OBJECTIVE: To examine the association between hypoglycemia and fall-related outcomes in older patients with type 2 diabetes mellitus (T2DM). METHODS: This retrospective cohort study used electronic medical records of T2DM patients (≥65 years) from the Veterans Integrated Service Network 16 (VISN 16) data warehouse (01/01/2004-06/30/2010). Patients in nonhypoglycemia group (non-HG) were 1:1 randomly matched with patients in hypoglycemia group (HG) by age (±5 years), sex, race, and medical center location. Fall-related events (i.e., fractures and head injuries) were identified, with a fall being the external cause within ±2 days. McNemar tests and generalized estimating equation (GEE) models were used to compare fall-related events in the 1-year outcome period after the index date (i.e., date of first hypoglycemic episode). We also examined fall-related healthcare utilization. RESULTS: A total of 4,215 patients in each group were studied, with the mean age of 76.5 years (SD: 5.85). The mean Charlson comorbidity index (CCI) scores were 5.73 (SD: 2.95) in the HG and 4.34 (SD: 2.40) in the non-HG. The HG had significantly higher rates of fall-related events than non-HG, 27 (0.64%) versus 1 (0.02%) and 89 (2.11%) versus 21 (0.50%) events within 30 days and 1 year, respectively. GEE models confirmed the elevated risk of fall-related events after controlling for sociodemographic and clinical characteristics, comorbidities, and medication use (adjusted odds ratio [aOR]: 2.70; 95% confidence interval [CI]: 1.64-4.47). The HG patients were more likely to have emergency department (ED) visits, hospital admissions, and long-term care placement compared to their counterparts. CONCLUSION: Hypoglycemia is associated with worse fall-related outcomes among the elderly veterans.
OBJECTIVE: To examine the association between hypoglycemia and fall-related outcomes in older patients with type 2 diabetes mellitus (T2DM). METHODS: This retrospective cohort study used electronic medical records of T2DM patients (≥65 years) from the Veterans Integrated Service Network 16 (VISN 16) data warehouse (01/01/2004-06/30/2010). Patients in nonhypoglycemia group (non-HG) were 1:1 randomly matched with patients in hypoglycemia group (HG) by age (±5 years), sex, race, and medical center location. Fall-related events (i.e., fractures and head injuries) were identified, with a fall being the external cause within ±2 days. McNemar tests and generalized estimating equation (GEE) models were used to compare fall-related events in the 1-year outcome period after the index date (i.e., date of first hypoglycemic episode). We also examined fall-related healthcare utilization. RESULTS: A total of 4,215 patients in each group were studied, with the mean age of 76.5 years (SD: 5.85). The mean Charlson comorbidity index (CCI) scores were 5.73 (SD: 2.95) in the HG and 4.34 (SD: 2.40) in the non-HG. The HG had significantly higher rates of fall-related events than non-HG, 27 (0.64%) versus 1 (0.02%) and 89 (2.11%) versus 21 (0.50%) events within 30 days and 1 year, respectively. GEE models confirmed the elevated risk of fall-related events after controlling for sociodemographic and clinical characteristics, comorbidities, and medication use (adjusted odds ratio [aOR]: 2.70; 95% confidence interval [CI]: 1.64-4.47). The HG patients were more likely to have emergency department (ED) visits, hospital admissions, and long-term care placement compared to their counterparts. CONCLUSION:Hypoglycemia is associated with worse fall-related outcomes among the elderly veterans.
Authors: Derek LeRoith; Geert Jan Biessels; Susan S Braithwaite; Felipe F Casanueva; Boris Draznin; Jeffrey B Halter; Irl B Hirsch; Marie E McDonnell; Mark E Molitch; M Hassan Murad; Alan J Sinclair Journal: J Clin Endocrinol Metab Date: 2019-05-01 Impact factor: 5.958