We recently showed that malnutrition is independently associated with a prolonged length of hospital stay.1
In a further analysis of our data, we investigated the association of malnutrition (nutritional risk index < 97.5) and age (≤ 65 years v. 65 years) with in-hospital mortality by a conditional logistic regression model stratified by diagnosis of cancer and hospital stay in a medical or surgical ward.
Older patients (44.4%) were more frequently at nutritional risk than younger patients (39.7% v. 49.7%, p < 0.001). The in-hospital mortality rate was 4.1% (n = 52; 95% confidence interval [CI] 3.0–5.2). The rate was 1.5% (95% CI 0.7–2.7) in patients with normal nutritional status and 7.1% (95% CI 5.1–9.4) in malnourished patients.
The crude odds ratio (OR) for inhospital death was 4.27 (95% CI 2.09–8.75, p < 0.001). There was a significant interaction between age and nutritional risk (p = 0.023), with age identified as a strong effect modifier. Indeed, in patients aged 65 years or less, mortality was 0.5% (95% CI 0.0–1.7) in the absence of malnutrition and 7.0% (95% CI 4.4–10.3) if malnutrition was present. The OR for death was 11.91 (95% CI 2.74–51.79), p < 0.001). Conversely, in patients aged more than 65 years, the difference in mortality was much lower (3.0%, 95% CI 1.3–5.7 v. 7.1%, 95% CI 2.5–10.2), and the OR for death was 2.19 (95% CI 0.91–5.24, p = 0.07).
Our data emphasize the importance of nutritional screening at the time of hospital admission, because malnutrition could represent a strong predictor of mortality, especially in nonelderly patients.
Further multivariable studies with adequate sample sizes are needed to investigate the association among malnutrition, age and in-hospital mortality.