The study detected no significant between-group difference in Type 1 ROP. Because the investigation was observational, the results establish an association rather than demonstrating that L-carnitine supplementation caused the lower incidence. The study examined whether exposure to L-carnitine during the neonatal period was associated with outcomes among preterm infants eligible for ROP screening.

The exposed group comprised 63 infants and the control group 204 infants. Investigators collected baseline demographic information, clinical-treatment data and outcome data, and designated overall ROP incidence, Type 1 ROP and Type 2 ROP as the primary outcomes. They also compared secondary neonatal complications.

Baseline characteristics were comparable between the groups in the reported statistical tests, with all P values above 0.05. Subgroup and interaction analyses were stratified by sex, gestational age and birth weight, while multivariate logistic regression assessed factors associated with Type 2 ROP. Beyond the lower incidence of overall and Type 2 ROP, the L-carnitine-exposed group had a lower blood-transfusion rate than the control group: 49.2% versus 66.2%, with P = 0.015.

No significant differences were detected between the groups for Type 1 ROP, sepsis, bronchopulmonary dysplasia, necrotizing enterocolitis, intraventricular hemorrhage or periventricular leukomalacia; all corresponding P values were above 0.05. Multivariate logistic regression found that early L-carnitine supplementation was independently associated with lower odds of Type 2 ROP, with an odds ratio of 0.253, a 95% confidence interval of 0.085–0.756 and P = 0.014. The reported analysis therefore identifies an association that persisted after adjustment for the factors entered into the model, but it does not remove the potential influence of unmeasured differences between infants or treatment decisions.

Interaction tests indicated that the association between L-carnitine exposure and ROP varied according to gestational age and birth weight, with both interaction P values below 0.05. The potentially stronger association was observed among infants born at 30–32 weeks’ gestation and weighing at least 1,500 grams. No significant interaction with infant sex was found, with P above 0.05.

These analyses do not establish that L-carnitine is specifically protective for the subgroup with the stronger observed association; rather, they show that the relationship differed according to these infant characteristics in the available data. The proposed biological rationale centers on energy metabolism and oxidative stress, both described in the study as key mechanisms in ROP. The findings are also limited to the studied population of preterm infants who met ROP screening criteria, and the group receiving supplementation was smaller than the control group.

The authors concluded that the results support an association between early prophylactic L-carnitine supplementation and reduced risks of overall and Type 2 ROP, with no detectable association for Type 1 ROP. They called for large-scale prospective randomized controlled trials to determine whether the supplement produces a genuine protective effect; the present findings alone cannot establish efficacy or causality.

Subject of Research: Association between early prophylactic L-carnitine supplementation and retinopathy of prematurity in preterm infants

Article Title: The effect of early prophylactic L-carnitine supplementation on retinopathy of prematurity: a retrospective cohort study

Article References: Bai, S., Sun, Z., Li, M., Yu, Z., Shi, C., Xu, F., & Sun, H. (2026). The effect of early prophylactic L-carnitine supplementation on retinopathy of prematurity: a retrospective cohort study. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07770-x

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07770-x

Keywords: L-carnitine, retinopathy of prematurity, preterm infants, neonatal eye disease, retrospective cohort, Type 2 ROP, fatty acid oxidation, gestational age, birth weight, observational study, effect, early