Serum Albumin and Mortality in the BUPA Study

Law M R (BUPA Epidemiological Research Group, Department of Environmental and Preventive Medicine, The Medical College of St Bartholomew's Hospital, Charterhouse Square, London EC1M 6BQ, UK), Morris J K, Wald N J and Hale A K. Serum albumin and mortality in the BUPA study. International Journal of Epidemiology 1994; 23: 38–41. In the British United Provident Association (BUPA) study, a prospective observational study of 21 520 men, the serum albumin of 877 men who died during 10 years of follow-up was compared with that of 877 controls, each matched to a case by age (within 1 year) and date of attendance (within 3 months). There was little overall difference (mean case-control difference=−0.11 g/l, P >0.2) despite the fact that other studies have reported a long-term association between low serum albumin and increased mortality. Cause-specific mortality data showed no association of low albumin with ischaemic heart disease or other circulatory diseases. An inverse association with cancer was confined to the first few years of follow-up and so attributable to pre-clinical cancer lowering both serum albumin itself and serum cholesterol, with which albumin was associated. There was an association of chronic respiratory, neurological, renal, liver and gut diseases with low serum albumin (case-control difference=−1.19 g/l, P <0.001) consistent with the effect of pre-clinical disease lowering serum albumin. Other causes of death showed no association with albumin. Our data do not support a cause and effect association of low serum albumin and mortality.