Cranial ultrasound prediction of disabling and nondisabling cerebral palsy at age two in a low birth weight population

Abstract

Objective: To employ multivariate analytic techniques to assess the association between neonatal cranial ultrasound (US) abnormalities and subsequent cerebral palsy (CP), defined as disabling CP (DCP) or nondisabling CP (NDCP) depending on the level of motor dysfunction.

Design: Prospective cohort study.

Subjects and methods: The Neonatal Brain Hemorrhage Study enrolled a geographically representative sample of 1105 newborns 501 to 2000 g and obtained follow-up data on 777 (86%) of the 901 survivors at age two. One hundred thirteen children (14.6%) had motor findings severe enough to classify them as having CP. The 61 (7.9%) of these children who were disabled by their motor impairment we classified as having DCP. The remaining 52 (6.7%) who had definite neurologic findings (usually mild spastic diplegia) but without evidence of interference with daily living, we classified as having NDCP.

Results: In a multivariate logistic regression model of perinatal and postnatal variables, the following factors were found to be significant risk factors for DCP: parenchymal echodensities/lucencies or ventricular enlargement (PEL/VE) on cranial US (OR = 15.4; 7.6, 31.1), germinal matrix/intraventricular hemorrhage (GM/IVH) (OR = 3.5; 1.7, 6.9) and mechanical ventilation (OR = 2.9; 1.2, 7.1). Fully 93.4% of infants were correctly classified as to presence or absence of DCP on the basis of this model. Birth weight, gestational age, length of hospital stay, gender, race, plurality, presence of labor and Apgar score were not significant independent predictors of DCP. For NDCP, the only risk factor significant in the multivariate model was PEL/VE (OR = 5.3; 2.2, 12.6).

Conclusions: Among perinatal and postnatal factors, cranial US abnormalities are by far the most powerful predictors of disabling CP in low birth weight infants. Although PEL/VE was the strongest predictor, GM/IVH also appeared to independently contribute to the risk of DCP. NDCP in low birth weight infants appears to have a different risk profile than DCP. In particular, it is less closely related to US evidence of perinatal brain injury.