The National Health and Nutrition Examination Survey (NHANES) conducts interviews and health examinations on a nationally representative sample of approximately 5,000 people each year. During each 2-year cycle through 2017-2018, NHANES visited 15 locations each year. No information about geographic location was released to protect the identification of NHANES participants. State and county information could be obtained through the NCHS Research Data Center (RDC). It is desirable, however, to use a finer level of geography to spatially analyze the NHANES data and to answer various important research questions about the effect of geography on health.
Therefore, National Center for Health Statistics (NCHS) entered into an agreement with U.S. Department of Housing and Urban Development (HUD) to use its geocoding services to assign geographic codes to the NHANES address data. HUD geocoded the 1999-2018 NHANES data to the 2010 census.
All survey participants have a geocoded record.
Addresses were collected for each household with at least one survey participant by trained interviewers using the Computer-Assisted Personal Interview (CAPI) system in the participant’s home or by telephone.
After geocoding the NHANES data, HUD provides QC documentation (see Appendix 1).
The NHANES participant address data were submitted to HUD and, geo-enabled with latitude/longitude coordinates and other legal, statistical, and administrative geographies including census, postal, and other attributes, using the HUD Geocode Service Center (GSC) system.
Due to disclosure concerns, this dataset is not available publicly. Access may be obtained through the NCHS Research Data Center.
Only 30 locations are included in each 2-year NHANES cycle, analysts are urged to combine cycles to achieve greater geographic coverage and to assess if additional cycles are needed to achieve adequate statistical power. Generally, obtaining stable estimates for rare outcomes is not possible for sub-national geographic areas.
Please note the Census year for the geocoded file. Geographic linkage is done using the most recent decennial Census data available. Although most analyses are likely to use the most recent data available, some analysts may choose geocoded data linked to older Census data. The documentation accompanying the file should be carefully reviewed because Census variable names differ between geographic linkage file versions.
| RC2KX | Frequency | Percent | Cumulative Frequency | Cumulative Percent |
|---|---|---|---|---|
| 35 | 0.03 | 35 | 0.03 | |
| 5 | 2718 | 2.68 | 2753 | 2.72 |
| 9 | 3056 | 3.02 | 5809 | 5.73 |
| S | 95507 | 94.27 | 101316 | 100 |
| LVL2KX | Frequency | Percent | Cumulative Frequency | Cumulative Percent |
|---|---|---|---|---|
| 35 | 0.03 | 35 | 0.03 | |
| 4 | 1400 | 1.38 | 1435 | 1.42 |
| B | 1656 | 1.63 | 3091 | 3.05 |
| R | 95507 | 94.27 | 98598 | 97.32 |
| T | 2718 | 2.68 | 101316 | 100 |