Association Analysis of Imaging‐Based Lean Mass Distribution and History of Gestational Diabetes Mellitus in
NHANES
2011–2018
Yanqiong Lu, Shuangjun Li, Xupeng Liu, Shuang Liang ABSTRACT
Aim
This work examines potential links between lean tissue composition and prior occurrences of gestational diabetes mellitus (GDM) within a nationally representative cohort.
Methods
This cross‐sectional study included 1807 participants from the National Health and Nutrition Examination Survey (NHANES) (2011–2018). GDM history was self‐reported. Body composition was assessed using dual‐energy X‐ray absorptiometry (DXA). Weighted multivariable logistic regression, adjusted for covariates, was used to assess the relationships between body composition distribution and GDM history. Nonlinear associations were evaluated using cubic spline functions, with subgroup and sensitivity analyses to assess robustness.
Results
The adjusted model revealed that a history of GDM was associated with higher android‐to‐gynoid lean mass ratio (AGLR), android lean mass index (ALMI), and android‐to‐gynoid fat ratio (AGFR). Multivariable restricted cubic spline (RCS) analysis showed a significant linear association between AGLR, AGFR, and GDM history (nonlinearity p > 0.05), while ALMI demonstrated a nonlinear relationship (nonlinearity p < 0.05). These associations were significantly influenced by economic status and blood pressure levels ( p < 0.05). Sensitivity analysis excluding participants whose last live birth occurred > 10 years ago yielded results consistent with the primary analysis.
Conclusions
Higher AGLR, ALMI, and AGFR are associated with increased odds of a history of GDM, with AGLR showing the strongest correlation. These findings underscore the importance of incorporating GDM history and regional lean mass distribution into risk assessment.