Abstract
Objective
To estimate the prevalence and factors associated with retinopathy of prematurity (ROP) in preterm infants at a tertiary hospital in Central-West Brazil.
Methods
This retrospective, observational cohort study analyzed the medical records of neonates with a gestational age (GA) ≤ 32 weeks and/or birth weight ≤ 1,500 g assisted between March 2022 and February 2024. Statistical analyses included the Mann–Whitney U test, Pearson’s chi-square test, and Fisher’s exact test, followed by multivariable logistic regression with Least Absolute Shrinkage and Selection Operator (LASSO) penalization.
Results
Among the 208 eligible preterm infants included in the study, 40 developed ROP, corresponding to a prevalence of 19.2%, where stage 2 was the most frequent, 17 (42.5%). Treatment was required in 25% of ROP cases, with no progression to stages 4–5. ROP was more frequent among extremely preterm infants and those with extremely low birth weight (p < 0.001). In the adjusted analysis, extreme prematurity and female sex were independently associated with ROP, whereas jaundice was inversely associated with the outcome.
Conclusions
Extreme prematurity and female sex were independently associated with ROP in this cohort of preterm infants, whereas jaundice showed an inverse association with the outcome. These findings reinforce the multifactorial nature of ROP and the importance of continuous surveillance in high-risk preterm newborns.
Keywords: Premature birth; Infant, premature, risk factors; Prevalence; Retinopathy of prematurity
Introduction
Retinopathy of prematurity (ROP) represents one of the leading causes of preventable childhood blindness worldwide and is the primary cause in preterm infants due to incomplete retinal vascular development in these patients [1]. Its pathogenesis is strongly related to supplemental oxygen exposure and dysregulation of vascular endothelial growth factor (VEGF) in the retina. Consequently, retinal metabolism increases, and in the presence of deficient vascularization, hypoxia and retinal detachment may occur [1, 2]. Advances in neonatal care have improved the survival of extremely low-birth-weight newborns, but have also been accompanied by a progressive increase in the incidence of this complication, giving rise to the so-called “third epidemic” of ROP in middle-income countries [2, 3].
The global prevalence of ROP varies significantly among different populations, with a recent meta-analysis demonstrating a pooled rate of 31.9% among preterm infants and 7.5% corresponding to severe forms [4]. An American study by Bhatnagar et al. [5] indicated a relative increase of 86% in ROP prevalence between 2003 and 2019, highlighting the need for continuous updates on the disease [6]. In the Brazilian context, previous multicenter studies reported prevalences between 21.4% and 33.9% [7–9], demonstrating regional variability that demands specific investigation for different national centers.
Classic risk factors for ROP include reduced gestational age, low birth weight, and prolonged oxygen exposure [1, 5, 6, 10]. However, emerging evidence suggests a growing role of maternal obstetric factors and perinatal inflammatory complications in the disease’s pathogenesis [11, 12]. The inflammatory paradigm in ROP has gained prominence, with studies demonstrating associations between neonatal infections, inflammatory mediators, and the development of more severe forms of the disease [12].
In Brazil, guidelines for the screening and treatment of retinopathy of prematurity have been established by the Brazilian Group for the Study of Retinopathy of Prematurity, which recommends screening newborns with a gestational age of ≤ 32 weeks or a birth weight of ≤ 1,500 g [13]. The epidemiological characterization of the disease across different healthcare settings is important for improving screening strategies and clinical follow-up. In this context, the present study aimed to estimate the prevalence of retinopathy of prematurity at a tertiary referral center in Brazil’s Central-West region, as well as to characterize the local epidemiological profile and identify obstetric and neonatal factors associated with the development of the disease.
Methods
Study design and population
This retrospective, observational cohort study was conducted at the State Women’s Health Clinic (HEMU), a tertiary referral center for high-risk maternal and neonatal care in Goiânia, Goiás, Brazil. The study included the medical records of newborns with a gestational age ≤ 32 weeks and/or a birth weight ≤ 1,500 g who underwent fundoscopic examination at approximately 28 days of life and received care between March 2022 and February 2024. Newborns, who died prior to undergoing ophthalmologic evaluation, were discharged, or were transferred before the examination. Moreover, those with incomplete ophthalmologic records were excluded from the research. The study was approved by the HEMU Research Ethics Committee (CAAE 82938124.8.0000.5080) and was conducted in accordance with Resolution No. 466/12 of the Brazilian National Health Council.
Obstetric variables included maternal comorbidities, pregnancy complications, alcohol or illicit drug use, peripartum corticosteroid therapy, and mode of delivery. Clinical and neonatal variables included sex, gestational age, Apgar score at birth, classification according to gestational age and birth weight, respiratory support, blood transfusions, and neonatal comorbidities. Neonatal sepsis, jaundice, and other neonatal infections were identified based on clinical and laboratory records. Bronchopulmonary dysplasia (BPD) was identified using clinical and radiologic data, whereas chorioamnionitis and premature rupture of membranes were identified from maternal obstetric records. Patent ductus arteriosus (PDA) was defined according to the available clinical and imaging findings. Neurological manifestations comprised abnormalities documented during clinical evaluation or ancillary testing, and cardiac conditions included congenital heart defects not dependent on ductal patency, primarily atrial septal defects and ventricular septal defects. The total duration of oxygen therapy, positive-pressure ventilatory support, and the number of blood transfusions were obtained from medical records.
ROP was diagnosed based on ophthalmologic examination performed at approximately 28 days of life. Disease classification with respect to stage, retinal zone involved, and the presence of plus disease followed the criteria of the International Classification of Retinopathy of Prematurity (ICROP). Treatment indications were determined according to the ICROP criteria in effect and adopted by the institution during the study period.
Gestational age was determined from information available in the medical records, and priority was given to estimates based on first-trimester ultrasonography and the date of the last menstrual period. In the absence of these data, clinical methods of neonatal assessment, including the Capurro and New Ballard scoring systems, were used.
Statistical analysis
Statistical analyses were performed using Python software, adopting a significance level of 5%. The normality of continuous variables was assessed using the Shapiro–Wilk test. Non-normally distributed data were summarized as median and interquartile range (IQR), and between-group comparisons were performed using the Mann–Whitney U test. Categorical variables were presented as absolute and relative frequencies and compared using Pearson’s chi-square test or Fisher’s exact test, as appropriate. For analyses involving multiple categories with a significant overall chi-square test result, post hoc analyses were conducted using adjusted standardized residuals with Bonferroni correction, and statistically significant categories were identified by the symbol “≠”.
To identify factors independently associated with ROP, a multivariable binary logistic regression model was constructed using Least Absolute Shrinkage and Selection Operator (LASSO) penalization for variable selection, aiming to minimize overfitting. The final model was subsequently refitted using maximum likelihood estimation to obtain odds ratios (ORs) and their corresponding 95% confidence intervals (95% CIs). Model performance was assessed using Nagelkerke’s coefficient of determination (R²), the area under the receiver operating characteristic curve (AUC) with bootstrap validation, and the Hosmer–Lemeshow goodness-of-fit test for calibration. A p-value < 0.05 was considered statistically significant.
Results
Among the 208 preterm neonates included in the analysis, 40 (19.2%) developed ROP. Among diagnosed cases, stage 2 disease was the most common, affecting 17 neonates (42.5%), followed by stage 1 disease in 16 cases (40.0%). Stage 3 disease was identified in 7 neonates (17.5%), whereas no cases of stage 4 or 5 disease were observed. Regarding retinal zone classification, zone 2 involvement predominated, occurring in 35 cases (87.5%), while zone 3 involvement was observed in 5 neonates (12.5%). No cases involved zone 1. Plus disease was present in 5 cases (12.5%), and treatment was indicated for 10 neonates (25.0%).
Neonates with ROP had a lower gestational age [27.2 weeks (IQR, 26.0–28.0) vs. 30.0 weeks (IQR, 28.7–31.3); p < 0.001] and lower birth weight [890.0 g (IQR, 780.0–992.5) vs. 1,272.5 g (IQR, 1,047.5–1,602.5); p < 0.001] than those without ROP. Extreme prematurity and lower birth weight categories were more frequent among neonates with ROP (p < 0.001). Female sex was also more common in the ROP group (65.0% vs. 47.6%; p = 0.048). No statistically significant differences were observed with respect to mode of delivery, method of gestational age estimation, multiple gestation, Intergrowth classification, or Apgar scores (Table 1).
Table 1.
Characterization of the epidemiological profile of the sample according to ROP prevalence (n = 208)
| ROP | Total | p | ||
|---|---|---|---|---|
| No n = 168 |
Yes n = 40 |
|||
| Gestational age | ||||
| Weeks (Median (IQR)) | 30.0 (28.7–31.3) | 27.2 (26.0–28.0) | 29.5 (28.0-31.1) | < 0.001a |
| Gestational age n (%) | ||||
| Moderate prematurity | 54 (32.1)≠ | 0 (0.0) | 54 (26.0) | < 0.001b |
| Very preterm | 91 (54.2)≠ | 13 (32.5) | 104 (50.0) | |
| Extremely preterm | 23 (13.7) | 27 (67.5)≠ | 50 (24.0) | |
| Sex n (%) | ||||
| Female | 80 (47.6) | 26 (65.0) | 106 (51.0) | 0.048b |
| Male | 88 (52.4) | 14 (35.0) | 102 (4.0) | |
| Type of delivery n (%) | ||||
| Cesarean | 114 (67.9) | 28 (70.0) | 142 (68.3) | 0.794b |
| Vaginal | 54 (32.1) | 12 (30.0) | 66 (31.7) | |
| Method n (%) | ||||
| Capurro | 10 (6.0) | 0 (0.0) | 10 (4.8) | 0.541b |
| LMP | 1 (0.6) | 0 (0.0) | 1 (0.5) | |
| New Ballard | 7 (4.2) | 1 (2.5) | 8 (3.8) | |
| 1st trimester US | 124 (73.8) | 32 (80.0) | 156 (75.0) | |
| Later US | 26 (15.5) | 7 (17.5) | 33 (15.9) | |
| Twin pregnancy n (%) | ||||
| No | 139 (82.7) | 29 (72.5) | 168 (80.8) | 0.140b |
| Yes | 29 (17.3) | 11 (27.5) | 40 (19.2) | |
| Birth weight | ||||
| Grams (Median (IQR)) | 1272.5 (1047.5-1602.5) | 890.0 (780.0-992.5) | 1172.5 (960.0-1480.0) | < 0.001a |
| Birth weight n (%) | ||||
| LBW | 49 (29.2)≠ | 0 (0.0) | 49 (23.6) | < 0.001b |
| ELBW | 36 (21.4) | 34 (85.0)≠ | 70 (33.7) | |
| VLBW | 83 (49.9)≠ | 6 (15.0) | 89 (42.8) | |
| Intergrowth n (%) | ||||
| AGA | 149 (88.7) | 33 (82.5) | 182 (87.5) | 0.361b |
| LGA | 1 (0.6) | 0 (0.0) | 1 (0.5) | |
| SGA | 18 (10.7) | 7 (17.5) | 25 (12.0) | |
| APGAR (Median (IQR)) | ||||
| 1st min | 7.0 (6.0–8.0) | 7.0 (6.0-7.2) | 7.0 (6.08.0) | 0.086a |
| 5th min | 9.0 (8.0–9.0) | 8.0 (8.0–9.0) | 9.0 (8.0–9.0) | 0.505a |
IQR Interquartile range, LMP last menstrual period, US ultrasound, LBW low birth weight, ELBW extremely low birth weight, VLBW very low birth weight, AGA appropriate for gestational age, LGA large for gestational age, SGA small for gestational age
a: Mann-Whitney U: b: Pearson´s chi-square test; n: absolute frequency; %: relative frequency
≠ Categories with significant adjusted standardized residuals in post hoc analysis after Bonferroni correction
As shown in Table 2, various maternal and obstetric complications were associated with ROP. Premature rupture of membranes and chorioamnionitis were more frequent among neonates with ROP (p < 0.05). Cervical insufficiency was also more common in this group, although the association was of borderline statistical significance and based on a small number of cases. No significant differences were observed between groups with respect to the remaining maternal comorbidities and obstetric complications.
Table 2.
Characterization of the maternal epidemiological profile according to ROP prevalence (n = 208)
| ROP | Total | p* | ||
|---|---|---|---|---|
| No n = 168 |
Yes n = 40 |
|||
| Maternal Comorbidities n (%) | ||||
| CAH | 17 (10.1) | 5 (12.5) | 22 (10.6) | 0.774c |
| Epilepsy | 3 (1.8) | 0 (0.0) | 3 (1.4) | 1.000c |
| T1DM | 1 (0.6) | 1 (2.5) | 2 (1.0) | 0.348c |
| T2DM | 6 (3.6) | 0 (0.0) | 6 (2.9) | 0.598c |
| Hypothyroidism | 4 (2.4) | 0 (0.0) | 4 (1.9) | 1.000c |
| Obesity | 3 (1.8) | 1 (2.5) | 4 (1.9) | 0.577c |
| Cervical incompetence | 2 (1.2) | 3 (7.5) | 5 (2.4) | 0.050c |
| Heart disease | 2 (1.2) | 1 (2.5) | 3 (1.4) | 0.475c |
| Others | 17 (10.1) | 3 (7.5) | 20 (9.6) | 0.771c |
| Maternal complications n (%) | ||||
| PIGH | 31 (18.5) | 4 (10.0) | 35 (16.8) | 0.199b |
| PE/HELLP syndrome | 35 (20.8) | 9 (22.5) | 44 (21.2) | 0.817b |
| GDM | 22 (13.1) | 7 (17.5) | 29 (13.9) | 0.470b |
| PROM | 34 (20.2) | 14 (35.0) | 48 (23.1) | 0.046b |
| Chorioamnionitis | 17 (10.1) | 9 (22.5) | 26 (12.5) | 0.033b |
| Infectious diseases | 59 (35.1) | 12 (30.0) | 71 (34.1) | 0.539b |
| Others | 56 (33.3) | 7 (17.5) | 63 (30.3) | 0.050b |
| Alcohol use n (%) | ||||
| No | 145 (86.3) | 38 (95.0) | 183 (88.0) | 0.177c |
| Yes | 23 (13.7) | 2 (5.0) | 25 (12.0) | |
| Smoking n (%) | ||||
| No | 145 (86.3) | 38 (95.0) | 183 (88.0) | 0.380c |
| Yes | 23 (13.7) | 2 (5.0) | 25 (12.0) | |
| Illicit drug use n (%) | ||||
| No | 159 (94.6) | 40 (100.0) | 199 (95.7) | 0.212c |
| Yes | 9 (5.4) | 0 (0.0) | 9 (4.3) | |
| Peripartum corticosteroid use n (%) | ||||
| No | 18 (10.7) | 5 (12.5) | 23 (11.1) | 0.780c |
| Yes | 150 (89.3) | 35 (87.5) | 185 (88.9) | |
| Peripartum corticosteroid doses n (%) | ||||
| 1 | 39 (26.4) | 8 (22.9) | 47 (25.7) | 0.196b |
| 2 | 85 (57.4) | 25 (71.4) | 110 (60.1) | |
| 3 a 4 | 24 (16.2) | 2 (5.7) | 26 (14.2) | |
IQR Interquartile range, CAH chronic arterial hypertension, GDM gestational diabetes mellitus, PE pre-eclampsia, PIGH pregnancy-induced gestational hypertension, PROM premature rupture of membranes, T1DM type 1 diabetes mellitus, T2DM type 2 diabetes mellitus
b: Pearson´s chi-square test; c:Fisher´s exact test; n: absolute frequency; %: relative frequency
Neonates with ROP had a higher frequency of bronchopulmonary dysplasia (BPD), neurological manifestations, patent ductus arteriosus (PDA), and neonatal sepsis, as well as a longer duration of oxygen therapy and a greater number of blood transfusions compared with those without ROP (p < 0.05). In contrast, jaundice was less frequent among neonates with ROP (p < 0.001). Cardiac abnormalities were observed only in the non-ROP group. However, this finding should be interpreted with caution given the small number of cases. No statistically significant differences were observed for the remaining neonatal variables analyzed (Table 3).
Table 3.
Characterization of neonatal risk factors according to ROP prevalence (n = 208)
| ROP | Total | p | ||
|---|---|---|---|---|
| No n = 168 |
Yes n = 40 |
|||
| Neonatal complications n (%) | ||||
| RDS | 147 (87.5) | 36 (90.0) | 183 (88.0) | 0.792c |
| Jaundice | 145 (86.3) | 24 (60.0) | 169 (81.3) | < 0.001b |
| Anemia | 85 (50.6) | 26 (65.0) | 111 (53.4) | 0.101b |
| Other infections | 69 (41.1) | 17 (42.5) | 86 (41.3) | 0.869b |
| BPD | 35 (20.8) | 23 (57.5) | 58 (27.9) | < 0.001b |
| Heart diseases | 18 (10.7) | 0 (0.0) | 18 (8.7) | 0.027c |
| Abdominal wall defects | 15 (8.9) | 5 (12.5) | 20 (9.6) | 0.550c |
| GHM | 24 (14.3) | 7 (17.5) | 31 (14.9) | 0.608b |
| Neurological manifestations | 42 (25.0) | 20 (50.0) | 62 (29.8) | 0.002b |
| Total oxygenation | ||||
| Time (Median (IQR)) | 25.5 (7.7–40.0) | 64.0 (44.0–91.0) | 28.0 (10.0-46.2) | < 0.001a |
| Oxygenation with PP n (%) | ||||
| No | 31 (18.5) | 5 (12.5) | 36 (17.3) | 0.371b |
| Yes | 137 (81.5) | 35 (87.5) | 172 (82.7) | |
| Blood transfusions | ||||
| Number (Median (IQR)) | 1.0 (0.0–2.0) | 2.0 (1.0–5.0) | 1.0 (0.0–2.0) | < 0.001* |
| IVH n (%) | ||||
| No | 128 (76.2) | 28 (70.0) | 156 (75.0) | 0.416b |
| Yes | 40 (23.8) | 12 (30.0) | 52 (25.0) | |
| IVH grade n (%) | ||||
| Grade I | 20 (50.0) | 4 (33.3) | 24 (46.2) | 0.249b |
| Grade II | 13 (32.5) | 5 (41.7) | 18 (34.6) | |
| Grade III | 7 (17.5) | 2 (16.7) | 9 (17.3) | |
| Grade IV | 0 (0.0) | 1 (8.3) | 1 (1.9) | |
| PDA n (%) | ||||
| No | 149 (88.7) | 24 (60.0) | 173 (83.2) | < 0.001b |
| Yes | 19 (11.3) | 16 (40.0) | 35 (16.8) | |
| Necrotizing enterocolitis n (%) | ||||
| No | 156 (92.9) | 37 (92.5) | 193 (92.8) | 1.000c |
| Yes | 12 (7.1) | 3 (7.5) | 15 (7.2) | |
| Neonatal sepsis n (%) | ||||
| No | 80 (47.6) | 7 (17.5) | 87 (41.8) | < 0.001b |
| Yes | 88 (52.4) | 33 (82.5) | 121 (58.2) | |
IQR Interquartile range, BPD bronchopulmonary dysplasia, GMH germinal matrix hemorrhage, IVH intraventricular hemorrhage, PDA patent ductus arteriosus, PP positive pressure, RDS respiratory distress syndrome
a: Mann-Whitney U; b: Pearson´s chi-square test;c:Fisher´s exact test; n: absolute frequency; %: relative frequency;
LASSO-penalized logistic regression identified female sex, extreme prematurity, neonatal jaundice, and neonatal sepsis as variables included in the final model. After adjustment, female sex and extreme prematurity remained independently associated with increased odds of ROP, whereas neonatal jaundice was inversely associated with the outcome. Neonatal sepsis was retained in the model but did not reach statistical significance. The model demonstrated moderate explanatory ability, high discriminative performance, and adequate calibration (Table 4).
Table 4.
Factors independently associated with the presence of ROP (n = 208)
| Predictor variable | b | Wald | OR (CI 95%) | p |
|---|---|---|---|---|
| Sex (Female) | 1.01 | 4.58 | 2.75 (1.09–6.95) | 0.032 |
| Extreme prematurity (Yes) | 2.43 | 19.02 | 11.33 (3.80–33.71) | < 0.001 |
| Neonatal jaundice (Yes) | -1.52 | 8.19 | 0.22 (0.08–0.62) | 0.004 |
| Neonatal sepsis (Yes) | 1.02 | 3.29 | 2.76 (0.92–8.29) | 0.070 |
Variables excluded before model fitting: congenital heart disease (perfect separation) and birth weight (quasi-collinearity with extreme prematurity). Events per variable (EPV) = 40/4 = 10.0; n = 208, events = 40
LASSO Least Absolute Shrinkage and Selection Operator, L1-regularized logistic regression with variable selection using the λ.1se criterion (10-fold stratified cross-validation), OR odds ratio, CI 95% confidence interval
b, logistic regression coefficient; Nagelkerke R² = 0.474; AUC, area under the receiver operating characteristic (ROC) curve = 0.882 (bootstrap 95% CI, 0.820–0.944); bootstrap-corrected AUC = 0.874; Hosmer–Lemeshow test: χ² = 5.39, df = 7, p = 0.612
Discussion
The prevalence of ROP observed in the present study (19.2%) was lower than that reported in Brazilian studies involving high-risk preterm neonates. Branco et al. reported a prevalence of 61.4%, whereas Theiss et al. [14] and Gonçalves et al. [15] found prevalences of 37.8% and 44.5%, respectively. Despite this variability across studies, lower gestational age and lower birth weight have consistently been associated with a higher occurrence of ROP in the Brazilian literature [14–16].
The characterization of ROP in the study population revealed a predominantly moderate disease profile, with most cases classified as stage 1 or 2, preferential involvement of zone 2, a low prevalence of plus disease, and no cases of advanced-stage disease. Although the prevalence observed in the present study was lower than that reported in part of the international literature [2], 25% of diagnosed cases required treatment, underscoring the clinical significance of ROP in this population. These findings highlight the importance of systematic screening and specialized follow-up for preterm neonates, particularly in tertiary referral centers that care for high-risk patients. The distribution of ROP stages also differed somewhat from that reported in certain international studies [4, 5], in which stage 1 disease predominated. In contrast, stage 2 disease was the most frequent presentation in our cohort, corroborating the findings reported by Blazon et al. [17].
In the bivariate analyses, premature rupture of membranes and chorioamnionitis were more frequent among neonates who developed ROP. These findings are consistent with the literature, which has reported an association between maternal inflammatory conditions and an increased risk of the disease. Meta-analyses [18, 19] have demonstrated a higher risk of ROP among neonates exposed to chorioamnionitis, particularly when accompanied by funisitis. However, one study [20] found that this association was attenuated after adjustment for gestational age, suggesting that part of the observed effect may reflect the greater prematurity and clinical vulnerability of these neonates. Similarly, premature rupture of membranes has been linked to intrauterine inflammatory exposure and adverse neonatal outcomes, including ROP [21, 22]. Thus, although our findings support a potential role of perinatal inflammatory mechanisms in the pathophysiology of ROP, these associations should be interpreted with caution given the close interplay among maternal inflammation, extreme prematurity, and neonatal illness severity.
Female sex remained independently associated with ROP in the adjusted model. However, the influence of sex on susceptibility to ROP remains controversial in the literature. A recent systematic review and meta-analysis [23] reported a higher risk of severe ROP among male neonates, particularly in high-income countries. Similarly, Hoyek et al. [24] observed a predominance of male infants among those requiring treatment for ROP. In contrast, Branco et al. [16], in a Brazilian cohort from the state of Maranhão, reported a higher prevalence of ROP among female neonates, suggesting that the influence of sex may vary across different populations and healthcare settings. Therefore, the association observed in the present study should be interpreted with caution, given the heterogeneity of findings reported in the literature.
Neonatal sepsis was significantly more frequent among neonates with ROP in the bivariate analysis, although it did not remain statistically significant after multivariable adjustment. This finding is consistent with previous evidence demonstrating an association between neonatal infectious processes and the development of the disease. In a systematic review and meta-analysis, Wang et al. [25] reported that sepsis was associated with both the occurrence of ROP at any stage and severe forms of the disease. Similarly, Boskabadi et al. [26] found a higher frequency of ROP among preterm neonates with sepsis compared with those without this condition. The biological plausibility of this association has been attributed to activation of the systemic inflammatory response, which may interfere with the regulation of growth factors and promote retinal vascular alterations involved in the pathophysiology of ROP [27]. However, sepsis frequently occurs in more premature and clinically vulnerable neonates and often coexists with established risk factors for ROP, such as lower gestational age and low birth weight [26]. Therefore, although sepsis was retained in the final model, the loss of statistical significance after adjustment suggests that part of its association with ROP may reflect its interaction with extreme prematurity and greater neonatal illness severity.
In our bivariate analysis, bronchopulmonary dysplasia (BPD), patent ductus arteriosus (PDA), longer duration of oxygen therapy, and a greater number of blood transfusions were more frequent among neonates with ROP. These findings are consistent with the literature, which has demonstrated an association between these conditions and an increased risk of developing the disease. BPD has been associated with severe ROP even after adjustment for gestational age and birth weight, suggesting shared pathophysiological mechanisms related to impaired vascular development in preterm neonates [28]. Similarly, PDA requiring therapeutic intervention has been identified as a factor associated with more severe forms of ROP [29]. In addition, prolonged oxygen exposure remains one of the principal factors implicated in the pathophysiology of ROP, particularly among neonates with significant respiratory compromise [30]. Likewise, the need for multiple blood transfusions has been associated with an increased risk of ROP, possibly reflecting greater clinical severity and neonatal immaturity [31]. However, these variables frequently coexist in extremely preterm neonates with more complex clinical courses, making it difficult to distinguish independent risk factors from markers of disease severity. Our findings support this interpretation, as these variables were associated with ROP in the bivariate analysis but were not retained as independently associated factors in the final adjusted model.
The inverse association between jaundice and ROP observed in the adjusted model should be interpreted with caution. Although previous studies have not consistently identified jaundice as an independent risk factor for ROP, some evidence suggests a potential antioxidant role of unconjugated bilirubin (UCB) in preterm neonates [32]. UCB has been described as a scavenger of reactive oxygen species and may attenuate oxidative stress at non-cytotoxic concentrations, particularly in vulnerable neonatal tissues [33]. Furthermore, physiological levels of UCB have been associated with increased antioxidant capacity in preterm neonates, whose antioxidant defense systems remain immature [34]. Nevertheless, the association observed in the present study should be considered exploratory, as it may have been influenced by factors related to neonatal illness severity, cumulative oxygen exposure, and other residual confounders not fully captured by the adjusted model. Additional studies are needed to clarify whether this finding reflects a true biological effect or residual confounding.
Several limitations should be acknowledged when interpreting and generalizing the findings of this study. First, the single-center design may not fully reflect the epidemiological diversity of Brazil, particularly given the substantial regional variations across the country. Although the sample of 208 preterm neonates was adequate for the primary analyses, it may have been insufficient to detect associations of smaller magnitude or to support stratified analyses within specific subgroups. In addition, the exclusion of neonates who died, were transferred, or were discharged before undergoing fundoscopic examination may have introduced selection and survival bias, potentially influencing both the estimated prevalence of ROP and the associations observed. Another limitation relates to the retrospective nature of the study, which restricted the availability of certain detailed clinical data. Specifically, the variable jaundice was identified from diagnostic information documented in the medical records, without standardized collection of parameters such as peak serum bilirubin levels, duration of phototherapy, or the need for exchange transfusion. Consequently, the inverse association observed between jaundice and ROP should be interpreted with caution and considered exploratory. Finally, limitations inherent to retrospective studies, including potential inaccuracies, inconsistencies, and incomplete data in medical records, cannot be entirely excluded.
Conclusions
Extreme prematurity and female sex were independently associated with the occurrence of ROP in this cohort of preterm infants from a tertiary referral center in Central-West Brazil, whereas neonatal jaundice showed an inverse association with the outcome. Lower gestational age and lower birth weight were also associated with higher ROP frequency in the bivariate analyses. These findings reinforce the multifactorial nature of ROP and highlight the importance of continuous surveillance in high-risk preterm newborns. Further longitudinal studies with larger samples are needed to better clarify the influence of neonatal and perinatal factors on the development and progression of ROP.
Acknowledgements
Not applicable.
Abbreviations
- ROP
Retinopathy of Prematurity
- ICROP
International Classification of Retinopathy of Prematurity
- GA
Gestational Age
- PROM
Premature Rupture of Membranes
- LBW
Low Birth Weight
- ELBW
Extremely Low Birth Weight
- VLBW
Very Low Birth Weight
- VEGF
Vascular Endothelial Growth Factor
- PDA
Patent Ductus Arteriosus
- IVH
Intraventricular Hemorrhage
- NEC
Necrotizing Enterocolitis
- BPD
Bronchopulmonary Dysplasia
- RDS
Respiratory Distress Syndrome
- GMH
Germinal Matrix Hemorrhage
- PP
Positive Pressure
- LMP
Last Menstrual Period
- LASSO
Least Absolute Shrinkage and Selection Operator
- US
Ultrasound
- AGA
Appropriate for Gestational Age
- LGA
Large for Gestational Age
- SGA
Small for Gestational Age
- CAH
Chronic Arterial Hypertension
- GDM
Gestational Diabetes Mellitus
- PE
Pre-eclampsia
- PIGH
Pregnancy-Induced Gestational Hypertension
- T1DM
Type 1 Diabetes Mellitus
- T2DM
Type 2 Diabetes Mellitus
- UCB
Unconjugated bilirubin
Authors’ contributions
VCJQ: Conceptualization, study design, data curation, formal analysis, data interpretation, writing – original draft, writing – review & editing. ALB: Data curation. MBFG: Writing – original draft, writing – review & editing. PSSC: Writing – original draft, writing – review & editing. All authors read and approved the final manuscript and agree to be accountable for all aspects of the work.
Funding
Paulo S. Costa received research scholarships from the Brazilian National Council for Scientific and Technological Development (CNPq).
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study was conducted in accordance with the ethical principles of the Declaration of Helsinki and in compliance with Resolution 466/12 of the Brazilian National Health Council. Ethical standards regarding respect for human subjects and confidentiality were strictly observed. The study was approved by the Research Ethics Committee of Hospital Estadual da Mulher (HEMU) under CAAE number: 82938124.8.0000.5080. Due to the retrospective nature of the study and the use of anonymized medical record data, the requirement for written informed consent was waived by the Research Ethics Committee.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Kim SJ, Port AD, Swan R, Campbell JP, Chan RVP, Chiang MF. Retinopathy of prematurity: a review of risk factors and their clinical significance. Surv Ophthalmol. 2018;63:618–37. 10.1016/j.survophthal.2018.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gilbert C, Malik ANJ, Nahar N, Das SK, Visser L, Sitati S, et al. Epidemiology of ROP update—Africa is the new frontier. Semin Perinatol. 2019;43:317–22. 10.1053/j.semperi.2019.05.005. [DOI] [PubMed] [Google Scholar]
- 3.Vinekar A, Gilbert C, Dogra M, Kurian M, Shainberg J, Shetty B, et al. The KIDROP model of combining strategies for providing retinopathy of prematurity screening in underserved areas in India using wide-field imaging, telemedicine, and financing incentives. Indian J Ophthalmol. 2014;62:89–97. 10.4103/0301-4738.126178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.García-Serrano JL, Ramírez-Estudillo MC, León-López KL, Protsyk O, Domech-Serrano T, Uberos Fernández J, et al. Global prevalence and severity of retinopathy of prematurity over the last four decades (1985–2021): a systematic review and meta-analysis. Arch Med Res. 2024;55:102967. 10.1016/j.arcmed.2024.102967. [DOI] [PubMed] [Google Scholar]
- 5.Bhatnagar A, Skrehot HC, Bhatt A, Herce H, Weng CY. Epidemiology of retinopathy of prematurity in the US from 2003 to 2019. JAMA Ophthalmol. 2023;141:479–85. 10.1001/jamaophthalmol.2023.0831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Graziano RM, Leone CR. Problemas oftalmológicos mais freqüentes e desenvolvimento visual do pré-termo extremo. J Pediatr (Rio J). 2005;81(Suppl 1):S95–100. 10.1590/S0021-75572005000200012. [DOI] [PubMed] [Google Scholar]
- 7.Graziano RM, Leone CR, Cunha SL, Pinheiro AC. Prevalência da retinopatia da prematuridade em recém-nascidos de muito baixo peso. J Pediatr (Rio J). 2007;83:47–54. 10.2223/JPED.1589. [DOI] [PubMed] [Google Scholar]
- 8.Fortes Filho JB, Eckert GU, Procianoy L, Barros CK, Procianoy RS. Incidence and risk factors for retinopathy of prematurity in very low and in extremely low birth weight infants in a unit-based approach in southern Brazil. Eye (Lond). 2009;23:25–30. 10.1038/sj.eye.6702924. [DOI] [PubMed] [Google Scholar]
- 9.Portes ALF, Baraúna H, Jeveaux G, Monteiro MLR. Perfil clínico e epidemiológico de recém-natos prematuros com muito baixo peso no Rio de Janeiro: estudo de 152 pacientes. Rev Bras Oftalmol. 2010;69:389–94. 10.1590/S0034-72802010000600008. [DOI] [Google Scholar]
- 10.Rashidian P, Karami S, Salehi SA. A review on retinopathy of prematurity. Med Hypothesis Discov Innov Ophthalmol. 2025;13(4):201–12. 10.51329/mehdiophthal1511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Almutairi M, Chechalk K, Deane E, Fox R, Janes A, Maguire-Henry T, et al. Biomarkers in retinopathy of prematurity: a systematic review and meta-analysis. Front Pediatr. 2024;12:1371776. 10.3389/fped.2024.1371776. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhang L, Buonfiglio F, Fieß A, Pfeiffer N, Gericke A. Retinopathy of prematurity-targeting hypoxic and redox signaling pathways. Antioxid (Basel). 2024;13(2):148. 10.3390/antiox13020148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Grupo Brasileiro de Estudo da Retinopatia da Prematuridade. Diretrizes brasileiras de triagem e tratamento da retinopatia da prematuridade. Arq Bras Oftalmol. 2019;82:128–45. 10.5935/0004-2749.20190027. [DOI] [Google Scholar]
- 14.Theiss MB, Grumann Júnior A, Rodrigues MRW. Perfil epidemiológico dos recém-nascidos prematuros com retinopatia da prematuridade no Hospital Regional de São José Dr. Homero de Miranda Gomes. Rev Bras Oftalmol. 2016;75(2):109–14. 10.5935/0034-7280.20160024. [DOI] [Google Scholar]
- 15.Gonçalves E, Násser LS, Martelli DRB, Alkmim IR, Mourão TV, Caldeira AP, et al. Incidence and risk factors for retinopathy of prematurity in a Brazilian reference service. Sao Paulo Med J. 2014;132:85–91. 10.1590/1516-3180.2014.1322544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Branco RCC, Menezes VJM, Avelar JFDC, Vale MS, Marques PF. Perfil epidemiológico de recém-nascidos com retinopatia da prematuridade do Hospital Universitário da Universidade Federal do Maranhão. Resid Pediatr. 2023;13(4):1–8. 10.25060/residpediatr-2023.v13n4-889. [DOI] [Google Scholar]
- 17.Blazon MN, Rezar-Dreindl S, Wassermann L, Neumayer T, Berger A, Stifter E. Retinopathy of prematurity: incidence, risk factors, and treatment outcomes in a tertiary care center. J Clin Med. 2024;13:6926. 10.3390/jcm13216926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Villamor-Martinez E, Cavallaro G, Raffaeli G, Mohammed Rahim OMM, Gulden S, Ghazi AMT, et al. Chorioamnionitis as a risk factor for retinopathy of prematurity: an updated systematic review and meta-analysis. PLoS ONE. 2018;13(10):e0205838. 10.1371/journal.pone.0205838. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.El Emrani S, Smit-Jansen E, Goeman J, Lopriore E, Termote J, Schalij-Delfos N, et al. Histological chorioamnionitis and funisitis as new risk factors for retinopathy of prematurity: a meta-analysis. Am J Perinatol. 2024;41:E3264–73. 10.1055/a-2215-0662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mitra S, Aune D, Speer CP, Saugstad OD. Chorioamnionitis as a risk factor for retinopathy of prematurity: a systematic review and meta-analysis. Neonatology. 2014;105:189–99. 10.1159/000357556. [DOI] [PubMed] [Google Scholar]
- 21.Matyas M. Preterm birth and inflammation. Current topics in caesarean section. IntechOpen; 2021. 10.5772/intechopen.96970. [DOI]
- 22.Decembrino L, Pozzi M, Falcone R, Stronati M. Chorioamnionitis and neonatal outcome: early vs late preterm infants. Ital J Pediatr. 2014;40(Suppl 2):A22. 10.1186/1824-7288-40-S2-A22. [DOI] [Google Scholar]
- 23.Hundscheid TM, Gulden S, Almutairi MF, Bartoš F, Cavallaro G, Villamor E. Sex differences in the risk of retinopathy of prematurity: a systematic review, frequentist and Bayesian meta-analysis, and meta-regression. World J Pediatr. 2024;20:293–303. 10.1007/s12519-023-00775-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hoyek S, Peacker BL, Acaba-Berrocal LA, Al-Khersan H, Zhao YD, Hartnett ME, et al. The male to female ratio in treatment-warranted retinopathy of prematurity: a systematic review and meta-analysis. JAMA Ophthalmol. 2022;140(11):1110–20. 10.1001/jamaophthalmol.2022.3988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wang X, Tang K, Chen L, Cheng S, Xu H. Association between sepsis and retinopathy of prematurity: a systematic review and meta-analysis. BMJ Open. 2019;9:e025440. 10.1136/bmjopen-2018-025440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Boskabadi H, Abrishami M, Shoeibi N, Kakavandi M, Moradi M, Zakerihamidi M. Incidence of retinopathy of prematurity in neonates with neonatal sepsis. Arch Iran Med. 2023;26:248–51. 10.34172/aim.2023.38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dammann O, Stansfield BK. Neonatal sepsis as a cause of retinopathy of prematurity: an etiological explanation. Prog Retin Eye Res. 2024;98:101230. 10.1016/j.preteyeres.2023.101230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Singh JK, Wymore E, Wagner BD, Thevarajah TS, Jung JL, Kinsella J, et al. Relationship between severe bronchopulmonary dysplasia and severe retinopathy of prematurity in premature newborns. J AAPOS. 2019;23:209. e1-209.e7. [DOI] [PubMed] [Google Scholar]
- 29.Kumar P, Sankar MJ, Deorari AK, Azad R, Chandra P, Agarwal R, et al. Risk factors for severe retinopathy of prematurity in preterm low birth weight neonates. Indian J Pediatr. 2011;78:812–6. 10.1007/S12098-011-0363-7. [DOI] [PubMed] [Google Scholar]
- 30.Freitas AM, Mörschbächer R, Thorell MR, Rhoden EL. Incidence and risk factors for retinopathy of prematurity: a retrospective cohort study. Int J Retina Vitreous. 2018;4:20. 10.1186/S40942-018-0125-Z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Thomas K, Shah PS, Canning R, Harrison A, Lee SK, Dow KE. Retinopathy of prematurity: risk factors and variability in Canadian neonatal intensive care units. J Neonatal Perinat Med. 2015;8(3):207–14. 10.3233/NPM-15814128. [DOI] [PubMed] [Google Scholar]
- 32.Fortes Filho JB, Fortes BGB, Tartarella MB, Procianoy RS. Incidence and main risk factors for severe retinopathy of prematurity in infants weighing less than 1000 grams in Brazil. J Trop Pediatr. 2013;59(6):502–6. 10.1093/tropej/fmt036. [DOI] [PubMed] [Google Scholar]
- 33.Endesfelder S, Schmitz T, Bührer C. Bilirubin exerts protective effects on alveolar type II pneumocytes in an in vitro model of oxidative stress. Int J Mol Sci. 2024;25:5323. 10.3390/ijms25105323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Perrone S, Lembo C, Cannavò L, Gitto E. Molecular mechanisms of oxidative stress-related neonatal jaundice. J Biochem Mol Toxicol. 2023;37(8):e23349. 10.1002/jbt.23349. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.