Serum osmolality was non-linearly associated with the risk of all-cause and cardiovascular mortality in patients with diabetes

Aims This study aimed to evaluate the relationship between both low and high osmolarity and the risk of all-cause and cause-specific mortality in diabetic population.Methods All participants were included from the National Health and Nutrition Examination Survey 1999-2014. Baseline serum osmolality was determined from laboratory tests and cause of death from national death records. HRs and 95% CIs for all-cause mortality and cardiovascular mortality in diabetes were estimated using Cox proportional regression analysis. The non-linear relationship was explored using restricted cubic splines regression.Results Among 7622 individuals with diabetes, 1983 (12.4%) died during a total of 3.26 thousand person-years of follow-up. Compared with the reference category (281-284 mmol/kg), the multivariable-adjusted HRs and 95% CIs for all-cause mortality were 1.27 (1.16-1.40; p<0.001) in the lowest osmolality category (<201 mmol/kg) and 1.18 (1.09-1.28; p<0.001) in the highest osmolality category (>312 mmol/kg). Restricted cubic splines results showed that serum levels of osmolality had a U-shaped association with the risk of all-cause mortality, and L-shaped relationship with the risk of cardiovascular death.Conclusions Both low osmolality and high osmolality were predictive of increased all-cause mortality in patients with diabetes, supporting a U-shaped relationship. Also, a lower serum osmolality increased the risk of cardiovascular mortality.