Patients and methods

All patients and controls were of French origin. The criteria for a definite diagnosis of Parkinson’s disease were the presence of a pronounced response to levodopa and at least two of the following: bradykinesia, rigidity, rest tremor, or asymmetric onset. Exclusion criteria were the presence of at least one of the following: ophthalmoplegia, pyramidal or cerebellar syndrome, apraxia, prominent and early postural instability, urinary incontinence, or dementia within the first two years of evolution. Familial cases, defined by the presence of definite Parkinson’s disease in at least two examined first degree relatives, were selected through a French clinical network (The French Parkinson’s Disease Genetics Study Group). Isolated cases with no family history of Parkinson’s disease among first and second degree relatives, verified by structured interviews of the patient and spouse, were recruited successively at the Fédération de Neurologie of the Hôpital de la Salpêtrière in Paris, France. Controls were age and sex matched selected subjects from a non-demented control sample.8

APOE genotypes were determined by polymerase chain reaction/restriction on DNA extracted from the blood of consenting patients.9APOE allele and genotype frequencies were compared by χ²test, with the Yates’ correction when appropriate.

Results

Mean age at onset and at examination were similar in isolated and familial Parkinson’s disease (table). The most frequent genotype in all groups was ε3/ε3. Only three patients were demented; they belonged to the familial Parkinson’s disease group and had the ε3/ε3 APOE genotype. No ε4 homozygotes were found. The allele distribution was significantly different in patients with familial and sporadic Parkinson’s disease compared with controls (χ²=8.16, P=0.017). The comparison was not significant between familial and sporadic Parkinson’s disease groups. The ε2 allele was significantly more frequent in sporadic Parkinson’s disease than in controls (odds ratio 2.45, 95% confidence interval 1.15–5.21), but did not differ between familial and sporadic patients.

Discussion

Thus in familial Parkinson’s disease, as in sporadic cases, there is no association with the APOE ε4 allele. The lack of association between APOE ε4 allele and Parkinson’s disease was already reported in previous studies that did not distinguish between sporadic and familial cases.2 3 6 7 Our data suggest that the APOE ε2 allele may be associated with patients with sporadic Parkinson’s disease. The association cannot be explained by the exclusion of patients with early dementia, possibly related to APOE ε4 allele, as demented subjects were also excluded from the controls.

Although no conclusion can be drawn concerning the relation between familial and isolated Parkinson’s disease, it is clear that an important risk factor for Alzheimer’s disease, the APOE ε4 allele, is less frequent in Parkinson’s disease, whether sporadic or familial.