Inverse relationship between ambient temperature and admissions for diabetic ketoacidosis and hyperglycemic hyperosmolar state: A 14-year time-series analysis

This study aimed to investigate the association of admissions for diabetic ketoacidosis (DKA) and hyperglycemic hyperosmolar state (HHS) with ambient temperature and season, respectively in patients with diabetes mellitus (DM), after excluding known co-morbidities that predispose onset of acute hyperglycemia events. This was a time series correlation analysis based on medical claims of 40,084 and 33,947 episodes of admission for DKA and HHS, respectively over a 14-year period in Taiwan. These episodes were not accompanied by co-morbidities known to trigger incidence of DKA and HHS. Monthly temperature averaged from 19 meteorological stations across Taiwan was correlated with monthly rate of admission for DKA or HHS, respectively, using the ‘seasonal Autoregressive Integrated Moving Average’ (seasonal ARIMA) regression method. There was an inverse relationship between ambient temperature and rates of admission for DKA (β = − 0.035, p < 0.001) and HHS (β = − 0.016, p < 0.001), despite a clear decline in rates of DKA/HHS admission in the second half of the study period. We also noted that winter was significantly associated with increased rates of both DKA (β = 0.364, p < 0.001) and HHS (β = 0.129, p < 0.05) admissions, as compared with summer. On the other hand, fall was associated with a significantly lower rate of HHS admission (β = − 0.016, p < 0.05). Further stratified analyses according to sex and age yield essentially similar results. It is suggested that meteorological data can be used to raise the awareness of acute hyperglycemic complication risk for both patients with diabetes and clinicians to further avoid the occurrence of DKA and HHS.