Birthplace as a risk factor in motor neurone disease and Parkinson's disease.

A proportional mortality study was conducted to determine if state of birth is a risk factor associated with motor neurone disease (MND) and Parkinson's disease (PD) using US death certificate information for 1981. State of birth was used as a surrogate variable for location of early childhood environment. A gradient of risk by geographical area in the US was found for MND and PD. Cerebral vascular accident (CVA) deaths served as a comparison group. Multiple sclerosis (MS) deaths were analysed to validate the methods used. A geographical relationship between latitude proportional mortality ratios for MND, PD and MS by state of birth, and a geographical relationship between latitude and proportional mortality ratios for MND and MS by state of birth were found. Statistical modelling was used to compare the deaths from MND, PD, MS, CVA to all other deaths in the US by state of birth. The resulting models were evaluated to determine if any individual states were not well represented by the model for each disease. As predicted, no pattern was evident for CVA. The geographical gradient observed for MS was as predicted by other studies. The geographical pattern found for MND has a northwest to southeast gradient and the pattern found for PD has a west to east gradient.