Key Concepts About Subsamples and Weights Associated with Environmental Chemical Data

For a various reasons, only some, but not all, participants are selected for environmental chemical measures. For example:

Although the majority of environmental chemicals are measured for subsamples, some are measured for the entire examined sample (within a targeted age or other demographic group). In these cases the corresponding MEC weight should be used in the analysis.

EXAMPLE:  Blood lead concentration in NHANES 1999–2000 participants aged 1 year and older
1999–2000 Lab 06 Nutritional Biochemistries documentation [PDF - 114 KB]
           

Subsampling strategy for environmental chemical measures

The subsamples selected for many environmental chemical components are chosen at random with a specified sampling fraction (for example, 1/2 or 1/3 of the total examined group) according to the protocol for that component. Each component subsample has its own designated weight, which accounts for the additional probability of selection into the subsample component, as well as the additional nonresponse.

Environmental chemical subsamples are most often 1/3 sample of a specified age group of the examined sample. In this case, the corresponding 1/3 subsample weight should be used in the analysis.

EXAMPLE:; Perchlorate concentrations are measured in the urine of a one-third subsample of participants ages 6 years and older in NHANES 2003–2004.
2003–2004 Lab 04 Urinary Perchlorate documentation [PDF - 3.3 MB]

EXAMPLE: Volatile organic compounds concentrations are measured in the blood of a one-half subsample of participants ages 20 to 59 years in NHANES 2003–2004. Lab 04 Volatile Organic Compounds in Blood and Water documentation [PDF - 80 KB]

 

Considerations if conducting an environmental analysis when variables originate from different subsamples and thus have different subsample weighting.

On occasion, an analyst may be interested in conducting an analysis that involves environmental chemical measures obtained in a subsample (such as a 1/3 subsample) AND measures obtained in another subsample (such as the morning fasting subsample). The reweighting of the sample weights in this situation is a special case that requires advanced statistical expertise. It is not a simple task and is beyond the scope of this tutorial.  Inexperienced analysts should seek appropriate statistical consultation in this case.

 

Pooled specimens (rarely used)

In order to have enough specimen volume to measure very low concentrations, some environmental chemical concentrations have been measured in specified demographic groups after pooling biologic specimens within specified age-race-gender examined subsamples with surplus serum or urine. Analysis of these pooled specimens will provide estimates for the age-race-gender categories sampled. These estimates are not representative of the United States population. For these analysis sample weights are not employed.

EXAMPLE:  Polyfluoroalkyl chemicals (PFCs) in pooled serum samples collected from participants 3-11 years old in NHANES 2001–2002.
Polyfluoroalkyl Chemicals (Pooled Samples) documentation [PDF - 116 KB]

 

NHANES III convenience sample for environmental chemical measures)

Some environmental chemical concentrations were measured in a convenience sample of NHANES III participants.  Analysis based on this convenience sample will not produce nationally representative estimates for these environmental chemicals. No sample weights are available for use with these laboratory results.

EXAMPLE Dialkylphosphates in surplus sera collected from participants > 18 years old in NHANES 1988–1994.
32A. Surplus Sera Laboratory Component: Dialkylphosphates documentation[PDF - 282 KB]

 

close window icon Close Window