Does accelerometry reflect hand function in infants at risk of unilateral cerebral palsy? A secondary analysis of BB-Bim results.
Wrist accelerometry bracelets don’t track real hand function in infants under 12 months—stick with structured assessments like HAI.
01Research in Context
What this study did
Six babies under 12 months wore small wrist bracelets that counted every arm twitch.
Each baby was already flagged as high risk for unilateral cerebral palsy.
Researchers compared the bracelet counts to a gold-standard hand test called the Hand Assessment for Infants (HAI).
What they found
The bracelet numbers did not match the HAI scores.
A baby could wave a lot yet still score low on real hand skills like grasping a toy.
The bracelets missed the difference between random flails and purposeful reaching.
How this fits with other research
Miltenberger et al. (2013) saw the same mismatch in older kids with autism.
Their accelerometers said activity levels were fine, but parents saw fewer play types.
Dembo et al. (2023) went the other way: hip-worn sensors worked great for adults with Down syndrome.
The key difference is placement and age.
Hip sensors in adults track big movements like walking, while wrist sensors in babies catch tiny, non-purposeful twitches.
Chen et al. (2015) showed that even tiny babies with stroke already move differently, so we need tests that spot quality, not just quantity.
Why it matters
Skip the wrist bracelets for now.
Keep using structured tests like HAI or simple reaching symmetry checks during routine play.
If you need objective data, wait until the child is older and use motion capture or video instead.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Put the bracelets away and run a five-minute HAI or watch for reaching symmetry during free play.
02At a glance
03Original abstract
BACKGROUND: Infants at high-risk of unilateral Cerebral Palsy (UCP) may have asymmetry in upper extremity movement and function, which should be identified as soon as possible for management. AIMS: To explore the feasibility of using two AX3 Axivity monitors in wrist-worn bracelets to quantify movements, and to identify whether accelerometry parameters are consistent with hand function. METHODS AND PROCEDURES: 6 infants at high risk of UCP (aged 3 to -12 months) were included in a Single-Case Experimental Design to explore the impact of an 8-week bimanual stimulation home program. OUTCOMES: Each week of the baseline (randomized duration 4-7 weeks) and 8-week program, the Hand Assessment for Infants (HAI) was performed and accelerometry parameters were collected during HAI and also during spontaneous activity, several times a week. RESULTS: Actimetry was analyzed during HAI and 238 spontaneous activity sessions (mean 42 ± 21 min). Actimetry ratios distribution and evolution show a high variability, especially for spontaneous activity. No strong correlation was found between HAI scores and accelerometry parameters, either collected during HAI, or during spontaneous activity times. CONCLUSION AND IMPLICATIONS: Despite its feasibility, using accelerometry bracelets looks unreliable for detecting and monitoring hand function in infants under one year.
Research in developmental disabilities, 2023 · doi:10.1016/j.ridd.2023.104549