ABSTRACT
BACKGROUND: Serum albumin has been suggested to be associated with insulin resistance. We evaluated the association between serum albumin concentration and insulin resistance. We also investigated whether serum albumin level has an independent effect on the development of diabetes. METHODS: In our study, 9,029 subjects without diabetes, who underwent comprehensive health check-ups annually for 5 years, were categorized into tertiles based on their serum albumin levels at baseline. The odds ratio (OR) for the prevalence of insulin resistance, defined as the top quartile of homeostasis model assessment of insulin resistance and the presence of impaired fasting glucose and nonalcoholic fatty liver disease, was evaluated cross-sectionally. Also, the hazard ratio (HR) for incident diabetes was estimated longitudinally, according to the baseline albumin tertiles using Cox proportional hazard analysis respectively. RESULTS: From the lowest to the highest tertile of albumin, the multivariable-adjusted ORs of insulin resistance increased significantly in both men and women. During the mean follow-up period of nearly 4 years, 556 (6.1%) subjects progressed to diabetes. The multivariable-adjusted HR (95% confidence interval [CI]) of diabetes in men were 1, 1.09 (95% CI, 0.86 to 1.40), and 1.10 (95% CI, 0.86 to 1.41), respectively, from the lowest to the highest tertiles of baseline albumin. Corresponding values for women were 1, 1.21 (95% CI, 0.66 to 2.21), and 1.06 (95% CI, 0.56 to 2.02), respectively. CONCLUSION: Our study showed that increased serum albumin level was associated with insulin resistance. However, serum albumin did not have an independent effect on the development of diabetes.