Abstract
OBJECTIVES: To define subaxial cervical spine vertebral body and canal dimensions in a paediatric cohort and to assess the influence of age and ethnicity. DESIGN: Retrospective radiological observational study. SETTING: Single-centre tertiary level trauma care setting in New Zealand. PARTICIPANTS: CT scans of children under 18 years of age were reviewed. A total of 111 participants were included (63 New Zealand European (NZE) and 48 Māori). Patients with cervical spine pathology or deformity were excluded. INTERVENTIONS: Not applicable. PRIMARY AND SECONDARY OUTCOME MEASURES: Primary outcomes were anteroposterior and transverse vertebral body and spinal canal dimensions measured at the mid-pedicle level from C3 to C7. Secondary outcomes included canal-to-vertebral body (canal:VB) ratios. Associations with age and ethnicity were assessed using correlation analysis and analysis of covariance (ANCOVA). RESULTS: Absolute vertebral body and canal dimensions were larger in NZE children compared with Māori. Canal:VB ratios were smaller in NZE children, reaching statistical significance at C7 (p=0.011). Age demonstrated a strong positive correlation with mean vertebral body anteroposterior diameter and a moderate correlation with canal:VB ratio. ANCOVA showed ethnicity (NZE) to be a significant predictor of vertebral body dimensions, particularly transverse diameter at C4-C6, while age had a lesser effect. Canal:VB ratios decreased with increasing age from C3 to C7, with low coefficients of determination indicating additional influencing factors. CONCLUSION: In this paediatric cohort, vertebral body dimensions were more strongly associated with age than spinal canal dimensions. Ethnicity was associated with modest differences in cervical spine morphology, particularly transverse vertebral body diameter. These findings suggest cervical spine development is multifactorial and may have implications for trauma assessment and spinal cord injury risk evaluation in adolescents. Further studies incorporating anthropometric and sex-specific variables are warranted.