An increased risk of Alzheimer disease (AD) has been reported in young mothers of Down syndrome (DS) probands. Subsequently, an increased frequency of the apolipoprotein E (apoE) allele epsilon 4 has been found in mothers (< or = 32 years) of DS children due to meiosis II (MII) errors providing a potential explanation for the increased risk of AD in DS mothers. In the present study we genotyped apoE and determined the origin of non-disjunction of 132 mothers and the corresponding fathers and DS children from Spain. Unexpectedly no epsilon 4 alleles have been detected in MII mothers of < or = 32 years of age (P = 0.02). Thus our study not only fails to find the effect previously reported, but it detects an opposite correlation. An increase in the epsilon 4 frequency (0.227) is detected in MI mothers <28 as compared to the epsilon 4 frequency present in MI mothers >28 years of age (0.089), although the differences are not significant if correction for multiple comparisons is applied. The simplest overall interpretation of the previously reported and present findings is that the detected associations are due to random statistical variation rather than to some real effect of the epsilon 4 allele. However the important potential implications of alternative explanations imply that this issue deserves further clarification in independent studies in other populations.