Pseudo-anticipation in Creutzfeldt-Jakob disease is due to a rhomboid-shaped artifact.

BACKGROUND AND PURPOSE: Previous studies have reported conflicting results regarding possible anticipation in familial E200K Creutzfeldt-Jakob disease (fCJD). Our objective was to use a large database to assess the age of disease onset (AODO) in CJD.

METHODS: The study population included 477 CJD patients [266 with fCJD, 145 with sporadic CJD (sCJD) and 66 patients of Libyan origin but negative family history] from the Israeli registry of CJD conducted since 1954. In all patients, AODO in relatives and family trees was documented. Comparison of AODO was done using a paired t test and regression using Pearson correlation for birth and year of onset.

RESULTS: The initial analysis in 52/73 families in which more than one generation was affected revealed an AODO of 63.30 ± 9.44 in the first generation compared to 56.96 ± 8.99 in the second generation (P < 0.001). However, inspection of individual AODO values plotted by year of birth showed a clear rhomboid methodological artifact generated by missing data of many young onset CJD patients who died before the database began to function in 1954 and of many late onset CJD patients missing at the present time since they will only develop the disease in the future. The 'generation' effect completely disappears if analysis is performed by year of disease onset or for the periods in which complete data are available.

CONCLUSIONS: In this very large dataset, true anticipation in fCJD patients was not detected. It is plausible that previous reports supporting the presence of anticipation are biased by a rhomboid-shaped data availability artifact.