RESEARCH. CONNECTED.

Explore the scholarly record.

Discover prefix ownership, publishers, journals and DOI metadata. Understand the data behind every publication.

Joint associations of triglyceride-cholesterol-body weight index and glycated hemoglobin with incident diabetes: a prospective study from the CHARLS cohort

Wei Liu, Haiwen Zhang, Lihong Zhang, Ru Zhang, Zuoxu Zhang, Xiawenyi Li, Hongwei Zhu

DOI10.1186/s41043-026-01479-4
PublisherSpringer Science and Business Media LLC
Journal / SourceJournal of Health, Population and Nutrition
Published2026-10-11
Metadata Deposited2026-10-11 (updated: 2026-10-11)
Subject—
Languageen
ISSN2072-1315
Typejournal-article
Volume / Issue / Pages— / — / —
Citations0
References deposited0
Access / license metadataOpen license identified License 1 ↗A reuse license does not by itself establish whether the full text is freely readable.

Abstract

Abstract Introduction The triglyceride–cholesterol–body weight index (TCBI) integrates lipid metabolism and adiposity, whereas glycated hemoglobin (HbA1c) reflects long-term glycemic status. However, their joint association with incident diabetes remains unclear. This study examined their joint and interactive associations with diabetes and quantified the proportion of the association between TCBI and diabetes that is mediated by HbA1c. Materials and methods We analyzed 7,975 participants free of diabetes at baseline from the China Health and Retirement Longitudinal Study (CHARLS). TCBI was natural log-transformed (ln-TCBI). TCBI and HbA1c were analyzed as continuous variables and tertiles, and joint categories were defined by median levels. Multivariable Cox models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). We tested additive and multiplicative interactions, examined dose–response relationships using restricted cubic splines, and performed an exploratory mediation analysis to assess the indirect association of ln-TCBI with incident diabetes through HbA1c. Sensitivity analyses were also conducted. Results Over a follow-up period of up to 9.0 years, 1,207 participants developed diabetes. Compared with the low TCBI/low HbA1c reference group, the high TCBI/high HbA1c group had the highest risk (HR 2.36, 95% CI 1.96–2.84), followed by the low TCBI/high HbA1c group (HR 1.78, 95% CI 1.47–2.16) and the high TCBI/low HbA1c group (HR 1.40, 95% CI 1.13–1.71). Neither multiplicative nor additive interaction was detected. HbA1c accounted for 18.9% of the association between ln-TCBI and diabetes. Sensitivity analyses confirmed the robustness of these associations. Conclusions In middle-aged and older Chinese adults, higher TCBI and higher HbA1c are each independently associated with an increased risk of incident diabetes; their concurrent elevation is associated with the highest risk, and HbA1c may partially mediate the association between TCBI and diabetes.