Autism Spectrum Disorder Screening at 18–36 Months in Infants with Moderate and Severe Neonatal Encephalopathy: Is Routine Screening Required?

Abstract

Background

Hypoxia, acidosis, and inflammation cause impairment of neuronal development due to lack of sufficient oxygen and nutrition, and there may be an increased risk of ASD in neonates with Neonatal Encephalopathy (NE). To evaluate the frequency of Autism Spectrum Disorder (ASD) in moderate and severe NE requiring therapeutic hypothermia at 18–36 months.

Methods

In this prospective study, infants with moderate and severe NE requiring hypothermia were included.

Results

Throughout the study period, 33 of 85 neonates with NE admitted to our unit were included in the study. M-CHAT results of six infants included in the study showed positive and results of twenty-seven patients were negative. Four of six infants with a positive screening test completed a psychiatric examination, and two infants did not complete the clinical examination. The results of the examination showed that one infant was diagnosed with ASD, one was diagnosed with attention deficit hyperactivity disorder (ADHD), and two infants were diagnosed with no psychiatric disorder.

Conclusion

Our results showed that the risk of ASD increases in infants with moderate and severe NE, and this relationship should be considered in long-term clinical follow-ups.

Introduction

Neonatal encephalopathy (NE) is a clinical syndrome and leading cause of 15%–20% of all neonatal deaths, with a 25% persistent deficit. The incidence of NE is 2–9 per 1000 live births.^1–3 Hypoxia and acidosis due to impaired gas exchange cause cellular energy metabolism deficits, decreased cerebral perfusion, insufficient cerebral glucose, and oxygenation.^4,5 The brain is more sensitive to metabolism deficits in the perinatal period than any other time in life, resulting in NE with variable severity due to hypoxia.^6–8 Many studies have been conducted to evaluate the long-term effects of brain injury on NE on neurological, cognitive, and behavioral development.^9–11 Most of the studies focused on the detection of structural brain damage and the effect of therapeutic hypothermia on the damage using imaging methods in moderate and severe NE. These studies have shown that structural damage causes cerebral palsy (CP), epilepsy, and hearing impairments.^12–14 Further evidence on neurological, cognitive, and behavioral long-term outcomes of NE has been reported in recent years. Furthermore, even in the presence of mild NE, the frequency of neurological deficits increased and the 5-year-old cognitive functions were more impaired than healthy peers.^15–18 However, it was reported that cognitive and behavioral deficits may present without motor deficits, and structural damage is not necessary for cognitive and behavioral deficits. Hence, there is evidence that the negative consequences of NE may cause cognitive impairment or behavioral deficits, not limited to neurological deficits.^19–21 There is evidence that the prevalence of autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD) increases in the presence of NE.^22–24 In addition, a recent meta-analysis showed that the risk of ASD has increased in neonates exposed to hypoxia.²⁵ A case-control cohort study of a large sample group with NE in Western Australia without CP showed a low development level than the healthy controls, using the Griffiths development scale.²⁶

Contrarily, ASD is a neurodevelopmental disorder and reportedly, its prevalence is increasing at 18,5 per 1000 children (1 of 54).²⁷ Studies have shown that pregnancy complications, preterm birth, cesarean delivery, breech presentation, fetal distress, perinatal hemorrhage, multiple pregnancies, perinatal asphyxia, infections, and induced and prolonged labor have been associated with ASD.^28–31 Few studies have reported on the relationship between NE and ASD. However, there is increasing evidence that cognitive and behavioral follow-up should be performed along with neurological development follow-up of patients with NE.

Thus, we hypothesize that hypoxia, acidosis, and inflammation cause impairment of neuronal development due to lack of sufficient oxygen and nutrition, and we believe there may be an increased risk of ASD in neonates with NE. In addition, placental and obstetric complications may cause bloodstream deficiency and toxic metabolites depot, eventually damaging neuronal development.^32–34 Therefore, we aimed to evaluate the related ASD features of neonates with NE requiring therapeutic hypothermia, according to the Neonatal Society Guidelines on Neonatal Encephalopathy.³⁵

Methods

Study Design and Population

Our study was designed prospectively and was approved by the ethics committee of our institution. In our unit, therapeutic hypothermia was applied to neonates with gestational age >35 weeks and birth weight of 1800 g with moderate and severe NE, according to Sarnat’s staging and umbilical cord blood pH ≤7, base deficit ≤–16.

After the approval of the ethics committee, the information of patients diagnosed with NE from 2015 to 2018 in our unit was collected from the hospital database. Children <18 months, >36 months, with mild NE, exitus, and when lack of medical and communication information were excluded from the study. The infants’ parents were contacted through the contact numbers collected from the hospital database. Infants whose parents could not be reached by phone or parents who did not agree to attend the control examination or who failed to attend the control examination were excluded from the study. Neonatologist (BK) performed the neurological examination of the patients. The infants with neurological deficits were excluded from the study and the parents of infants without neurological deficits were informed of the ASD screening test, M-CHAT. Written consent was obtained from the parents who agreed to participate in the study. All interviews were held in the same place with the same neonatologist. The questions on the scale were asked to the parents with some necessary information by the neonatologist.

ASD Screening Test

Modified Checklist for Autism in Toddlers (M-CHAT) is a screening scale consisting of 23 questions with yes/no responses, developed for the scanning of autism symptoms in 18–36-month-old infants. M-CHAT is a scale usually filled by parents. However, in the Turkish version of the scale, it was stated that it would be more proper to apply it with the help of health professionals.³⁷ In the Turkish cultural adaptation of the M-CHAT test, if minimum 2 questions numbered 2, 5, 7, 9, 10, 14, 15, 17, 19, and 21 or 3 of the 23 questions are marked, the screening test result is considered positive and the infant is considered high risk for ASD.³⁷

Clinical Examination

As a result of the M-CHAT screening scale, infants with positive screening tests were examined by a child and adolescent psychiatrist (BS), according to Diagnostic And Statistical Manual Of Mental Disorders, Fifth Edition (DSM-5).³⁹

Sociodemographic Data Collection

The data pertaining to gestational age, birth weight, gender, maternal age, mode of delivery, delivery room resuscitation, Apgar scores at first and fifth min, umbilical cord blood pH and base deficit, history of seizures on amplitude-integrated electroencephalography, Sarnat stage, received surfactant, invasive ventilation duration, frequency of pulmonary hypertension, and hospitalization duration were collected from the hospital database.

Statistical Analysis

Statistical analysis was performed using SPSS version 22. Groups were compared using Student’s t-test or the Mann–Whitney U test as appropriate. Correlation analyses were performed using Pearson’s correlation test for parametric variables and Spearman’s test for nonparametric variables.

Results

Throughout the study period, 33 of 85 neonates with NE admitted to our unit were included in the study. The study algorithm is shown in Figure 1. M-CHAT test results of six children included in the study showed positive and results of twenty-seven patients were negative (Figure 2). Four of six patients with a positive screening test completed a clinical examination, and two patients did not complete the clinical examination. The results of the clinical examination showed that one infant was diagnosed with ASD, one was diagnosed with ADHD, and two infants were diagnosed with no psychiatric disorder.

Figure 1.

The Study Flowchart

Figure 2.

Flowchart of the Screening Tests

The demographic and clinical characteristics of the infants with a positive and negative screen on the M-CHAT were shown in Table 1. Infants with positive M-CHAT had lower birth weight, higher maternal age, more resuscitation requirement, lower Apgar score at first and fifth min, lower umbilical cord blood pH and base deficit values, and higher seizure frequency and a higher Sarnat stage than those with negative M-CHAT screening test. The infants’ perinatal risk factors were summarized in Figure 3.

Table 1. Demographic Information for M-CHAT Screening Test.

	N		AGE RANGE		AGE MEAN (SD)		MALE N (%)		FEMALE N (%)
M-CHAT positive		6		19–36 months		30,3 months (3,2)		5 (83,3%)		1 (16,6%)
M-CHAT negative		27		18–36 months		32,1 months (4,4)		14 (51,8%)		13 (48,1%)

M-CHAT, Modified Checklist for Autism in Toddlers; SD, standard deviation.

Figure 3.

Demonstration of Perinatal Risk Factors

1: Cephalopelvic Disproportion, 2: Abruptio Placenta, 3: Preeclampsia, 4: Diabetes Mellitus, 5: Chorioamnionitis, 6: Umbilical cord compression, 7: Meconium aspiration syndrome.

The sociodemographic characteristics of the infants according to the screening test results were shown in Table 2. Comparing the groups with positive and negative screening test results, the male gender was higher in the positive group (83,3%); however, there was no statistically significant. The aberrant answers with positive results were shown in Figure 4. The questions with the highest aberrant response rate were determined as 2nd, 5th, 8th, and 23rd questions.

Table 2. The Characteristic Features of the Study Population.

		SCREENING NEGATIVE (N = 27)		SCREENING POSITIVE (N = 6)		P VALUE
Gestational age (weeks)		38,7 ± 1,48		37,3 ± 1,86		0,06
Birth weight (g)		3280 ± 474		2747 ± 569		0,022
Male		14 (51,8)		5 ± (83,3)		0,969
Maternal age		27,7 ± 4,7		32,6 ± 6,9		0,042
Cesarean delivery		13 (48,1)		5 (83,3)		0,103
Delivery room resuscitation		26 (96,2)		6		0,019
Intubation		25		2
Chest compression		1		3
Epinephrine		0		1
Apgar scores (min)
first		2,96 (1–5)		1,33 (0–4)		0,009
fifth		5,14 (3–6)		3,16 (2–5)		0,000
Umblical cord blood pH		6,97 ± 0,07		6,8 ± 0,12		0,018
Umblical cord blood base deficit (mmol/L)		–15,5 ± –2,6		–19,8 ± –4,5		0,004
Seizures		14		5		0,030
Sarnat stage						0,030
2		13		1
3		14		5
Received surfactant						0,099
0		17		2
1		7		2
2≤		0		2
Invasive ventilation duration (days)		3,8 ± 1,4		10,1 ± 6,7		0,071
Pulmonary hypertension		5		3		0,44
Hospitalization duration (days)		10,4 ± 2,5		21,8 ± 16,9		0,162

Results are expressed as n (%), mean ± standard deviation, or median [IQR].

Figure 4.

Comparison of Specific Questions from the Modified Checklist for Autism in Toddlers (M-CHAT) Questionnaire Failed in M-CHAT Positive and Negative Infants

The correlation analysis related to the hypothesis that the M-CHAT scores of the infants may be correlated to Apgar scores (1st and 5th min), umbilical cord blood gas pH and base deficit value, invasive ventilation, and duration of hospitalization is shown in Table 3. The highest correlation coefficient with M-CHAT score were 5-min Apgar score (r:–0,614), pH (r:–0,567), invasive ventilation duration (r:–0,534), base deficit (r:–0,530), and duration of hospitalization (–0,453), respectively. The correlations were statistically significant (p < 0.01). The correlation coefficient with the M-CHAT score Apgar score at first minute was r = –0.436 (p:0,011).

Table 3. Pearson Correlation Analyses between M-CHAT Scores and Neonatal Clinic and Laboratory Findings.

		M-CHAT SCORE		APGAR AT FIRST MINUTES		APGAR AT FIFTH MINUTES		PH		BASE DEFICIT		INVASIVE VENTILATION DURATION		HOSPITALIZATION DURATION
M-CHAT Score
r		1		–,436*		–,614**		–,567**		–,530**		,534**		,453**
p value				,011		,000		,001		,002		,001		,008
Apgar at first minutes
r		–,436*		1		,764**		,323		–,205		–,389*		–,338
p value		,011				,000		,067		,253		,025		,055
Apgar at fifth minutes
r		–,614**		,764**		1		,394*		–,349*		–,467**		–,423*
p value		,000		,000				,023		,047		,006		,014
pH
r		–,567**		,323		,394*		1		–,423*		-,423*		–,450**
p value		,001		,067		,023				,014		,014		,009
Base deficit
r		,530**		–,205		–,349*		–,423*		1		,264		,284
p value		,002		,253		,047		,014				,137		,110
Invasive ventilation duration
r		,534**		–,389*		–,467**		–,423*		,264		1		,916**
p value		,001		,025		,006		,014		,137				,000
Hospitalization duration
r		,453**		–,338		–,423*		–,450**		,284		,916**		1
p value		,008		,055		,014		,009		,110		,000

*Correlation is significant at the 0.05 level.

**Correlation is significant at the 0.01 level.

Discussion

We aimed to evaluate the relationship between NE and ASD in infants with moderate and severe NE requiring therapeutic hypothermia at 18–36 months. We found that the M-CHAT positivity rate of infants with moderate and severe NE between at 18–36 months was higher than the general population. In this study, the M-CHAT positivity rate reported in the general population was 4.4%–9.4%, whereas in infants with moderate and severe NE, it was 18.1% (six of thirty-three).^36–38 Although 66% (four of six) of the infants with positive screening tests were clinically evaluated, the ASD frequency was 3% (one of thirty-three).

According to the latest Center for Disease Control and Prevention data, the prevalence of ASD has increased with time, and the latest prevalence is 1,8% (one in fifty-four). This study suggested that infants with moderate and severe NE were at a higher risk for ASD than the healthy population.^27,40 ASD prevalence in infants with NE found in our study was consistent with the study conducted by Badawi et al.⁴¹ Badawi et al. suggested a strong relationship between moderate to severe NE and ASD development in the population-based case-control study. They found that ASD frequency was 5% in infants with moderate and severe NE. It was also reported that infants with NE were six times more likely to be diagnosed with ASD than healthy controls.⁴¹ Furthermore, the results of our study are consistent with studies that reported a higher risk of ASD in infants with moderate and severe NE.^42–46

In addition to ASD, there is evidence of increased ADHD incidence, tractability, aggression, passivity and anxiety, and cognitive dysfunction on the neurodevelopmental outcomes of NE. Previous studies have reported an increase in the frequency of cognitive dysfunction, tractability, aggression, passivity, and anxiety in mild NE,^18,21,47 and cognitive dysfunction and ADHD in moderate and severe NE.^{11,13,14,20,21} Despite a study based on parents’ observations, infants with NE were not diagnosed with ASD and ADHD, and it was found that the cognitive functions of these infants were impaired and tractability, aggression, passivity, and anxiety problems were reported more than the control group.²¹ In our study, 3% (one of thirty-three) ADHD was diagnosed and this result was consistent with studies that reported an increased ADHD prevalence in infants with NE.

Contrarily, many perinatal risk factors with NE have been identified for ASD. Many studies and meta-analyses have reported prematurity, higher parental age, prenatal maternal drug use, bleeding during pregnancy, gestational diabetes, preeclampsia, cesarean delivery,²⁸ hypertension, hypoxia,^42–45 seizure,⁴⁶ and low Apgar score,^25,48 with evidence of increased ASD risk. In addition, Ravi et al. reported that a positive ASD screening test was most common in children who had seizures in the neonatal period.³⁸ Here, we observed a statistically significant difference in the frequency of seizures in infants with positive and negative M-CHAT results (p:0.03). Many studies also reported that the incidence of ASD following prematurity, hypoxia, and complicated pregnancies is higher in males than females.^44,49–52 We observed that positive ASD screening test was higher in males than females. However, there was no statistically significant (p:0,969).

Some theories have proposed that hypoxia causes neurodevelopmental disorders. Reportedly, one study had divided moderate NE neonates into low and high severity groups based on the severity of the neonatal symptoms, followed by a comparison between the groups’ cognitive functions.²⁰ In this study, Marlow et al. observed that the school success of the low-symptom severity group was higher than the high-symptom severity group. They further argued that cognitive functions deteriorated with an increasing degree of hypoxia in moderately severe NE, and hypoxia has a dose-response effect on cognitive functions. In addition, Low et al. reported that there was a threshold for asphyxia in brain damage, and when the threshold exceeded, minor or major deficits might be seen.⁴⁷

Some studies suggested that the hippocampus and striatum may be affected in the NE-related brain damage.^53–57 These structures have been associated with cognitive functions and were believed to play a role in the pathogenesis of ASD, ADHD, and schizophrenia.^58–62

Contrarily, ASD-related perinatal risk factors are the same environmental factors in ASD etiology. They are also risk factors for NE. These factors may lead to both NE and ASD, and NE is the first sign of ASD in the neonatal period. Future research on the causes of ASD and common risk factors that can lead to both ASD and NE is warranted.

Finally, in our study, we did not aim to specify any results of ASD etiology. However, our results showed that the risk of ASD increases in children with NE, and this relationship should be considered in long-term clinical follow-ups. NE may be a high-risk profile for autism. Therefore, early diagnosis and interventions are important for these children. Early studies showed that earlier interventions for ASD can lead to a better prognosis.

Perinatal risk factors and hypoxia can increase the frequency of ASD. There is evidence that early diagnosis and treatment of ASD can improve the prognosis.^63–65 The diagnosis of ASD can be established reliably before the age of 2.^64–68 APA suggested that all children undergo a screening test for ASD between the ages of 18–24 months, regardless of the risk factors.⁶⁹

In conclusion, as the aim is to diagnose early between the ages of 18–24 months, we believe that the M-CHAT screening test is suitable for use in the screening of this age group, adaptable to all countries, and can be applied by all healthcare workers with sufficient sensitivity (92%), and it will be the first step for a consultation to child psychiatry clinics.⁷⁰ Contrarily, regardless of the severity, children with NE should be monitored for cognitive functions and psychiatric symptoms even if they are not diagnosed with ASD. Otherwise, in the general community, children with these conditions are likely not diagnosed or mislabeled as behavioral problems or mental disabilities.

This study has two limitations. First, the study sample size is small. Second, there is no control group. We believe that these limitations reduce the reliability of our data as a psychiatric examination of all children with NE could not be performed and two patients who were found positive for ASD screening test refused to psychiatric examination.