Task 3a: How to Estimate Mean Nutrient Intakes from Supplements Using SUDAAN

This section describes how to use SUDAAN to estimate mean nutrient intakes from supplements, along with standard errors.  To illustrate this, consumption of supplemental calcium by females ages 20 and older is used as an example.

 

Step 1: Sort Data by Strata and PSU

Before running any SUDAAN procedure, sort the data by strata and PSU, using the PROC SORT procedure.  In the sample code below, DEMOOSTS is the dataset that was previously created for this analysis with the appropriate variables of interest.

 

Step 2: Compute Properly Weighted Estimated Means and Standard Errors

To compute properly weighted estimated means and standard errors, use the PROC DESCRIPT procedure in SUDAAN.  This procedure includes a required nest statement that identifies the variables for strata and PSU. 

In the sample code below, note that the weight variable being used is for the MEC subsample (WTMEC2YR).  If you were only using dietary supplements data, then it might be appropriate to use the interview weight, as supplement use is collected during the household interview.  However, when supplement data are combined with other variables from the MEC exam, then you would use the MEC weight.  In this case, the program below is part of a larger program that uses variables from the MEC exam.

The subgroup statement indicates that the results will be reported by age group (AGEGRP), which has three levels.  DAILYAVG is a variable that was previously created to represent total calcium intake from all calcium-containing supplements (see full program in appendix to see how this variable was created).  The SUBPOPN statement identifies the subset of people that will be included in the analysis; INCOHF20 is a variable name that signifies “in the cohort.”  In this case, these are females ages 20 and older with complete and reliable recall data.

 

Use SUDAAN to Estimate Mean Intake of Calcium, in Milligrams, from Supplements among Females Ages 20 years and Older

Sample Code

*-------------------------------------------------------------------------;
* Use the PROC SORT procedure to sort the data by strata and PSU.         ;
*                                                                         ;
* Use the PROC DESCRIPT procedure in SUDAAN to estimate mean intakes of   ;
* calcium from supplements using complex survey design factors (strata    ;
* and PSU).                                                               ;
*-------------------------------------------------------------------------;

proc sort data =DEMOOSTS;
by SDMVSTRA SDMVPSU;
run ;

proc descript data =DEMOOSTS;
  nest SDMVSTRA SDMVPSU;
  weight WTMEC2YR;
  subgroup AGEGRP;
  levels 3 ;
  var DAILYAVG;
  tables AGEGRP;
  subpopn INCOHF20= 1 ;
  rformat AGEGRP AGEGRP. ;
run

 

Output of Program


Number of observations read    :   9643    Weighted count :286222757     
Number of observations skipped :    479           
(WEIGHT variable nonpositive)                     
Observations in subpopulation  :   2467    Weighted count:106855170  
Denominator degrees of freedom :     15                                 
        
Variance Estimation Method: Taylor Series (WR)     
For Subpopulation: INCOHF20 = 1                                         
by: Variable, Age of subject.     
     
--------------------------------------------------------------------------------------------------
|                 |                  |                                                           | 
| Variable        |                  | Age of subject                                            |
|                 |                  | Total        | 20-39        | 40-59        | >= 60        |  
--------------------------------------------------------------------------------------------------
|                 |                  |              |              |              |              |
| DAILYAVG        | Sample Size      |         2462 |          885 |          679 |          898 |
|                 | Weighted Size    | 106551618.00 |  39975812.51 |  40488299.08 |  26087506.41 |
|                 | Total            | ************ | ************ | ************ | ************ |
|                 | Lower 95% Limit  |              |              |              |              | 
|                 |  Total           | ************ | ************ | ************ | ************ | 
|                 | Upper 95% Limit  |              |              |              |              | 
|                 |  Total           | ************ | ************ | ************ | ************ |   
|                 | Mean             |       251.05 |       135.86 |       251.80 |       426.39 |  
|                 | SE Mean          |        16.89 |        17.04 |        23.64 |        26.58 |
|                 | Lower 95% Limit  |              |              |              |              | 
|                 |  Mean            |       215.05 |        99.55 |       201.41 |       369.75 | 
|                 | Upper 95% Limit  |              |              |              |              |  
|                 |  Mean            |       287.04 |       172.17 |       302.20 |       483.04 | 
--------------------------------------------------------------------------------------------------         
 

 

Highlights from the output include:

  • There are 2,462 females in this analysis, 885 ages 20-39, 679 ages 40-59, and 898 age 60 and older.
  • The Weighted Size is the sum of the weights for the observations used in this analysis, which is the denominator for computing the mean.
  • The Total, which is the numerator for computing the mean, is the weighted sum of all supplemental calcium reported.  Note that, in this case, because these numbers are so large, the default SUDAAN output displays a row of asterisks.  It is possible to display these numbers by using SUDAAN options to control the output format.  These values, as well as their respective confidence limits, are part of the default output and are not generally relevant to these types of analyses when used alone.
  • The supplemental calcium intake increases with age group among females.  Females ages 20-39 reported 136 mg, while those ages 40-59 reported 252 mg, and those ages 60 and older reported 426 mg.  These are estimates of the population mean intake of supplemental calcium on a given day, which is equivalent to the mean usual intake of supplemental calcium for these age groups in the population.
  •  
  • The Lower 95% Limit Mean and the Upper 95% Limit Mean are the bounds of the confidence intervals.  See “Module 16: Test Hypotheses” for more information on confidence intervals.

 

close window icon Close Window to return to module page.