************************************************************************
* Program:  C:\NHANES\.sas                                          *
* Proposal:                                                          * 
************************************************************************;
LIBNAME NH "C:\NHANES\DATA";
OPTIONS NODATE NOCENTER;
options ls=72;

proc format;                                       
                                                   
                                                 
  VALUE sexfmt  1 = 'Male'                          
                2 = 'Female'                        
                ;                                  
  

  VALUE racefmt  1 = 'NH White'
                 2 = 'NH Black'
			     3 = 'Mexican American'
			     4 = 'Other'
			      ;
                       
  VALUE agefmt                             
                1 = '20-39'                               
                2 = '40-59'                               
				3 = '60+'       
				;         

 run;   
data ANALYSIS_DATA;
set NH.ANALYSIS_DATA; 

if ridstatr = 2;  ***examined  ;
 
age = .;
if 20 LE ridageyr LE 39 then age=1;
if 40 LE ridageyr LE 59 then age=2;   
if ridageyr GE 60 then age=3;  

race=.;
if ridreth1=3 then race=1;
if ridreth1=4 then race=2; 
if ridreth1=1 then race=3;
if ridreth1=2 or ridreth1=5 then race=4;

LABEL                           
       age  = 'AGE GROUP'    
	   race = 'Race Ethnicity'	
       riagendr = 'Gender' 
          ;                                                       
                                                                                                              
RUN;




proc sort; by riagendr age; 
run;


PROC UNIVARIATE NOPRINT;

where ridageyr >= 20;

by riagendr age;    

VAR lbxtc;         
 
freq wtmec4yr; 

FORMAT age AGEFMT. riagendr SEXFMT.  race RACEFMT. ; 

output out=sasdataset mean=mean q1=p_25 median=median q3=p_75;


proc print data=sasdataset;
title "Distribution of cholesterol: NHANES 1999-2002";           
run;