Prevalence and risk factors for referable diabetic retinopathy in patients with type 1 diabetes: a nationwide study in Brazil.

Abstract

Nowadays, there is a tendency to individualize the screening for diabetic retinopathy (DR), based on clinical data such as glycated haemoglobin and duration of diabetes (DCCT/EDIC 2017). These methods allow the identification of patients that require referral to a specialist. Currently, in the epidemiological literature, DR has also been graded as referable and nonreferable (Abràmoff et al. 2013). This letter reports a multicentre, nationwide study, conducted in Brazil, a country with a highly admixed population. The study included 1760 patients with type 1 diabetes that were submitted to a standard questionnaire, clinical and laboratorial tests. The screening for DR was assessed by mydriatic binocular indirect ophthalmoscopy and then categorized as referable (moderate nonproliferative DR, severe nonproliferative DR, proliferative DR and macular oedema) and nonreferable (DR absent and mild DR). Of the 1760 patients recruited at baseline, a total of 1644 (93.4%) performed fundoscopy and were included in the study. The distribution of patients by subgroups according to DR categories was as follows: 296 (18%) referable DR and 1348 (82%) nonreferable DR. The exploratory analyses were performed by t-test or chi-square test and showed that patients with referable DR were older (p < 0.001), had fewer years of formal education (p = 0.01), lower socioeconomic status (p = 0.03), longer duration of diabetes (p < 0.001), higher HbA1c levels (p = 0.005), BMI (p < 0.001), serum uric acid (SUA) (p < 0.001) and triglycerides (p =0.008) and were more likely to be users of an ACE inhibitor (p < 0.001), current smokers (p = 0.03) and hypertensive (p < 0.001); they also presented more frequently macrovascular disease (p < 0.001) or chronic kidney disease (CKD) (p < 0.001) than patients without referable DR. There were no associations with gender, HDL cholesterol or LDL cholesterol. The final adjusted multivariate binomial logistic regression model was statistically significant (p < 0.001). The model explained 31.2% (Nagelkerke R2) of the variance for the presence of referable DR. The highest odds were 2.09 for the use of ACE inhibitor (CI 95% 1.44–3.04; p < 0.001), 1.95 for current smoking (CI 95% 1.09–3.48; p = 0.025) and 1.44 for chronic kidney disease (CI 95% 1.84–1.27; p = 0.001) (Table 1). Longer diabetes duration, poor glycemic control and hypertension are well-stablished risk factors for the development of DR as has been shown in several studies (ADA 2017). One result that needs to be highlighted and discussed is the role of SUA, which has been shown to be a potential marker of diabetic microvascular disease (Krizova et al. 2015). Possible mechanisms that may explain this association include the pro-inflammatory and pro-oxidant effects of SUA, which can lead to endothelial damage. Our results show an association between referable patients with cigarette smoking. Despite the systemic effects of cigarette smoking, which are related to coagulation and inflammatory pathways and lead to hypoxia and endothelial damage, data regarding the relationship between smoking and DR are conflicting. Some studies have shown no association or a negative association between smoking and DR (Thorlund et al. 2013), whereas other studies have shown a positive association. However, our sample included only a small number of smokers (5.6%), and this might have influenced our results. In the current study, the presence of referable DR showed a significant association with CKD. Retinopathy and nephropathy are microvascular complications that may reflect similar vascular changes sharing common baseline mechanisms. In conclusion, our data show a high prevalence of referable retinopathy. Referable retinopathy was associated with traditional risk factors such as duration of diabetes, levels of HbA1c, use of ACE inhibitor, current smokers, and presence of hypertension and CKD but also with serum uric acid. Further studies are needed to confirm and explain this association. It is important to highlight the urgent need for screening programs aiming the prevention of this incapacitating comorbidity.