MON-034 Impact of Race and Obstructive Sleep Apnea on Glucose and Insulin Regulation in Women with PCOS

Abstract

The prevalence of prediabetes and diabetes is substantially higher in PCOS women with obstructive sleep apnea (OSA) compared to PCOS women without OSA1,2,3. Prior studies, however, did not examine the complex interaction between race and OSA on metabolic function in PCOS. We sought to determine if the impact of OSA on glucose and insulin metabolism is affected by race. We studied non-Hispanic white (NHW) (n=53) and African-American (AA) (n=48) women with PCOS. Following an overnight polysomnogram (PSG), PCOS women (NHW without OSA n=40; NHW with OSA n=13; AA without OSA n=36; AA with OSA n=12) had a 2-h 75-g oral glucose tolerance test (OGTT) with blood sampling every 30 minutes for measurement of glucose, insulin, and C-peptide concentrations. OSA severity was measured by the Apnea-Hypopnea Index (AHI). Only women without OSA (AHI < 5) or with moderate-to-severe OSA (AHI > 15) were included in these analyses; women with mild OSA were excluded. Insulin secretion rates (ISR) during the OGTT were derived by deconvolution of C-peptide levels 4. Area under the curve (AUC) response to the glucose challenge was calculated using the trapezoidal method. BMI and age did not differ between races in PCOS women without OSA (BMI [kg/m2]: 36.3±1.2 vs. 37.2±1.1, p=0.58; Age [yr]: 27.7±0.8 vs. 27.2±0.8, p=0.65; for NHW and AA respectively), or in PCOS women with OSA (BMI [kg/m2]: 42.8±1.7 vs. 44.7±2.0, p=0.50; Age [yr]: 31.4±1.6 vs. 28.6±1.6, p=0.18; for NHW and AA respectively). OSA severity was similar in NHW and AA PCOS women without OSA (AHI: 1.5±0.2 vs 2.1±0.2, p=0.076), and PCOS women with OSA (AHI: 32.0±4.9 vs. 28.3±4.4, p=0.26). Higher glucose responses during the OGTT were observed in NHW PCOS women with OSA compared to both NHW (AUC: 18,965±648 vs. 15,797±371, p=0.0004) and AA (AUC: 18,965±648 vs. 15,801±497, p=0.0005) PCOS women without OSA. Glucose responses did not differ significantly between AA PCOS women with OSA and AA PCOS women without OSA (AUC: 17,104±965 vs. 15,801±497, p=0.15). Similarly, ISR was higher in NHW PCOS women with OSA compared to both NHW (AUC: 5,648±488 vs. 3,907±231, p=0.0006) and AA (AUC: 5,648±488 vs. 3,981±235, p=0.0011) PCOS women without OSA. ISR did not differ significantly between AA PCOS women with OSA and AA PCOS women without OSA (AUC: 4,827±461 vs. 3,981±235, p=0.09). CONCLUSIONS: OSA has a greater impact on glucose and ISR during an oral glucose challenge in NHW compared to AA women with PCOS. Future studies would benefit from including race when evaluating metabolic outcomes in women with PCOS.