The distributions of HLA-A, -B, -(Bw4/Bw6), -C, -DR, -(DRw52/DRw53), and -DQ genes in 19 type I diabetics, 37 type II diabetics, and 13 nondiabetics of the Yoruba tribe in southwestern Nigeria were studied. Because no associations between type II diabetes and HLA were detected in the current study and such associations are not known to exist in most ethnoracial groups, type II diabetics plus nondiabetics were used as a group of controls for the group of type I diabetics. Trends toward associations between increased DR3 (53 versus 30% of controls) and decreased DR2 (21 versus 46% of controls) and type I diabetes were found (0.1 > P >0.05). The strongest HLA association with type I diabetes in Caucasians is usually with DR4. The percent of DR4-positive type I diabetes (11%) was not significantly greater that that in the controls (4%). Because the strong HLA associations with type I diabetes in American Blacks are the same as in Caucasians (i.e., increased DR3 and DR4 and decreased DR2), the genetic contribution (i.e., the lack of an association with DR4) to susceptibility to type I diabetes in most Nigerian Blacks may be different from that in most Caucasians and American Blacks.
Onset of diabetes in most of the type I subjects was after age 20, and type I diabetics were difficult to recruit for the study, in keeping with reports on the rarity of type I diabetes among Blacks in western Africa and reports of DR4, but not DR3, being correlated with an early age of onset in Caucasians. Most cases of type I diabetes in American Blacks have an onset before age 20, and the incidence of type I diabetes in American Blacks is between that in American Blacks and that in Caucasians. Thus, the current data provide additional support for the theory that most type I diabetes occurring in American Blacks is the result of the admixture of a Caucasian-derived susceptibility gene into the American Black gene pool and, in addition, suggest that the susceptibility determinant is associated with DR4.