Key Concepts about Identifying, Recoding, and Evaluating Missing Data
NHANES study participants were carefully screened by questionnaire and measurements of resting heart rate and resting blood pressure before they performing the treadmill test. High-risk (e.g., symptomatic or reporting selected chronic conditions or diseases) individuals were excluded so that only apparently healthy individuals underwent treadmill testing. A complete list of reasons for possible exclusion can be found in the Cardiovascular Fitness Procedures Manual.1
Approximately 70% of the eligible NHANES participants performed a treadmill test during the two survey cycles between 1999 and 2002, and exclusion rates increased with age. In the 2003-2004 dataset, women had a much higher exclusion rate than men due to pregnancy (23% of the exclusions were among women; pregnant women are one of the subpopulations oversampled in the NHANES 2003-2004). Data users need to be aware of the exclusion rate and be cautious when making inferences from the analyses as the exclusions may affect the representativeness of the sample.
For this tutorial, we assess only those study participants with an estimated VO2 max. For your analysis, you may wish to further describe the population of study participants who were eligible for the exam but were excluded.