Low Albumin Levels Are Associated with Mortality Risk in Hospitalized Patients.

BACKGROUND: The aim of this study was to investigate the association of albumin levels on admission and change in levels during hospitalization with hospitalization outcomes.

METHODS: Historical prospective data of patients hospitalized between 2011 and 2013 were collected. Levels of albumin were classified as marked hypoalbuminemia (<2.5 mg/dL), mild hypoalbuminemia (2.5-3.5 mg/dL), normal albumin (3.5-4.5 mg/dL), and hyperalbuminemia (>4.5 mg/dL). Main outcomes were length of hospitalization, in-hospital mortality, and long-term mortality.

RESULTS: The cohort included 30,732 patients (mean age 67 ± 18 years, 51% male). Most patients had normal albumin levels on admission (n = 20,124, 65%), 29% of patients had hypoalbuminemia, mostly mild (n = 7,334, 24%), and 5% of patients had marked hypoalbuminemia (n = 1436). Hyperalbuminemia on admission was evident in 6% of the patients (n = 1838). Follow-up (median ± standard deviation) was 1675 ± 325 days. Compared with in-hospital mortality with normal albumin on admission (2%), mortality was higher with mild (12%) and marked hypoalbuminemia (34%) and lower with hyperalbuminemia (0.3%). Mortality rate at the end of follow-up was 29% with normal albumin levels, 67% and 83% with mild and marked hypoalbuminemia, respectively. Patients with hyperalbuminemia on admission and before discharge have the best short- and long-term survival. This pattern was similar when analyzed separately in different age groups. In patients with hypoalbuminemia on admission, normalization of albumin levels before discharge was associated with better short- and long-term survival, compared with patients with hypoalbuminemia before discharge.

CONCLUSIONS: Low albumin levels on admission are associated with increased short- and long-term mortality. Normalization of albumin levels before discharge was associated with lower mortality risk, compared with hypoalbuminemia before discharge.