Hypoglycemia in hospitalized nondiabetic older patients.

OBJECTIVE

To analyze the clinical characteristics, associated risk factors, and outcome of hypoglycemia in nondiabetic hospitalized older patients.

METHODS

A retrospective case control study.

METHODS

Sixty patients, aged 65 years and older, in the acute medical and geriatric wards who developed hypoglycemia. A control group was composed of 83 older patients, sex and age matched, in orthopedic and surgery wards who were undergoing corrective surgery for hip fracture or hernioplasty.

METHODS

For all patients, data for the following variables were abstracted from the charts: age, sex, degree of hypoglycemia, clinical presentation of hypoglycemia, number and duration of hypoglycemia episodes, nutritional state, and blood chemistry analysis. Risk factors were defined as nutritional state, heart failure, renal or liver disease, malignancy, and infection or sepsis.

RESULTS

Mean blood glucose in hypoglycemic cases was 38.9 +/- 7 mg/dL. Symptoms and signs of hypoglycemia were noted in only 38.4% (23/60) of patients. All identified risk factors except cachexia were found significantly more frequently in the hypoglycemic patients than in the control group. Mean total number of risk factors was greater in the hypoglycemic group than in the control group, 2.97 +/- 1.1 versus 1.64 +/- .8, respectively (P < .001). In a multivariant logistic model, low plasma albumin level, liver disease, malignancy, and congestive heart failure were significant predictors of hypoglycemia. In-hospital mortality rate was higher among the hypoglycemic patients, 48% versus 18.1% (P < .001), and was independent of the degree of hypoglycemia or the number of hypoglycemic episodes. Mortality was correlated significantly with the number of risk factors (3.4 +/- 1.1 vs 2.5 +/- 1.1; P = .006). Hypoglycemia remained a significant predictor of mortality (OR = 3.67; 95% CI, 1.2-11.2) even after the adjustment for other risk factors.

CONCLUSIONS

Hypoglycemic episodes occur even among nondiabetic hospitalized older patients. Symptoms and signs of hypoglycemia were noted in only two-fifths of the patients. Albumin less than 3.0 g%, liver disease, renal insufficiency, malignancy, congestive heart failure, and sepsis were statistically significant predictors of developing hypoglycemia. The overall mortality rate was significantly higher among the hypoglycemic patients and was independent of hypoglycemia levels. Mean total number of risk factors was significantly higher among those who died compared with hypoglycemic patients who survived. Based on the present study, the estimated odds of mortality in an older patient with hypoglycemia were 3.67 times higher than in those without hypoglycemia.