Children's Height and Injury Risk With Child Restraint Systems

The transition from child restraint systems (CRSs) to seat belts (SBs) is often based on fixed age or height thresholds. However, the association between height and injury patterns among injured children remains insufficiently characterized within clinical trauma settings.

To examine the association between height and regional injury patterns among children involved in motor vehicle crashes evaluated at trauma centers and to assess how these patterns vary between CRS and SB users.

This retrospective cohort study of children 6 to 11 years old who were involved in motor vehicle crashes evaluated at trauma centers used inverse probability of treatment weighting with gradient boosting machine-derived propensity scores and restricted cubic spline modeling to evaluate height-dependent associations between protective device and injury pattern. Data from the American College of Surgeons-Trauma Quality Improvement Program (2017-2020) were used. Statistical analyses were conducted between June 1, and August 31, 2025.

Prevalence of severe regional injuries (Abbreviated Injury Scale score ≥3).

A total of 6503 children aged 6 to 11 years (mean [SD] age, 8.48 [1.69] years; 3251 male [50.0%], 3252 female [50.0%]) were identified from the database; 1066 (16.4%) had used a CRS. After weighting, CRS use was associated with a higher prevalence of severe head and neck injury (13.9% vs 10.5%; odds ratio, 1.41; 95% CI, 1.06-1.88, P = .02) and severe spinal cord injury (2.9% vs 1.8%; odds ratio, 1.70; 95% CI, 1.02-2.84, P = .04) compared with SB use. Height-dependent analysis revealed a nonlinear association between restraint type and injury patterns. Among children measuring 110 to 140 cm, CRS use was associated with a higher prevalence of severe spinal and thoracic injuries, whereas the prevalence of abdominal injury was lower within the same height range.

In this cohort study, the association between protective devices and injury patterns was height dependent and nonlinear. A shift in regional injury anatomy was observed across the 110- to 140-cm spectrum, suggesting that a uniform threshold may not adequately characterize growth-specific vulnerabilities.