SAS codes and format? Directions: Use either SAS to answer the following questions. Include the code, plots, and comments under each question, then submit the completed document on Canvas.

In this assignment, you will investigate the relationship between weight and the incidence of cardiovascular disease (CVD) in a longitudinal study of 453 adults. The outcome variable of interest is time to the occurrence of CVD.

Data=hw7.comprisk

The variables in the dataset are:

ID - participant identifier number

Age - age in years

Sex - 0=Male, 1=Female

BMI - body mass index

Time - follow up time in days

Status - 1=CVD, 0=censored

BMI.1 - 1=Underweight, 0=other

BMI.2 - 1=Overweight, 0=other

BMI.3 - 1=Obese, 0=other

Body Mass Index (BMI) is a continuous measure of body fat based on height and weight that applies to both adult men and women. The Center for Disease Control has discretized BMI and assigned descriptive labels to the different categories:

Underweight:

Normal weight:

Overweight:

Obesity:

1. Create a Kaplan-Meier curve to estimate the distribution of the time-to-onset of CVD for the subjects in this study. Find the 25th percentile of the distribution of the time to onset of CVD and interpret, including confidence intervals.
2. Create and compare the Kaplan-Meier survival curves for males versus females. Use the log-rank test to determine whether the distribution of time-to-onset of CVD is the same or different by sex.
3. Create and compare the Kaplan-Meier survival curves for the categories of BMI using the categorical indicator variables:BMI.1,BMI.2,BMI.3. Use the log-rank test to determine whether the time to onset of CVD is the same or different by BMI category. Interpret your results in the context of the problem.
4. Fit a Cox proportional hazards model in which the continuous BMI variable (i.e.,BMI), age (i.e.,Age), and sex (i.e.,Sex) are predictors of the hazard rate for the time to the occurrence of CVD. Interpret the results in the context of the problem. Include test statistics and 95% confidence intervals in your interpretation.
5. Evaluate the proportional hazards assumption for the model you fit in question 4 both globally and for each variable individually. What do you conclude?
6. Check for influential observations using DFBETAS for the model you fit in question 4. What do you conclude?
7. Refit the model but instead of using BMI as a continuous variable, use the indicator variables for the CDC's descriptive categories of weight. Interpret the results in the context of the problem. Include test statistics and 95% confidence intervals in your interpretation. In this model with indicator variables for BMI, test for the significance of BMI.
8. Evaluate the proportional hazards assumption for the full model you fit in question 7. Interpret the results for each of the variables individually and globally. What do you conclude?