Prealbumin an independent predictor of mortality/morbidity

Caroline Cassels

December 20, 2005

Dec 20, 2005

Davis, CA - Lower prealbumin (transthyretin) levels are an independent predictor of mortality and hospitalization?particularly due to infection, a new study has found [ 1 ].

Dr George Kaysen (University of California, Davis) and colleagues conducted an epidemiologic analysis of 7815 hemodialysis patients to determine whether prealbumin was independently associated with mortality and morbidity, including all-cause, cardiovascular, infection-related, and vascular-access-related hospitalization.

"We found that as prealbumin levels declined there was a gradual increase in mortality risk independent of serum albumin and other clinical characteristics. We also found that patients with higher prealbumin levels had a greater risk of clotting of vascular-access grafts, Kaysen told renal wire .

In addition, the study found increased mortality and morbidity risk was seen in levels below 40 mg/dL, suggesting optimal prealbumin levels are above 30 to 40 mg/dL.

Their report is published in the December 2005 issue of Kidney International.

Better snapshot of events

Previous research has suggested that the serum prealbumin concentration is independently associated with mortality in hemodialysis patients even with adjustment for serum albumin and other nutritional parameters.

According to Kaysen, the prevailing wisdom is that low albumin levels are a powerful predictor of mortality because of their ability to indicate both nutritional status and the presence of inflammation in end-stage renal-disease (ESRD) patients. He said that serum albumin levels don't usually drop unless there is a combination of malnutrition and inflammation.

"If you're measuring serum albumin, you're measuring a serum protein that is reporting malnutrition, inflammation, or a combination or both," Kaysen said. "Much of albumin's predictive power as it relates to mortality is removed if you include the presence of other acute-phase proteins such as C-reactive protein and interleukin-6. So while serum albumin is associated with mortality and all-cause hospitalization, it was not associated with hospitalization due to infection."

Another potential issue with serum albumin is that it is a long-lived protein and therefore may be reporting events that occurred up to three weeks previously. Prealbumin, on the other hand, is a short-lived negative acute-phase protein and provides a more recent of snapshot of events.

"Albumin turns over very slowly, whereas prealbumin levels change fairly quickly because the concentration is fairly low and the half-life is fairly short," he said.

The study cohort came from a large database of 40¿538 dialysis patients on hemodialysis three times a week as of January 1, 1998. Patients had at least one determination of serum phosphorus and calcium during the last three months of 1997. Of these, 7815 participants, ultimately included in the study cohort, had at least one serum prealbumin concentration measured during this three-month period.

Prealbumin levels were categorized into six categories of 5-mg/dL increments of <15, 15-20, 20-25, 25-30, 35-40 and >40.

Cardiovascular hospitalization causes included acute pulmonary edema, diseases of the circulatory system, fluid overload, symptoms involving the cardiovascular system, chest pain, syncope and collapse, and sudden death.

Infection-related hospital admissions included infection and parasitic disease, meningitis and encephalitis, endocarditis, pneumonia, infections of the kidney, infections of the skin and subcutaneous tissue, and bacteremia. In addition, noninfection vascular-access-related hospitalizations were calculated.

The mean serum prealbumin level was 32.0 mg/dL and was significantly higher in younger patients, men, African Americans, and nondiabetic patients.

The authors report that unadjusted mortality results showed a significantly increased relative risk for all categories of prealbumin below 40 mg/dL. However, even when adjusted for age, sex, race, diabetes, time from initiation of dialysis, and other laboratory findings, "all categories of prealbumin retained a significantly higher relative risk than the referent category."

Adjusted relative mortality risk

Prealbumin levels (mg/dL)
< 15
Relative risk of death

Of 4805 recorded hospitalizations, lower prealbumin concentrations were also associated with an increased risk of all-cause hospitalization. Other predictors of all-cause hospitalization included older age, white race, diabetes, lower body weight, lower hemoglobin, lower bicarbonate and urea reduction ratio (URR), and higher ferritin levels.

A total of 1180 cardiovascular hospitalizations were recorded. The authors report that the unadjusted results of the association between prealbumin levels and cardiovascular hospitalization were of "marginal statistical significance" and the adjusted data showed no significance.

However, lower prealbumin levels were associated with an increased risk of the 567 infection-related hospitalizations.

Adjusted relative risk of hospitalization due to infection

Prealbumin levels (mg/dL)
< 15
Relative risk of infection


A surprise finding of the study, said Kaysen, was that lower prealbumin levels were associated with a decreased risk of vascular-access-related hospitalization?a result that was in direct contrast to albumin, where decreased levels were shown to increase vascular-access morbidity.

"We don't know the reason for this finding, but the association was quite high and likely to be reproducible. At this point it is pure speculation, but it could be that prealbumin levels are more specifically affected by liver disease. So it is possible that these patients are not producing clotting factors at a normal rate, and this may somehow be protecting them against vascular-access-related morbidity," he said.

In fact, he said, following the study, the researchers went back and looked at transaminase levels and found that they were increased in patients with lower prealbumin levels.

Adjusted relative risk of vascular-access-related hospitalization

Prealbumin levels (mg/dL)
< 15
Relative risk of vascular-access-related hospitalization

To download tables as slides, click on slide logo below

Next steps

The next steps in terms of research, said Kaysen, are already under way. In an National Institutes of Health (NIH)-sponsored initiative known as the Comprehensive Dialysis Study, Kaysen and colleagues will develop a database of information on acute-phase proteins based on serum samples from 900 incident dialysis patients.

"This research will give us a longitudinal analysis over a year that will provide us with a better understanding of how prealbumin interacts with long-lived and short-lived acute-phase proteins, correlates with a nutritional questionnaire, and how that relates to clinical outcomes," he said.


  1. Chertow G, Goldstein-Fuchs, Lazarus J, et al. Prealbumin, mortality, and cause-specific hospitalization in hemodialysis patients. Kidney Int 2005; 68: 2794-2800.