Abstract
Purpose:
To delineate racial differences in the incidence and time course of ROP in a large cohort of premature infants.
Methods:
The secondary analysis of data from the two Postnatal Growth and ROP Studies (G-ROP-1 and G-ROP-2) that were collected in 41 hospitals in North America from 2006 to 2017. According to self-reported maternal race, premature infants were classified into 3 groups: White (N=5580), Black (N=3252), and Asian (N=353). Incidence, severity, and time course of ROP; plus disease; and postnatal weight gain rate were compared among racial groups.
Results:
Black infants had significantly smaller BW (mean 1035 vs. 1131 vs.1144 grams, P<0.001) and lower GA (28.2 vs. 28.6, vs. 29.1 weeks, P<0.001) than White and Asian infants. However, Black infants had lower incidences of severe ROP (11.1% vs. 12.4% vs. 11.9%), ROP (42.1% vs. 43.2% vs. 30.6%), and plus disease (3.6% vs. 6.3%, vs. 5.9%) than White and Asian infants (BW and GA adjusted risk ratio for Black vs. White 0.69 for severe ROP, 0.83 for ROP, 0.44 for plus disease, all P<0.0001). Mean daily-weight-gain on days of life 11-20 and 21-30 were similar across groups (P>0.05), but lower in Black and Asian infants on days 31-40 (P<0.001). There were no differences in the timing of severe ROP and ROP across racial groups.
Conclusions:
Despite relatively lower GA, BW, and daily-weight-gain, Black preterm infants had lower incidences of ROP and plus disease than White preterm infants. The mechanisms for these differences require further investigation.
INTRODUCTION
Retinopathy of prematurity (ROP), a sight-threatening retinal disease, is a significant cause of childhood visual impairment and blindness in developed countries.1 Among perinatal and postnatal risk factors for ROP in a large, racially and geographically diverse cohort of infants undergoing ROP screening examinations in North America,2 we observed that maternal race was a significant and intriguing risk factor for ROP. Although African-Americans are only 13.6% of the American population3, they constitute 30% to 40% of preterm-born infants.4 In 2021, the rate of preterm birth among African-American women (14.8%) was about 50% higher than the rate of preterm birth among white women (9.5%)5. With a higher incidence of prematurity, one might postulate that Black infants would develop severe ROP at a higher rate than White infants. However, maternal race may be independently associated with ROP6-10, and in three large multi-center ROP studies (CRYO-ROP, ETROP, and e-ROP studies), Black infants had significantly lower incidence of ROP, pre-threshold ROP, or plus disease than White infants.4 As lower gestational age (GA) and smaller birthweight (BW) consistently have been associated with higher incidences of ROP and severe ROP2,10,11, these findings are counterintuitive.
Thus, we sought to delineate the association between race and ROP. According to findings from the previous studies4, we hypothesize that there are significant racial differences in ROP incidence in preterm infants. We specifically aimed to determine 1) giving the same BW and GA, whether Black race is independently associated with a lower incidence of ROP compared to White race; 2) whether postnatal factors such as postnatal weight gain rate may explain differences in ROP incidence among racial groups. Poor postnatal weight-gain is an established postnatal risk factor for ROP development.12 We performed a secondary analysis of the rich datasets of the Postnatal Growth and ROP (G-ROP) Studies, which included large, racially and geographically diverse cohorts of infants in the United States and Canada.
METHODS:
This is a secondary analysis of data from the G-ROP Studies (G-ROP-1 and G-ROP-2). Major features of the G-ROP Studies related to this paper are described below, and further details on the study design and procedures of G-ROP Studies have been reported previously.13-15 G-ROP-1 was a retrospective study of 7483 infants from 29 hospitals, born between January 1, 2006, and December 31, 2011, who underwent ROP examinations and had a known ROP outcome, to develop postnatal weight gain-based modified ROP screening criteria. The G-ROP-2 study was a prospective cohort study of 3,980 premature infants from 41 hospitals between 2015 and 2017 to validate the G-ROP screening criteria. Institutional Review Board approval for the studies was obtained at Children’s Hospital of Philadelphia and all study sites (Appendix), and a waiver of informed consent was granted at each center. All sites adhere to the guidelines of the Declaration of Helsinki.
All infants receiving ROP examinations were eligible for inclusion, without restriction by BW or GA. However, the ROP screening criteria used during that time at the study sites were typically BW less than 1501 g, or GA less than 30 weeks, or an unstable clinical course, as determined by the neonatologist, in an infant with larger BW and/or older GA.14 All included infants had a known ROP outcome defined as Early Treatment of ROP Study Type 1 ROP, Type 2 ROP, or ROP treatment in either eye; or retinal vasculature maturity, immature vasculature extending into zone III without prior disease in zone I or II, or regression of ROP not reaching criteria for Type 1 or 2 ROP in both eyes.14 Furthermore, we defined “severe ROP” as Type 1, Type 2, and “Acute zone III” disease in either eye. “Acute zone III” was defined as stage 3 in zone 3 with or without plus disease.
Statistics Analysis:
Analyses were restricted to three racial groups based on self-reported maternal race recorded in the medical record: White/Caucasian (White), Black/African American (Black), and Asian/Asian American (Asian). Infants not falling into one of these groups were excluded due to unknown racial information or a small number. We compared the incidence and timing of any stage and of severe ROP among the three groups. We used logistic regression models to compare the incidence of ROP and severe ROP and used generalized linear regression models to compare the timing of ROP and severe ROP (in terms of postmenstrual age) among the three groups without and with adjustment by BW, GA and gender. We performed subgroup analyses stratified by GA less than 28 weeks and GA greater than or equal to 28 weeks, because 28 weeks of gestational age is critical for ROP development.16-18 Although BW is another critical factor for ROP development, due to the high correlation between GA and BW, we did not perform stratified analysis by BW. We assessed racial differences in postnatal weight gain by plotting mean daily-weight-gain rate up to 40 days after birth (the critical period we used for ROP risk prediction in the G-ROP studies) stratified by the three groups.13,14 We compared mean daily-weight-gain during three time periods used in developing and validating the G-ROP modified ROP screening criteria 13,14 of the 11th to 20th days, 21st to 30th days and 31st to 40th days of postnatal life across racial groups using generalized linear models both without and with adjustment by BW, GA and gender. We compared incidences of retinal detachment and post-hospital-discharge treatment across racial groups. All statistical analyses were performed using SAS v9.4 (SAS Institute Inc., Cary, NC) and two-sided P<0.05 were considered statistically significant.
RESULTS
A total of 9185 premature infants were analyzed including 5580 White, 3252 Black, and 353 Asian infants (Table 1). Compared to White and Asian infants, Black infants had significantly smaller mean BW (1131 vs. 1144 vs. 1035 grams, P<0.001), lower mean GA (28.6 vs. 29.1 vs. 28.2 weeks, P<0.001), and higher percentage of females (46.7% vs. 47.3% vs. 50.7%, P=0.001). Even when infants were stratified by GA <28 weeks and GA ≥28 weeks, Black infants had smaller mean BW than White and Asian infants with GA<28 weeks (786 vs. 830 vs. 805 respectively, P<0.001), and in infants with GA ≥28 weeks (1256 vs. 1328 vs. 1307 respectively, P<0.001).
Table 1.
Comparison of baseline characteristics among Caucasian, African American and Asian American groups (N=9185)
| White (N=5580) |
Black (N=3252) |
Asian (N=353) |
P | |
|---|---|---|---|---|
| Gestational age (weeks) | ||||
| Mean (SD) | 28.6 (2.6) | 28.2 (2.7) | 29.1 (2.8) | <0.001 |
| Median | 28.7 | 28.0 | 29.1 | |
| Q1, Q3 | 26.7, 30.6 | 26.0, 30.3 | 27.0, 31.1 | |
| Range | (22.0-37.6) | (22.3-38.1) | (22.4-35.1) | |
| 22 | 23 (0.4%) | 7 (0.2%) | 1 (0.3%) | |
| 23 | 160 (2.9%) | 135 (4.2%) | 6 (1.7%) | |
| 24 | 350 (6.3%) | 303 (9.3%) | 23 (6.5%) | |
| 25 | 427 (7.7%) | 325 (10.0%) | 28 (7.9%) | |
| 26 | 578 (10.4%) | 365 (11.2%) | 26 (7.4%) | |
| 27 | 664 (11.9%) | 397 (12.2%) | 31 (8.8%) | |
| 28 | 697 (12.5%) | 406 (12.5%) | 38 (10.8%) | |
| 29 | 730 (13.1%) | 341 (10.5%) | 51 (14.4%) | |
| 30 | 863 (15.5%) | 385 (11.8%) | 40 (11.3%) | |
| 31 | 584 (10.5%) | 321 (9.9%) | 49 (13.9%) | |
| ≥32 | 504 (9.0%) | 267 (8.2%) | 60 (17.0%) | |
| Birth weight (g) | ||||
| Mean (SD) | 1131 (366) | 1035 (332) | 1144 (338) | <0.001 |
| Median | 1110 | 1010 | 1188 | |
| Q1, Q3 | 849, 1389 | 760, 1280 | 866, 1381 | |
| Range | (310-2961) | (360-2480) | (380-2230) | |
| ≤500 | 82 (1.5%) | 65 (2.0%) | 5 (1.4%) | |
| 501-750 | 857 (15.4%) | 713 (21.9%) | 49 (13.9%) | |
| 751-900 | 742 (13.3%) | 521 (16.0%) | 47 (13.3%) | |
| 901-1000 | 527 (9.4%) | 301 (9.3%) | 21 (5.9%) | |
| 1001-1100 | 559 (10.0%) | 311 (9.6%) | 35 (9.9%) | |
| 1101-1250 | 757 (13.6%) | 452 (13.9%) | 46 (13.0%) | |
| ≥1251 | 2056 (36.8%) | 889 (27.3%) | 150 (42.5%) | |
| Sex | 0.001 | |||
| Female | 2604 (46.7%) | 1649 (50.7%) | 167 (47.3%) | |
| Male | 2976 (53.3%) | 1603 (49.3%) | 186 (52.7%) |
SD=Standard deviation, Q1=1st quartile, Q3=3rd quartile
Race and Incidence of Severe ROP
Table 2 shows the incidence of severe ROP for each of the three racial groups among all infants, infants with GA<28 weeks, and infants with GA ≥28 weeks. The incidence of severe ROP was lower in Black than White and Asian infants considering all infants (11.1% vs. 12.4% vs. 11.9%), infants with GA<28 weeks (22.5% vs. 28.8% vs. 33.0%), and infants with GA≥28 weeks (0.9% vs. 1.7% vs. 1.7%). After adjustment by GA, BW and gender, Black infants had a significantly lower risk of severe ROP than White infants for all infants (adjusted risk ratio [aRR]=0.69, P<0.001), for infants with GA<28 weeks (aRR=0.71, P<0.001), and for infants with GA ≥28 weeks (aRR=0.45, P=0.005). There was no significant difference in the risk of severe ROP between Asian and White infants (all P>0.65). Figure 1A shows the incidence of severe ROP for each racial group stratified by GA. For most GA levels, Black infants had a lower rate of severe ROP than White infants. For example, the rates of severe ROP among White, Black and Asian infants were 59.4%, 52.6%, and 33.2% respectively for GA 23 weeks, and 49.1%, 42.6%, and 65.2% respectively for GA 24 weeks.
Table 2.
Association between race and incidence of severe ROP, incidence of any ROP, and incidence of plus disease
| Incidence of Severe ROP | Incidence of Any ROP | Incidence of Plus disease | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Maternal race | N of infants |
N of severe ROP (%) |
Adjusted risk ratio* (95%CI) |
P | N of any ROP (%) |
Adjusted Risk ratio* (95% CI) |
P | N of plus disease (%) |
Adjusted Risk ratio* (95% CI) |
P |
| All infants | <0.001 | <0.001 | <0.001 | |||||||
| White | 5580 | 694 (12.4%) | 1.00 | 2411 (43.2%) | 1.00 | 350 (6.3%) | 1.00 | |||
| Black | 3252 | 360 (11.1%) | 0.69 (0.62, 0.77) | <0.001 | 1370 (42.1%) | 0.83 (0.80, 0.87) | <0.001 | 118 (3.6%) | 0.44 (0.37, 0.54) | <0.001 |
| Asian | 353 | 42 (11.9%) | 1.05 (0.83, 1.34) | 0.68 | 108 (30.6%) | 0.77 (0.68, 0.87) | <0.001 | 21 (5.9%) | 1.04 (0.70, 1.53) | 0.85 |
| Gestational age <28 weeks | <0.001 | <0.001 | <0.001 | |||||||
| White | 2202 | 635 (28.8%) | 1.00 | 1678 (76.2%) | 1.00 | 322 (14.6%) | 1.00 | |||
| Black | 1532 | 345 (22.5%) | 0.71 (0.64, 0.78) | <0.001 | 1063 (69.4%) | 0.87 (0.84, 0.90) | <0.001 | 114 (7.4%) | 0.46 (0.37, 0.56) | <0.001 |
| Asian | 115 | 38 (33.0%) | 1.06 (0.83, 1.35) | 0.65 | 77 (67.0%) | 0.86 (0.76, 0.96) | 0.01 | 21 (18.3%) | 1.14 (0.77, 1.68) | 0.51 |
| Gestational age ≥28 weeks | 0.005 | <0.001 | 0.001 | |||||||
| White | 3378 | 59 (1.7%) | 1.00 | 733 (21.7%) | 1.00 | 28 (0.8%) | 1.00 | |||
| Black | 1720 | 15 (0.9%) | 0.45 (0.25, 0.79) | 0.005 | 307 (17.8%) | 0.76 (0.67, 0.85) | <0.001 | 4 (0.2%) | 0.25 (0.09, 0.71) | 0.009 |
| Asian | 238 | 4 (1.7%) | 1.08 (0.40, 2.94) | 0.88 | 31 (13.0%) | 0.64 (0.47, 0.88) | 0.006 | 0 (0.0%) | ND | |
CI=Confidence interval.
Adjusted by birth weight, gestational age and gender.
Figure 1.
Risk for severe ROP (the top row) and any stage ROP (the bottom row) are plotted against gestational age (the left column), postmenstrual age (the middle column) and postnatal age (the right column) for each racial group. Blue indicates White infants group, red indicates Black infants, and green indicates Asian infants. Error bars demonstrate standard errors for proportions were calculated as sqrt(p*(1-p)/n).
When stratified by BW, Black infants had a lower rate of severe ROP than White and Asian infants. For example, for infants with BW 501 to 750g, the incidence rates of severe ROP among White, Black, and Asian infants were 41.5%, 32.3%, and 46.9% respectively. The severe ROP rates for each racial group stratified by combinations of BW and GA are shown in Online Supplementary Tables 1A, 1B and 1C.
Race and Incidence of Any ROP
Table 3 shows the incidence of any stage ROP for the three groups. The incidence rate of any stage ROP was 42.1% in Black infants, 43.2% in White infants, and 30.6% in Asian infants (P<0.001). After adjustment by GA, BW, and gender, the risk of any ROP was significantly lower among Black and Asian infants than White infants for all infants (aRR=0.83, 0.77, respectively, all P<0.001), for infants with GA<28 weeks (aRR=0.87, 0.86 respectively, all P<0.05) and for infants with GA ≥28 weeks (aRR=0.76, 0.64, respectively, all P<0.01). Figure 1B shows the incidence of ROP for each group stratified by GA. For most of the GA weeks, Black infants had a lower rate of ROP than White infants.
Table 3.
Comparison of daily-weight-gain in grams (mean (SD) during 3 intervals among races.
| White (N=5580) |
Black (N=3252) |
Asian (N=353) |
Adjusted P* |
|
|---|---|---|---|---|
| All infants | ||||
| DOL 11-20 | 19.7 (14.5) | 18.3 (12.9) | 19.9 (12.6) | 0.81 |
| DOL 21-30 | 25.1 (14.7) | 23.1 (19.3) | 23.9 (13.4) | 0.09 |
| DOL 31-40 | 26.7 (15.4) | 24.8 (14.4) | 22.9 (14.6) | <0.001 |
| GA <28 weeks | N=2202 | N=1532 | N=115 | |
| DOL 11-20 | 13.7 (11.1) | 12.9 (10.4) | 14.2 (10.1) | 0.44 |
| DOL 21-30 | 17.4 (12.7) | 16.7 (11.1) | 15.7 (11.8) | 0.52 |
| DOL 31-40 | 21.9 (14.8) | 21.0 (13.7) | 17.0 (12.7) | 0.003 |
| GA ≥28 weeks | N=3378 | N=1720 | N=238 | |
| DOL 11-20 | 23.4 (15.1) | 22.7 (13.1) | 22.6 (12.7) | 0.56 |
| DOL 21-30 | 30.0 (13.7) | 28.6 (22.9) | 28.0 (12.3) | 0.16 |
| DOL 31-40 | 30.1 (14.9) | 28.7 (14.1) | 26.3 (14.6) | 0.003 |
DOL=Day of life; SD=standard deviation; GA=gestational age.
Adjusted by birth weight, gestational age and gender
Race and Incidence of Plus Disease
Table 4 shows the incidence of plus disease for the three groups. Plus disease occurred in 350 (6.3%) White infants, 118 (3.6%) Black infants, and 21 (5.9%) Asian infants. After adjustment by GA, BW, and gender, the risk of plus disease was significantly lower among Black infants versus White infants for all infants (aRR=0.44, P<0.001), for infants with GA<28 weeks (aRR=0.46, all P<0.001), and for infants with GA ≥28 weeks (aRR=0.25, P=0.009).
Table 4A.
Association between race and ROP Zone 1.
| Maternal race | N of babies |
N of Zone 1 ROP (%) |
Risk ratio (95%CI) |
P | Overall P |
Risk ratio (95%CI)* |
P* | Overall P* |
|---|---|---|---|---|---|---|---|---|
| White/Caucasian | 2406 | 329 (13.7%) | 1.00 | 1.00 | ||||
| Black/African American | 1370 | 194 (14.2%) | 1.04 (0.88, 1.22) | 0.68 | 0.78 | 0.82 (0.70, 0.95) | 0.008 | 0.02 |
| Asian/Asian American | 108 | 13 (12.0%) | 0.88 (0.52, 1.48) | 0.63 | 0.77 (0.46, 1.26) | 0.30 |
CI=Confidence interval.
Adjusted by birth weight, gestational age and gender.
Race and timing of ROP
We compared the timing of severe ROP and timing of any stage ROP among the racial groups in terms of postmenstrual age (PMA) (Supplementary Table 2). Overall, there were no significant differences among groups in the timing of first diagnosis of either any stage ROP or severe ROP. When stratified by GA, Black infants had slightly earlier PMA at the first diagnosis of any stage ROP than White infants (difference = −0.39 weeks, 95% CI: −0.68 to −0.10 weeks, P=0.008), but the difference was not clinically significant. Figure 1C shows the incidence of severe ROP by PMA for each group. Figure 1D shows the incidence of any stage ROP by PMA for each group. Overall, across most PMA, Black infants had lower risk of severe ROP and any ROP than White infants, and their ROP risk decreased with PMA and dropped to very low risk (<1%) after 46 weeks PMA.
We plotted the risks of severe ROP (Figure 1E) and any ROP (Figure 1F) among the three groups across postnatal age. Overall, across most postnatal ages, Black infants had lower risk of severe ROP and any ROP than White infants. The risk of severe ROP (Figure 1E) peaked at 11-12 weeks of postnatal age and dropped to less than 4% after 19 weeks of postnatal age. The risk of any ROP (Figure 1F) peaked at <9 weeks of postnatal age and dropped to approximately 10% at 15-16 weeks of postnatal age; after 15-16 weeks, Black infants had higher risk compared with White infants.
Race and daily-weight-gain by day of life
Mean daily-weight-gain by day of life for each racial group is shown for all infants (Figure 2A), for infants with GA<28 weeks (Figure 2B) and for those with GA≥28 weeks in Figure 2C. On average, Black infants showed a trend of lower daily-weight-gain than White infants. In analysis adjusted by BW, GA and gender (Table 3), no difference was identified in postnatal daily-weight-gain at days of life 10-20 and 21-30 among the racial groups. However, at days of life 31-40, postnatal daily-weight-gain was significantly lower among Black and Asian infants than White infants for all infants (P<0.001), for infants with GA<28 weeks (P=0.003), and infants with GA≥28 weeks (P=0.003).
Figure 2.
Mean daily-weight-gain across day of life by race for all infants (A), and for infants with GA < 28 weeks (B) and infants with GA ≥ 28 weeks (C). Blue indicates White infants group, red indicates Black infants, and green indicates Asian infants. Each data point presented in the figure represents the mean daily-weight-gain on a specific day of life for a specific racial group.
Race and ROP Zone 1 disease
Table 4A shows the incidence of Zone 1 disease for the three groups among infants who ever developed ROP. Zone 1 disease occurred in 329 (13.7%) White (N=2406), 194 (14.2%) Black (N=1370) and 13 (12.0%) Asian (N=108) infants. After adjustment by GA, BW, and gender, the risk of Zone 1 disease was significantly lower among Black infants versus White infants for all infants (aRR=0.82, P=0.008); there is no difference between Asian infants versus White (aRR=0.77, P=0.30).
Race, retinal detachment and post-discharge treatment
Retinal detachment data up to 50 weeks PMA were collected for all infants. Retinal detachment occurred in 36 (0.8%) White (N=5580), 12 (0.4%) Black (N=3253) and 0 (0.0%) Asian (N=353) infants, and no difference was identified (P=0.10).
Among the infants who developed ROP (Table 4B), retinal detachment occurred in occurred in 36 (1.5%) of the White (N=2411), 12 (0.9%) of the Black (N=1370) and 0 (0.0%) of the Asian (N=108), and no difference was identified (P=0.17).
Table 4B.
Distribution of the rate of retinal detachment in three race groups
| Infants with ROP | ||||
|---|---|---|---|---|
| White/Caucasian (N=2411) |
Black/African American (N=1370) |
Asian/Asian American (N=108) |
P | |
| Retinal detachment | 0.17 | |||
| No | 2375 (98.5%) | 1358 (99.1%) | 108 (100.0%) | |
| Yes | 36 (1.5%) | 12 (0.9%) | 0 (0.0%) | |
Among the infants discharged before 40 weeks PMA, post-discharge ROP treatment occurred in 20 (0.8%) of the White (N=2411), 9 (0.4%) of the Black (N=1451) and 3 (2.4%) of the Asian (N=124), and their difference is statistically significant (P=0.003).
DISCUSSION
In this study, we evaluated racial differences in the characteristics of ROP development. On average, Black infants not only had lower GA and smaller BW, but also had significantly lower mean postnatal daily-weight-gain than White infants, especially after the 30th day of life. Despite the presence of these risk factors, Black infants still had lower risk of developing any ROP, severe ROP, plus disease, Zone 1 disease, retinal detachment and post-discharge treatment than White infants.
Previously, we reported that lower GA and smaller BW were the most important perinatal risk factors for ROP development in the G-ROP-1 study.2 In this study, we further investigated the effect of maternal race on ROP incidence, and found that Black infants had lower incidence of any ROP, severe ROP, and plus disease than White infants. For example, among infants with GA 24 weeks or less, White infants had a 7% to 9% higher incidence of severe ROP than Black infants; among infants with BW lower than 750g, White infants had a 10% higher incidence of severe ROP than Black infants. These results agree with previous reports that White race is associated with a higher risk of ROP.19-22 Furthermore, White race was also suggested as an independent risk factor for plus disease and ROP treatment.23 Consistent with this, we found that Black infants had a lower risk than White infants for plus disease (aRR of 0.44), severe ROP (aRR of 0.69), and any stage ROP (aRR 0.83). Similar to our study, data from CRYO-ROP suggested black infants are less likely to develop severe ROP although they did not report Zone 1 disease.9 Port et al also reported that the black race was less likely for developing treatment-requiring ROP (based on 239 white and 73 black infants who developed any ROP).24
We did not find substantial racial difference in that timing of ROP, which is consistent with a previous study that showed that timing of ROP with peak risk is similar at postnatal week 12 regardless of GA but no specific pattern by GA was seen for peak risk in terms of postmenstrual age.11 We found that both White and Black infants develop severe ROP at a similar postnatal age of 11-12 weeks; furthermore, at the peak risk time, White infants had higher risk than Black infants. Our results showed that risk of ROP decreases over postmenstrual age for all racial groups.
Previously, Janevic et al reported that, after excluding infants born before 24 weeks’ gestation, Asian infants had an increased risk for ROP compared with White infants.25 However, in our study, Asian infants showed a significantly lower risk for any ROP, but not for severe ROP, when compared to White infants. This observation may be related to the lower birth-weight-gain of Asian infants with GA<28 weeks. The sample size of Asian infants in our study was relatively smaller than White and Black infants, which may limit our comparisons for the Asian infants.
We found that Black infants had a lower risk to develop Zone 1 disease compared with White infants. Our findings disagree with the previous secondary analysis of the ET-ROP database, which showed that Black infants (N=64) were actually at higher risk of developing Zone I disease when compared to White infants (N=230) after controlling for birth weight and gestational age.26 The difference might be related to the difference in sample size or geographical distribution.
White infants had significantly higher postnatal weight-gain rate than Black and Asian infants irrespective of their GA. Figure 2 showed a different weight gain pattern in infants with GA<28 weeks and infants with GA≥28 weeks. The different patterns of weight gain over time were probably due to the nature of the growth curve, which is characterized by rapid growth at an earlier time and slowly reaches a plateau.
It is unknown why preterm-born Black infants have a lower risk of ROP than White infants. Racial differences in retinal pigmentation are suggested as a cause.9 Large amounts of melanin in the retinal pigment epithelium or choroid associated with dark fundi may protect very low birth weight infants from developing ROP. Darker fundus was reported to be associated with less severe ROP after controlling for other factors such as race, GA, BW, and length of oxygen therapy.27 However, another study of premature infants found dark fundus was not associated with ROP risk.28 The role of ocular pigmentation in the development of ROP is still unclear and needs further investigation.
This study provided evidence to clinicians that, among infants with similar GA or BW, Black infants have a lower risk of severe ROP, any stage ROP, and plus disease than White infants. The racial difference in daily-weight-gain by day of life found here might provide a new reference for clinical practice. Applying these racial differences may be meaningful in clinical practice with regard to ROP screening examinations or ROP risk stratification. This study does not intend to use our findings to modify ROP care across races, and we were sensitive to the impracticality of racial-based screening protocols. Because such racial disparities for ROP may contribute to inequalities in health and development later in the child's life, further investigation on the mechanism for racial differences is important. Understanding these racial differences may lead to better pathophysiological understandings about ROP development. In the future, genetic or anatomical studies on racial differences in preterm infants might bring us insights into what mechanisms protect Black infants from ROP development.
Strengths of this study include the large sample size of infants from many clinical centers of North America, representative of preterm infants at risk of ROP. The data were based on examinations performed by pediatric ophthalmologists and retinal specialists with clinical expertise in ROP management, using standard International Classification of ROP terminology with regard to the diagnosis of ROP. The comprehensive data on postnatal weight gain and ROP characteristics allowed us to evaluate racial differences.
Our study had several limitations:
1) The G-ROP datasets did not contain other socioeconomic determinants of health data, so such factors, which may influence prenatal care and nutrition, were not considered in our analysis. However, given the generally recognized socioeconomic inequities favoring White infants over Black infants29 failure to adjust for such confounders would likely bias the study results in the direction opposite of our study findings. 2) Our classification of racial groups of infants was not comprehensive and based solely on self-reported maternal race, without consideration of paternal race or multiethnic backgrounds. Racial groups might be better classified based on genetic testing results30,31, which were not available. Our analyses are restricted to Black, White, and Asian, and the Asian group had relatively fewer infants than other groups. Infants not falling into one of these racial groups were excluded due to unknown racial information or a relatively small number (Supplementary Table 3A). 3) Our analysis did not consider ethnicity because infants of Hispanic ethnicity were relatively small (Supplementary Table 3B). Previously, ethnicity was suggested as a risk factor for severe ROP.8 However, our sensitivity analysis among non-Hispanic infants showed similar differences among three race groups (Supplementary Table 3C). 4) In the G-ROP-1 study (not in G-ROP-2), we collected mortality data from infants who were not eligible for the G-ROP study due to incomplete ROP data (i.e., infants died before the ROP outcome was known, or other reasons, such as transfer to a non-study hospital) and whether infants died during ROP follow-up. We compared the rate of mortality among racial groups enrolled in G-ROP-1. The proportion of death in the three groups was similar (P=0.19) although the white infants had the highest rate of death (5.9%) (Supplementary Table 3D). Therefore, we do not think there was not selection bias due to differential survival between these racial groups. 5) Our comparisons of ROP risks among three race groups did not account for many other factors that we did not collect. Thus, the observed differences in ROP risk across race groups could be due to unmeasured confounders, such as differences in oxygen monitoring protocols or clinical diagnosis of ROP across institutions.
In conclusion, in this secondary analysis from a large, racially and geographically diverse cohort of infants undergoing ROP screening in North America, we found that while Black infants had lower GA, smaller BW and less mean daily-weight-gain by day of life than White infants, all strong risk factors for ROP, Black infants had much lower risk for developing ROP, severe ROP, Zone 1 disease, and plus disease than White infants.
Supplementary Material
Funding:
Supported by NIH grants R01EY021137-01A (Binenbaum), R21EY029776 (Ying), R21EY026664 (Wang), and the Richard Shafritz Chair in Ophthalmology Research.
Footnotes
Conflict of interest’s disclosure: All authors had no conflict of interests with materials presented in this article.
Statement: This submission has not been published anywhere previously and that it is not simultaneously being considered for any other publication.
References
- 1.Gilbert C Retinopathy of prematurity: a global perspective of the epidemics, population of babies at risk and implications for control. Early Hum Dev. 2008;84(2):77–82. [DOI] [PubMed] [Google Scholar]
- 2.Ying GS, Bell EF, Donohue P, Tomlinson LA, Binenbaum G, Group GRR. Perinatal Risk Factors for the Retinopathy of Prematurity in Postnatal Growth and Rop Study. Ophthalmic Epidemiol. 2019;26(4):270–278. [DOI] [PubMed] [Google Scholar]
- 3.United States cencus bureau. Quickfacts. 2022. https://www.census.gov/quickfacts/fact/table/US/RHI225221 (accessed Nov. 22, 2022)
- 4.Quinn GE, Barr C, Bremer D, et al. Changes in Course of Retinopathy of Prematurity from 1986 to 2013: Comparison of Three Studies in the United States. Ophthalmology. 2016;123(7):1595–1600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Centers for Disease Control and Prevention. Preterm birth. 2022. www.cdc.gov/reproductivehealth/maternalinfanthealth/pretermbirth.htm#:~:text=In%202021%2C%20the%20rate%20of,%25%20and%2010.2%25%20respectively).&text=Preterm%20births%20may%20also%20take,a%20financial%20burden%20for%20families (accessed Nov 22 2022).
- 6.Bhudia R Role of Maternal Race on Algorithms Predicting Retinopathy of Prematurity. JAMA Ophthalmol. 2018;136(2):221. [DOI] [PubMed] [Google Scholar]
- 7.Husain SM, Sinha AK, Bunce C, et al. Relationships between maternal ethnicity, gestational age, birth weight, weight gain, and severe retinopathy of prematurity. J Pediatr. 2013;163(1):67–72. [DOI] [PubMed] [Google Scholar]
- 8.Aralikatti AK, Mitra A, Denniston AK, Haque MS, Ewer AK, Butler L. Is ethnicity a risk factor for severe retinopathy of prematurity? Arch Dis Child Fetal Neonatal Ed. 2010;95(3):F174–176. [DOI] [PubMed] [Google Scholar]
- 9.Saunders RA, Donahue ML, Christmann LM, et al. Racial variation in retinopathy of prematurity. The Cryotherapy for Retinopathy of Prematurity Cooperative Group. Arch Ophthalmol. 1997;115(5):604–608. [DOI] [PubMed] [Google Scholar]
- 10.Yang MB, Donovan EF, Wagge JR. Race, gender, and clinical risk index for babies (CRIB) score as predictors of severe retinopathy of prematurity. J AAPOS. 2006;10(3):253–261. [DOI] [PubMed] [Google Scholar]
- 11.Pivodic A, Hard AL, Lofqvist C, et al. Individual Risk Prediction for Sight-Threatening Retinopathy of Prematurity Using Birth Characteristics. JAMA Ophthalmol. 2020;138(1):21–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lofqvist C, Hansen-Pupp I, Andersson E, et al. Validation of a new retinopathy of prematurity screening method monitoring longitudinal postnatal weight and insulinlike growth factor I. Arch Ophthalmol. 2009;127(5):622–627. [DOI] [PubMed] [Google Scholar]
- 13.Binenbaum G, Tomlinson LA, de Alba Campomanes AG, et al. Validation of the Postnatal Growth and Retinopathy of Prematurity Screening Criteria. JAMA Ophthalmol. 2020;138(1):31–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Binenbaum G, Tomlinson LA. Postnatal Growth and Retinopathy of Prematurity Study: Rationale, Design, and Subject Characteristics. Ophthalmic Epidemiol. 2017;24(1):36–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Binenbaum G, Bell EF, Donohue P, et al. Development of Modified Screening Criteria for Retinopathy of Prematurity: Primary Results From the Postnatal Growth and Retinopathy of Prematurity Study. JAMA Ophthalmol. 2018;136(9):1034–1040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Provis JM, Diaz CM, Dreher B. Ontogeny of the primate fovea: a central issue in retinal development. Prog Neurobiol. 1998;54(5):549–580. [DOI] [PubMed] [Google Scholar]
- 17.Provis JM, Hendrickson AE. The foveal avascular region of developing human retina. Arch Ophthalmol. 2008;126(4):507–511. [DOI] [PubMed] [Google Scholar]
- 18.Wang J, Spencer R, Leffler JN, Birch EE. Critical period for foveal fine structure in children with regressed retinopathy of prematurity. Retina. 2012;32(2):330–339. [DOI] [PubMed] [Google Scholar]
- 19.Darlow BA, Hutchinson JL, Henderson-Smart DJ, et al. Prenatal risk factors for severe retinopathy of prematurity among very preterm infants of the Australian and New Zealand Neonatal Network. Pediatrics. 2005;115(4):990–996. [DOI] [PubMed] [Google Scholar]
- 20.Ying GS, Quinn GE, Wade KC, et al. Predictors for the development of referral-warranted retinopathy of prematurity in the telemedicine approaches to evaluating acute-phase retinopathy of prematurity (e-ROP) study. JAMA Ophthalmol. 2015;133(3):304–311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Wheatley CM, Dickinson JL, Mackey DA, Craig JE, Sale MM. Retinopathy of prematurity: recent advances in our understanding. Arch Dis Child Fetal Neonatal Ed. 2002;87(2):F78–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nodgaard H, Andreasen H, Hansen H, Sorensen HT. Risk factors associated with retinopathy of prematurity (ROP) in northern Jutland, Denmark 1990–1993. Acta Ophthalmol Scand. 1996;74(3):306–310. [DOI] [PubMed] [Google Scholar]
- 23.Tsui I, Ebani E, Rosenberg JB, Angert RM, Lin J, Mian U. Trends in retinopathy of prematurity over a 5-year period in a racially diverse population. Ophthalmic Surg Lasers Imaging Retina. 2014;45(2):138–142. [DOI] [PubMed] [Google Scholar]
- 24.Port AD, Chan RV, Ostmo S, Choi D, Chiang MF. Risk factors for retinopathy of prematurity: insights from outlier infants. Graefes Arch Clin Exp Ophthalmol. 2014;252(10):1669–1677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Janevic T, Zeitlin J, Auger N, et al. Association of Race/Ethnicity With Very Preterm Neonatal Morbidities. JAMA Pediatr. 2018;172(11):1061–1069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ishikiryama M, Levison A, Deiner M, de Alba Campomanes AG. Risk Factors for Development of Zone I versus Zone II Retinopathy of Prematurity: A Retrospective Analysis of the Early Treatment of Retinopathy of Prematurity Study (ET-ROP). Invest Ophthalmol Vis Sci 2014;55(13):5909. [Google Scholar]
- 27.Monos T, Rosen SD, Karplus M, Yassur Y. Fundus pigmentation in retinopathy of prematurity. Pediatrics. 1996;97(3):343–348. [PubMed] [Google Scholar]
- 28.Arnold RW, Kesler K, Avila E. Susceptibility to retinopathy of prematurity in Alaskan Natives. J Pediatr Ophthalmol Strabismus. 1994;31(3):192–194. [DOI] [PubMed] [Google Scholar]
- 29.Boen C The role of socioeconomic factors in Black-White health inequities across the life course: Point-in-time measures, long-term exposures, and differential health returns. Soc Sci Med. 2016;170:63–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bonham VL, Callier SL, Royal CD. Will Precision Medicine Move Us beyond Race? N Engl J Med. 2016;374(21):2003–2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Royal CD, Dunston GM. Changing the paradigm from 'race' to human genome variation. Nat Genet. 2004;36(11 Suppl):S5–7. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.