SAT-535 Association Between Adiponectin and Bone Mineral Density in Mexican Postmenopausal Women with Normal Glucose Tolerance and Type 2 Diabetes Mellitus

Background: Adiponectin’s effect on bone health is not well dilucidated and literature reports are controversial. Human studies on the relationship between adiponectin and bone mineral density, have obtained variable results. Prevalence of osteoporosis and type 2 diabetes mellitus (T2DM) has significantly increased in our population, making it useful to identify a diagnostic and prognostic marker that can predict skeletal outcomes in these patients. Objective: To stablish the association between adiponectin levels and bone mineral density (BMD) in Mexican postmenopausal women with normal glucose tolerance (NGT) and T2DM. Material and methods: Cross-sectional study. We reviewed all bone densitometries since January 2014, identifying all postmenopausal women. We excluded patients with impaired fasting glucose, impaired glucose tolerance and concurrent diseases or therapies that could impact BMD. Glucose, HbA1c, adiponectin, body composition, and anthropometric parameters were measured in all participants. Considering an association of 0.30, allowing an α-error of 0.05 and β-error of 0.20 we estimated a sample size of 170 women. Results: 69 women have been recruited until October 2018 (NGT n= 35; T2DM n= 34). Mean age was 64.2 ± 9.6 years with time since menopause onset of 15.5 ± 8.3 years. Mean BMI was 26.5 ± 4.3 kg/m(2), mean body fat 39.4 ± 5.8% and median waist circumference 85 [78.3-93] cm. In women with T2DM median time from diagnosis was 12 years and median HbA1c of 7% (6.2-8.0%). There were no significant differences in adiponectin concentration between groups (NGT: 11.7 [8.4-20.4] μg/mL; T2DM: 9.7 [5.7-17.3] μg/mL, p= 0.192). In all participants we observed a positive correlation between body weight and BMD in the femoral neck (ρ= 0.435, p< 0.0001), total hip (ρ= 0.405, p= 0.001) and lumbar spine (ρ= 0.387, p= 0.001). A negative association between adiponectin and BMD was demonstrated only in the femoral neck (ρ = -0.349, p = 0.004). Analyzing by subgroups women with NTG preserve this association (ρ = - 0.348, p = 0.044), while women with T2DM only show a tendency (ρ = -0.320, p = 0.075). When adjusting by body weight, the association was lost (NGT ρ = -0.252, p = 0.157; T2DM ρ = -0.075, p = 0.689). Conclusion: There is no association between adiponectin levels and BMD in Mexican postmenopausal women with NGT and T2DM. The negative association observed before the adjustment was mediated by body weight. To clarify these findings is necessary to complete the study’s sample size.