Recent updates in factors associated with incidence and screening of diabetic eye disease

Abstract

Diabetic eye disease remains a major cause of visual impairment and blindness in many countries, and presents a significant public health challenge worldwide. The global prevalence of retinopathy among people with diabetes is approximately 22%, including 6% of vision-threatening diabetic retinopathy, and 4% of macular edema cases, and the burden is expected to remain high until 2045¹. Similar to other complications of diabetes mellitus, numerous studies have been conducted worldwide to assess the risk factors for diabetic eye disease. Several major factors have been identified, as well as validated in a recent study². The two factors, glycemic control and diabetes duration, are commonly reported in most studies.

In terms of glycemic control, most previous studies have used the levels of fasting plasma glucose or glycosylated hemoglobin as indices. However, recent advances in continuous glucose monitoring have made it possible to analyze the daily trajectory of glucose levels, including postprandial and nocturnal glucose levels. Standardized ‘time in range’³ and glycemic variability⁴ have been previously associated with retinopathy. However, since the influence of glycemia depends on both the duration and the glucose levels, indices that reflect both variables are expected to predict retinopathy better than the conventional glycemic indices. A recent attempt using the ‘area under the curve in range’ was unable to demonstrate a significant association with retinopathy⁵. Nevertheless, further research of similar indices for detailed glycemic status using the new technology is expected.

Factors related to the duration of diabetes should also be further detailed, including the effects of aging. A recent study that divided participants into early- or late-onset diabetes demonstrated that an earlier onset of diabetes is more predictive of microvascular complications, including retinopathy, than late-onset diabetes⁶. In addition to the age of diabetes onset, aging of the entire population is also related to the phenotype and the pathophysiology of diabetic retinopathy. In particular, diabetic macular edema is becoming increasingly prevalent in the older population of many developed countries; however, its treatment in clinical settings is insufficient⁷, which needs to be addressed in the future.

The association between diabetic retinopathy and other diabetic complications, as well as other morbidities and mortality, has been widely investigated⁸. Interestingly, a recent meta-analysis⁹ revealed a lack of significant association between non-alcoholic fatty liver disease (NAFLD) and retinopathy, despite sharing common pathophysiological backgrounds, such as insulin resistance. Recent studies on NAFLD and metabolic dysfunction-associated fatty liver disease (MAFLD) revealed that MAFLD has a stronger association with cardiovascular complications than NAFLD¹⁰; this might be helpful when reanalyzing the association between fatty liver and diabetic eye disease.

Even though vascular endothelial growth factor (VEGF) has long been established as an etiology and a major therapeutic target, other humoral or hormonal factors are being investigated as unknown residual factors that affect the initiation and progression of the disease. Recently, serum levels of progesterone¹¹, parathyroid hormone¹², and vaspin¹³ have been reported to be significantly associated with retinopathy in Chinese people with type 2 diabetes. Since all these studies were cross-sectional, further longitudinal investigation with careful adjustment for potential confounders is needed to elucidate the detailed pathogenesis of retinopathy to aid in treating patients that could not be sufficiently controlled with current anti-VEGF therapy. Progesterone¹¹, as we demonstrated nearly three decades ago, stimulates VEGF secretion in vitro within the physiological range during pregnancy, which is not the case for estrogen¹⁴, suggesting that progesterone could play a role in the clinical worsening of retinopathy during pregnancy. The previous reports indicate that low blood 25-hydroxyvitamin D levels are associated with an increased risk of macrovascular and microvascular complication events in type 2 diabetes¹⁵. Considering this, the preliminary finding that elevated parathyroid hormone levels, even within the normal range, were associated with a higher incidence of diabetic retinopathy¹² suggests a potential impact of calcium regulation or metabolism on diabetic eye disease.

Screening for risk factors is another essential step in the prevention of vision loss due to diabetic eye disease. Although screening using artificial intelligence-based image analysis has already been put to practical use¹⁶, a more affordable device that screens high-risk populations could still be of help in diabetes management. For example, crossing capillaries in the fingernail fold were recently shown to be associated with retinopathy in persons with diabetes¹⁷, although validation with many more patients is required to determine sensitivity and specificity. Simultaneously, sociological interventions that promote screening, such as utilizing financial incentives for patients¹⁸ and care providers¹⁹, should also be pursued in daily practice. In addition, the role of the doctor–patient relationship in reducing the risk of retinopathy has been reaffirmed¹⁹.

In the future, in addition to conventional cohorts and patient registries, the utility of large real-world data, such as health insurance claim data, has greater potential in the search for new risk factors or predictors²⁰, as well as in developing efficient and effective strategies for screening to reduce the number of people suffering from vision impairment.

The author declared no conflict of interest.