Serum Albumin as a Marker of Microvascular Burden in Diabetic Retinopathy: Mediation by Diabetic Nephropathy in Uncontrolled Diabetes.

Abstract

Purpose: This study evaluated whether low serum albumin functions as an independent risk marker for diabetic retinopathy (DR) or primarily reflects renal microvascular damage, which frequently co-occurs with DR. We also re-examined the neutrophil-to-lymphocyte ratio (NLR) as a candidate inflammatory marker.

Methods: Cross-sectional analysis of 1107 adults with diabetes from the National Health and Nutrition Examination Survey (NHANES) 2005-2008. Three sequential logistic regression models were fitted: Model 1 (unadjusted), Model 2 (metabolic covariates), and Model 3 (Model 2 plus log-transformed urinary albumin-to-creatinine ratio [UACR]). A mediation analysis with 2000 bootstrap resamples estimated the proportion of the albumin-DR association attributable to the UACR pathway in participants with HbA1c ≥ 7.0%. Sensitivity analyses applied the NHANES sampling weights, strata, and primary sampling units.

Results: DR was present in 356 of 1107 participants (32.2%; 29.0% after weighting to the U.S. population). Each 1 g/dL increase in serum albumin was associated with a 39% reduction in the odds of DR in the unadjusted model (OR 0.61, 95% CI 0.41-0.89). The association attenuated after adjustment for metabolic and renal covariates (Model 3 OR 0.82, p = 0.399). In the HbA1c ≥ 7.0% subgroup (N = 502), Model 2 yielded an OR of 0.45 per 1 g/dL (95% CI 0.23-0.87, p = 0.019), which attenuated to 0.56 (p = 0.107) once UACR was added. Weighted models gave the same pattern (Model 2 OR 0.41, 95% CI 0.17-0.99, p = 0.047; Model 3 OR 0.44, p = 0.078). The indirect association through UACR was statistically significant (OR 0.79, 95% CI 0.61-0.92; bootstrap p < 0.001), giving a proportion mediated of 34.4%. NLR was not associated with DR in any adjusted model.

Conclusions: In poorly controlled diabetes, low serum albumin behaved as a downstream marker of systemic microvascular leakage, with much of its association with DR mediated by concurrent renal damage. The design is cross-sectional, so the temporal ordering assumed by mediation analysis cannot be verified and these estimates remain associational. Low albumin in this setting should still prompt assessment of the kidney and the retina at the same visit.