Objectives.—Determine the relationship between liver function test (LFT) results (GGT, alkaline phosphatase, AST, ALT and albumin) and all-cause mortality in life insurance applicants.
Method.—By use of the Social Security Master Death File, mortality was examined in 15,272,955 insurance applicants for whom blood samples were submitted to the Clinical Reference Laboratory. There were 268,593 deaths observed in this study population, after an average follow-up time of 10.9 years. Results were stratified by sex and by age less/greater than 60, creating 4 groups. Liver function test values were grouped using percentiles of their distribution within these age/ sex groups – so as to update the results generated in prior publications. Additional models were fit using different exclusions and percentile groups within single year age groups. Also, LFTs were treated as continuous variables and included in Cox models with age and smoking status.
Results.—Using the risk of the middle 50% of the population by distribution as a reference, relative mortality observed for GGT and alkaline phosphatase was linear with a steep slope from very low to high values. AST showed a J-shaped association with mortality. ALT showed a low-magnitude inverse correlation with mortality. Albumin demonstrated a higher-magnitude inverse correlation with mortality, especially at values below the median. The overall risk associated with LFTs was durable over at least 10 years of follow-up.
Conclusion.—Liver function tests show a strong and durable correlation to mortality in a large group of insurance applicants. The durability over time suggests that even older values of LFTs found in medical records could be of use in mortality risk prediction.