Objectives: to determine the contribution of the two acute phase proteins alpha 1-acid glycoprotein, and C-reactive protein to the prediction of overall mortality in hospitalised elderly patients.
Design: prospective cohort study.
Setting: a department of geriatric medicine of Charles Richet Hospital, in the Paris-Ile de France area.
Subjects: 433 consecutive patients (mean age 84+/-7 years) admitted for rehabilitation in a department of geriatric medicine.
Methods: clinical and laboratory examinations were performed at baseline. Follow-up ended at hospital discharge or death. Prognostic factors of survival were identified using Cox proportional hazards regression model.
Results: compared with the survivor group, the mean serum levels of both C-reactive protein and alpha 1-acid glycoprotein at baseline were higher in the deceased group (44+/-51 mg/l versus 22+/-34 mg/l and 1691+/-69 mg/l versus 1340+/-456 mg/l respectively; P<0.001 for each). Baseline levels of albumin and prealbumin were significantly lower in patients who died than in patients who survived. In multivariate analysis, alpha 1-acid glycoprotein (but not C-reactive protein), previous stroke, previous heart failure, and age emerged as the only parameters significantly and independently related to overall mortality.
Conclusion: this study shows that the increase in plasma levels at admission of two acute-phase proteins, alpha 1-acid glycoprotein and C-reactive protein, were associated with in-hospital mortality in a population of hospitalised elderly patients. Furthermore, associations of overall mortality with inflammation differed among the markers and only alpha 1-acid glycoprotein entered the multivariate prediction model. Our findings support the hypothesis that alpha 1-acid glycoprotein may be superior to C-reactive protein in mortality risk assessment strategies for elderly patients.