National Health and Nutrition Examination Survey

In NHANES 2009-2010, Arthritis Body Measurements were obtained as a part of a 2-year data collection effort to obtain U.S. prevalence estimates for certain Spondylarthritis (SpA) related conditions. Data collection included a Household Interview questionnaire on Inflammatory Back Pain and SpA related conditions (ARQ_F) and also data for the HLA-B27 biomarker (B27_F_R) (Dillon & Hirsch, 2011). The U.S. nationally representative prevalence of these disorders had not been previously reported (Helmick et al., 2008).

Spinal arthritis or Spondyloarthritis is a common form of chronic arthritis among U.S. adults. It can affect the spine at any level including the neck, the upper back, the mid-back, the low back or the sacroiliac joints. Spondyloarthritis is often associated with physical impairment due to decreased spinal mobility and can lead to early mortality.

The Arthritis Body Measures were designed to assess the degree of a participant's spinal mobility at each of three levels: the neck the mid-back and the low back. These represent spinal mobility in the cervical spine, thoracic spine and the lumbo-sacral spine, respectively. Three physical measures were obtained:

• Occiput-to-Wall Distance (OWD) measurement,
• Chest Expansion (CE) measurement
• Active Lumbar Flexion (LF) measurement (also called the modified Schober’s test).

These three arthritis body measures are standard tests in clinical medicine, but have not been previously used in large scale population based surveys. The aim of the current NHANES 2009-2010 arthritis body measures data collection is to establish U.S. population based reference ranges for these clinical indicators of spinal mobility.

 

The Arthritis Body Measures were obtained on a full sample of male and female participants ages 20 to 69 years.

There were no specific medical, safety, or other exclusions for the Arthritis Body Measurement protocol. Measurements were obtained on all participants who were able to stand unassisted, even if only for a short period of time (e.g. those using wheelchairs). Although current back or spinal pain was not a specific exclusion, the Health Technicians used their discretion to obtain the Arthritis Body Measures when a participant had current spinal or back pain.

 

Data for the three arthritis body measurements were collected by the same trained Health Technicians who performed the NHANES anthropometric examinations: the three measures were included in the 2009-2010 NHANES anthropometry protocol because of their similarity to the body measures already being routinely obtained by NHANES. The detailed protocols for obtaining the three measurements are documented in the NHANES 2009-2010 Anthropometry Procedures Manual .

The Occiput to Wall Distance is a routine clinical test for cervical spine mobility that has been in use for many years. The OWD is measured by having an examinee stand with the back against a wall keeping the posture as straight as possible and with the heels, buttocks and shoulders touching the wall. While looking forward, the examinee also attempts to have the back of the head (the occiput) touch the wall as well. In most normal individuals in this standard position, the occiput will touch the wall and the OWD measurement will be zero. If the occiput does not touch the wall, then the OWD is measured with a goniometer. A value greater than 2 cm. is considered to be abnormal. The OWD may be abnormal in kyphosis (forward curvature of the upper thoracic spine) due to either spondyloarthritis or osteoporosis, as well as in other conditions such as postural instability, congenital spinal deformity and marked obesity.

Chest Expansion measurement is a clinical measure to assess limitation of thoracic spine mobility. It measures the range of motion of joints that connect the posterior ends of the ribs to the thoracic spine. These small joints are usually activated only with very deep breathing. In normal breathing, it is only the muscular diaphragm that varies in position. However with deep inspiration, the posterior joints between the ribs and the thoracic spine become active pivoting the rib cage upward resulting in chest "expansion." This allows additional volume for the lungs to expand in a deep inspiration. In spondyloarthritis, these small posterior joints between the ribs and the spine may become affected by arthritis resulting in reduced chest expansion. Two chest circumference measurements are made: the first at full exhalation when the lungs are "empty" and a second after a maximal or full inspiration. The Chest Expansion measurement is taken as the difference in cm between the measured chest circumference in full inspiration minus the chest circumference in full exhalation.

The Lumbar Flexion measurement (modified Schober's test) is a clinical measure to assess lumbosacral spine mobility. It is performed as follows: with the examinee standing upright, a mark is placed at a point on the skin over the lower lumbosacral spine and a second mark is placed 10 cm above this to provide a baseline for measurement. The examinee then bends forward as much as possible and attempts to touch the toes. The knees are kept straight. With the examinee still bending forward, the distance between the two marks is again measured. Normally, the measured distance between the two marked points will increase due to flexion of the spine when bending forward. A reduction in the measured distance may reflect a reduction in lumbar spine mobility due to stiffness in spinal arthritis or other causes.

 

The 2009-2010 NHANES arthritis data was collected as a part of the ongoing NHANES Body Measures examination, which provides U.S. national-level reference anthropometry data on height, weight, waist and limb circumferences, and skinfolds. All NHANES Mobile Examination Center (MEC) health technicians received professional training in obtaining body measurements, as well as periodic refresher training. Prior to conducting the 2009-2010 Arthritis body measures exam, most health technicians had already had 5-10 years of anthropometry data collection experience in the MEC.

The three NHANES arthritis body measures, like other NHANES anthropometry data, were obtained by a team of two technicians working together, rather than by a single technician. One technician performed the measurement and the second verified that the measurement technique and value were correct and then recorded the data. In the MEC, the Chief Health Technician directly monitored staff performance, which included the arthritis body measures/anthropometry staff. Health technician performance was also monitored during the QC process for data analysis of all incoming data. This occurred on an ongoing basis, and during regularly scheduled site visits by NCHS staff and a professional Anthropometrist directly observing data collection procedures. The NHANES Body Measures Procedures Manual provides more detailed descriptions of measurement techniques and the quality assurance and quality control measures used in the NHANES anthropometry component.

During the 2009-2010 survey, inter-observer reliability statistics for the three arthritis body measures were assessed on a periodic basis using a Gold Standard Examiner (GSE). These compared the health technicians’ measurements to the GSE’s measurements on the same participants. During each exam, the health technicians and the GSE separately performed a single measurement on the participant, so inter-examiner (not intra-examiner) reliability was measured. If needed, a health technician also received retraining at that time to help improve technique. Bland Altman plots for Chest Circumference and Lumbar Flexion measurements are presented in Appendix A. Gold Standard examination summary statistics for the three arthritis body measures including correlation coefficients and Technical Error of the Measurements (Norton & Olds, 2001; WHO Study Group, 2006) are as follows:

Measures	Variable	# Pairs	Difference Mean (cm)	Range(cm)	Pearson's r	Individual ICC	Average ICC	Absolute TEM (cm)	Relative TEM %
Occiput-to-Wall distance (cm)
	ARXO2WD	113	0.01	3.6 (-1.3 to 2.3)	0.99	0.99	0.99	0.19	61.4
Chest Circumferences (cm)
Inspiration	ARXCCIN	108	-0.16	9.7 (-6.4 to 3.3)	0.99	0.99	1	1.07	1.1
Exhalation	ARXCCEX	108	-0.99	9.7 (-6.4 to 3.3)	0.99	0.99	0.99	1.42	1.5
Men
Inspiration	ARXCCIN	47	-0.3	4.7 (-2.2 to 2.5)	1	1	1	0.69	0.7
Exhalation	ARXCCEX	47	-0.79	5.3 (-3.3 to 2.0)	0.99	0.99	1	1.01	1
Women
Inspiration	ARXCCIN	61	-0.05	9.7 (-6.4 to 3.3)	0.98	0.98	0.99	1.29	1.4
Exhalation	ARXCCEX	61	-1.14	9.7 (-6.4 to 3.3)	0.98	0.98	0.99	1.67	1.9
Lumbar Flexion Test length (cm)
	ARXLFTL	115	-0.22	3.5 (-2.1 to 1.4)	0.84	0.83	0.91	0.51	3.6

Notes: The Difference Mean is the mean of health technician measured values in centimeters minus the Gold Standard Examiner measured values in centimeters. ARXLFTL is the original data collection variable.

Abbreviations: ICC= Intra-class Correlation Coefficient to test inter-examiner reliability; TEM= Technical Error of the Measurement.

For the Occiput-to-Wall distance measurement, 95% of sample measurements were zero. All the Occiput-to-Wall distance inter-rater measurement pairs were concordant: i.e. there were no discordant pairs such that if the Occiput-to-Wall distance for one examiner=0 then the Occiput-to-Wall distance for the paired examiner was > 0.

The 2009-2010 Arthritis Body Measures data were reviewed for unusual and erroneous values. Range checks were employed including flagging values for review that were above the 99th percentile or below the 1st percentile. A very few values that were determined to be data recording errors were corrected in this file. None of the other original body measures data were changed and there are no imputed values in this file.

Derived Variables:
1. A final Arthritis Body Measures component status code variable (ARDEXSTS) provides analysts with a quick method of identifying survey participants with complete or partial arthritis body measurement data.

2. For the Chest Expansion measurements, the variable ARDDINEX provides the difference in cm between the chest circumference measurement in full inspiration and the chest circumference measurement in full exhalation. This is the difference value that is typically evaluated by clinicians.

3. For the Lumbar Flexion measurement, a single derived variable ARDLFTL is provided. This represents the difference in cm between the lumbar flexion measurement taken with the examinee bending forward in full lumbar flexion minus the 10 cm baseline measurement. For the Lumbar Flexion Test, ARDLFTL is the difference value that is typically evaluated by clinicians.

 

Each of the three NHANES Arthritis Body Measures were clinical examination tests that had been in long term use in Medicine, primarily in the field of arthritis evaluation and treatment. However they had not been previously used in large scale population based surveys. The intent of the NHANES survey was to field the three tests in their original clinical forms, however for the Chest Expansion test this was not practical. In particular, the originally specified protocol for the Chest Expansion test obtains chest circumference measurements at the level of the 2nd intercostal space. In clinical settings, the measurement is primarily obtained in males as opposed to females. Initial NHANES feasibility tests showed that performing chest circumference measurements at the level of the 2nd intercostal space was unreliable or in fact not feasible due to breast tissue for a large percentage of women that were examined. The population based survey setting required a single chest circumference measurement protocol that would provide valid comparison data between genders. A lower level on the chest was needed to obtain valid and comparable chest circumference measurements in all examinees.

The NHANES 2009-2010 survey obtained chest circumference measurements at the level of the xiphoid notch. However even at this lower level, chest circumference measurement was often difficult to obtain, especially among women. In such instances, the NHANES protocol was to take the chest circumference measurement at the next lowest level on the chest where a circumference measurement could be practically obtained. For these instances, ARXXDIST is the variable that documents the distance between this level and the level of the xiphoid notch. ARXXDIST was 1 cm or greater for over 20% of women examined. In the NHANES protocol, all chest circumference measurements were obtained by a team of two Health Technicians: one Health Technician to obtain the primary measurement and a second Health Technician to verify that the measuring tape was positioned correctly and properly aligned horizontal to the exam room floor and maintained at the same position on the chest as the particpant went from full exhalation to full inspiration. In NHANES 2009-2010, a metal measuring tape was used instead of a cloth measuring tape, as is the more common practice in clincial settings (Bockenhauer 2007).

In clinical medicine, it is considered that the normal chest expansion between complete exhalation to complete inspiration ranges from 3 to 7 centimeters. A value less than 2.5 centimeters is often considered abnormal (Weiss et al., 2010). Reduced chest expansion is often seen in spinal arthritis such as SpA, but can also be seen in lung disease such as chronic obstructive pulmonary disease (COPD), in chest wall deformities, or with chest wall pain. It can also be seen in smaller individuals or it can simply be the result of a poor effort in exhalation or inspiration. Chest Expansion measurements are therefore effort dependent and measurement validity depends on the participant's efforts. The NHANES Health Technicians spent considerable time coaching participants to make maximal inspirations and exhalations. They were well versed in this because they were also administering Spirometry tests to this same set of survey participants.

In Spondyloarthritis cases, an abnormality in the Occiput to Wall Distance test most likely reflects a more advanced stage of disease. For example, the OWD is correlated with positive spinal radiographic findings in the neck in Ankylosing Spondylitis, which is the prototype disorder for SpA. The OWD test also has relatively high interobserver reliability (Viitanen, et al. 1995, 2000). In the NHANES 2009-2010 data collection, the great majority of examinees had a normal OWD (i.e they were able to touch the back of their head to the measurement stadiometer when standing straight) and only about 5% of examinees had an OWD measurerment greater than zero. NHANES 2009-2010 did not specifically examine participants for kyphosis (forward curvature of the upper spine), however in the general NHANES Body Measures data collection (BMX_F), the variable BMIHT is available which has a code indicating that the examinee's spine was not straight (i.e the examinee had either kyphosis or scoliosis). This variable can be used to help interpret the OWD test results.

The Lumbar Flexion measurement obtained in NHANES 2009-2010 is similar to standard clinical protocols. There is variation among different published studies as to where best to place the lower boundary mark for the baseline LF test measurement and whether the baseline measurement distance should be 10 or 15 cm. In the original description of the LF test, a 10 cm. distance upward beginning from the lumbosacral junction was specified (Schober, 1937). Subsequently a 15 cm baseline distance was specified measured from 5 cm below to 10 cm above the lumbosacral junction (Macrae and Wright, 1969; Moll and Wright, 1972). This is the approximate level of the superior margins of the sacroiliac joints in the pelvis. More recently a 15 cm baseline measurement was specified that begins with marking a point midway along a line level with the iliac crests (a line thought to mark the the L4/5 junction), then marking points 5 cm below and 10 cm above this (Gladman et al. 2007; Jenkinson et al. 1994). Other research indicates this line is at the level of the L4 spinous process or the L4–5 interspinous spaces (Chakraverty et al., 2007).

The NHANES LF was performed similarly to the above LF tests except that the initial point for the baseline LF distance was marked midway along a line level with the superior margin of the lateral iliac crests and a then second mark was placed 10 cm above it. The superior margin of the lateral iliac crests is the same landmark used in the NHANES Anthropometry exam to measure waist circumference. This landmark was selected for reliability as it is easily palpable, because the Health Technicians already had considerable experience locating it, and because of participant privacy concerns of using a lower landmark in the population based survey. The measuring tape was always maintained directly against the skin when measuring and marking the 10 cm baseline LF distance. For example in examinees with marked lumbar lordosis, the 10 cm measurement followed the course of the inward curve of the lordosis. The one exception to this was in markedly overweight examinees who had large fat folds.

Sample Weights: The NHANES examination sample weights should be used to analyze the arthritis body measurement data. Please refer to the NHANES Analytic Guidelines and the on-line NHANES Tutorial for further details on the use of sample weights and other analytic issues.