A Cox proportional hazards model is estimated, relating time to development of myocardial infarction (heart attack) over 10 years to high diastolic blood pressure (DBP), defined as DBP $ 90 mmHg. The unadjusted association is estimated first, then the association is adjusted for age and sex, and then for additional clinical risk factors. The parameter estimates and significance levels for diastolic blood pressure for each model are shown in Table 11–29.
a. Is there a statistically significant association between high DBP and time to MI? Discuss significance in each of the models.
b. What is the unadjusted hazard ratio for MI in participants with high versus low DBP?
c. What is the age- and sex-adjusted hazard ratio for MI in participants with high versus low DBP?
d. Compute the hazard ratios, adjusting for the additional clinical risk factors, and compare them to the unadjusted hazard ratio from (b). Are there differences? If so, what would account for the differences?