Abstract Body (Do not enter title and authors here): Introduction: Visceral abdominal (VA) fat is an established risk factor for type 2 diabetes. It is still unclear, however, whether subcutaneous abdominal (SA) fat independently also increases diabetes risk.
Aims: We aimed to 1) describe the distribution of abdominal fat composition by sex, and 2) determine the independent association between SA fat and diabetes.
Methods: We included U.S. adults 20-59 years from the 2011-2018 National Health and Nutrition Examination Survey (NHANES) with whole body DEXA scans. DEXA scans were performed using Hologic Discovery model A densitometers. Visceral & subcutaneous fat mass (in lbs) were estimated via Hologic software (APEX v4.0) at the L4-L5 interspace. We used multivariable logistic regression to determine the association between SA, VA, & SA-to-VA fat ratio and diabetes (A1c ≥6.5%, fasting glucose ≥126 mg/dl or known diabetes diagnosis). All analyses were stratified by sex, accounted for the complex survey design to obtain nationally representative estimates, and were adjusted for potential confounders.
Results: Among 7,125 adults (48% female, 61% non-Hispanic White), 33% had BMI ≥30, 49% had high waist circumference (>88cm for female, >102cm for male), and 7% had diabetes. Females had a median 0.9 lbs (IQR, 0.5-1.3) of VA fat, median 3.9 lbs (2.9-5.2) of SA fat, and a SA-to-VA ratio of 4.6 (3.5-6.2). Males had a median 1.0 lbs (IQR, 0.7-1.5) of VA fat, median 2.7 lbs (1.8-3.5) of SA fat, and a SA-to-VA ratio of 2.3 (1.8-3.0). After adjusting for VA and confounders, every 1 lb increase in SA fat was associated with a lower, albeit insignificant, odds of diabetes in females and males (OR [95% CI]; female: 0.84 [0.65,1.08], male: 0.87 [0.71,1.07]). In contrast, every 1 lb increase in VA fat was associated with a >2.5x higher odds of diabetes (female: 3.36 [2.26,4.99], male: 2.58 [1.69,3.98]). Every 1 unit increase in SA-to-VA ratio was associated with a significant 33-35% lower odds of diabetes (female: 0.67 [0.57,0.78], male: 0.65 [0.53,0.80]).
Conclusion: U.S. females typically have four-to-sixfold more SA than VA fat, whereas males have one-to-threefold more SA than VA fat. In both females and males, SA was independently associated with lower likelihood of diabetes, suggesting subcutaneous abdominal adiposity may not contribute to, and potentially protect against, diabetes risk, akin to appendicular (arm/leg) adiposity.