Triglyceride–HDL cholesterol–glucose body index and cardiometabolic multimorbidity: evidence from a national longitudinal cohort and a subplateau health-screening population
4 affiliations
- 1Department of Pharmacy, Zhongshan Hospital of Xiamen University, School of Medicine, Xiamen University
- 2Department of Pharmacy, Foshan University Affiliated Third Hospital
- 3Department of Pulmonary and Critical Care Medicine, Northeast Yunnan Central Hospital; Northeast Yunnan Hospital of Kunming Medical University; Northeast Yunnan Clinical School of Kunming Medical University
- 4Kunming Medical University
Abstract
Background: Cardiometabolic multimorbidity (CMM) is an increasing health burden, but accessible composite markers integrating glycemic, lipid, and adiposity-related abnormalities remain insufficiently evaluated across heterogeneous populations. We examined the association between the triglyceride–high-density lipoprotein cholesterol–glucose body index (TyHGB) and CMM in two distinct Chinese populations.
Methods: This study included 8,116 participants without baseline CMM from the China Health and Retirement Longitudinal Study (CHARLS) and 10,092 adults undergoing routine health examinations in a southwestern China subplateau region. Cox proportional hazards models were used to evaluate incident CMM in CHARLS, whereas logistic regression was used for prevalent CMM in the health-screening population. TyHGB was analyzed per standard deviation and by quartiles. Nonlinear associations, subgroup consistency, discriminatory performance, exploratory mediation, and robustness were further assessed.
Results: During follow-up, 632 CHARLS participants developed CMM (7.8%), while 941 participants in the health-screening population had prevalent CMM (9.3%). Compared with the lowest quartile, the fully adjusted hazard ratio for the highest TyHGB quartile was 2.92 (95% CI, 2.19–3.89) in CHARLS, and the corresponding odds ratio was 11.62 (95% CI, 8.82–15.3) in the health-screening population (both P for trend <0.001). Nonlinear associations were observed in both populations, with inflection points of 9.139 and 10.470, respectively. TyHGB showed modestly higher AUCs than TyG in CHARLS (0.689 vs 0.648) and the health-screening population (0.762 vs 0.710; both P < 0.001). Associations were consistent across major subgroups and sensitivity analyses. In CHARLS, exploratory mediation analysis estimated that HbA1c accounted for 48.15% of the association.
Conclusions: Higher TyHGB was consistently associated with greater CMM burden across longitudinal and cross-sectional settings. As a low-cost index derived from routine measurements, TyHGB may serve as a useful adjunct to established cardiometabolic risk indicators in population-level screening, and its clinical role warrants further prospective evaluation; its thresholds and performance should be interpreted within the target population.
Related
- The relationship between C-reactive protein-triglyceride-glucose index and incident cardiometabolic multimorbidity in postmenopausal women: insights from the China Health and Retirement Longitudinal Study (CHARLS).
- Dose-Response Relationship Between Triglyceride Glucose-Waist Height Ratio Index and Cardiometabolic Multimorbidity Among Chinese Middle-Aged and Older People.
- Waist-centered triglyceride-glucose markers and inflammation in relation to incident cardiometabolic multimorbidity: a multinational pooled analysis of five aging cohorts
- Association between C-reactive protein-triglyceride-glucose index and one-year all-cause mortality in patients with cardiometabolic multimorbidity
- Association of the remnant cholesterol inflammatory index and frailty with cardiometabolic multimorbidity among Chinese middle-aged and elderly populations: a nationwide cohort study.
BibTeX
@article{su2026triglyceridehdl,
title = {{Triglyceride–HDL cholesterol–glucose body index and cardiometabolic multimorbidity: evidence from a national longitudinal cohort and a subplateau health-screening population}},
author = {Meiqin Su and Yu Du and Jie Wu and Huanqiong Fan and Guosong Jiang},
journal = {Frontiers in Endocrinology},
year = {2026},
month = nov,
doi = {10.3389/fendo.2026.1948613},
url = {https://www.frontiersin.org/articles/10.3389/fendo.2026.1948613/full},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Triglyceride–HDL cholesterol–glucose body index and cardiometabolic multimorbidity: evidence from a national longitudinal cohort and a subplateau health-screening population
AU - Meiqin Su
AU - Yu Du
AU - Jie Wu
AU - Huanqiong Fan
AU - Guosong Jiang
PY - 2026
DA - 2026/11/03
JO - Frontiers in Endocrinology
DO - 10.3389/fendo.2026.1948613
UR - https://www.frontiersin.org/articles/10.3389/fendo.2026.1948613/full
AB - Background: Cardiometabolic multimorbidity (CMM) is an increasing health burden, but accessible composite markers integrating glycemic, lipid, and adiposity-related abnormalities remain insufficiently evaluated across heterogeneous populations. We examined the association between the triglyceride–high-density lipoprotein cholesterol–glucose body index (TyHGB) and CMM in two distinct Chinese populations. Methods: This study included 8,116 participants without baseline CMM from the China Health and Retirement Longitudinal Study (CHARLS) and 10,092 adults undergoing routine health examinations in a southwestern China subplateau region. Cox proportional hazards models were used to evaluate incident CMM in CHARLS, whereas logistic regression was used for prevalent CMM in the health-screening population. TyHGB was analyzed per standard deviation and by quartiles. Nonlinear associations, subgroup consistency, discriminatory performance, exploratory mediation, and robustness were further assessed. Results: During follow-up, 632 CHARLS participants developed CMM (7.8%), while 941 participants in the health-screening population had prevalent CMM (9.3%). Compared with the lowest quartile, the fully adjusted hazard ratio for the highest TyHGB quartile was 2.92 (95% CI, 2.19–3.89) in CHARLS, and the corresponding odds ratio was 11.62 (95% CI, 8.82–15.3) in the health-screening population (both P for trend <0.001). Nonlinear associations were observed in both populations, with inflection points of 9.139 and 10.470, respectively. TyHGB showed modestly higher AUCs than TyG in CHARLS (0.689 vs 0.648) and the health-screening population (0.762 vs 0.710; both P < 0.001). Associations were consistent across major subgroups and sensitivity analyses. In CHARLS, exploratory mediation analysis estimated that HbA1c accounted for 48.15% of the association. Conclusions: Higher TyHGB was consistently associated with greater CMM burden across longitudinal and cross-sectional settings. As a low-cost index derived from routine measurements, TyHGB may serve as a useful adjunct to established cardiometabolic risk indicators in population-level screening, and its clinical role warrants further prospective evaluation; its thresholds and performance should be interpreted within the target population.
SN - 1664-2392
ER -