BIOL PSYCHIATRY
1991;29:721-729
721
COMMENT
Marker Genotyping Errors in Old Data on
X-Linkage in Bipolar Illness
Elliot S. Gershon
Investigations of linkage markers of the X-chromosome colorblindness region in bipolar
manic-depressive illness (BP) have yielded inconsistent results, with linkage accepted in
some and rejected in ott,'er studies. Although genetic heterogeneity has b, en proposed
as the reason for differences, other possibilities exist, including systematic procedural
errors. Statistical evidence for linkage between the markers, Xg and colorblindness, is
present in a series of papers on bipolar illness reported in 1972-1975. The linkage
implied by this reanalysis is spurious, since the two markers are at opposite ends of the
X chromosome. The presumptive reason for this spurious linkage is that it is a result of
systematic genotypi~s c~ors. The support provided by these data to the X-linkage hy-
pothesis in BP illness is :thus diminished. That is, the linkage to illness may depend on
systematic errors in marker genotyping. In general, the possible causes of inconsistency
between linkage reports may be divided into statistical and systematic causes. Statistical
causes wouM generally consist of chance differences in sampling, such as might occur
under genetic heterogeneity. If this occurs, the reports rejecting linkage may be fi,',~e
negatives, or the reports detecting linkage may be false-positive results. Systematic caases
of differences among reports couM include systematic errors (or variations) in procedures,
including ascertainment, diagnosis, genotyping, or analysis. Consistency of the marker
map in a particular study with the known marker map is one test for systematic errors
in genotyping.
lllLl UUU~ LLILILI
Bipolar manic-depressive (BP) illness is a common disease, in the sense that its population
prevflence is greater than 1%, and several times that if one takes into account the
repeatedly observed comorbidity of major depressive illness and schizoaffective disorder
in families of manic-depressives (Gershon et al 1982). Twin and adoption studies suggest
it is an inherited illness, but segregation analyses do not identify Mendelian inheritance
(Goldin et al 1983). As in other common diseases, there has been a great deal of interest
in detecting single locus susceptibility through linkage markers. Linkage has been reported
to various loci, most notably to loci on the X chromosome, where the first positive linkage
From the Clinical Neurogcnetics Branch, National Institute of Mental Health, Building 10, Room 3N218, 9000 RockvilIe Pike,
Bethesda, MD 20892.
Address reprint ~equests to Elliot S. Gershon, Clinical Neurogenetics Branch, National Lnstitute of Mental Health, Buil~ag
10, Room 3N218, 9000 Rockvilie Pike, Bethesda, MD 20892.
Received March 22, 1.o90; revised July 11, 1990.
© 1991 Society of Biological Psychiatry 0006-3223/9!/$03.50