Objective: Determine the relationship between albumin levels and all-cause mortality in life insurance applicants.
Method: By use of the Social Security Death Master File, mortality was determined in 1,704,566 insurance applicants for whom blood samples were submitted to Clinical Reference Laboratory. There were 53,211 deaths observed in this healthy adult population during a median follow-up of 12 years. Results were stratified by 6 age-sex groups: females: ages 20 to 49, 50 to 69 and 70+; and males: ages 20 to 49, 50 to 69 and 70+. The middle 50% of albumin values specific to each group was used as the reference band for that group. The mortality in bands representing other percentiles of albumin values higher and lower than the middle 50% were compared to the mortality in the reference band for each age-sex group. The highest percentile bands represent the lowest albumin values.
Results: Relative risk exceeded 150% of each age- and sex-specific reference band for all groups between the 90th and 95th percentile of albumin values. This translates into 150% risk thresholds at approximately 3.8 mg/dL for all females and for males 70+, and 4.1 mg/dL for males ages 20 to 69. Conversely, the highest 25% of albumin values were associated with approximately a 20% reduction in risk in males and a variable 10% reduction in risk in females when compared to the middle 50% of albumin values. Excluding those with total cholesterol < or = 160 mg/dL, or with AST, GGT or alkaline phosphatase elevations, had little impact on relative risk except at the lowest 0.5% of albumin values.
Conclusion: When stratified by age and sex, albumin discriminated between all-cause mortality risks in healthy adults at all ages and across a wide range of values independent of other laboratory tests.