[Conditional logistic regression analysis and path analysis of risk factors of cervical cancer].

A case-control study of 125 pairs was carried out in Jinan during 1987-1988. 39 variables were analysed with the single factor analysis, 14 factors were confirmed as the suspicious risk factors. Conditional Logistic regression analysis of the above 14 factors was used. Results showed that the following five risk factors were associated with cervical cancer, i.e., cervical erosion, low family income, poor personal hygiene during menstrual period, sexual factors and early age at marriage. The causal model with 9 risk factors of cervical cancer was set up and studied with path analysis. The result indicated that 4 risk factors had direct and indirect influence on the occurrence of cervical cancer respectively.