Objective: To examine the significance of crossover analysis in gene-environmental interaction studies.
Methods: Through elaboration of a case-control study on the increased risk of venous thrombosis in oral-conceptive users who were carriers of factor V Leiden mutation, core information from 2 x 4 crossover table were analyzed and compared with stratified analysis and 'case only' study.
Results: Different models (additive or multiplicative) in analyzing gene-environmental interaction yielded different results. The result of interaction based on multiplicative model was 1.35 (P > 0.05), compatible with that of stratified analysis and case only study. Calculated by crossover analysis based on additive model, synergy index S(S), attributable proportion of interaction (AP) and relative excess risk of interaction (RERI) appeared to be 3.90, 72.24%, 25.08 (P > 0.05) respectively.
Conclusion: Crossover analysis should further be applied in gene-environmental interaction studies.