Methodological issues in linkage analyses for psychiatric disorders: secular trends, assortative mating, bilineal pedigrees. Report of the MacArthur Foundation Network I Task Force on Methodological Issues.

A Task Force was assembled to address three data problems in genetic linkage analyses: (1) a secular trend, e.g. cohort effect; (2) positive assortative mating, and (3) bilineal pedigrees. All are cited as reasons for failure to replicate genetic linkage reports. However, we knew of no work demonstrating that these factors could invalidate or bias linkage analyses, nor that they were complications (e.g., variable age of onset). The Task Force concluded that these factors can reduce the power of linkage analyses and result in bias in the estimate of the recombination frequency due to the fact that they represent 'noise' in the system. There was little evidence that the three factors would invalidate a linkage analysis or be directly responsible for negating a linkage finding.