Infant motor behavior serves as a window into the integrity of neural systems and later cognitive development. Deviations in early motor development predict a wide range of later outcomes in toddlers and school-aged children, including but not limited to language, cognition, academic achievement, adaptive functioning, self-regulation, and social skills.^1-4 Recently, to support developmental surveillance efforts, the American Academy of Pediatrics published a revised set of evidence-informed milestones.⁵ While these guidelines and milestone checklists are important steps toward the early identification of motor delay, historically there have been discrepancies between guidelines and their application in clinical practice. Because only a minority of pediatricians who perform routine screening use standardized tools, approximately two-thirds of developmental delays are not identified until school age, resulting in delayed initiation of early intervention services.^6-10

Similar patterns of under-identification of developmental delay are seen in high-risk populations born preterm. Despite the published guidelines for screening,¹¹ half of infants born preterm with moderate-to-severe delays and roughly three-quarters of those with mild delay fail to receive early intervention services by 2 years of age.¹² Early and routine screening of motor development is particularly important for infants born preterm because neuromotor abnormalities contribute significantly to the ‘hidden disability’ of preterm birth associated with poorer outcomes in middle and late childhood.^{13, 14} Although very early neuromotor development is not well understood in this population,¹⁵ infants born preterm demonstrate differences in their motor trajectories compared to their peers born at term; early neuromotor abnormalities and variations (e.g. posturing, dystonia, persistence of primitive reflexes, atypical general movements) are detected in a sizable portion of infants born preterm during their first year of life.^{16, 17} A meta-analysis found that children born preterm are on average − 0.57 to −0.88 SD behind their peers born at term with regard to motor development.¹⁸ These early motor difficulties can entail risk for significant lifelong motor impairments, such as those seen in cerebral palsy (7–20% of infants born preterm) or more subtle deficits such as developmental coordination disorder and motor overflow (19–41% of infants born preterm). Motor delays can occur in the absence of significant cerebral damage or cognitive deficits and have considerable effects on long-term functional outcome.^{16, 18-23}

The reasons for these discrepancies and under-identification of delay are multifactorial. They are, in part, due to inconsistent or informal methods for assessing milestones, lack of a systematic method to collect and synthesize information about development from caregivers, and the inability to assess developmental trajectories using current screening tools.^{5, 7, 9, 24} Several examiner-administered assessment measures with relatively good psychometric properties can be used to assess infant neuromotor development, including the Alberta Infant Motor Scale, Hammersmith Infant Neurological Examination, Prechtl Qualitative Assessment of General Movements, Bayley Scales of Infant Development, Mullen Scales of Early Learning,^25-27 and the Developmental Assessment of Young Children, Second Edition.²⁸ These instruments are rarely used by primary care physicians during routine clinical care, require advanced and specific training, and can be time consuming, costly to administer, and unavailable in less-resourced regions.^{27, 29, 30}

Relying on direct assessment to perform developmental surveillance is not viable in most regions and can lead to disparities in access to care. In contrast, caregiver-completed developmental screening tools can be used to efficiently and systematically track early development. They are less costly, can be administered digitally, provide information on current levels of functioning, or can be used to track developmental trajectories through serial administration.^{31, 32} Caregivers can be dependable and accurate reporters of early motor development.^{32, 33} At present, there are only a few caregiver report instruments that allow for multidimensional assessment of early development. The most commonly used and recommended parent-completed developmental screening tools in the primary care setting include the Ages and Stages Questionnaire, Third Edition, the Child Development Inventory, and the Parents' Evaluations of Developmental Status.^{8, 10, 34} These measures have been validated in typically developing infants and, to a lesser extent, in infants born preterm;^{10, 35, 36} however, each has limitations. For example, the Ages and Stages Questionnaire, Third Edition identifies risk based on pre-established cutoff scores while none of the instruments examines deviations in developmental trajectories in infants at risk for neurodevelopmental disorders.^{36, 37}

The current longitudinal study examined motor development in infants born at term and preterm from birth through to 12 months based on caregiver ratings of the PediaTrac v3.0 motor domain items from which estimates of the latent trait of motor ability were modeled. In psychometric theory, a latent trait is an unobservable trait or ability (e.g. intellectual ability) that can be measured using observed behaviors or responses. An initial psychometric study of the PediaTrac motor domain at the newborn/term-equivalent time point using item response theory (IRT) graded response modeling demonstrated that the latent trait of motor ability could be estimated reliably.³⁸ Those findings provided the foundation for the current follow-up longitudinal study that compares term and preterm development. To this end, we used the motor scale of PediaTrac, a Web-based caregiver report instrument^{31, 32} at the newborn, 2-, 4-, 6-, 9-, and 12-month time points. We hypothesized that: (1) caregiver PediaTrac motor ratings would provide reliable estimates of the latent trait of motor ability at each time point; and (2) infants born preterm would have lower scores on the motor domain than infants born at term across the newborn and term-equivalent and subsequent time points. We also examined the main and interaction effects for biological sex at each time point.

METHOD

RESULTS

As illustrated in Figure S1, based on caregiver ratings of motor ability, the motor domain was highly reliable at each time point, ranging from 0.93 to 0.99 across the five time points, with reliability based on the estimate closest to the mean of θ.

Values of θ, an index of the latent trait of motor ability, were estimated separately for participants at each time point, based on caregiver ratings of the motor domain items. Mean θ values and item difficulty parameter estimates can be thought of as being on a scale similar to a z-score, with a distribution centered at zero with an SD metric.⁴⁴ The mean and SD θ values according to the time point for the groups born at term and preterm and according to infant biological sex are shown in Table 2.

At the newborn time point, infants born preterm had significantly higher motor θ scores than infants born at term (p = 0.002, partial η² [_pη²] = 0.016). At the 2- and 4-month time points, group differences were not significant; at the 6- (p = 0.007, _pη² = 0.016), 9- (p = 0.002, _pη² = 0.019), and 12-month (p < 0.000, _pη² = 0.028) time points, infants born at term had higher motor θ scores than infants born preterm. This ‘cross-over’ in motor domain scores reflects lower rates of growth in motor development across the first year in infants born preterm, as illustrated in Figure 1. The main effect for infant biological sex, and the interaction between term status and biological sex, were not significant at any time point.

DISCUSSION

The findings of this study demonstrate that caregiver longitudinal PediaTrac ratings provide highly reliable estimates of infant motor ability that distinguish between the motor development of infants born at term and preterm over the first 12 months of life. At each time point, the parameter estimates of the motor domain items could be reliably modeled. The latent trait of motor ability was reliably estimated by θ at each time point.

There were significant term status group differences in motor function at the newborn, 6-, 9-, and 12-month time points. Although the motor scores of infants born preterm at the term-equivalent newborn period were higher than those of infants born at term, by the 6-month time point the motor scores of infants born preterm were lower than those of infants born at term. These findings suggest that the motor development of infants born preterm in the post-birth period up to term equivalence is temporarily advantageous. Studies of the development of infants born preterm in the initial post-birth phase are limited; however, in those infants born preterm for whom there are fewer problems associated with preterm birth, it is possible that early exposure to the extrauterine environment may be advantageous, at least initially.^{45, 46} More specifically, Bosworth and Dobkins^{47, 48} found that in infants born very to moderately preterm (i.e. ≥30 weeks gestational age) chromatic contrast sensitivities tended to be highest in infants who were born most preterm (i.e. those with more visual experience outside the womb). Palmer et al.⁴⁵ found that by 40 weeks postmenstrual age, infants born preterm (27–35 weeks gestational age) had less flexor tone and better auditory and visual orientation and alertness than their peers born at term. Neuroimaging studies support this, with findings of accelerated white matter development in infants born preterm (28–34 weeks gestational age), as measured by increases in fractional anisotropy, in the right and left sagittal stratum and in the posterior thalamus compared to full-term controls.⁴⁹

While there is an early advantage of extrauterine exposure for motor development, that advantage does not persist beyond the newborn time point. Lower values of the latent motor trait for infants born preterm compared to infants born at term by 6 months in the present study are consistent with findings obtained by direct measurement and with the effects of preterm birth on brain growth and development.^{50, 51} Biological sex differences in early caregiver-reported motor functioning and interactions with term status were not significant, although follow-up longitudinal IRT analyses are needed to determine if developmental trajectories differ according to group or biological sex.

The recent revision of the Centers for Disease Control and Prevention developmental surveillance checklists clearly was an important step toward improving early identification of developmental risk.⁵ The milestones identified in the surveillance checklists at 2, 4, 6, 9, and 12 months are largely identified as reliable items in the PediaTrac motor domain, although the wording differs in a few instances. Surveillance milestones were loaded on exploratory factor analysis-identified factors, with the exception of bringing hand to mouth at 4 months. Thus, study findings provide further psychometric support for the Centers for Disease Control and Prevention surveillance milestones as measured by PediaTrac. That said, there are important differences between the Centers for Disease Control and Prevention surveillance checklists and the PediaTrac scales. PediaTrac is intended as an online assessment instrument based on caregiver report rather than clinician interview and observation. Surveillance checklists are intended to assess risk at a specific time rather than measuring developmental trajectories per se. The PediaTrac approach is based on IRT-derived scales, with overlap in items at consecutive assessments, and provides the foundation for eventual computation of developmental trajectories.