Background: Low serum albumin levels are associated with cardiovascular disease and mortality risk. This study evaluated the predictive value of low serum albumin for all-cause-mortality in a large Viennese patient cohort and investigated sex differences in the association between serum albumin and mortality.
Materials and methods: Serum albumin concentrations of 285 930 patients, who attended the General Hospital Vienna between 1992 and 2002, were evaluated and linked with the Austrian Death Registry. The median observation period was 7.4 +/- 4.0 years and the death rate was 16.8%. For Cox regression analysis, albumin levels were divided into deciles, the highest category served as reference value. To analyse associations between albumin and mortality independent of liver function, results were adjusted for cholinesterase, which indicates protein synthesis capacity of the liver.
Results: Hazard ratios for all-cause-mortality increased linearly with decreasing albumin levels from 1.05 in the 9th to 2.98 in the 1st decile. Adjusted for cholinesterase, the relative risk for mortality was still 1.91 in the lowest category. Compared with women, men had an average 50% increased risk of death in almost every decile, adjusting for cholinesterase reduced the sex difference to a 10-20% higher mortality risk for men. In critically ill patients, hazard ratios for all-cause-mortality ranged from 4.5 in the 9th decile to 9.5 in the lowest albumin category.
Conclusion: This study demonstrates a strong inverse association between serum albumin and mortality in a large patient cohort. The predictive value of low albumin was remarkably higher in men than in women.