Perinatal risk factors and neurocognitive outcomes in children and adolescents with sickle cell disease

Abstract

Background:

The literature on cognitive and academic outcomes for children with sickle cell disease (SCD) who experience perinatal risk factors is limited. We aimed to evaluate if low birthweight (LBW), gestational age, and history of neonatal intensive care unit (NICU) admission were associated with neurocognitive functioning, grade retention, or receipt of early intervention or formal educational support in children with SCD.

Procedures:

This prospective birth cohort study included 336 participants, ages 8–18, with SCD, who received cognitive testing as part of standard of care and whose caregivers completed behavioral rating scales. Multivariable generalized linear regression models were used to examine associations between perinatal risks and outcome variables, after adjusting for demographic and medical covariates.

Results:

The prevalence of NICU admission and LBW were 12.03% and 13.50%, respectively. Lower birthweight, earlier gestational age, and NICU admission were associated with worse working memory performance and receipt of early intervention services. Lower birthweight and NICU admission were also associated with slower processing speed. History of NICU admission was associated with caregiver ratings of hyperactivity and emotional dysregulation. The effects of perinatal risk factors on neurocognitive, academic, or educational outcomes were not dependent on SCD genotype.

Conclusions:

History of LBW or NICU admission was associated with worse cognitive outcomes and increased use of early intervention services among children with SCD. Early identification of perinatal risk factors will help identify children who will benefit from formal developmental or neuropsychological evaluations to manage the comorbidity of SCD and perinatal risks and facilitate increased intervention.

1 |. INTRODUCTION

In the general population, preterm birth and low birthweight (LBW; <2500 g) are risk factors for poor outcomes, including delayed developmental milestones, slowed neurocognitive and academic development, and increased risk of developing attention-deficit/hyperactivity disorder (ADHD).^1–3 Importantly, term infants can be LBW due to fetal growth restriction, maternal chronic health conditions (e.g., diabetes, high blood pressure, and cardiac, renal, or respiratory problems), placental abruption, and exposure to toxins (e.g., tobacco, alcohol, or pollutants).⁴ Developmental trajectories also differ for neonatal intensive care unit (NICU) graduates, as those who experience more medical complications after delivery exhibit greater future difficulties in cognition and academic achievement.⁵ Epidemiological data indicates nearly half of all NICU admissions are now for children with normal birthweight or those born at ≥37 weeks gestation.⁶

Sickle cell disease (SCD) is the most common inherited genetic disorder in the United States, affecting one in every 365 Black or African American births.⁷ Compared to the general population, children with SCD generally demonstrate worse cognitive and academic outcomes.^8,9 Children whose mothers have SCD are nearly three times more likely to be born preterm compared to mothers without SCD.¹⁰ In general, Black and African American mothers have higher rates of preterm birth and infants who are LBW (14.75% and 14.66%, respectively) compared to White mothers (9.50% and 7.03%, respectively).¹¹ Together, these data suggest that children with SCD, who are most likely to be Black or African American in the United States, may be at increased risk for poor neurodevelopmental outcomes due to preterm birth and/or LBW.

The literature on cognitive and academic outcomes for children with SCD who experience perinatal risk factors is limited and has only considered gestational age to date.^12,13 This is problematic as (i) perinatal risk factors represent early medical events that are distinct from hematological disease course and are capable of impacting the child’s neurodevelopmental trajectory, (ii) findings on impact of gestational age in SCD have been inconsistent, and (iii) the impact of other perinatal risk factors, such as LBW and NICU admission, remains unknown in this population. Understanding the impact of perinatal risk factors on cognitive and academic outcomes in children with SCD can facilitate earlier referrals to multidisciplinary specialties, and enhance access to clinical and community-based intervention services, which may improve long-term outcomes in this population.^14,15

The first objective of this prospective birth cohort study was to evaluate the effects of birthweight, gestational age, and NICU admission on cognitive and academic outcomes in children with SCD, including interactions between genotype and perinatal risks. A secondary objective was to examine if perinatal risk factors were associated with receipt of early intervention or formal educational support services in this population (i.e., Section 504 Plan or Individualized Education Program [IEP]). We hypothesized that history of LBW, early gestational age, or NICU admission would be associated with worse neurocognitive and academic outcomes, as well as greater use of early intervention and formal educational services.

2 |. METHODS

This study was approved by the Institutional Review Board (IRB) at St. Jude Children’s Research Hospital. Written informed consent was obtained from parents or guardians in accordance with hospital policy.

2.1 |. Participants

Participants were children and adolescents with SCD enrolled in the Sickle Cell Clinical Research and Intervention Program (SCCRIP) who completed neurocognitive evaluations at St. Jude Children’s Research Hospital. Briefly, SCCRIP is an observational cohort study of clinical outcomes in SCD, including hematological, genetic, cardiac, renal, orthopedic, neurocognitive, and psychosocial outcomes.¹⁶ All children with SCD were referred for systematic neurocognitive evaluations as standard of care, irrespective of disease severity, cerebrovascular accident (CVA), or specific clinical concern. These evaluations generally occurred at school age (8–9), early adolescence (12–13), and late adolescence (16–17), and included performance-based measurement of cognitive and academic abilities, along with caregiver ratings of mood, behavior, and adaptive function. Similar to traditional neuropsychological evaluations, these targeted evaluations assessed key cognitive domains in the absence of a clinical referral question. We provided diagnoses as appropriate and developed tailored recommendations to meet patient needs. Although all children with SCD are scheduled for systematic evaluations, some do not receive an evaluation due to missed appointments. Documentation to support need for school services (i.e., Section 504 Plan or IEP) was provided by the primary medical team or the neuropsychologist after targeted evaluations. Contingent upon interest, school advocacy coordinators assisted caregivers in establishing or updating formal school services.

A total of 448 children and adolescents between the ages of 8−18 with SCD completed neurocognitive evaluations. Given the design of SCCRIP, some participants completed more than one neurocognitive evaluation and when this occurred, data from the earliest visit were selected. Exclusion criteria included history of stroke or curative therapy for SCD. Participants with missing perinatal data (n = 112) were excluded from analysis. Data were collected between 2013 and 2021.

2.2 |. Neurocognitive measures

Psychological examiners and clinical trainees administered these tests, and visits were supervised by licensed psychologists. The Wechsler Abbreviated Scale of Intelligence, Second Edition was used to measure general cognitive ability (Full-Scale Intelligence Quotient [FSIQ]), verbal comprehension (Verbal Comprehension Index [VCI]), and perceptual reasoning (Perceptual Reasoning Index [PRI]).¹⁷ The Wechsler Intelligence Scale for Children, Fourth or Fifth Edition or Wechsler Adult Intelligence Scale, Fourth Edition were used to assess working memory (Digit Span subtest) and processing speed (Coding subtest).^18,19 Complex attention was measured with the Conners Continuous Performance Test, Second or Third Edition.²⁰ Executive function was assessed using the Delis Kaplan Executive Function System (Verbal Fluency subtest),²¹ and caregiver rating on the Behavioral Rating Inventory of Executive Functioning, First or Second Edition.^22,23 Academic skills were measured with the Woodcock–Johnson Tests of Achievement, Third or Fourth Edition (Letter-Word Identification and Math Fluency subtests).²⁴ The Behavior Assessment System for Children, Second or Third Edition was completed by caregivers, and provided information on emotional and behavioral functioning.^25,26

2.3 |. Demographic, medical, and treatment variables

Demographic, medical, and treatment variables were abstracted from the SCCRIP database. Variables for gestational age, birthweight, and NICU admission were collected from the participant’s medical chart. Patients are referred to our hospital after a positive newborn screening for SCD, thus our medical records do not contain comprehensive information from pregnancy, labor, or delivery. Information on early intervention services, educational services (e.g., Section 504 Plan or IEP), and grade retention was collected from caregivers during clinical interviews at the neurocognitive evaluation. Provision of early intervention or educational services (e.g., Section 504 Plan or IEP) was determined by the providing agency or each child’s individual school district. The Social Vulnerability Index (SVI) was calculated for each participant, which measures socioeconomic disadvantage at the neighborhood level, including housing data, education, and poverty, with higher scores indicating greater risk.²⁷ National Heart, Lung, and Blood Institute (NHLBI) guidelines were followed for treatment with hydroxyurea.²⁸ Guidelines on timing for initiation of hydroxyurea therapy evolved over the course of our study. Lab values were collected on the day of neurocognitive evaluation or, if unavailable, the closest blood draw within 6 months prior to testing.

2.4 |. Statistical analyses

Birthweight was analyzed as a continuous and categorical (<2500 or ≥2500 g) variable. Birthweight of less than 2500 g was considered low, and birthweight ≥2500 g was considered normal.²⁹ Gestational age was analyzed as a continuous variable. NICU admission was categorized as yes/no. Educational support services were analyzed according to past or current receipt of Section 504 Plan or IEP (yes/no). History of grade retention or participation in early intervention services was also categorized as yes/no. Two group differences were evaluated with the two-sample t-test, Wilcoxon rank sum test, or Fisher’s exact test.

For the first objective, we used multivariable linear regression models to evaluate the effects of birthweight, gestational age, and NICU admission on neurocognitive functioning. Age, SVI, genotype, and hydroxyurea treatment were examined as covariates in primary data analysis, given associations with disease progression and cognitive outcomes.³⁰ Sex was also included as a covariate in models based on differences in cognitive outcomes in the SCD and perinatal literature.^31,32 Interactions with genotype were tested in models with main effects. For our second objective, we evaluated the effect of birthweight, gestational age, and NICU admission on early intervention, educational support services, and grade retention using logistic regression models. Logistic regression models were adjusted for age, sex, SVI, genotype, and hydroxyurea treatment. Given differences in disease pathology and progression based on genotype, we evaluated genotype as a binary covariate (HbSS/HbSβ⁰ thalassemia and HbSC/HbSβ⁺ thalassemia/Others) across both objectives. As an exploratory analysis, we also evaluated interaction effects between genotype and perinatal risk factors on the neurocognitive outcomes, educational support services, and grade retention, by including an interaction term in the model.

False discovery rate-adjusted p-values or q-values were calculated across all perinatal risk factors to correct for multiple comparisons. q-Value less than .05 was considered significant in our primary analyses. Otherwise, p-value less than .05 was considered statistically significant.

3 |. RESULTS

3.1 |. Demographic and clinical characteristics

A total of 336 participants received a neurocognitive evaluation, and the mean age at time of testing was 11.73 (standard deviation [SD] = 3.42) years. Males represented 50.60% of the sample. All participants identified as Black. The HbSS thalassemia genotype was most prevalent (53.57%). Approximately 64.58% and 14.58% of children initiated hydroxyurea treatment or received chronic transfusion prior to the time of neurocognitive evaluation, respectively. Mean gestational age and birthweight for the overall sample were 38.75 weeks (SD = 2.56) and 3047.55 g (SD = 595.26), respectively. See Table S1 for additional information.

Approximately 13.50% of participants were born with LBW (<2500 g; n = 44), and 12.03% of participants were admitted to the NICU (n = 38; Table 1, Table S1). Relatedly, 9.72% of participants received early intervention services, and 20.96% and 51.76% reported current or past receipt of an IEP or Section 504 Plan, respectively. Approximately 19.29% of participants repeated a grade. As expected, the LBW group (<2500 g) was born at a significantly earlier gestational age, more likely to be admitted to the NICU after birth, and more likely to participate in early intervention services compared to the normal birthweight group (≥2500 g). No other significant differences for demographic or disease characteristics were observed between the two groups (all p > .05; Table 1). There were few statistically significant differences in demographic and medical variables between participants included and excluded from analysis due to missing perinatal data and exclusion criteria (Table S2).

TABLE 1.

Participant characteristics by birthweight.

	Low birthweight (n = 44)	Average birthweight (n = 282)
	N (%)	N (%)	p
Sex			.42
Male	25 (56.82)	141 (50.00)
Female	19 (43.18)	141 (50.00)
Genotype			.23
SS	18 (40.91)	154 (54.61)
Sβ0	2 (4.55)	9 (3.19)
SC	15 (34.09)	87 (30.85)
Sβ+	8 (18.18)	29 (10.28)
Othera	1 (2.27)	3 (1.06)
Hydroxyurea prior to time of evaluation	24 (54.55)	185 (65.60)	.18
Chronic transfusion prior to time of evaluation	6 (13.64)	40 (14.18)	1.00
NICU admission	24 (57.14)	12 (4.43)	<.001
Current or past IEP	11 (29.73)	44 (19.13)	.19
Current or past 504 Plan	17 (45.95)	125 (52.08)	.60
History of grade retention	11 (27.50)	46 (17.56)	.13
History of early intervention services	8 (19.51)	20 (8.40)	.044
	Mean (SD)	Mean (SD)
Gestational age, weeks	34.61 (4.52)	39.34 (1.36)	<.001
Birthweight, g	1987.09 (503.35)	3213.01 (409.25)	<.001
Age at HU initiation, years	9.19 (6.05)	7.42 (4.92)	.21
Age at chronic transfusions, years	8.34 (8.83)	8.72 (6.99)	.83
Hemoglobin, %	10.47 (1.68)	10.11 (1.72)	.23
WBC count, ×109/L	8.44 (3.62)	8.50 (3.38)	.81
Platelet count, ×109/L	328.19 (209.30)	338.75 (167.25)	.32
Fetal hemoglobin, %	12.01 (10.49)	15.19 (10.16)	.08
Social Vulnerability Indexb	0.62 (0.25)	0.65 (0.26)	.27
Age at evaluation, years	12.32 (3.65)	11.60 (3.35)	.20

Note: p-Values based on Wilcoxon test for continuous variables or Fisher’s exact test for categorical variables.

Abbreviations: HU, hydroxyurea; IEP, Individualized Education Program; n, sample size; NICU, Neonatal Intensive Care Unit; SD, standard deviation; WBC, white blood cell.

^aOther genotypes include: S/HPFH, S/O-Arab, S/gamma delta beta (γδβ), S/α.

^bClassifies individuals based on social vulnerabilities at the neighborhood level (e.g., housing data, poverty, and education); higher values indicate higher social vulnerability.

3.2 |. Performance across neurocognitive domains

In general, participants performed within the low average to average range across neurocognitive domains (Table S3).³³ Caregiver ratings of attention problems, hyperactivity, and executive functioning were within the average range for children with and without a history of LBW or NICU admission. The effects of all perinatal risk factors on neurocognitive, caregiver rating, and educational outcomes were not dependent on SCD genotype across analyses.

3.3 |. Birthweight and neurocognitive performance

Participants who were born with LBW (<2500 g) had lower scores on working memory, compared to the normal birthweight group, after adjustment for covariates (q = .016; Table 2). When analyzed continuously, increases in birthweight were associated with increased performance on processing speed (q = .025) and working memory (q = .016) measures. Interactions between genotype and main effects were nonsignificant (q > .05).

TABLE 2.

Multivariate models for birth-related risk factors with cognitive and academic performance.

			Birthweight (<2500 or ≥2500 g)			Birthweight (continuous)			Gestational age (continuous)			NICU stay (yes/no)
Domain	Measure	Scorea	Estimateb	SE	q c	Estimateb	SE	q c	Estimateb	SE	q c	Estimateb	SE	q c
Intellectual functioning	WASI-II	FSIQ	−1.49	1.95	.45	2.22	1.10	.17	0.30	0.27	.34	−2.97	2.10	.32
		VCI	−3.59	2.39	.18	2.25	1.42	.18	0.10	0.40	.80	−5.48	2.88	.18
		PRI	0.82	1.99	.68	2.53	1.14	.11	0.41	0.27	.18	−3.54	2.10	.18
Processing speed	WISC-IV/V or WAIS-IV	Coding	−0.90	0.44	.059	0.62	0.25	.025	0.05	0.06	.40	−1.35	0.47	.018
Working memory	WISC-IV/V or WAIS-IV	Digit span	−1.03	0.43	.016	0.57	0.24	.016	0.15	0.06	.016	−1.09	0.45	.016
Attention	CPT-3	HRT	0.99	4.30	.87	−0.46	2.83	.87	0.38	0.75	.87	11.80	4.93	.084
		Omissions	4.02	6.47	.54	−5.54	4.19	.33	−2.68	1.08	.07	9.15	7.84	.33
Executive function	DKEFS	Letter fluency	0.001	0.51	1.00	0.03	0.29	1.00	−0.002	0.07	1.00	0.18	0.54	1.00
		Category fluency	−0.16	0.48	.75	0.18	0.27	.75	−0.02	0.06	.75	−0.26	0.51	.75
Academic achievement	WJ-III	Letter word	−4.96	2.73	.14	3.25	1.54	.14	0.46	0.37	.27	−2.14	2.98	.47
		Math fluency	−4.75	2.81	.19	2.92	1.56	.19	0.15	0.38	.69	−4.16	3.12	.25

Note: Bold values statistically significant at q < .05.

Abbreviations: CPT-3, Conners Continuous Performance Test, Third Edition; DKEFS, Delis Kaplan Execution Function System; FSIQ, Full-Scale Intelligence Quotient; HRT, hit reaction time; NICU, neonatal intensive care unit; SE, standard error; PRI, Perceptual Reasoning Index; VCI, Verbal Comprehension Index; WAIS-IV, Wechsler Adult Intelligence Scale, Fourth Edition; WASI-II, Wechsler Abbreviated Scale of Intelligence, Second Edition; WISC-IV/V, Wechsler Intelligence Scales for Children, Fourth and Fifth Editions; WJ-III, Woodcock–Johnson Test of Achievement, Third Edition.

^aScaled scores with mean (SD) of 10 (3): coding, digit span total, letter fluency, category fluency; standard scores with a mean (SD) of 100 (15): FSIQ, VCI, PRI, letter word, math fluency; T-scores with a mean (SD) of 50 (10): HRT, omissions.

^bEstimates for the effect of birth-related risk factors on neurocognitive performance from linear regression model: neurocognitive measure = participant age at neurocognitive assessment, sex (ref = female), use of hydroxyurea prior to neurocognitive evaluation, sickle cell genotype, social vulnerability, and target perinatal variable (i.e., categorical birthweight, continuous birthweight [per 1000 g), gestational age, NICU admission).

^cq-Values are calculated to correct for multiple corrections to control the false discovery rate across all perinatal variables.

3.4 |. NICU admission, gestational age, and neurocognitive performance

Participants who were admitted to the NICU after birth had lower scores on measures of processing speed (q = .018) and working memory (q = .016), compared to those who did not require NICU admission, after adjustment for covariates. Increased gestational age was associated with higher working memory scores, but no other cognitive or academic outcome variables, after adjustment for covariates (Table 2). Interactions between genotype and main effects were nonsignificant (q > .05).

3.5 |. Birth-related risk factors and caregiver rating scales

After accounting for demographic and medical factors, NICU admission was associated with emotional regulation (q = .043), as well as hyperactivity (q = .044), on caregiver rating scales. There were no significant associations between birthweight or gestational age and caregiver ratings of behavior or executive functioning (all q > .05; Table 3).

TABLE 3.

Multivariate models for birth-related risk factors with parent rating scales.

			Birthweight (<2500 or ≥2500 g)			Birthweight (continuous)			Gestational age (continuous)			NICU stay (yes/no)
Domain	Measure	Scorea	Estimateb	SE	q c	Estimateb	SE	q c	Estimateb	SE	q c	Estimateb	SE	q c
Parent Ratings	BRIEF-2	Behavior Regulation Index	−0.52	2.06	.80	−2.06	1.18	.17	−0.35	0.29	.32	4.73	2.14	.11
		Emotional Control Index	−0.63	1.98	.75	−1.91	1.14	.19	−0.40	0.28	.21	5.23	2.03	.043
		Metacognition Index	−0.93	2.02	.65	−1.57	1.17	.42	−0.15	0.29	.65	2.66	2.12	.42
	BASC-3	Hyperactivity	−0.67	1.75	.70	−2.07	1.01	.08	−0.38	0.25	.19	4.75	1.85	.044
		Attention problems	−0.28	1.88	.88	−1.39	1.09	.45	−0.26	0.27	.45	2.09	2.01	.45

Note: Bold values statistically significant at q < .05.

Abbreviations: BASC-3, Behavior Assessment System for Children, Third Edition; BRIEF-2, Behavior Rating Inventory of Executive Function, Second Edition; NICU, neonatal intensive care unit; SE, standard error.

^aT-scores with a mean (SD) of 50 (10).

^bEstimates for the effect of birth-related risk factors on parent ratings from linear regression model: parent rating measure = participant age at neurocognitive assessment, sex (ref = female), use of hydroxyurea prior to neurocognitive evaluation, sickle cell genotype, social vulnerability, and target perinatal variable (i.e., categorical birthweight, continuous birthweight [per 1000 g], gestational age, NICU admission).

^cq-Values are calculated to correct for multiple corrections to control the false discovery rate across all perinatal variables.

3.6 |. Birth-related risk factors and educational supports

Lower birthweight (analyzed continuously and categorically), earlier gestational age, and history of NICU admission were associated with a greater probability of participation in early intervention services (all q < .05). The associations between perinatal risk factors and the probability of grade retention or receipt of an IEP or Section 504 Plan were nonsignificant (all q > .05; Table 4).

TABLE 4.

Multivariate models for birth-related risk factors with educational services.

	Birthweight (<2500 or ≥2500 g)			Birthweight (continuous)			Gestational age (continuous)			NICU stay (yes/no)
Outcome	ORa	95% CI	q b	ORa	95% CI	q b	ORa	95% CI	q b	ORa	95% CI	q b
Current or past IEP	1.70	0.71–3.90	.25	0.74	0.45–1.21	.25	0.92	0.83–1.02	.25	1.70	0.70–4.10	.25
Current or past 504 Plan	0.68	0.32–1.45	.55	0.80	0.53–1.22	.55	1.00	0.92–1.12	.73	0.71	0.32–1.60	.55
History of grade retention	1.90	0.86–4.40	.15	0.56	0.35–0.91	.08	0.93	0.84–1.03	.18	2.10	0.91–4.90	.15
History of early intervention services	3.30	1.20–9.20	.019	0.30	0.15–0.58	.0015	0.85	0.74–0.96	.013	4.00	1.45–11.10	.013

Note: Bold values statistically significant at q < .05.

Abbreviations: CI, confidence interval; IEP, Individualized Education Program; NICU, neonatal intensive care unit; OR, odds ratio.

^aOR and 95% CI for the effect of birth-related risk factors on educational services from logistic regression model: education service = participant age at neurocognitive assessment, sex (ref = female), use of hydroxyurea prior to neurocognitive evaluation, sickle cell genotype, social vulnerability, and target perinatal variable (i.e., categorical birthweight, continuous birthweight [per 1000 g], gestational age, NICU admission).

^bq-Values are calculated to correct for multiple corrections to control the false discovery rate across all the perinatal variables.

4 |. DISCUSSION

In this cohort study of children with SCD, perinatal risk factors were significant predictors of cognitive outcomes and early intervention support. Although CVA and disease progression are leading causes of poor cognitive outcomes in children with SCD, our results demonstrate that early medical events represent additional risk factors for cognitive difficulties in this population that deserve clinical consideration. Specifically, lower birthweight, earlier gestational age, and NICU admission were associated with worse working memory performance and receipt of early intervention services. Lower birthweight and NICU admission were also associated with slower processing speed. Interactions between genotype and perinatal risk factors were nonsignificant. Perinatal risk was not associated with formal academic services in school-aged children and adolescents, suggesting that supports fade as these patients age despite ongoing risk for cognitive impact. Caregiver ratings indicated children with SCD who received care in the NICU were more likely to be hyperactive and struggle with emotional regulation. More broadly, the prevalence of LBW in our study (13.50%) was consistent with national rates of LBW in Black children, demonstrating an early age of onset for health disparities in this population.¹¹ The overall percentage of children with SCD in our study who received early intervention services was 9.72%, slightly less than the national average of 12.22% for Black children in the United States.³⁴

This is the first study to describe the prevalence rate of NICU care for children with SCD, which was 12.03% in our cohort study. NICU admissions are likely a proxy for neonatal complications. Consistent with previous studies in the general population, our findings indicate that history of NICU admission was an identifiable risk factor for future neurocognitive difficulties in children with SCD, including hyperactivity, and emotional dysregulation.^35,36 In addition to the novel findings related to NICU admission, our results demonstrate that birthweight was generally more sensitive to neurocognitive and educational outcomes, compared to gestational age. Our findings are consistent with prior studies that have reported worse neurodevelopmental, behavioral, and educational outcomes for term babies who were small for gestational age or those with fetal growth restriction.^37–40 These findings may also reflect increased objectivity with birthweight data compared to gestational age, the latter which can be estimated using different methods and is subject to variability.⁴¹

The current results demonstrate the importance of considering perinatal risk factors when monitoring and evaluating cognitive outcomes in SCD. Brief medical record review or discussion with caregivers can identify children with a history of perinatal risk factors who would benefit from additional follow-up with subspecialties such as neurology, neuropsychology, and occupational, speech, and/or physical therapy. This informal option may be preferred by hematologists until more targeted and applicable screening measures are available for young children with SCD.⁴² Children with a known history of perinatal risk factors should be referred for a developmental and/or neuropsychological evaluation, depending on the age of the child, to (i) establish a baseline in order to track future cognitive development; (ii) support differential diagnosis for cognitive, academic, and behavioral concerns; and (iii) facilitate intervention support.⁴³

Identifying perinatal risk factors in children with SCD during routine clinical care can offer another route to access intervention or support services during early and middle childhood in clinical, school, and community settings. Currently, only five states offer explicit early intervention services for SCD, and additional advocacy is needed to close the equity gap.⁴⁴ In contrast, availability of resources for perinatal risk factors may be more accessible from the perspective that having a medical history alone often qualifies children to receive a number of free services (e.g., state-based early childhood intervention, formal school programming to support learning needs) or services automatically covered by insurance (e.g., cognitive intervention or occupational therapy). Our findings demonstrate that children with SCD who had perinatal risk factors were likely to receive early intervention services, but not formal school support services later in childhood. This may reflect children responding positively to intervention and no longer requiring services, attenuation of risk factors after exposure to intervention, or a misunderstanding among parents and providers regarding the continued risk for cognitive and behavioral difficulties beyond early childhood for children with early gestational age, LBW, or history of NICU admission. The latter is problematic from the standpoint that early intervention gains can be short-term, and these services do not eliminate future cognitive risk.⁴⁵ Continued intervention and/or academic support for perinatal risk factors beyond early childhood in the general population has shown promise and may be of benefit for children with SCD.^46–49

This study has several strengths, including a large cohort of children with a prevalence of LBW that parallels national statistics for Black children in the United States. However, several limitations should be considered. Participants with missing perinatal history were excluded from analysis, and future studies should consider prospective collection of these variables. Detailed perinatal information beyond birthweight, gestational age, and NICU admission status was unavailable via medical records. Further research is needed to examine additional perinatal data (e.g., number of days in the NICU or specific medical intervention(s) provided while receiving care) to clarify the relationship between SCD and perinatal risk factors with neurocognitive outcomes. This study was focused on effects of LBW, and NICU admission in children and adults were excluded from analysis. Genotype in our sample was heterogenous, with only 56.84% of participants having the HbSS/Sβ⁰ thalassemia genotype. The rate of silent cerebral infarcts was unknown in our sample, as all children did not receive imaging as standard of care, thus the potential impact is unknown. Likewise, we did not include a non-SCD control group with a history of prematurity, LBW, or NICU stay. As such, it is uncertain if these perinatal risk factors act in a way that is different from the general population. Future studies should consider including parent genotype as a covariate to examine potential influence on utilization of available intervention resources for children with SCD and perinatal risk factors.

5 |. CONCLUSION

The impact of health-related disparities in SCD is present at birth. In our study, all participants with SCD were Black and the rate of LBW was 13.50%, which is consistent with national statistics for Black children, and higher than prevalence rates of LBW in White children. History of LBW, earlier gestational age, and NICU admission was associated with worse neurocognitive outcomes and history of early intervention services, regardless of SCD genotype or disease management. Perinatal risk factors represent early medical events for children with SCD, and should be considered clinically and in research when examining cognitive and academic outcomes. The cumulative neurocognitive risk associated with both SCD and perinatal adversities highlights the importance of screening for comorbid neurodevelopmental risk factors in children with SCD. Identification of perinatal risk factors in children with SCD can help expand access to intervention services, and provides an opportunity to educate caregivers on the needs for long-term intervention support throughout childhood.

Abbreviations:

CVA

cerebrovascular accident

IEP

Individualized Education Program

LBW

low birthweight

NICU

neonatal intensive care unit

SCD

sickle cell disease

SVI

Social Vulnerability Index

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author.