Sex difference in the association between malnutrition and hypoglycemia in hospitalized patients.

OBJECTIVE To examine the difference between males and females regarding association between malnutrition risk and hypoglycemia. DESIGN Sub-analysis of a cross-sectional study of newly admitted patients to internal medicine departments. METHODS Malnutrition risk, assessed with Nutritional Risk Screening 2002 (NRS2002), and serum albumin were measured upon admission. Logistic regression was applied to men and women separately, to test the effect of malnutrition and hypoalbuminemia on incidence of hypoglycemia. RESULTS Included were 1186 patients (50.4% males, 39.2% with positive NRS2002). Rate of positive NRS2002 was similar across sexes (36.5% vs. 41.2% in males and females respectively, p=0.204). Among females, NRS2002 was associated with higher incidence of hypoglycemia (9.5% vs. 2.4% in NRS2002 negative females, p<0.001). Among males, no such difference was noted (9.2% compared to 7.1% in NRS2002 positive & negative males respectively, p=0.520). The weight loss/decreased food intake criterion of the NRS2002 was significantly higher in the hypoglycemic group within females (p=0.03). Logistic regression showed that serum albumin was inversely associated with hypoglycemia in both females (OR 0.477, 95% CI 0.282-0.806, p=0.006) and males (OR 0.532, 95% CI 0.355-0.795, p=0.002). However, increased malnutrition risk was associated with hypoglycemia only among females (OR 2.007, 95% CI 1.058-3.809, p=0.033). Diabetes status was associated with hypoglycemia (OR 1.907, 95% CI 1.056-3.445, p=0.032) only in males; this association did not occur in females. CONCLUSIONS Malnutrition risk, as measured by the NRS2002, is associated with significantly increased incidence of hypoglycemia in women alone. Females who lose weight prior to hospitalization have an increased risk to develop hypoglycemia.