Skip to main content
. 2021 Dec 8;45(2):348–356. doi: 10.2337/dc21-1692

Figure 1.

Figure 1

Associations between 4-year change in lifestyle in mid-life and weight change, including AHEI (A), physical activity (B), alcohol consumption (C), and smoking status (D), stratified by history of GDM. Least squares (LS) means of 4-year weight change were modeled in multivariable marginal models with generalized estimating equations adjusting for follow-up period, race (White or non-White), marital status (ever married or other), family history of diabetes (yes or no), age (years), BMI (underweight <18.5, normal 18.5–24.9, overweight 25–29.9, or obese ≥30.0 kg/m2), oral contraceptive use (current, past, or never), menopausal and hormone replacement therapy (HRT) use status (premenopausal never HRT use, postmenopausal never HRT use, postmenopausal current HRT use, postmenopausal past HRT use, or missing), sleep duration (≤6, 7–8, or >8 h), and concurrent changes in the other lifestyle factors (continuous change in AHEI score including alcohol, continuous change in physical activity [MET-h/week], and categorical change in smoking status) depending on the model. LS means were calculated as marginal means with adjusted covariates set as mean for continuous covariates and prevalence for categorical covariates. P heterogeneity (het) was calculated by the generalized score statistics for the interaction term between each lifestyle change and history of GDM.