OR31-5 Serum 25-hydroxyvitamin D is independently associated with peripheral insulin sensitivity, visceral adiposity and adiponectin concentrations in postmenopausal women with abnormal glucose tolerance

Program: Abstracts - Orals, Featured Poster Presentations, and Posters
Session: OR31-Disorders of Vitamin D Metabolism & Action
Monday, June 17, 2013: 11:15 AM-12:45 PM
Presentation Start Time: 12:15 PM
Room 130 (Moscone Center)
Vanessa Tardio1, Anne-Sophie Morisset*2, S John Weisnagel2, Jean Bergeron3, Simone Lemieux1 and Claudia Gagnon2
1Laval University, Quebec City, Canada, 2CHU de Québec Research Centre, Quebec City, QC, Canada, 3CHU de Québec Research Centre, Quebec City, Canada
Background: It is unclear whether the relationship between serum 25-hydroxyvitamin D (25(OH)D), insulin sensitivity and insulin sensitivity markers differs according to the glucose tolerance status.

Objectives: To determine whether the glucose tolerance status influences the pattern of associations between serum 25(OH)D and (1) euglycemic-hyperinsulinemic clamp (EHC)- and OGTT-based insulin sensitivity measures, (2) anthropometric and CT-derived adiposity measures and (3) adipokine/cytokine levels among a sample of 112 postmenopausal women with normal glucose tolerance (NGT) or abnormal glucose tolerance (AGT).

Methods: Anthropometry (waist circumference, weight, height) and total, visceral and subcutaneous adiposity (CT) were assessed. Fasting serum was analyzed for glucose, insulin, hs-CRP, TNF-α, IL-6, adiponectin, 25(OH)D and PTH. Indices of peripheral insulin sensitivity (glucose disposal rate (GDR), Matsuda index) and hepatic insulin sensitivity (HOMA-IS) were derived from a 2-h EHC and a 75g-OGTT. Glucose tolerance status was based on the 2-h post-OGTT glucose: NGT (<7.8 mmol/L) (n=65) and AGT (≥7.8 mmol/L) (n=47). Energy expenditure was calculated from a physical activity questionnaire.

Results: 23% of women with NGT vs. 32% of women with AGT had serum 25(OH)D levels ≤50 nmol/L (P=NS). In women with NGT, no association was seen between serum 25(OH)D and any of the variables studied. In contrast, in women with AGT, serum 25(OH)D was positively associated with GDR, Matsuda index, HOMA-IS and adiponectin concentrations (r=0.38, r=0.38, r=0.46 and r=0.44, respectively, all P ≤0.01) and negatively correlated with visceral adipose tissue area and IL-6 (r=-0.42 and r=-0.56, respectively, all P ≤0.05). Serum 25(OH)D was not associated with other measures of adiposity, hs-CRP or TNF-a in both the NGT and AGT groups. In a multivariate regression model including visceral adipose tissue area, age, PTH, season of blood sampling and energy expenditure, 25(OH)D levels remained significantly associated with GDR (partial r2=16%, P ≤0.01) and Matsuda index (partial r2=33%, P=0.001) but not with HOMA-IS in women with AGT. Serum 25(OH)D levels also significantly predicted adiponectin levels in the same model (partial r2=20%, P ≤0.002). The association between 25(OH)D and GDR was no longer significant when adiponectin or IL-6 was included in the model. However, the association between serum 25(OH)D and Matsuda index persisted after adding adiponectin or IL-6 to the model.

Conclusions: Serum 25(OH)D concentrations are independently associated with EHC- and OGTT-based measures of peripheral insulin sensitivity (GDR and Matsuda index), visceral adipose tissue area and adiponectin levels only in postmenopausal women with AGT. Our results suggest that in these women, the association between 25(OH)D and GDR could be mediated by a concomitant variation in adiponectin and/or IL-6.

Nothing to Disclose: VT, ASM, SJW, JB, SL, CG

*Please take note of The Endocrine Society's News Embargo Policy at http://www.endo-society.org/endo2013/media.cfm