SAT-229 Diabetes Mellitus Is Associated With Lower Fracture Risk: A US Population Survey

Abstract Y. Chen: None. R. Karnchanasorn: None. H. Ou: None. L. Chuang: None. L. Polgreen: None. A.G. Gianoukakis: None. K.C. Chiu: None. It is well-recognized that Diabetes Mellitus (DM) is associated with increased fracture risk, likely attributed to a lower bone mineral density from poor glycemic control. However, it is unclear whether glucose metabolism status is also associated with fracture risk. This study compares the fracture rates among the three glucose tolerance states in a representative US population. In the National Health and Nutrition Examination Survey, reporting of osteoporosis (OP) was available in the 2005-2006, 2007-2008, 2009-2010, 2013-2014, and 2017-2018 surveys. We included subjects aged 20 years and older, with reported BMI and defined states of glucose tolerance, which were categorized into normal glucose tolerance (NGT), abnormal glucose tolerance (AGT), and DM based on established diabetes, HbA1c, fasting, and/or 2-hour post-challenged glucose levels. Reported fractures of the wrist, hip, spine, and other were included after excluding those related to trauma. The sample-weighted prevalence of fracture was compared among the 3 states of glucose tolerance. The odds ratio of all fractures and the individual fracture types were all computed for unadjusted analysis (OR) and adjusted (AOR) for age, gender, BMI, race/ethnicity, alcohol, tobacco use, established OP, steroid use, parental OP, and OP treatment. This study included 21,849 subjects with a total fracture rate of 9.66%, wrist fracture rate of 1.54%, hip fracture rate of 0.56%, spine fracture rate of 0.46%, and other fracture rate of 7.53%. Weighted-all fracture rates in the NTG, AGT, and DM groups were 8.58%, 10.54%, and 11.03% respectively with a significant increasing trend from NGT to DM groups (p<0.0001). Unadjusted OR for all fracture of AGT [1.255 (95CI: 1.254-1.256)] and DM [1.321 (95%CI: 1.319-1.323)] confirmed the increased fracture risk in DM. However, the trend was reversed after adjustment for covariates, with an AOR of 0.923 (95%CI: 0.922-0.925) in the AGT group and 0.788 (95%CI: 0.787-0.789) in the DM group. Within the DM group, reduced fracture risk was observed at the wrist (AOR: 0.631; 95%CI: 0.629-0.633), hip (AOR: 0.780; 95%CI: 0.775-0.785), and other fracture (AOR: 0.806; 95%CI: 0.805-0.808). Only spine fracture risk remained increased (AOR: 1.164; 95%CI: 1.158-1.171). The reversed trends occurred only after adjustments for age but not for BMI, gender, or race/ethnicity. In summary, in the US population, we observed a reduced fracture risk in the DM vs. NGT groups, except for spine fractures. The type of DM was not defined in the NHANES but based on population statistics, more than 90% of DM are expected to represent type 2 diabetes, and thus our observations are most applicable to type 2 diabetes. It is interesting that the increased fracture risk observed in the DM population disappeared, except for spine fracture, after adjustment for the covariate of age. The paradox of this observation remains to be elucidated. Saturday, June 1, 2024