Less Than One Hypoglycemic Event a Year Can Increase Risk of Diabetic Retinopathy

Diabetes comes with many challenges, and one major risk factor is the possibility of losing sight. Diabetic retinopathy prevention is vital to keep vision intact. A study recently published in JAMA Ophthalmology analyzed one investigation’s data—the Japan Diabetes Optimal Integrated Treatment Study for 3 Major Risk Factors of Cardiovascular Diseases (J-DOIT3)—to look for retinopathy events and their associations. Hypoglycemia may stress the diabetic retina, with transient low glucose stabilizing hypoxia-inducible factor 1a and upregulating VEGF/ANGPTL4. In turn, this promotes inner blood-retinal barrier breakdown and vascular leakage. Photo: Julie Torbit, OD. Click to enlarge. The J-DOIT3 was a multicenter, open-label, parallel-group, randomized clinical trial examining efficacy of intervention on cardiovascular outcomes and mortality in patients with type 2 diabetes aged 45 to 69 with hypertension and/or dyslipidemia; retinopathy events were a secondary endpoint. Of the 2,540 participants (5,080 eyes) who were either randomly assigned to intensive therapy or conventional therapy, mean age was 59 and 38% were women. Participants were monitored for a median of 8.5 years, and it was found that intensive therapy was associated with risk reduction in retinopathy onset but not retinopathy progression. After one year of randomization, HbA1c was linked with onset of retinopathy but with no clear threshold observed. What’s more, risk of onset was elevated by 25% in individuals who had 0.5 hypoglycemic episodes per year or fewer than those without any hypoglycemic episodes; this risk was even greater in those who had more than one episode a year.In their paper, the researchers relay how the increased risk of onset in those with less than one hypoglycemic episode a year is about equivalent to an increase of HbA1c of 1%, which would elevate risk by 31%. Due to this finding, the authors emphasize the importance of achieving strict, yet safe, glycemic management, pointing to the fact that rapid improvement in glycemic management—classified as decreased HbA1c by ≥1% during first year after treatment initiation—could elevate hypoglycemia risk. This reflects the study itself, in which retinopathy incidence was relatively higher in the first year after randomization.  See this feature to know how to react when you diagnose diabetic retinopathy. Thus, the authors believe that “the HbA1c target should be individualized, but glycemic management with an HbA1c target of 7% or lower may be recommended in a certain proportion of people with type 2 diabetes.”1In an invited commentary on the paper, also included in JAMA Ophthalmology, commentators note how the findings of this study may be limited in generalizability. As they explain, retinopathy in this trial was a secondary endpoint. Consequently, issues of multiplicity and limited power, especially for progression—which only had 107 events—raised the likelihood of false negative or borderline signals. Another factor is that retinopathy was assessed via dilated fundus exams, which are less sensitive than standard 7-field stereoscopic fundus photographs. Also, since all participants were Japanese and between ages 45 to 69 with dyslipidemia or hypertension, care patterns and risks may not reflect other populations or settings around the world.However, they do generally agree that “taken together, these findings highlight both the importance and complexity of glycemic management in protecting against diabetic retinopathy.”2 Click here for the journal source and here for the commentary. 1. Effect of a multifactorial intervention on retinopathy in people with type 2 diabetes: a secondary analysis of the J-DOIT3 randomized clinical trial. Sasako T, Ueki K, Miyoshi K, et al. JAMA Ophthalmol. October 23, 2025. [Epub ahead of print].2. Beyond hemoglobin A1c—hypoglycemia, glycemic patterns, and risk of diabetic retinopathy. Giacalone JC, Barkmeier AJ. JAMA Ophthalmol. October 23, 2025. [Epub ahead of print]. This article was developed by the editorial staff in conjunction with experts in the field. In the process, AI may have been among the editorial tools used to meet the goals of human editors, who approved all content.