Published in Research

Researchers establish link between DR and serum albumin

This is editorially independent content
4 min read

A recent study, published in BMC Ophthalmology reevaluated the relationship between serum albumin (sALB) and the prevalence of diabetic retinopathy (DR).

Give me some background first.

As one of the leading causes of preventable blindness, DR’s prevalence is projected to reach 160.5 million individuals by 2045.

A recent study identified a significant negative correlation between sALB and DR, indicating sALB as a possible biomarker of DR. A prominent protein in plasma, sALB plays an important role in regulation and function.

Go on…

Low sALB levels may increase oxidative stress and inflammatory response, both known factors of DR development. Further, previous studies on this relationship have utilized linear analysis to draw conclusions.

Thus, researchers set out to reevaluate this relationship and investigate if it is nonlinear, which would more closely match outcomes in the clinical research.

Who was included in the study?

Data came from the Dryad database—where users can download raw data from literature—and was utilized from a previously published paper for a secondary analysis.

This was a retrospective, cross-sectional study that included 426 participants, all diagnosed with type 2 diabetes mellitus.

Each patient had visited Guangdong Provincial People’s Hospital’s endocrine department (in Guangzhou, Guangdong Province, China) and received an ophthalmology consultation from December 2017 to November 2018.


Of the participants, 167 were diagnosed with DR, with a prevalence of 39.20%. In terms of association between DR and sALB, a negative correlation was found (p = 0.0008).

Fully adjusted, DR prevalence decreased by 8% for every additional 1 g/L of sALB (p = 0.0474). Following sensitivity analysis however, the increase trend was not significant, pointing to a nonlinear relationship.

Tell me more.

Nonlinear analysis was further examined, and an inflection point was found at 38.10 g/L of sALB, meaning that sALB was negatively associated with DR prevalence below this level.

The association was then not significant when sALB was above 38.10 g/L.

Expert opinion?

Per the study authors, “We discovered a non-linear relationship between sALB and the prevalence of DR with a saturation point close to normal levels after fully adjusting for covariates, which implied that we need to pay much attention to sALB levels in type 2 diabetic patients.”


The data from this research came from a cross-sectional study, and the level of sALB that preceded DR was uncertain. In addition, this study tested for association instead of causation.

Lastly, there was also a possible influence of liver diseases and gastrointestinal bleeding on the association. Analysis took this into account, but the authors suggested caution when applying findings to populations that have either of these conditions.

Take home.

While future studies will be necessary to confirm these findings and address certain limitations, this study found new notions in the nonlinear aspect of the relationship between DR and sALB.

As a negative association was found in line with past studies, the authors suggested that for patients with type 2 diabetes mellitus, sALB levels should be closely monitored.