijing and Shanghai showed lower prevalence of metabolically unhealthy obesity and sarcopenic obesity than other regions.

The researchers note that these regional analyses are descriptive and do not identify the mechanisms behind the observed gradients. The prospective results sharpened the picture. Using Cox proportional hazards regression models fully adjusted for potential confounders, the team found that metabolically unhealthy obesity was associated with a higher hazard of incident diabetes (hazard ratio 2.22, 95 percent confidence interval 1.65 to 2.98) and a higher hazard of incident stroke (hazard ratio 1.43, 95 percent confidence interval 1.04 to 1.94).

Sarcopenic obesity, in which excess fat mass coexists with reduced muscle strength, told a different story: it was associated with a higher hazard of incident cardiovascular disease (hazard ratio 1.39, 95 percent confidence interval 1.09 to 1.77). Technically, the study classified phenotypes by combining anthropometric thresholds with markers of cardiometabolic health, systemic inflammation, and muscle strength, and it assessed outcomes using self-reported diagnoses alongside clinical measures. This design allows a more granular view than conventional BMI-based studies, but it remains observational.

The authors are explicit that the associations between phenotypes and incident disease do not establish causality. The study also addresses gaps in current knowledge. According to the authors, the heterogeneous nature of obesity phenotypes and their relationship to disease in older adults, particularly in China, remains poorly understood, and this analysis quantifies multiple phenotypes simultaneously in a nationally representative aging cohort.

Given that roughly three in four obese participants in this cohort were metabolically unhealthy and nearly half had sarcopenic obesity, the phenotypes linked to incident disease in this study are not rare edge cases but common among older Chinese adults with obesity. For now, the message from this large longitudinal dataset is measured but consequential: in older Chinese adults, the type of obesity, not merely its presence, was associated with which diseases appeared during follow-up. Metabolically unhealthy obesity was associated with incident diabetes and stroke, and sarcopenic obesity with incident cardiovascular disease.

Subject of Research: Obesity phenotypes and late-life disease risk in older Chinese adults

Article Title: Heterogeneous obesity phenotypes and late-life health in China

Article References: Zhang, K., Zhao, Z., Li, J., Gong, M., Yao, X., Liang, W., & Mang, J. (2026). Heterogeneous obesity phenotypes and late-life health in China. BMC Geriatrics. https://doi.org/10.1186/s12877-026-08562-9

Image Credits: AI Generated

DOI: 10.1186/s12877-026-08562-9

Keywords: obesity phenotypes, metabolically unhealthy obesity, sarcopenic obesity, older adults, China, diabetes, stroke, cardiovascular disease, cohort study, BMC Geriatrics, Heterogeneous, obesity