Key points are not available for this paper at this time.
Introduction: Body mass index (BMI) and waist circumference (WC) capture related but distinct measures of adiposity (overall vs central adiposity). It is unclear if changes in WC are associated with incident diabetes mellitus (DM) and whether the relation is similar between BMI groups. Methods: We included 3,234 CARDIA participants (61% Black, 63% females, mean age 24.7 3.7 years at baseline) without prevalent DM from baseline (1985-1986) through exam Year 20 (Y20; 2005-2006). Incident DM after Y20 was defined as fasting blood glucose ≥126 mg·dl -1 , post-load blood glucose ≥200 mg·dl -1 , HbA1c ≥6.5%, or use of antidiabetic medication through exam Year 35 (2022-2023). WC (cm) and BMI (kg·m -2 ) were measured at baseline and Y20. WC change was calculated as the difference between Y20 and baseline examinations. We categorized participants into three WC 20-year change groups: no change (±≤6 cm), decrease (>6cm loss), and increase (>6cm gain). We used Cox proportional hazards regression to estimate HR (95% CIs) for incident DM over 15 years. The model was adjusted for baseline age, sex, race, field center, education, smoking status, alcohol intake, diet score, physical activity, HDL, LDL, and BMI status ( Table footnote). We assessed interactions of WC change with BMI status and sex, respectively. Results: The 15-year incidence of DM was 30% (973 cases, 40,037 person-years). Compared with no change in WC, WC decrease was not associated with incident DM (0.84 0.66, 1.06) (see Table) , but WC increase was associated with a greater 15-year risk of DM (1.21 1.02, 1.44). Interactions of WC change with sex ( p =0.35) and BMI status ( p =0.68) were null. Associations between 20-year WC change and subsequent DM risk were consistent between 20-year BMI status change groups. Conclusion: An increase in central adiposity during young and middle adulthood portends subsequent DM risk. Investigation of more accurate physiological measures of total adiposity and its distribution may help inform adipose tissue-specific pathways underlying DM risk.
Booker et al. (Tue,) studied this question.