Component Description
Balance disorders constitute a major public health concern, especially for older persons, who are an increasing segment of the U.S. general population. For example, balance disorders are thought to be causally related to some 200,000 hip fractures that occur annually in Americans over the age of 65. The objectives of the Balance Component are
- to obtain accurate prevalence data on balance problems among a sample of U.S. adults;
- to examine the relationship between balance problems and specific covariates such as certain medical conditions and indicators of health status; and
- to help characterize normal and disordered balance and spatial perception.
The NHANES 2003–2004 Balance Component consists of 2 parts:
- A pre-exam screening questionnaire. This is a series of questions that identify conditions which would constitute either practical or safety exclusions to performing Balance testing. Please note that more extensive survey questions relating to balance problems are contained in the BAQ section of the NHANES household survey questionnaire; and
- Balance Testing, consisting of a modified Romberg Test of Standing Balance on Firm and Complaint Support Surfaces.
Eligible Sample
The 2003–2004 Balance Component tested a full sample of U.S. adults 40 years of age and older. There were five principal exclusion criteria applied in the Balance testing protocol. These were:
- at the time of the exam, an examinee felt unable to stand on his or her own;
- at the time of the exam, an examinee was currently having dizziness sufficient to cause unsteadiness;
- weight over 275 lbs;
- an examinee needed a leg brace in order to stand unassisted;
- an examinee had an amputation of their feet or legs.
In addition, subjects who were totally blind or visually impaired sufficiently to require assistance in finding the exam room were disqualified, as were those who had a waist circumference either too small or too large to accommodate proper fitting of any standard sized safety gait belt. Also, the Mobile Examination Center (MEC) physician could exclude subjects if it was considered medically contraindicated to perform testing.
Protocol and Procedure
The Romberg Test of Standing Balance on Firm and Compliant Support Surfaces examines the ability of the examinee to stand unassisted under four test conditions, ordered in increasing level of difficulty.
In Test Condition 1, the subject stands making use of all the sensory inputs that contribute to balance: the central vestibular system, vision, and proprioception (leg muscle position sense).
Test Condition 2 tests balance when only vestibular and proprioceptive information is available (the subject closes the eyes to eliminate visual input).
In Test Condition 3, the subject must maintain balance on a foam-padded surface, which reduces the proprioceptive input, leaving only visual and vestibular cues.
Finally, in Test Condition 4 (also using a foam pad), the eyes are again closed (visual input is removed) and the subject's ability to maintain balance is tested using only his vestibular system.
All Balance Testing was scored on a pass/fail basis, with Test Conditions 1 and 2 conducted for a total of 15 seconds, and Test Conditions 3 and 4 conducted for 30 seconds each. Test failure was defined as:
- a subject needed to open the eyes in an eyes-closed Test Condition (#2 or #4);
- the subject moved the arms or feet in order to achieve stability; or
- the subject began to fall or required MEC Technician intervention to maintain balance.
In the standard Balance Testing protocol, each subject is eligible for an initial test and one re-test to pass a specific Test Condition. Within each Test Condition, the protocol for retesting is the same as for the primary examination. Because each successive Test Condition from 1 to 4 is progressively more difficult than the Condition preceding it, the Balance Testing Component is ended whenever a subject fails to pass a Test Condition.
Balance testing was conducted with the utmost consideration for subject safety. The examining MEC Technician stood immediately to the side of the participant prepared to stabilize the subject by the use of a safety gait belt should significant unsteadiness occur. Also, the examination was conducted in a small exam room, with well-padded walls and floor.
For the procedures relevant to this component, please refer to the Survey Operations Manual, Consent Documents, Brochures on the NHANES website.
Quality Assurance & Quality Control
All data from each of the sections of the Balance Component were directly entered by MEC technicians into ISIS, a computerized database information system. Technicians selected for the Mobile Examination Centers were professionally trained in the Modified Romberg Testing protocol by an expert consultant in Balance and Vestibular disorders. In addition, Audiologic Consultants from the National Institute of Occupational Safety & Health (NIOSH) and contractor staff provided performance monitoring of each Technician on a regular, periodic basis. Additionally, NCHS staff visited the MECs approximately twice per year to observe the Balance examinations and verify that standard testing procedures were being strictly followed by technicians. All MEC Balance technicians also received annual retraining and updates by the NIOSH consultants. Finally, exclusion rates and outcome statistics were continually monitored by NCHS statisticians according to technician and MEC, and retraining was conducted as necessary to maintain test quality.
For details on the QA/QC process for this component, please refer to the Interviewer Procedure Manuals on the NHANES website.
Data Processing and Editing
As mentioned previously, all data were captured in ISIS, a computerized database. NCHS statisticians performed data reviews on an ongoing basis using a computerized data editing program to check for logical inconsistencies in the data and technician errors, as well as to cross-check other issues affecting data quality. Back-end edits of the data were performed as necessary where errors were detected. The 2003–2004 Balance Examination data were re-reviewed for errors and logical consistency and verified against the primary examination data database prior to public release.
Analytic Notes
Because the NHANES 2003–2004 Balance examination was administered in the MEC, the full sample MEC examination sample weight should be used for data analysis. For information on the use of sample weights in NHANES data analysis, refer to the NHANES 1999–2000 Addendum to the NHANES III Analytic Guidelines. Analysts who wish to combine Balance data from NHANES 1999–2000 with the NHANES 2001–2002 or 2003–2004 Balance examination data should be aware that the sampling frame for NHANES 1999–2000 Balance Examination data differed from that for the 2001–2004 Balance data. Specifically, NHANES 1999–2000 Balance data were collected for a one-half sample of U.S. adults ages 40–69 years, whereas NHANES 2001–2004 Balance was collected for a full sample of U.S. adults ages 40 and older. Special half-sample weights are available on the NHANES 1999–2000 Balance data set and are to be used when analyzing the data for those years. For combined 1999–2000 and 2001–2002 data analysis, special 4-year Balance subsample weights are provided (WTSBA4YR) on the 2001-2002 Balance dataset. Six-year data analysis (1999–2004) is performed by combining the 4-year Balance subsample weights with the 2003–2004 MEC examination weights according to the NHANES Analytic Guidelines.
Data analysts should ideally be familiar with Balance and Vestibular disorders, or have a specialist available for consultation. The systematic differences in testing length (15 seconds for Conditions 1 and 2; 30 seconds for Conditions 3 and 4) may need to be taken into account for the analysis. The variable BAAEXCMT provides coded data summarizing the reasons that an examination was incomplete (partial) or not done.
Please refer to the Analytic Guidelines for further details on the use of sample weights and other analytic issues at https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx.
No data related to this component are in the Research Data Center.