A multistate model for coronary heart disease--an application to different prevention strategies.

"The discrete-time stochastic model presented in this paper pays attention [to] the association between physiological risk factors and [coronary heart disease] morbidity as well as mortality trends. Transition probabilities are modeled as a polychotomous logistic function of risk factors. A series of separate simple logistic regression analyses...are performed as a replacement for estimating polychotomous logistic regression parameters. Data from the Finnish North Karelia Project is used to test the model." (SUMMARY IN KOR)