Overview
In this assignment, you will be working with the Framingham Study data to answer the research question, is high blood pressure associated with risk of death? Follow the steps below. Submit your responses to the questions below, as well as your SAS log, via Canvas. Complete the following steps. Step 1 (4 points): Import frmgham2.sas7bdat file, examine the meta data and first 10 observations of the data file. Subset the dataset to only include observations where PERIOD=1. Run a PROC FREQ to examine the frequencies of data for Death (DEATH), Incident MI or Stroke (CVD), Sex (SEX), and Anti-Hypertensive Meds (BPMEDS). Run a PROC MEANS to examine data for Age at exam (AGE), systolic blood pressure (SYSBP) and diastolic blood pressure (DIABP). Step 2 (3 points): Create a binary high blood pressure variable where high blood pressure is defined as systolic blood pressure ?130 mm Hg OR diastolic blood pressure ?80 mm Hg. Run a PROC FREQ to examine the prevalence of high blood pressure in the sample. Step 3 (4 points): Create a 2x2 table comparing your newly created high blood pressure variable against the outcome of death and then run a PROC LOGISTIC statement to compute the unadjusted odds ratio and 95% confidence interval for this relationship. Interpret the odds ratio. Provide a one sentence interpretation of the association between high blood pressure and deth. Step 4 (3 points): Consider the role of Age in the association you tested in Step 3. Determine whether you should consider age as a confounder, effect modifier or mediator in this association. If you determine age to be a confounder, then estimate and present age-adjusted odds ratios. If you determine age to be an effect modifier, then estimate and present age-stratified odds ratios. If you determine age to be a mediator, it should not be included in your multivariable regression model. Provide 1-2 sentence rationale for your determination as well as a one sentence interpretation of the association between high blood pressure and death.
Step 5 (3 points): Consider the role of Sex in the model you tested in Step 3. Determine whether you should consider sex a confounder, effect modifier or mediator in this association. If you determine sex to be a confounder, then estimate and present sex- adjusted odds ratios. If you determine sex to be an effect modifier, then estimate and present sex-stratified odds ratios. If you determine sex to be a mediator, it should not be included in your multivariable regression model. Provide 1-2 sentence rationale for your determination as well as a one sentence interpretation of the association between high blood pressure and death. Step 6 (3 points): Consider the role of anti-hypertensive medication use in the model you tested in Step 5. Determine whether anti-hypertensive medication use serves as a confounder, effect modifier or mediator in this association and present the appropriate analyses. Provide a 1-2 sentence rationale for your determination as well as a one sentence interpretation of the association between high blood pressure and death. Step 7 (Bonus + 5): Re-run your regression models from prior steps but estimate risk ratios (and 95% confidence intervals) rather than odds ratios. Describe how the risk ratios compare to your previously estimated odds ratios.