Editor’s Note: Dr. Preston Klein (he/him/his) is a resident physician in child neurology at the University of Virginia. He is interested in neonatal neurology and medical education. -Rachel Y. Moon, MD, Associate Editor, Digital Media, Pediatrics
Achievement of age-appropriate developmental milestones is a routine measurement at well-child appointments, and assessment of gross motor development is a key part of this evaluation. Assessing gross motor milestones (GMM) often relies on parental completion of questionnaires and direct observation of the infant during the appointment. However, parent questionnaires are inherently subjective. It can also be difficult for a pediatrician to evaluate these milestones based on observation alone during the short time period of an appointment, particularly given that infants may behave differently in a clinic setting than when they are at home.
These difficulties demonstrate the need for an objective measure of motor development, as this could allow for earlier detection of delays, and earlier intervention.
Dr. Manu Airaksinen and colleagues at the University of Helsinki and Helsinki University Hospital sought to determine if at-home wearable multi sensors could be used to objectively quantify and track an infant’s gross motor development in their paper, entitled “Assessing Infant Gross Motor Performance with an At-Home Wearable,” which, in conjunction with a video abstract, is being early released this week in Pediatrics (10.1542/peds.2024-068647).
The authors independently enrolled two prospective cohorts. Cohort 1 consisted of 97 typically developing children aged 4–19 months, and cohort 2 consisted of 37 children aged 4–22 months attending pediatric neurology follow-up due to diagnosed or suspected developmental differences. Home motor measurements were obtained using a standardized Motor Assessment of Infants with a Jumpsuit (MAIJU) wearable. A GMM-specific algorithm was developed from the wearable data based on postures and movements that were behaviorally relevant for the performance of specific milestones. This automated algorithm was then used to predict achievement of milestones based on the at-home wearable data. Accuracy of the GMM algorithms in predicting milestone achievement were compared to data from parental questionnaires.
The authors’ analysis demonstrated that the GMM algorithm had high accuracy in correctly detecting the achievement of milestones:
- Independent walking and four-limb crawling had a prediction accuracy of 94.3%–96.8%
- Standing had a prediction accuracy of 90.9%–92.4%
- Sitting had a prediction accuracy of 93.1%–95.8%
- Prone crawling had a prediction accuracy of 94.3%–95.5%
Additionally, the algorithm could be used to calculate an infant's time spent in different postures, such as prone, crawling, or standing positions and provide longitudinal tracking of gross motor development.
Data from at-home wearables eliminates the subjective nature of parent questionnaires, allows for information to be obtained from the infant’s normal environment, and allows for a longer period of monitoring than a typical clinic appointment allows. Additional research to validate the generalizability of this model is needed, but the authors demonstrate promising potential for at-home wearables in providing objective data to monitor an infant’s gross motor development.
