Neurobehavior in Low-Risk Very Preterm Infants with Low Medical Risk and Full-Term Infants

Abstract

Objective:

To describe differences in neurobehavior among very preterm infants with low medical risk at term equivalent age and full-term infants.

Study Design:

One-hundred eighty-six (136 infants born ≤ 32 weeks gestation with low medical risk at term equivalent age and 50 full-term infants within 4 days of birth) had standardized neurobehavioral assessments. Low medical risk was defined by ventilation < 10 days and absence of significant brain injury, necrotizing enterocolitis, patent ductus arteriosus, and retinopathy of prematurity.

Results:

Very preterm infants with low medical risk at term equivalent age demonstrated more sub-optimal reflexes (p<0.001; ß=1.53) and more stress (p<0.001; ß=0.08) on the NICU Network Neurobehavioral Scale compared to their full-term counterparts. Very preterm infants with low medical risk also performed worse on the Hammersmith Neonatal Neurological Examination (p=0.005; ß=−3.4).

Conclusion:

Very preterm infants at term equivalent age continue to demonstrate less optimal neurobehavior compared to full-term infants.

Introduction

Despite evolving standardization of NICU care and improved survival among preterm infants born at early gestational ages, very preterm infants (born < 32 weeks gestation) remain at risk for neurodevelopmental (1), social (2), and academic challenges (3). These infants are also at an increased risk for medical complications and may have socioeconomic risk factors, which can impact neurodevelopmental outcome (4). Further, research has demonstrated that preterm infants with cerebral injury have a high risk of neurobehavioral morbidity, cognitive impairment, and developmental delay (5). However, late preterm infants with no acute, severe, or chronic illness and no identified cerebral injury also have a higher risk of long term developmental challenges (6). Therefore, a better understanding of how immaturity at birth, in the absence of medical comorbidities, impacts early outcomes of very preterm infants is important.

The developmental differences in infancy and early childhood among infants born late preterm compared to full-term, as well as among very preterm infants compared to full-term infants, has been described (7-10). Many of these studies assess neurodevelopmental outcome after NICU hospitalization in late infancy and early childhood (1, 10, 11). However, there are several neurobehavioral assessments that can be used to identify early alterations in neurobehavioral outcomes during the neonatal period prior to NICU discharge (12-16). It is well-defined that preterm infants have alterations in early neurobehavior during the neonatal period, when compared to full-term infants (17-19). While studies on late preterm infants may include healthy samples, studies of very preterm infants include samples of infants with multiple medical complications that occur alongside preterm birth. Although many of these studies control for medical complications during statistical analysis, it remains unclear whether preterm infants without major medical complications and/or cerebral injury also exhibit altered neurobehavior.

By understanding developmental differences in a very preterm sample who did not experience significant medical complications compared to full-term infants, we can better understand the early developmental trajectory without the influence of medical comorbidities. This study aimed to define neurobehavior in very preterm infants with low medical risk at term equivalent age and investigate differences with neurobehavior in full-term infants. We hypothesized that although a cohort of very preterm infants had low medical risk and had reached the developmental age of a full-term infant, there would still be differences in neurobehavioral outcome across both motor and behavioral scores.

Methods

This study was approved by the University of Southern California Office for the Protection of Research Subjects. This study included infants from several prospective longitudinal cohorts (multi-cohort study) of infants enrolled within the past 10 years from 2011-2018 (20-25). The original studies were approved by the Washington University Human Research Protection Office. Parents signed informed consent, which included permission for neurobehavioral assessment at term equivalent age.

Very preterm infants with low medical risk

One-hundred and thirty-six infants enrolled within 7 days of life from 2011-2018 and who met the study inclusion criteria were included in the very preterm group with low medical risk. From the overarching cohorts, consecutive admissions during the study enrollment periods of preterm infants were recruited if they were born ≤ 32 weeks estimated gestational age (EGA) and were excluded if they had a congenital anomaly and/or were not expected to survive according to the physicians’ opinion within the first 24 hours of life. EGA was defined by dates (from the mother’s last menstrual period) and confirmed by Ballard exam. Considering available data across each cohort, to achieve a sample with low medical risk, infants were further excluded from this study if they had the following medical conditions during their NICU stay: grade III or IV intraventricular hemorrhage (IVH), cystic periventricular leukomalacia (cPVL), ventilation for greater than 10 days, confirmed necrotizing enterocolitis (NEC; all stages), patent ductus arteriosus (PDA; treated with medications or surgical ligation), or retinopathy of prematurity (ROP; any stage) by 34 weeks postmenstrual age (PMA). PMA was defined as the EGA plus the number of completed weeks since birth. Infants with the aforementioned complications were excluded due to their known relationships with outcome (26-29), as the intent of this study was to compare a sample of preterm infants with low medical risk to a group of full-term infants. Further, infants who did not have neurobehavioral testing in the NICU prior to 45 weeks PMA were also excluded from the overarching studies, which enrolled consecutive admissions during the study time period.

All preterm infants received standard of care in the level IV NICU at St. Louis Children’s Hospital. This NICU had 75-beds during the initial enrollment periods and expanded to 132-beds for later cohorts. Two of the cohorts that made up the sample were embedded in randomized clinical trials, one evaluating alternative positioning and the other evaluating the effect of a sensory-based intervention (20, 23).

Preterm infants had neurobehavioral testing conducted at term-equivalent age (range 34-45 weeks PMA; average PMA of 36.0 ± 1.9 weeks). Assessments were done as early as 34 weeks PMA, when infants were ready for discharge home or as late as 45 weeks for those who did not get discharged until later. Assessments in the NICU were done at the infant’s bedside shortly before a feeding or care time. Infants had to be on < 2 liters of oxygen and able to tolerate handling without physiological consequences to have the neurobehavioral assessments.

Full-term infants

The full-term group consisted of fifty consecutive admissions of full-term infants born at ≥ 37 weeks gestation enrolled from May 2016 to August 2016 as part of an overarching study that enrolled a full-term comparison group. Mother-infant dyads were hospitalized on the 35-bed labor and delivery floor of Barnes-Jewish Hospital in St. Louis, MO. Infants were excluded if they were admitted to the NICU, had an identified cerebral injury, needed respiratory support or intravenous medications at birth, and/or if their mother was less than 18 years or was not fluent in English. The full-term infants had neurobehavioral testing in the mother’s room within 4 days of birth (range 37 to 41 weeks PMA; average PMA of 38.7 ± 1.0 weeks).

Medical and sociodemographic factors

Medical and sociodemographic information were extracted from the electronic medical record. Social factors collected included infant sex, infant race (Caucasian/White or not Caucasian/White), maternal age, maternal marital status (married or single), and insurance type (public or private). Medical factors collected included EGA, birthweight, length of stay, and mode of delivery. The study site NICU largely admitted infants who define themselves as White/Caucasian or Black/African-American, with < 5% representation from other racial/ethnic groups. Information on the other ethnic and racial groups was not consistently captured across all cohorts. Medical factors were collected largely to ensure appropriate inclusion, in order to define that the infant was low medical risk. However, differences across groups were also explored and included in the statistical model, when appropriate. PMA at the time of neurobehavioral testing was also recorded.

Outcome measures

Neurobehavioral testing using the NICU Network Neurobehavioral Scale (NNNS) and Hammersmith Neonatal Neurological Exam (HNNE) was performed by a trained and certified examiner. There are many items that overlap between the NNNS and HNNE, therefore the NNNS items were completed first followed by the 3 items unique to the HNNE. There were approximately 4 different trained and certified evaluators across the different cohorts. They did not have information about the current study’s hypothesis, but understood that the full-term group would be used as a comparison across many different measures. For all infants, assessments were completed up to 60 minutes before a scheduled or anticipated feeding or care time. All evaluations were videotaped to aid scoring when needed. While scoring was done from memory of the in-person performance on each item, due to the significant number of items on the assessment(s), the videos were used when an item (such as palmar grasp) could not be remembered by the examiner. As all infants were in the hospital during the time of assessment, evaluators could not be blinded as to whether infants were preterm or full-term.

NNNS:

The NNNS is a 115-item test that can be administered in approximately 20-25 minutes. Raw scores on the NNNS are calculated into 13 summary scores: Habituation, Orientation, Hypertonicity, Hypotonicity, Excitability, Arousal, Lethargy, Sub-optimal Reflexes, Asymmetry, Stress, Self-regulation, Quality of Movement, and Handling. The habituation items on the NNNS require a quiet environment and were not captured across many of the cohorts due to the inability to control this with ease. Therefore, Habituation scores were not included in this study. Higher summary scores on the NNNS indicate more of that construct, meaning some high scores are considered better neurobehavior (higher self-regulation scores reflect more regulated behaviors; higher attention scores reflect more attention), and other high scores are considered worse neurobehavior (higher hypertonia scores indicate more hypotonia; higher stress scores indicate more stress). The NNNS has been used extensively to evaluate preterm infants and has good predictive validity and reliability (30).

HNNE:

The HNNE is a 34-item assessment that takes approximately 10 minutes to administer by a trained evaluator. For the optimality score, each item is assigned a value of 0, 0.5, or 1 based on the infant’s performance, with some scoring adjustments available based on the infant’s PMA (31). When items resulted in asymmetry, the least optimal side was scored. Scores for each of the 34 items were summed and a total score was achieved, with higher scores indicating better performance (31). Items on the assessment are grouped into 6 categories: tone, tone patterns, reflexes, movements, abnormal signs, and behavior. Subscale scores for each of these categories can also be achieved. The HNNE has excellent reliability and good validity (32).

Data analysis

Statistical analyses were conducted using SPSS Version 28 (IBM, Chicago). Independent samples t-tests and chi-square analyses were used to explore differences in biomedical and social factors. Although some factors were expected to differ across preterm and full-term groups (such as length of stay, EGA, birthweight), others that differed across groups were considered for inclusion in the statistical model. Those that differed (p<0.05) were explored for co-linearity prior to inclusion. Independent samples t-tests were used to determine differences in group means on the NNNS summary scores and HNNE total score across groups (preterm infants with low medical risk compared to full-term infants). Multivariate regression models were then used to investigate differences in neurobehavior across groups, while controlling for PMA at the time of testing, as PMA has previously been reported to impact neurobehavioral outcome (19) in addition to other factors identified to be different across groups. All analyses were conducted with an α < 0.025 (Bonferroni correction α =0.05/2, with correction for the number of outcome measures used) to balance the risk of a Type I and Type II error. Analyses were also re-run controlling for treatment assignment to account for participants who received an intervention as part of a randomized clinical trial, to ensure the findings remained consistent. Finally, analyses were re-run controlling for cohort to ensure findings remained consistent.

We also graphed neurobehavioral scores across infants with different EGAs and reported test statistics in the figures for investigations of relationships between EGA and neurobehavioral scores using multivariate linear regression controlling for PMA at the time of testing and other medical and sociodemographic factors that differed across groups.

Results

See Figure 1 for a flow diagram. Four hundred thirteen infants who were enrolled across cohorts from 2011-2018 were screened for inclusion. Two hundred twenty seven infants were excluded due to having 1 or more of the identified factors for exclusion, with some excluded for multiple criteria. There were 57 (14%) who did not have testing during the range of PMA specified or who had testing at home after discharge. There were 56 (14%) who had Grade III-IV IVH or PVL, 87 (21%) who had ventilatory support for > 10 days, 28 (7%) with NEC, 108 (26%) with PDA, and 78 (19%) with ROP. After excluding infants for one or more reasons, there were 186 infants who were included in this study (136 very preterm infants with low medical risk and 50 full-term infants).

Figure 1.

Flow diagram of enrolled and excluded infants.

The mean PMA of the low-risk very preterm infant group at the time of testing was 36.0 ± 1.9 weeks; the mean PMA of the full-term infant group was 38.7 ± 1.0 weeks (p < 0.001).

Table 1 identifies the social and medical characteristics of infants in the two groups with differences across groups reflecting expected differences in EGA, birthweight, and days of hospitalization. The preterm group had more mothers who were married (p=0.02), White/Caucasian (p=0.04), had a Caesarean delivery (p < 0.001), and had public insurance (p=0.01). None of these factors were related to each other; therefore, the multivariate model that investigated differences in preterm and full-term infant neurobehavior controlled for PMA at the time of testing, insurance type, mode of delivery, race, and marital status.

Table I.

Maternal and social characteristics of the sample.

Baseline Factors (N=186)	Full-term (n=50) Mean (SD) or N (%)	Low-risk Preterm (n=136) Mean (SD) or N (%)	P-value
Continuous Factors, M (SD)
Maternal age, y	27.5 (6.0)	27.7 (6.7)	0.83
Estimated gestational age, weeks	38.6 (1.0)	29.8 (2.0)	< 0.001
Birthweight, grams (n=176)	3315.0 (444.8)	1410.4 (384.5)	< 0.001
Length of stay, days	2.6 (1.0)	54.1 (25.7)	< 0.001
PMA at time of testing, weeks (n=185)	38.7 (1.0)	36.0 (1.9)	< 0.001
Endotracheal intubation, days (n=176)	NA	1.2 (1.8)
Categorical Factors, N (%)
Infant sex, female	29 (58%)	77 (57%)	0.87
Infant race, Caucasian/White	11 (22%)	57 (42%)	0.04
Insurance type, public (n=182)	33 (66%)	59 (45%)	0.01
Marital status, married (n=170)	13 (26%)	54 (45%)	0.02
Mode of delivery, Casearean (n=176)	17 (34%)	102 (75%)	< 0.001

^*n=186 unless otherwise specified; p-value is from investigating differences across groups using independent samples t-tests and chi-square analysis.

Table 2 identifies neurobehavioral scores among very preterm infants with low medical risk and full-term infants with associated percentiles, ranges, means, and standard deviations.

Table 2.

NNNS summary scores and HNNE score distributions.

	Full-term Group				Preterm Group
	M (SD)	Range	Percentiles (10, 25, 50, 75, 90)	* Percentile of mean score, based on full-term normative comparison	M (SD)	Range	Percentiles (10, 25, 50, 75, 90)	* Percentile of mean score, based on full-term normative comparison
NNNS
Orientation	4.1 (1.3)	2.0-7.0	2.6, 3.0, 4.0, 5.0, 6.0	5	3.9 (1.2)	1.0-7.1	2.0, 3.0, 3.9, 5.0, 5.7	5
Hypertonicity	0.9 (0.9)	0.0-4.0	0.0, 0.0, 1.0, 1.0, 2.0	*	0.9 (1.2)	0.0-5.0	0.0, 0.0, 0.0, 2.0, 3.0	*
Hypotonicity	0.4 (0.6)	0.0-2.0	0.0, 0.0, 0.0, 1.0, 1.0	50-75	0.7 (0.9)	0.0-6.0	0.0, 0.0, 1.0, 1.0, 2.0	50-75
Excitability	4.3 (3.0)	0.0-10.0	1.0, 2.0, 4.0, 6.3, 9.0	50	4.5 (2.7)	0.0-12.0	1.6, 3.0, 4.0, 6.0, 8.4	50
Arousal	3.9 (1.0)	1.7-5.6	2.6, 3.3, 4.0, 4.6, 4.9	25	3.9 (0.8)	1.9-6.3	2.9, 3.3, 3.9, 4.3, 5.0	25
Lethargy	5.8 (3.3)	0.0-14.0	2.0, 3.0, 5.0, 8.0, 11.0	50	6.5 (2.9)	0.0-14.0	3.0, 4.0, 6.0, 8.0, 11.0	75
Non-Optimal Reflexes	5.0 (1.9)	1.0-9.0	2.1, 4.0, 5.0, 6.3, 7.0	75	6.6 (1.9)	2.0-11.0	4.0, 5.0, 7.0, 8.0, 9.0	90
Asymmetric Reflexes	1.2 (1.1)	0.0-4.0	0.0, 0.0, 1.0, 2.0, 2.9	50-75	2.4 (1.8)	0.0-7.0	0.0, 1.0, 2.0, 4.0, 5.0	75-90
Stress	0.2 (0.1)	0.04-0.4	0.1, 0.1, 0.2, 0.2, 0.2	90	0.3 (0.1)	0.1-0.5	0.1, 0.2, 0.3, 0.4, 0.4	95
Self-Regulation	4.7 (1.0)	2.9-7.1	3.1, 4.0, 4.9, 5.4, 6.0	25	4.6 (0.9)	2.2-6.5	3.4, 4.0, 4.7, 5.3, 5.9	10
Quality of Movement	3.9 (0.6)	2.5-5.7	3.2, 3.5, 3.8, 4.2, 4.8	25	3.6(0.9)	1.3-5.8	2.3, 3.0, 3.5, 4.3, 4.8	10
Handling	0.5 (0.4)	0.0-1.0	0.0, 0.0, 0.4, 1.0, 1.0	90	0.6 (0.3)	0.0-1.0	0.0, 0.5, 0.6, 0.9, 1.0	90
HNNE Optimality Score	21.9 (4.3)	10.5-30.5	16.3, 18.8, 22.5, 25.4, 27.3	Suboptimal	18.0 (4.8)	8.0-29.0	11.5, 14.0, 18.0 21.5, 24.5	Suboptimal

Full term infants were ≥ 37 weeks gestation without an identified cerebral injury, needed respiratory support or intravenous medications at birth. Mothers were > 18 years and fluent in English. The full-term infants had neurobehavioral testing in the mother’s room on the labor and delivery floor within 4 days of birth. Preterm infants were born < 32 weeks estimated gestational age (EGA) with no congenital anomaly. This sample of preterm infants had low medical risk with no evidence of grade III or IV intraventricular hemorrhage (IVH), cystic periventricular leukomalacia (cPVL), ventilation for greater than 10 days, confirmed necrotizing enterocolitis (NEC; all stages), patent ductus arteriosus (PDA) (treated with indomethacin or surgical ligation), or retinopathy of prematurity (ROP; any stage) by 34 weeks postmenstrual age (PMA). Preterm infants were evaluated in the NICU between 34-45 weeks postmenstrual age (PMA). Abbreviations: NNNS: NICU Network Neurobehavioral Scale, HNNE: Hammersmith Neonatal Neurological Evaluation

^*The percentile of each mean score, based on a full-term normative sample, are provided for context to compare each mean score on the NNNS to a norm. Full term percentile comparisons were derived from Fink, N. S., Tronick, E., Olson, K., & Lester, B. (2012). Healthy newborns' neurobehavior: norms and relations to medical and demographic factors. The Journal of pediatrics, 161(6), 1073-1079.

^**No scores reported for hypertonicity in the previous article.

^***HNNE scores for both the preterm and full term groups fell in the ‘sub-optimal’ range, when compared to full-term healthy cohorts derived from Dubowitz L, Mercuri E, Dubowitz V. An optimality score for the neurologic examination of the term newborn. J Pediatr. 1998;133(3):406-16.

Table 3 identifies results from investigating differences between neurobehavioral performance of very preterm infants with low medical risk and full-term infants. Very preterm infants with low medical risk at term equivalent age demonstrated more hypotonia (p=0.01), more asymmetrical reflexes (p < 0.001), and better quality of movement (p=0.005) on univariate analysis (however, these relationships were no longer significant on multivariate analysis). Very preterm infants with low medical risk at term equivalent age also demonstrated more sub-optimal reflexes (p<0.001; ß=1.5) and more stress (p<0.001; ß=0.08), which remained significant after controlling for PMA at the time of testing, insurance type, mode of delivery, marital status, and race. Very preterm infants with low medical risk also performed worse on the HNNE (p=0.005; ß=−3.41) which remained significant after controlling for PMA at the time of testing, insurance type, mode of delivery, marital status, and race. No other significant relationships were found.

Table 3.

Group mean differences in neurobehavioral scores in preterm infants with low medical risk and full-term infants.

		Univariate Analysis		Multivariate Regression Analysis
	p-value *	Mean Difference	95% Confidence Interval	p-value **	Beta	95% Confidence Interval of Beta
NNNS
Orientation	0.39	0.22	[−2.8, 0.723.0]
Hypertonicity	0.72	−0.06	[−0.40, 0.27]
Hypotonicity	0.01	−0.35	[−0.62, 0.09]	0.17	0.28	[−0.18, 0.67]
Excitability	0.69	−0.19	[−1.1, 0.73]
Arousal	0.80	0.04	[−0.25, 0.32]
Lethargy	0.18	−0.68	[−1.67, 0.31]
Sub-Optimal Reflexes	<0.001	−.1.58	[−2.2, −1.0]	<0.001	1.53	[0.65, 2.42]
Asymmetry	<0.001	−1.11	[−1.55, −0.68]	0.03	0.86	[0.09, 1.63]
Stress	<0.001	−0.11	[−0.14, −0.09]	<0.001	0.08	[0.04, 0.13]
Self-Regulation	0.57	0.09	[−0.23, 0.41]
Quality of Movement	0.005	0.35	[0.11, 0.59]	0.65	−0.09	[−0.48, 0.30]
Handling	0.11	−0.12	[−0.28, 0.03]
HNNE Optimality Score	<0.001	3.86	[2.25, 5.48]	0.005	−3.41	[−5.80, −1.03]

^*p-value is from investigating group mean differences in neurobehavioral scores in preterm infants with low medical risk and full-term infants using independent samples t-tests.

^**p-value is from investigating differences in neurobehavioral scores in preterm infants with low medical risk and full-term infants using multivariate regression analysis while controlling for PMA at the time of testing, infant race, maternal marital status, mode of delivery, and insurance type.

Bolded values met the threshold for significance (0.05/2=0.025)

After controlling for the assigned intervention among those who were part of a randomized clinical trial and after controlling for cohort, the findings remained largely unchanged with no change in significance across any of the outcome variables.

See Figure 2 for a scatterplot of HNNE scores across infants born at different EGAs. Higher EGA at the time of birth was related to higher HNNE scores (p <0.001) after controlling for PMA at the time of testing, insurance type, mode of delivery, marital status, and race.

Figure 2.

Scatterplot of HNNE scores across infants born at different EGA (p-value is from investigating relationships between EGA and HNNE scores using multivariate linear regression while controlling for PMA at the time of testing, mode of delivery, insurance type, marital status, and race).

See Figure 3 for a scatterplot of NNNS summary scores across infants born at different EGA. EGA at the time of birth was related to NNNS sub-optimal reflexes (p <0.001; lower EGA, more sub-optimal reflexes), stress (p <0.001; higher EGA, less stress), and hypertonia (p = 0.005; higher EGA, less hypertonia). There were no other relationships between NNNS summary scores and EGA observed.

Figure 3.

Scatterplot of NNNS scores across infants born at different EGA (p-value is from investigating relationship between EGA and NNNS summary score using multivariate linear regression while controlling for PMA at the time of testing, mode of delivery, insurance type, marital status, and race).

Discussion

The key finding of this study was that very preterm infants with low medical risk demonstrated alterations in early neurobehavior at term-equivalent age when compared to their full-term counterparts. Specifically, very preterm infants with low medical risk exhibited more stress and more sub-optimal reflexes on the NNNS and had lower scores on the HNNE. Such understanding elucidates the developmental differences that exist among low medical risk preterm infants who have reached the same developmental age as their full-term counterparts. Further, infants born at earlier gestational ages show more stress, tonal alterations (hypotonicity and hypertonicity), and fewer optimal reflexes.

Studies reporting neurobehavioral comparisons of full-term infants and preterm infants during the neonatal period differ in their findings. This may relate to variations in the study populations, including the infant’s gestational age at birth and whether infants with preterm comorbidities were included. Our previous findings that neurobehavioral performance at term-equivalent age was worse in very preterm infants compared to their full-term counterparts included preterm infants with significant medical challenges (19). Further, this investigation compared a socially disadvantaged cohort of preterm infants to standardized normative scores of full-term infants, which likely reflected a more socially advantaged full-term comparison. Other investigators have studied differences in moderately preterm, late preterm, and full-term infants (including those with medical challenges) and found that healthy full-term infants had better scores for stress, lethargy, sub-optimal reflexes, excitability, hypotonicity, quality of movement, attention, arousal, handling, and asymmetrical reflexes (33). However, still others have reported that preterm infants can reach the same neurobehavioral performance as their full-term counterparts by 36 weeks PMA for arousal, lethargy, and asymmetry; by 38 weeks PMA for quality of movements; by 40 weeks PMA for attention, sub-optimal reflexes, and stress with persisting differences with lower scores in the domains of excitability and hypotonicity continuing beyond the neonatal period (18). Our findings differ from these findings, as our study not only found that infants did not ‘catch up’ in stress and reflex development as they reached term-equivalent age, but that these differences exist in preterm infants without brain injury or medical complications related to prematurity. Research has also demonstrated relationships between medical factors and neurobehavioral performance, attributing neurodevelopmental/neurobehavioral outcomes to medical consequences associated with prematurity (4, 11, 19, 34), rather than prematurity alone. Other studies have controlled for medical complications, yet none (that we know of) have excluded the infants with significant medical complications to focus the investigation on identifying differences in neurobehavior only in very preterm infants with low medical risk compared to their full-term counterparts.

Our findings highlight that there are likely other contributors, beyond medical factors and cerebral injury, that impact developmental outcomes in preterm infants and warrant attention. Significant changes in brain development have been identified in the perinatal period, specifically with increases in gray matter volume occurring from 34-40 weeks PMA, which highlights the significant amount of change occurring during a critical and vulnerable period of time (35). Studies of preterm infants have demonstrated alterations in cortical folding, specifically in the temporal and parietal regions, white matter abnormalities, and decreased structural connectivity (36-38). In addition to other genetic or nutritional influences, the role of the environment on early brain development could be a contributing factor. Preterm birth removes the infant from the protective environment of the mother’s womb during the last trimesters of pregnancy. For preterm infants, this brain development occurs in the NICU environment which has chaotic sensory exposures, significant stressors, and pain that all have been related to altered functional neuronal connectivity as well as abnormalities on neurobehavioral examination (39). Altered sensory and motor experiences may disrupt the trajectory of infant development, impacting long term neurological functioning (40).

Very preterm infants with low medical risk at term equivalent age demonstrated more sub-optimal reflexes and more stress. Further, infants born at earlier gestational ages show increased stress, hypotonicity and hypertonicity, and less optimal reflexes. Reflex development is the foundation for later motor control. Reflexes rely on sensory experiences that elicit them and are flexion dominated during this time period. The NICU environment takes the infant out of the uterine environment that provides timed sensory exposures and natural positioning into flexion and puts the infant in a chaotic sensory environment with more extended positioning. This could explain why preterm infants have more sub-optimal reflexes compared to their full term peers. Further, lack of appropriate flexion and sensory and motor experiences can drive tonal differences that were observed in infants born at earlier gestational ages. Finally, the NICU environment is known to have lights and sounds that the preterm infant is not able to tolerate, in addition to the preterm infant undergoing procedural touch and painful experiences. This could be reflected in the higher stress scores observed in the preterm group in this study. Understanding that areas of function in relation to stress and reflexes are different in preterm and full term infants can provide important targets for interventions that can optimize these early areas of function.

The very preterm infants in this sample differ from their local comparisons in this study, but they differ even more to a previously reported cohort of full-term infants who were more socioeconomically advantaged (41). Comparatively, our sample scored at the 90^th percentile on NNNS scores that reflect increased stress, need for handling assistance, and suboptimal reflexes compared to the previous reports of a socioeconomically advantaged sample of full terms infants assessed at similar ages.

There are several limitations in this study. Determining and isolating the effects of medical factors on neurobehavior is difficult, especially when preterm infants are inherently medically complex. It is possible that the exclusion criteria did not account for all infants with medical risk. With advanced medical technology, preterm infants are able to stabilize faster and are ready to be discharged sooner, despite early PMA. Due to the nature of the discharging process at the study site NICU, there was variability in the timing of neurobehavioral testing of preterm infants, as they were assessed right before NICU discharge between a varying range of 34-45 weeks PMA and not necessarily at term-equivalent age. Thus, neurobehavioral performance could have been confounded by the infant’s PMA or maturity at the time of the assessment, as there are rapid neurodevelopmental changes in the final 6 weeks prior to term equivalent age (42). Multivariate analysis enabled us to control for the timing (PMA) of neurobehavioral testing. Further, the full-term infants were assessed within 4 days of birth, a time period which may not be ideal for early assessment. Waiting until the full-term infants were at least one week after birth could have resulted in more differences observed, however, was impractical. We were unable to blind evaluators on whether infants were full-term or preterm. There were several summary scores and 2 assessment outcomes used, which can increase the risk of Type II error. Bonferroni correction for each measure used aided in balancing risk of Type I and II errors. Data were collected over a course of the years 2011-2018 from multiple historic cohorts. Further, there was not consistent reporting of medical factors and socio-demographic factors across all the cohorts, making exclusion of important factors such as bronchopulmonary dysplasia not possible and resulting in the inability to report factors such as grade I and II IVH in the sample. Also, this data was collected from a specific demographic region, with a certain grouping of socioeconomic factors and may limit generalizability.

Conclusion

This study focused on the influence of immaturity in the absence of medical risk factors. The findings of the current study affirm the importance of developmental surveillance for very preterm infants, regardless of medical complications. This study also demonstrated the sensitivity of neurobehavioral assessment in early identification. The identification of early alterations in neurobehavioral performance presents important information for practices, programs, and policy makers to enable early targeted interventions to address early neurobehavioral deficits. Interventions that target stress and handling along with ensuring symmetry in movement patterns while optimizing positioning and opportunities for reflex development are important areas for future research. Further, intentional adaptation of the NICU environment is a modifiable factor, and additional investigation into its influence is warranted. Future research is also needed on the long-term neurodevelopmental outcomes of low-risk very preterm infants compared to their full-term counterparts.