Estimating the relative recurrence risk ratio for leprosy in Karonga District, Malawi.

ABSTRACT

Leprosy is a chronic disease caused by infection with Mycobacterium leprae. Susceptibility to leprosy is influenced by both genetic and non-genetic factors and the disease is known to cluster in families. One measure of genetic effect is the relative recurrence risk ratio, lambdaR. Estimates of this parameter can be inflated if environmental risk factors which also cluster in families, such as household contact, are not properly accounted for. We present the results of fitting a cross ratio model that allows estimation of the odds ratio of disease conditional on disease or no disease in a given relative, given measured covariates. From this model we can predict fitted values for lambdaR that represent the familial risk not accounted for by other covariates including observed household contact. If all covariates could be measured, this would be the 'genetic relative risk ratio'. We find lambdaR > 1 for all relative pairs except grandparent-grandchild, and lambdaR > 2 for siblings. Though not in itself evidence for a strong genetic susceptibility to leprosy, this result is consistent with much other evidence which suggests susceptibility to leprosy is under the control of many factors, the strongest of which may be non-genetic, with host genetics playing a small but significant role.