Space-Time Clustering Analyses of Occurrence of Cerebral Palsy in Northern England for Births 1991 to 2003

Purpose

The etiology of congenital cerebral palsy is unclear. Recent studies have suggested that maternal infection is involved. Indirect, but supportive, evidence for an infectious hypothesis would be provided by the finding of space-time clustering.

Methods

We analyzed all 906 cases born during the period 1991–2003 using a population-based registry from Northern England. We applied the K-function method for testing global space-time clustering and used residential addresses at the time of birth. The Knox method was used to determine the spatio-temporal range over which global clustering occurred. Tests were repeated using nearest neighbor thresholds to allow for variable population density. Additionally, Kulldorff's scan statistic identified specific space-time clusters.

Results

There was statistically significant evidence of global space-time clustering (p ≤ 0.05). Clustering occurred over a range of close distances, but was most pronounced for cases born within 1–4 months of one another. A specific space-time cluster was identified during 1994.

Conclusions

The application of spatial statistics for the study of cerebral palsy is novel. There is tentative evidence for the involvement of spatially and temporally varying environmental etiological factors. These findings may support a role for infections or similar agents. However, the role of chance cannot be excluded.

Key Words: Etiology, Cerebral Palsy, Environment, Infections, Space-time Clustering

Selected Abbreviations and Acronyms: HHV, human herpes virus, O, number of pairs of cases observed to be in close proximity, E, number of pairs of cases expected to be in close proximity, S, strength of clustering, NN, nearest neighbor