neighborhood control sample and any sharing of risk with related phenotypes (e.g., other substance dependence or externalizing disorders) more prevalent in this group (than the non-dependent controls) could be contributing to the lack of significant differences found in this comparison. Although not significant, the point estimates (ORs 1.09) for rs2236855 and rs2236857 in the comparison of cases to neighborhood controls were greater than unity suggestive of shared risk. Interestingly, the respective ORs for comparisons of cases to neighborhood controls, with and without substance dependence diagnoses, were 1.21 and 1.00 respectively. Although neither value is significant, the estimate for the comparison to substance dependent neighborhood controls is fairly similar to that found for the case-non-dependent control comparison, a finding more suggestive of specificity of risk for opioid dependence. It is possible that other factors are protective against substance dependence in non-dependent neighborhood controls.
