Insulin, Hyperglycemia, and Severe Retinopathy of Prematurity in Extremely Low-Birth-Weight Infants

• Published in: American journal of perinatology Vol. 33; no. 4; p. 393
• Main Authors: Lee, Jan Hau, Hornik, Christoph P, Testoni, Daniela, Laughon, Matthew M, Cotten, C Michael, Maldonado, Ramiro S, Belcastro, Marc R, Clark, Reese H, Smith, P Brian
• Format: Journal Article
• Language: English
• Published: United States 01.03.2016
• Subjects: Apgar Score; Blood Glucose - analysis; Databases, Factual; Female; Gestational Age; Humans; Hyperglycemia - epidemiology; Infant, Extremely Low Birth Weight; Infant, Newborn; Infant, Premature; Insulin - adverse effects; Logistic Models; Male; Multivariate Analysis; Odds Ratio; Retinopathy of Prematurity - epidemiology; Retrospective Studies; Risk Factors
• ISSN: 1098-8785
• DOI: 10.1055/s-0035-1565999

This study aims to determine the association between hyperglycemia, insulin therapy, and severe retinopathy of prematurity (ROP) in extremely low-birth-weight (ELBW) infants. In this retrospective database study, we included all ELBW infants who were ≤ 32 weeks gestational age (GA). We excluded infants without any ophthalmology evaluation and infants who died before 28 days of life. A multivariable model was constructed to determine the association between hyperglycemia, insulin use, and severe ROP. We defined hyperglycemia as blood glucose (BG) > 180 mg/dL. Covariates were GA, small for GA status, discharge year, sex, Apgar score at 5 minutes, mechanical ventilation, oxygen use, bacteremia, and postnatal steroid exposure. We defined severe ROP as ROP requiring bevacizumab, cryotherapy, laser therapy, or vitrectomy. Sensitivity analysis using BG > 150 mg/dL and > 200 mg/dL was performed. A total of 24,548 infants were included; 2,547 (10%) had severe ROP. Hyperglycemia alone was not associated with severe ROP (odds ratio [OR], 0.88; 95% confidence interval [CI], 0.66-1.17). Hyperglycemia and insulin use were not associated with severe ROP (OR, 1.43; 95% CI, 0.91-2.23). BG > 150 mg/dL and insulin use were associated with severe ROP (OR, 1.34; 95% CI, 1.02-1.76). Hyperglycemia alone was not associated with severe ROP in ELBW infants. However, we did observe a possible trend between the use of insulin and severe ROP.