Abstract
Non–object-oriented exploratory behaviors infants perform with their bodies and surfaces have been proposed to be key precursors of infants’ object exploration, early learning, and future cognitive development. Little is known about the developmental trajectories of these behaviors, especially for infants born preterm.
The purpose of the study was to longitudinally compare non–object-oriented exploratory behaviors performed by full-term and preterm infants.
The study followed 24 full-term and 30 preterm infants (6 with significant brain injury) performing non–object-oriented exploratory behaviors in prone, supine, and sitting from birth through 24 months.
Infants were observed without objects or direct social interaction for 3 minutes in prone and supine (0 through 9 months) and in sitting (3 through 24 months). Behavioral coding produced data that were analyzed using hierarchical linear modeling. Developmental trajectories of behaviors were compared among full-term infants, preterm infants without significant brain injury, and preterm infants with significant brain injury.
Compared to full-term peers, preterm infants showed poorer postural control (less head lifting in prone), midline behavior (holding the head in midline, holding both hands in midline), hand-to-mouth and visual-motor behaviors against gravity, and more asymmetrical one-handed fisting. Preterm infants performed fewer bouts of non–object-oriented exploratory behaviors, and their behaviors were less variable with fewer combinations.
There was a limited sample of infants born preterm with significant brain injury.
Non–object-oriented exploratory behaviors are important for early perceptual-motor development. Key differences were noted in these behaviors for infants born preterm. These differences may lead to impaired reaching, object exploration, and cognition. Early intervention programs should utilize assessments and interventions that target these very early non–object-oriented exploratory behaviors.
Infants born preterm are at risk for delays in the development of gross and fine motor skills,1 hand-eye coordination,2 attention, learning, memory,3-8 language,3,9 and problem solving abilities7 that can negatively impact their school readiness and academic performance.10 These risks are even greater for infants born preterm with significant brain injuries.10-12 Early perceptual-motor abilities may serve as key early identifiers of these cognitive disabilities. For instance, poor cognitive outcomes observed in children born preterm have been shown to be preceded by and related to delays in object exploration earlier in development.12-16 For example, infants born preterm perform less visual-haptic multimodal exploration, have impaired bimanual abilities, display reduced variability of exploratory behaviors, and are less able to understand the affordances of different objects and to match their exploratory strategies to the properties of objects compared with their peers born full-term.16 Delays in object exploration in preterm infants at 7 months of age have been related to poorer cognitive outcomes at 24 months.12
These delays in object-oriented exploration and other cognitive abilities may be argued to stem from delays in non–object-oriented exploratory behaviors early in development. Non–object-oriented exploratory behaviors are operationally defined here as the behaviors infants engage in when not directly interacting with portable objects. These behaviors allow infants to explore the action capabilities of their bodies, surrounding surfaces, and their potential interactions. Examples of non–object-oriented exploratory behaviors include exploration of head movement against gravity, of head and hand movement toward midline, of more open hand postures, of the hands via looking or mouthing, and of surfaces or the body via touching. It has been proposed that the information infants gather via exploration of their bodies and surfaces is important for early learning and forms the foundation for future object-oriented exploration and other cognitive abilities.5,15,17-20 For instance, infants’ early experiences with touching surfaces and the body provide them with proprioceptive and haptic feedback, increase their body awareness, and are believed to be the precursors of future reaching and grasping behaviors.21-24 Similarly, early hand-to-mouth movements enable the development of infants’ coordination that later develops into controlled arm movements for object-directed reaching, grasping, and self-feeding.25,26 Furthermore, when infants begin exploring objects, their object exploration behaviors develop directly from their repertoire of non–object-oriented behaviors. Specifically, they do not engage in novel behaviors with objects but rather utilize their ongoing non–object-oriented exploratory behaviors in greater amounts and in novel combinations when presented with objects.27
Exploration is not only critical for infants to learn how to engage in social interactions and to learn about objects; it is also key for infants to learn to control their own bodies so they can perform perceptual-motor behaviors like lifting their heads against gravity, reaching, or moving their hands to midline.28,29 For example, when 1-month-olds were provided opportunities to independently control their heads, their head control improved within just 2 weeks relative to control infants, suggesting that it may be lack of experience and movement exploration rather than lack of strength that limits head control for young infants and that exploration of body affordances is critical for the development of perceptual-motor skills.29 Similarly, when 2-month-olds were provided opportunities to engage in general movement of their arms, they had advanced emergence of reaching behaviors.28 Control infants who did not receive enhanced opportunities to explore their action capabilities showed poorer performance in both of these studies.
Despite the fact that developmental theory and a growing body of research suggest that non–object-oriented exploratory behaviors form the foundation for important future perceptual-motor and cognitive skills and that early identification and intervention are key to minimizing future delays, there are no articles documenting these behaviors comprehensively and longitudinally in typical populations or in populations at risk for delays. The purpose of this study was to estimate developmental trends of spontaneous non–object-oriented exploratory behaviors in full-term and preterm infants throughout the first 2 years of life. We aimed to document which non–object-oriented exploratory behaviors infants performed during this time frame, to quantify the level of performance across time and body positions (prone, supine, and sitting), and to determine if and how these behaviors differed for full-term and preterm infants. We were interested in behavioral differences in relation to gestational age as well as in relation to the presence of structural and physiological changes to the nervous system due to a prior brain injury.
Methods
Participants
Twenty-four typically developing healthy infants born full-term (FT) (14 males; 37–42 weeks gestational age, mean = 39.4; SD = 1.1) were recruited from the community using public birth records. An additional 30 infants born preterm (10 males; 22–30 weeks gestational age, mean = 26.5; SD = 1.7) were recruited from a regional level 3 neonatal intensive care unit. Six of the preterm infants (2 males) had periventricular leukomalacia (PVL) and/or Grade III or Grade IV intraventricular hemorrhage and were identified as “preterm with significant brain injury” (PTI), whereas the remaining 24 preterm infants (8 males) were born without these brain injuries and were identified as “preterm without significant brain injury” (PT). The ethnic composition of the entire sample was 56.4% Caucasian, 30.9% African American, 12.7% Asian, and 9.1% Hispanic. Additional demographic and health-related information about the participants is provided in the Table. Recruitment of participants, informed consent, and data collection were done in accordance with the regulations set by the University of Delaware's and Christiana Care Health System's Institutional Review Boards. Participants received monetary compensation for their participation in the study.
Participants’ Demographic and Health-Related Information (Mean ± SD)
| Criteria | Infants Born Full-Term | Infants Born Preterm | Infants Born Preterm With Brain Injury |
|---|---|---|---|
| African American | 20.8% | 41.7% | 16.7% |
| Caucasian | 70.9% | 45.8% | 50% |
| Asian | 8.3% | 12.5% | 33.3% |
| Hispanic | 0% | 12.5% | 16.7% |
| Gross household income $0–14,999 | 4.2% | 16.7% | 16.7% |
| Gross household income $15,000–24,999 | 0% | 20.8% | 0% |
| Gross household income $25,000–34,999 | 4.2% | 0.0% | 0% |
| Gross household income $35,000–44,999 | 4.2% | 8.3% | 16.7% |
| Gross household income $45,000–59,999 | 8.3% | 12.5% | 0.0% |
| Gross household income $60,000–79,999 | 25% | 16.7% | 33.3% |
| Gross household income greater than $80,000 | 54.1% | 25% | 33.3% |
| Maternal drug use | 0% | 8.3% | 0% |
| Weeks preterm relative to expected date of birth | 0.6 ± 1.1 | 13.4 ± 1.6 | 12.2 ± 2.0 |
| Birth weight (grams) | 3366.7 ± 651.2 | 893.8 ± 259.0 | 1034.3 ± 291.2 |
| Multiple gestation | 0% | 25% | 50% |
| Conceived via in vitro fertilization | 0% | 4.2% | 16.7% |
| Weeks in the neonatal intensive care unit | 0.2 ± 0.5 | 12.5 ± 5.5 | 11.3 ± 4.8 |
| Weeks on ventilation | 0 | 5.4 ± 12.2 | 3.0 ± 3.8 |
| Grade III or IV intraventricular hemorrhage | 0% | 0% | 66.7% |
| Periventricular leukomalacia | 0% | 0% | 100% |
| Necrotizing enterocolitis | 0% | 16.7% | 0% |
| Chronic lung disease | 0% | 50% | 33.3% |
| Respiratory distress syndrome | 0% | 100% | 100% |
| Bronchopulmonary dysplasia | 0% | 66.7% | 50% |
| Sepsis | 4% | 29.2% | 100% |
| Retinopathy of prematurity | 0% | 50% | 66.6% |
| Patent ductus arteriosus | 0% | 50% | 50% |
| Receiving early intervention services during the study | 0% | 50% | 50% |
| Criteria | Infants Born Full-Term | Infants Born Preterm | Infants Born Preterm With Brain Injury |
|---|---|---|---|
| African American | 20.8% | 41.7% | 16.7% |
| Caucasian | 70.9% | 45.8% | 50% |
| Asian | 8.3% | 12.5% | 33.3% |
| Hispanic | 0% | 12.5% | 16.7% |
| Gross household income $0–14,999 | 4.2% | 16.7% | 16.7% |
| Gross household income $15,000–24,999 | 0% | 20.8% | 0% |
| Gross household income $25,000–34,999 | 4.2% | 0.0% | 0% |
| Gross household income $35,000–44,999 | 4.2% | 8.3% | 16.7% |
| Gross household income $45,000–59,999 | 8.3% | 12.5% | 0.0% |
| Gross household income $60,000–79,999 | 25% | 16.7% | 33.3% |
| Gross household income greater than $80,000 | 54.1% | 25% | 33.3% |
| Maternal drug use | 0% | 8.3% | 0% |
| Weeks preterm relative to expected date of birth | 0.6 ± 1.1 | 13.4 ± 1.6 | 12.2 ± 2.0 |
| Birth weight (grams) | 3366.7 ± 651.2 | 893.8 ± 259.0 | 1034.3 ± 291.2 |
| Multiple gestation | 0% | 25% | 50% |
| Conceived via in vitro fertilization | 0% | 4.2% | 16.7% |
| Weeks in the neonatal intensive care unit | 0.2 ± 0.5 | 12.5 ± 5.5 | 11.3 ± 4.8 |
| Weeks on ventilation | 0 | 5.4 ± 12.2 | 3.0 ± 3.8 |
| Grade III or IV intraventricular hemorrhage | 0% | 0% | 66.7% |
| Periventricular leukomalacia | 0% | 0% | 100% |
| Necrotizing enterocolitis | 0% | 16.7% | 0% |
| Chronic lung disease | 0% | 50% | 33.3% |
| Respiratory distress syndrome | 0% | 100% | 100% |
| Bronchopulmonary dysplasia | 0% | 66.7% | 50% |
| Sepsis | 4% | 29.2% | 100% |
| Retinopathy of prematurity | 0% | 50% | 66.6% |
| Patent ductus arteriosus | 0% | 50% | 50% |
| Receiving early intervention services during the study | 0% | 50% | 50% |
Participants’ Demographic and Health-Related Information (Mean ± SD)
| Criteria | Infants Born Full-Term | Infants Born Preterm | Infants Born Preterm With Brain Injury |
|---|---|---|---|
| African American | 20.8% | 41.7% | 16.7% |
| Caucasian | 70.9% | 45.8% | 50% |
| Asian | 8.3% | 12.5% | 33.3% |
| Hispanic | 0% | 12.5% | 16.7% |
| Gross household income $0–14,999 | 4.2% | 16.7% | 16.7% |
| Gross household income $15,000–24,999 | 0% | 20.8% | 0% |
| Gross household income $25,000–34,999 | 4.2% | 0.0% | 0% |
| Gross household income $35,000–44,999 | 4.2% | 8.3% | 16.7% |
| Gross household income $45,000–59,999 | 8.3% | 12.5% | 0.0% |
| Gross household income $60,000–79,999 | 25% | 16.7% | 33.3% |
| Gross household income greater than $80,000 | 54.1% | 25% | 33.3% |
| Maternal drug use | 0% | 8.3% | 0% |
| Weeks preterm relative to expected date of birth | 0.6 ± 1.1 | 13.4 ± 1.6 | 12.2 ± 2.0 |
| Birth weight (grams) | 3366.7 ± 651.2 | 893.8 ± 259.0 | 1034.3 ± 291.2 |
| Multiple gestation | 0% | 25% | 50% |
| Conceived via in vitro fertilization | 0% | 4.2% | 16.7% |
| Weeks in the neonatal intensive care unit | 0.2 ± 0.5 | 12.5 ± 5.5 | 11.3 ± 4.8 |
| Weeks on ventilation | 0 | 5.4 ± 12.2 | 3.0 ± 3.8 |
| Grade III or IV intraventricular hemorrhage | 0% | 0% | 66.7% |
| Periventricular leukomalacia | 0% | 0% | 100% |
| Necrotizing enterocolitis | 0% | 16.7% | 0% |
| Chronic lung disease | 0% | 50% | 33.3% |
| Respiratory distress syndrome | 0% | 100% | 100% |
| Bronchopulmonary dysplasia | 0% | 66.7% | 50% |
| Sepsis | 4% | 29.2% | 100% |
| Retinopathy of prematurity | 0% | 50% | 66.6% |
| Patent ductus arteriosus | 0% | 50% | 50% |
| Receiving early intervention services during the study | 0% | 50% | 50% |
| Criteria | Infants Born Full-Term | Infants Born Preterm | Infants Born Preterm With Brain Injury |
|---|---|---|---|
| African American | 20.8% | 41.7% | 16.7% |
| Caucasian | 70.9% | 45.8% | 50% |
| Asian | 8.3% | 12.5% | 33.3% |
| Hispanic | 0% | 12.5% | 16.7% |
| Gross household income $0–14,999 | 4.2% | 16.7% | 16.7% |
| Gross household income $15,000–24,999 | 0% | 20.8% | 0% |
| Gross household income $25,000–34,999 | 4.2% | 0.0% | 0% |
| Gross household income $35,000–44,999 | 4.2% | 8.3% | 16.7% |
| Gross household income $45,000–59,999 | 8.3% | 12.5% | 0.0% |
| Gross household income $60,000–79,999 | 25% | 16.7% | 33.3% |
| Gross household income greater than $80,000 | 54.1% | 25% | 33.3% |
| Maternal drug use | 0% | 8.3% | 0% |
| Weeks preterm relative to expected date of birth | 0.6 ± 1.1 | 13.4 ± 1.6 | 12.2 ± 2.0 |
| Birth weight (grams) | 3366.7 ± 651.2 | 893.8 ± 259.0 | 1034.3 ± 291.2 |
| Multiple gestation | 0% | 25% | 50% |
| Conceived via in vitro fertilization | 0% | 4.2% | 16.7% |
| Weeks in the neonatal intensive care unit | 0.2 ± 0.5 | 12.5 ± 5.5 | 11.3 ± 4.8 |
| Weeks on ventilation | 0 | 5.4 ± 12.2 | 3.0 ± 3.8 |
| Grade III or IV intraventricular hemorrhage | 0% | 0% | 66.7% |
| Periventricular leukomalacia | 0% | 0% | 100% |
| Necrotizing enterocolitis | 0% | 16.7% | 0% |
| Chronic lung disease | 0% | 50% | 33.3% |
| Respiratory distress syndrome | 0% | 100% | 100% |
| Bronchopulmonary dysplasia | 0% | 66.7% | 50% |
| Sepsis | 4% | 29.2% | 100% |
| Retinopathy of prematurity | 0% | 50% | 66.6% |
| Patent ductus arteriosus | 0% | 50% | 50% |
| Receiving early intervention services during the study | 0% | 50% | 50% |
Procedure
Infants’ non–object-oriented exploration behaviors were assessed in their home at visits just after expected date of birth, and at 1.5, 3, 4, 6, 9, 12, 18, and 24 months of corrected age. The aim of this project was to collect snapshots of infants’ typical daily behaviors, and their natural home environment was expected to produce more valid data on infants’ behaviors as compared to the lab setting. For all visits through 9 months, infants were assessed in prone (lying on the stomach) and supine (lying on the back) on a blanket on the floor. After 9 months, infants were no longer evaluated in these postures because most were mobile and would not remain stationary for the assessment durations. At all visits from 3 through 24 months, infants were assessed in sitting (supported in an infant seat). Assessment in sitting began at 3 months because it is typical for infants to be supported upright at that age. Infants were placed in a high-back upright infant seat before the emergence of independent sitting (typically 6 months), and in a low-back booster seat thereafter. In each posture, infants’ spontaneous movements in the absence of direct social interaction or portable objects were observed for 3 minutes in a safe space with the parent(s) and experimenter present. For infants not yet mobile, the experience of being positioned on the floor or in a seat, near others but not the focus of their attention, is common. It was this experience we wanted to capture. All of the assessments were video recorded by 2 synchronized cameras using the frontal and side views of the child (eFigs. 1–5, available at https://academic.oup.com/ptj/). Infants were in a positive or neutral behavioral state for all testing procedures. Testing was stopped if participants showed continued signs of discomfort like crying and was resumed within the same day or week when participants had returned to a positive or neutral behavioral state; this situation occurred occasionally for younger participants, who have short periods of wakefulness, and rarely for older participants.
All assessments were performed by a trained experimenter with expertise in child development, video recorded using 2 views for optimal coding, and coded by experimenters blind to participants’ age and study group. Behavioral coding of the exploration assessments was performed using OpenSHAPA software. Intrarater agreement was 87.5±2.8%, and interrater agreement with a primary coder was 86.8±0.4% for 20% of the coded data based on the equation [Agreed/(Agreed + Disagreed)] * 100.
Measures
The exploratory behaviors coded were for ones that provided infants opportunities to learn about the affordances and action potentials of their bodies and of their bodies in relation to surfaces. The following behaviors were coded in prone, supine, and sitting: 1) Head in midline, when the head was not turned more than 30 degrees to the right or left; 2) One hand in midline, when only one hand was positioned in line with the infant's torso, such as resting on the torso, resting on the face, elevated anterior to the torso or head but not positioned laterally with respect to the torso; 3) Both hands in midline, when both hands were in the midline position; 4) One hand fisted, when at least 4 fingers were flexed/bent completely into the palm of the hand on only one hand; 5) Both hands fisted, when at least 4 fingers were flexed/bent completely into the palm of the hand on both hands; 6) Mouthing the hands, when any part of either hand was in contact with the infant's mouth, tongue, and/or lips; 7) Looking at the hands, when the infant's eyes were directed at either hand; 8) Touching the body, when the infant's hand contacted part of his body, such as his head, trunk, or arm; 9) Touching surfaces, when the infant's hand contacted surfaces around the infant, ie, the floor or infant seat; and in prone we coded: 10) Head up, when no part of the head, chin, or face was touching the support surface.
Using Filemaker Pro Advanced custom programs (Filemaker, Inc, Santa Clara, CA), the coded data were then converted to percentages of assessment time, dividing each behavior's duration by the total assessment time. The programs also allowed us to detect overlapping occurrences of different behaviors and to summarize the data to obtain the following additional variables: 1) Looking at hands while acting, when the infant was looking at the hand(s) while performing another behavior with the hand(s) (multimodal behavior); 2) Bouts of exploration, number of times the infant switched from performing one behavior to another per minute (measure of intensity of behavioral performance); 3) Separate behaviors performed, of the separate behaviors observed across the entire longitudinal dataset, the percentage of those the infant was observed performing at each assessment; and 4) Combined behaviors performed, of the combinations of behaviors (2 or more behaviors performed simultaneously) observed across the entire longitudinal dataset, the percentage of those the infant was observed performing at each assessment. Separate and combined behaviors serve as measures of variability of behavioral performance.
Statistical Analyses
All statistical analyses were conducted using Hierarchical Linear Modeling30 (HLM), which is the most appropriate and recommended technique for longitudinal designs, since it allows for a hierarchical data structure in which observations across age are nested within participants, thus accounting for nonindependence of multiple observations of the same participants. The 2-level models allow estimation of the variability in outcome scores both within and between participants, enabling flexible modeling for an overall average trend across all participants based on the individual participant-specific growth curves. In all statistical models, linear, quadratic, and cubic trends of change with age were tested. To assess the development of non–object-oriented exploration behaviors among groups (FT, PT, and PTI), all multilevel models were tested including the dummy-coded status variables FT/PT (comparing FT to PT infants) and PT/PTI (comparing PT to PTI infants), PT infants being the reference group. Age by status variables interaction terms (Age*FT/PT, Age*PT/PTI, (Age)2*FT/PT, (Age)2*PT/PTI, (Age)3*FT/PT, (Age)3*PT/PTI) were also included in all models to address the extent to which growth trajectories differed among the infant status groups. When analyses suggested no statistically significant differences between any 2 or among all 3 groups, their data were displayed in combination in the figures. Therefore, all individual group trajectories displayed in the figures represent significant group differences, whereas any one trajectory representing combined groups signifies no statistically significant differences among these groups.
To ensure that the observed effects in the current study were not only statistically significant, but also meaningful, we calculated effect sizes for each variable using Cohen's d parameter, with 0.2 considered small, 0.5 medium, 0.8 large, 1.2 very large, and 2.0 huge.
Results
Final multilevel models for all statistical analyses along with effect size estimates are provided in eTable 1 (available at https://academic.oup.com/ptj/), whereas the statistical parameters for any group comparisons excluded due to statistical nonsignificance are presented in eTable 2 (available at https://academic.oup.com/ptj/). Although all infants showed improved ability to raise the head in prone throughout the first 9 months, PT infants, and especially PTI infants, performed this behavior consistently less than FT infants across all ages. This difference increased with age. For instance, the difference in holding the head up between FT and PTI was 23.1% at 1 month, 40.5% at 5 months, and 57.8% at 9 months. The multilevel analysis showed significant differences in developmental trajectories of the percent time infants held their heads up while in prone among FT, PT, and PTI infants (Fig. 1A, note that the growth curves displayed on all figures represent the estimated/modeled values).
Estimated/modeled growth curves for holding the head up in prone (A) and holding the head in midline in prone (B), supine (C), and sitting (D); FT = full-term, PT = preterm; PTI = preterm with brain injury; PT & PTI = no significant differences were found between the trajectories of the PT and PTI groups, so they were represented by a single line; FT, PT, & PTI = no significant differences were found among the trajectories of the 3 groups, so they were represented by a single line.
A steady increase in the ability to hold the head in midline was detected in all 3 postures throughout the first 24 months of life. However, PTI infants showed midline head positioning less often than FT and PT infants during the first 9 months in prone and supine. In prone, significant differences were identified among FT, PT, and PTI infants (Fig. 1B). In supine, significant differences were found between FT and PT infants, but not between PT and PTI infants (Fig. 1C). No significant differences were observed in sitting among FT, PT, and PTI infants (Fig. 1D).
Preterm infants (both PT and PTI) held one hand in midline less frequently than FT infants in prone and sitting, but not in supine. In prone, significant differences were identified between FT and PT infants, but not between PT and PTI infants (Fig. 2A). In supine, no significant differences were detected among FT, PT, and PTI infants (Fig. 2B). In sitting, significant differences were found between FT and PT infants, but not between PT and PTI infants (Fig. 2C).
Estimated/modeled growth curves for holding one (A, B, C) or both (D, E, F) hands in midline in prone, supine, and sitting; FT = full-term, PT = preterm; PTI = preterm with brain injury; PT & PTI = no significant differences were found between the trajectories of the PT and PTI groups, so they were represented by a single line; FT & PT = no significant differences were found between the trajectories of the FT and PT groups, so they were represented by a single line.
Preterm infants (both PT and PTI) held both hands in midline less in sitting compared to FT infants, whereas in prone only PTI infants showed difficulty holding both hands in midline. In prone, significant differences were found between PT and PTI infants, but not between FT and PT infants (Fig. 2D). In supine, no significant differences were observed among FT, PT, and PTI infants (Fig. 2E). Significant differences were detected in sitting between FT and PT infants, but not between PT and PTI infants (Fig. 2F).
Asymmetrical hand fisting decreased with age in all 3 postures. In supine and prone, infants born with brain injury showed more one-handed fisting compared to infants born without brain injury. In sitting, all preterm infants (PT and PTI) showed more asymmetrical fisting compared to their FT peers. In prone, significant differences were detected between PT and PTI infants, but not between FT and PT infants (Fig. 3A). In supine, significant differences were observed between PT and PTI infants, but not between FT and PT infants (Fig. 3B). Significant differences were found in sitting between FT and PT infants, but not between PT and PTI infants (Fig. 3C).
Estimated/modeled growth curves for holding one (A, B, C) or both (D, E, F) hands fisted in prone, supine, and sitting; FT = full-term, PT = preterm; PTI = preterm with brain injury; PT & PTI = no significant differences were found between the trajectories of the PT and PTI groups, so they were represented by a single line; FT & PT = no significant differences were found between the trajectories of the FT and PT groups, so they were represented by a single line; FT, PT, & PTI = no significant differences were found among the trajectories of the 3 groups, so they were represented by a single line.
The percent time infants spent holding both hands fisted steadily decreased throughout the first year in all 3 postures, with both hands becoming unfisted in the majority of infants by the age of approximately 7 months in supine and prone, and by 12 months in sitting. Significant differences were observed between FT and PT infants only in prone, with FT infants spending more time fisting both hands as compared with their preterm peers during the first 6 months (Fig. 3D). No significant differences among the 3 groups were identified in supine (Fig. 3E) or sitting (Fig. 3F).
In prone, PTI infants showed more hand mouthing than their peers, especially during the 1.5- to 6-month period, whereas in supine, PTI infants performed less mouthing of the hands throughout the entire 0- to 9-month period. Mouthing of the hands steadily decreased from 3 to 15 months (8.5% to 1.9%) for all infants in sitting. In prone and supine, significant differences were detected between PT and PTI infants, but not between FT and PT infants (Figs. 4A and 4B). No significant differences were found among full-term, PT, and PTI infants in sitting Fig. 4C).
Estimated/modeled growth curves for mouthing of the hand(s) (A, B, C), touching the body (D, E, F), and touching surfaces (G, H, I) in prone, supine, and sitting; FT = full-term, PT = preterm; PTI = preterm with brain injury; FT & PT = no significant differences were found between the trajectories of the FT and PT groups, so they were represented by a single line; FT, PT, & PTI = no significant differences were found among the trajectories of the 3 groups, so they were represented by a single line.
Looking at the hands and looking at the hands while acting did not often occur, especially in prone; this produced 92.5% of zeroes in the prone data and did not allow for reliable statistical analysis. As a result, only data in supine and sitting were analyzed. In supine, PTI infants looked at their hands less compared to other infants (0.14% vs 0.87% looking at hands consistently across all ages; 0.15% vs 0.61% looking at hands while acting consistently across all ages); significant differences between PT and PTI infants, but not between FT and PT infants, were detected in developmental trajectories of looking at the hands and looking at the hands while acting. In sitting, no significant differences among FT, PT, and PTI infants were identified in developmental trajectories of looking at the hands (quadratic trajectory increasing from 0.76% at 3 months to 2.32% at 15 months, and decreasing thereafter to 0% at 24 months) or looking at the hands while acting (quadratic trajectory increasing from 0.47% at 3 months to 1.82% at 12 months, and decreasing thereafter to 0% at 24 months).
In prone, touching the body decreased for the majority of infants (FT and PT) from 0 through 9 months, with PTI infants performing significantly less body touching compared with other infants (Fig. 4D), whereas touching surfaces increased during the same age period with no differences among the groups (Fig. 4G). In supine, touching the body steadily decreased uniformly for all 3 groups (Fig. 4E); touching surfaces peaked at 9 months, with more affected infants (PTI) showing more surface touching than their less affected FT and PT peers (Fig. 4H) In sitting, touching the body decreased from 3 to 18 months with no significant differences detected among the 3 groups (Fig. 4F), whereas a steady increase in touching surfaces was observed during the same age period for all infants (Fig. 4I).
Infants born with significant brain injury performed fewer bouts of exploration per minute than other infants from 6 to 9 months in prone and from 6 to 24 months in sitting. For example, in sitting, infants born without brain injury on average performed 1.8 bouts (or 7.0%) more of exploration per minute than PTI infants at the age of 6 months and 10.3 bouts (or 74.6%) more at the age of 18 months. A significant difference was found between PT and PTI infants, but not between FT and PT infants in prone (Fig. 5A) and sitting (Fig. 5C), whereas no differences were detected among the 3 groups in supine (Fig. 5B).
Estimated/modeled growth curves for the number of exploration bouts per minute (A, B, C), variability of separate behaviors (D, E, F), and variability of combined behaviors (G, H, I) performed in prone, supine, and sitting; FT = full-term, PT = preterm; PTI = preterm with brain injury; PT & PTI = no significant differences were found between the trajectories of the PT and PTI groups, so they were represented by a single line; FT & PT = no significant differences were found between the trajectories of the FT and PT groups, so they were represented by a single line; FT, PT, & PTI = no significant differences were found among the trajectories of the 3 groups, so they were represented by a single line.
With age, all infants increased the variety of separate behaviors they performed, with FT infants showing greater variability than PT infants in prone and with infants without significant brain injury showing greater variability than infants with brain injury in sitting. In prone, significant differences were detected between FT and PT infants, but not between PT and PTI infants (Fig. 5D). In supine, no significant differences were identified among FT, PT, and PTI infants (Fig. 5E). Significant differences were observed in sitting between PT and PTI infants, but not between FT and PT infants (Fig. 5F).
Finally, FT infants performed a greater variety of combined behaviors in prone and sitting as compared to all PT infants, and PT infants also performed a greater variety of combined behaviors than PTI infants in sitting. In prone, significant differences were detected between FT and PT infants, but not between PT and PTI infants (Fig. 5G). In supine, no significant differences were observed among FT, PT, and PTI infants (Fig. 5H). In sitting, significant differences were found among FT, PT, and PTI infants (Fig. 5I).
Discussion
Summary of the Findings
The goal of this study was to quantify the performance of non–object-oriented exploratory behaviors from 0 to 24 months of corrected age in prone, supine, and sitting postures for infants born full-term, preterm without significant brain injury, and preterm with significant brain injury. Infants across groups performed a variety of non–object-oriented exploratory behaviors beginning in the first months of life and continuing throughout the 2-year study. Significant differences (very large effect sizes) were observed among the 3 groups in the developmental patterns of most non–object-oriented exploratory behaviors. A limitation of this study was that there were only 6 infants born preterm with significant brain injury. Below we discuss the key findings of this study.
Infants born preterm had poorer head control and midline head and hand behaviors than infants born full-term. All preterm infants, especially infants born with brain injury, held their heads up less often than their full-term peers in prone. Preterm infants also had difficulties holding their heads in midline in prone and supine. In addition, all preterm infants held their hands in midline less often than full-term infants in prone and sitting. Holding the head up against gravity and holding it in midline, as well as holding the hands in midline, are important markers of postural control that facilitate exploration of the environment and learning.18,27,33,34 For example, these abilities can result not only in improved reaching skill,35,36 but also in enhanced visual tracking and visual-motor coordination.34,37 This can allow fixation of gaze on objects and faces for word learning and social interaction, further facilitating the development of attention, joint attention, language acquisition, and social-emotional and cognitive development.38-42 Consequently, delayed postural control in preterm infants might result in missed opportunities to explore and learn, and may put them at risk for developmental motor, language, social, and cognitive delays.
Infants born preterm showed more asymmetrical fisting of one hand than infants born full-term. Hand fisting decreased with age in all infants across the 3 postures. Differences between full-term and preterm infants were not observed in the developmental trajectories of bimanual fisting, but asymmetrical one-handed fisting was observed more often in preterm infants than in their full-term peers throughout the 2-year study. The ability to maintain both hands open is an important milestone that allows infants opportunities to grasp objects, which further leads to object manipulation and exploration.24,27,43 Object manipulation and exploration results in information gathering that facilitates word learning, object discrimination and categorization, and other key cognitive abilities.5,12,15,17-20 The observed asymmetrical hand fisting may be a soft sign of neurologic impairment or may reflect poorer ability to control movement across the many degrees of freedom in the upper extremity.44 This behavior may impair information gathering and learning for infants born preterm.
Some non–object-oriented exploratory behaviors showed different patterns across the 3 postures, and the interpretation of these behaviors might also differ. For example, full-term infants performed more mouthing of their hands in supine but less in prone than preterm infants. Mouthing of the hands in prone is more often observed in infants who cannot hold their heads up against gravity because the hands are typically positioned between the face and the floor. In contrast, mouthing of the hands in supine demands the infant lift the arms against gravity and accurately reach the hand toward the mouth as a target. Therefore, the observed differences might suggest that full-term infants had better head control in prone and better hand-to-mouth behavior in supine.
Furthermore, looking at the hands is more challenging in supine than in prone. In prone, the hands are more directly in view, while in supine infants must move the arms against gravity and control the upper extremities to bring the hands within the field of vision. Consequently, in supine, infants born preterm with significant brain injury looked less at their hands and looked less at their hands while performing actions with their hands. Looking at the hands, especially while engaging in other behaviors with the hands, provides visual and proprioceptive feedback that stimulates the development of reaching45-48 and visual-motor coordination.49,50 Previous research demonstrated that preterm infants performed less visual-haptic multimodal exploration of objects than their full-term peers.16 We suggest that these object exploration differences likely stem from insufficient multimodal hand-eye coordination experiences via early non–object-oriented exploratory behaviors.
The number of bouts of exploration per minute allowed us to estimate the intensity of infants’ non–object-oriented exploration. Infants born without significant brain injury performed more bouts of exploration per minute (more frequent switching between behaviors) than infants born with brain injury in prone and sitting. The difference in the intensity of exploration increased from a mere 7% at the age of 6 months to a disturbing 74.6% at the age of 18 months. This behavioral snapshot represents a significant gap in performance that is likely amplified across each hour, day, and month throughout at least the first 2 years of life. The result is that infants born with a significant brain injury likely gather much less information about their bodies, the surrounding environment, and their action possibilities than their peers. This diminished physical activity level also likely equates to less daily natural movement training and negative sequelae such as impaired strength, coordination, and motor control. These impairments in non–object-oriented exploration likely translate into the delays in intensity and quality of object exploration that have been reported for preterm infants.16
Variability is a hallmark of typical development, and infants born preterm have been shown to demonstrate less variability in their spontaneous movements,51 center of pressure distribution in supine and sitting,52 and in their object exploration behaviors.16 In this study, infants born with brain injury performed a lesser variety of separate non–object-oriented exploratory behaviors than infants without brain injury (FT and PT) in sitting, and all preterm infants (PT and PTI) showed less variability relative to full-term infants in prone. Thus, preterm infants showed less variability of their exploratory behaviors, which likely results in a lesser variety of daily opportunities for sensorimotor exploration and learning.7,20,53,54
Combined behaviors represent an infant's ability to perform at least 2 behaviors simultaneously. In this study, preterm infants performed a lesser variety of combined behaviors than full-term infants in sitting and prone throughout the first 2 years of life. Therefore, they likely had fewer opportunities every day to engage in multimodal exploration, which would allow for the uptake and comparison of information via more than one sensory system, the mapping of object properties using proprioceptive, somatosensory, oral, tactile, and visual information, rich information gathering, better preparation for future object exploration, and enhanced future learning outcomes.40,43,55–59
Clinical Implications
These results highlight important points for early intervention practice and research for infants born preterm. In summary, compared to full-term peers, preterm infants showed signs of impaired antigravity head control, less midline head and hand behavior, more frequent asymmetrical one-handed fisting, poorer antigravity hand-to-mouth and visual-motor coordination, and lower intensity and variability of their non–object-oriented exploratory behaviors. These impairments in early non–object-oriented exploration are likely to cascade into delays in reaching and object exploration, which in turn will result in future motor and cognitive delays.17,20 Thus, these aforementioned behaviors might serve as early predictors of future motor and cognitive development. The behavioral differences observed in this study likely result from a complex interaction of individual abilities, risk factors, differences in brain structure and physiology, caregiver-infant interactions, and perceptual-motor experiences.1–10,60,61
The results also suggest that non–object-oriented exploratory behaviors should be a focus of early intervention for preterm infants. Future research should expand upon this work by assessing early non–object-oriented exploratory behaviors in larger samples of preterm infants born with brain injury and in other at-risk groups. It should also aim to determine the types of interventions that can positively impact infants’ abilities to explore their action capabilities so they can optimally explore their bodies and surfaces, especially in the first months of life before they demonstrate sustained interest in objects. Potential interventions should focus on improving antigravity head control, midline head and hand behaviors, hand posture and control for grasp, visual-motor coordination, hand-to-mouth behavior, and intensity and variability of behavioral performance across postures including prone, supine, sidelying, and sitting. Future studies should provide these interventions using experimental designs to determine whether such interventions in early months can impact immediate as well as future developmental abilities.
Author Contributions and Acknowledgements
Concept/idea/research design: I. Babik, J.C. Galloway, M.A. Lobo
Writing: I. Babik, M.A. Lobo
Data collection: M.A. Lobo
Data analysis: I. Babik, M.A. Lobo
Project management: J.C. Galloway, M.A. Lobo
Fund procurement: J.C. Galloway, M.A. Lobo
Providing facilities/equipment: J.C. Galloway, M.A. Lobo
Providing institutional liaisons: J.C. Galloway
Consultation (including review of manuscript before submitting): I. Babik
The authors acknowledge the infants and their parents for their participation, time, and patience in making the current longitudinal study possible. The authors also thank the research assistants who helped with the coding of data.
Ethics Approval
Recruitment of participants, informed consent, and data collection were done in accordance with the regulations set by the University of Delaware's and Christiana Care Health System's Institutional Review Boards.
Funding
This research was supported by the National Institutes of Health, Eunice Kennedy Shriver National Institute of Child Health & Human Development (1R21HD076092-01A1, M.A. Lobo, principal investigator; 1R01HD051748, J.C. Galloway, principal investigator).
Disclosures/Presentations
The authors completed the ICJME Form for Disclosure of Potential Conflicts of Interest and reported no conflicts of interest.
Participants received monetary compensation for their participation in the study.
Information in this manuscript was presented in part at an education session titled “High- and Low-Tech Approaches to Assessment and Intervention in Infants” at APTA's 2017 Combined Sections Meeting on February 18, 2017, and at a Society for Research in Child Development (SRCD) oral symposium held on April 7, 2017, at the SRCD conference in Austin, Texas.
References
Comments