A guide to interpreting the low-back pain (LBP) study poster

A randomized controlled trial of Neuro Emotional Technique (NET) for low back pain.

Peter BABLIS/Henry POLLARD/Rod BONELLO

• Department of Chiropractic, Faculty of Science, Macquarie University, Ryde, NSW Australia
• Completed in partial fulfilment of the degree PhD in the Dept. of Chiropractic, Macquarie University
• Dept. of Chiropractic, Macquarie University, Sydney, Australia
• Dept of Medicine, University of Notre Dame, Sydney, Australia

This project was funded by a grant from Foot Levelers Inc., USA, through The ONE Research Foundation.

Dr. Peter Bablis (Chiropractor) pb@universalhealth.com.au

Low-back pain is one of the most common afflictions of western society.  It is the most common presentation to chiropractors, and chronic LBP is the most common form of LBP treated by chiropractors.

Chronic LBP is associated with:

• pain and disability (Heymans et al 2010, Becker et al 2007, Reuben et al 2002)
• inflammation (Kallewaard et al 2010)
• discs (Ludeke et al 2010, Li et al 2010, Cuellar et al 2010) and facet joints, and associated inflammation (van Kleef et al 2010, Igarushi et al 2004).

The outcomes of chronic LBP are not predicted by physical characteristics (van Nieuwenhuyse et al 2009, Christiansen et al 2010), but by psychosocial factors (Melloh et al 2009, Du Bois et al 2009).

NET is a stress-relieving technique, using exposure therapy to affect psychosocial variables associated with many pain states.  It is considered a mind/body technique.

This study utilised the latest in blood- and paper-based assessments to demonstrate large improvements in mind and body measures of pain, anxiety, depression, and indicators of inflammation in chronic LBP sufferers using a new stress reduction technique called Neuro Emotional Technique (NET).  In our search of the literature we have not been able to discover any therapy study that has better outcome assessments of inflammation in chronic low-back pain patients than this study.

The paper was presented at the recent meeting of the World Federation of Chiropractic in Rio de Janeiro, Brazil.  It will shortly become a manuscript for submission to a high-quality, peer-reviewed medical or chiropractic journal.

To our knowledge, this study represents the first therapy study to look at all of the paper-based and the blood-based outcomes at the same time, which, in our opinion, makes it the most thorough study of its kind.

It is a randomized control trial (RCT). It is a medium-sized study that has good statistical power (i.e., the statistics can be considered a reliable indication of the true outcome).  The patients were “blinded” or unaware of which treatment they were receiving. This is important so that a patient’s expectation (or bias) of what treatment they are receiving does not alter the outcome of the study.  The practitioners (one each for the treatment and the control groups) were not hidden to the true allocation (blinded) of their intervention.  But, in order to minimize the effect of practitioner outcome (bias) on the treatment being rendered, the practitioner did not record any of the outcomes. A research assistant performed this function without having knowledge of which group each patient was in.

Bias in research is an important problem.  The best type of research occurs when the patient, the practitioner, and the assessor are all blinded to the intervention, as each has the potential to influence the results (whether they know it or not).  In a therapy study it is impossible for the practitioner to be blinded.  So, we instead blind the assistant doing the assessments.  It is the next best thing.  Many RCTs do not do this and are therefore listed as being single-blinded only.

The trial is the first to report strong improvements (P<0.001) in both subjective paper-based and objective blood-based outcomes:

1. McGill Pain Questionnaire (MPQ), which measures pain levels and is a reliable and valid tool for use in research into pain.

2. Oswestry Disability Index (ODI) which measures low-back related disability. It is a widely used, valid, and reliable measure of low-back pain related disability.

3. The Distress and Risk Assessment Method (DRAM) to measure anxiety and depression in low-back pain patients. It is a valid and reliable measure of anxiety and depression in various populations. It is made up of two measures: the modified ZUNG Depression Index and the Modified Somatic Perception Questionnaire. The questionnaire attempts to predict which patients with psychological distress would require multidisciplinary psycho-social treatment in addition to physical-modality treatment.

4. The SF-36 measures general health and depression levels in various populations. It is commonly used for these purposes in health research and is considered reliable and valid.  The SF-36 consists of eight scaled scores, which are the weighted sums of the questions in their section.  Each scale is directly transformed into a 0-100 scale on the assumption that each question carries equal weight.  The eight sections of the SF-36 are:

• vitality
• physical functioning
• bodily pain
• general health perceptions
• physical role functioning
• emotional role functioning
• social role functioning
• mental health

Pro-inflammatory markers:

• TNF alpha
• Interleukin-1
• Interleukin-6
• C-reactive protein (CRP)

Anti-inflammatory marker:

Interleukin-10

Tumor necrosis factor (TNBF) is a cytokine involved in the acute inflammatory phase.  The primary role of TNF is in the regulation of immune cells.

A protein produced by various immune cells, including macrophages, interleukin-1 (IL-1) raises body temperature, spurs the production of interferon, and stimulates growth of disease-fighting cells, among other functions.  Interleukin-1 possesses a wide spectrum of metabolic and physiological activities and plays a central role in the regulation of the immune responses.

Interleukin 6 (IL-6) is one of the most important mediators of fever and of acute inflammation.  It is capable of crossing the blood/brain barrier and initiating synthesis of prostaglandin E2 in the hypothalamus, thereby changing the body’s temperature set-point upwards.

C-reactive protein (CRP) is a protein found in the blood.  Its levels rise in response to inflammation.  CRP is a member of the class of acute-phase proteins that rise dramatically during the inflammatory processes occurring in the body.  The rise in the CRP is related to the rise in the blood plasma concentration of IL-6.

Interleukin-10 (IL-10), also known as human cytokine synthesis inhibitory factor (CSIF), is an anti-inflammatory cytokine.  It is capable of inhibiting the synthesis of pro-inflammatory cytokines.  Its release reduces the level of inflammation in the body.

Whilst it is possible that patients may have altered or biased their responses in paper-based assessments that ask questions and require interpretive answers, it is not possible for patients to change their blood chemistry at will. Utilising the blood-based measures significantly strengthens the findings of this study, by removing the possibility of unintended patient effects (bias) altering the outcomes of the study. The fact that the subjective and objective findings were consistent across all the tests also indicates a good result.

The following paragraphs will outline a brief explanation of the statistics used in this report.

A result is called statistically significant if it is unlikely to have occurred by chance.  The significance level is usually denoted by the Greek symbol a (lowercase alpha).  The p-value is the probability of obtaining a statistic by chance compared with the one that was observed in the research.  When the p-value is: p < 0.05 or p < 0.01 or p < 0.001, this corresponds to a relative chance of obtaining a significant result of 5% or 1% or 0.1% respectively.

In this study, we found very significant findings at the “p = 0.001” level for almost all of the variables measured.  We found no significant changes at the “p = 0.05” level in the control group.  These findings indicate that treatment significantly improved the low-back pain but the sham treatment did not improve the pain status in the control group.  (A sham treatment is a treatment that looks exactly like the treatment intervention but one that has no biological effect).

If we, in our paper, argue that “there is only one chance in a thousand this could have happened by coincidence,” we are implying a p = 0.001 level of statistical significance.  Therefore, the lower the significance level, the stronger the evidence required.  Our values were significant at the p = 0.001 level for most of our outcomes.  By contrast the values in the control group were p > 0.05 or even p > 0.5 in some cases.  This means that there is a greater than 5% or even 50% probability of the result happening by chance.

Randomization is the process of making something truly random so that there are no ordering effects in the sample being tested.  Also, it means that patients in a study are not subject to any form of “stacking,” thereby reducing bias in how the subjects were recruited, selected, or allocated to groups to alter the possible outcomes.

An intention to treat (ITT) analysis (sometimes also called intent to treat) is a requirement of high-quality research.  It is an analysis based on the initial treatment intent, not on the treatment that is eventually administered.  ITT analysis is intended to avoid various misleading artifacts that can arise in clinical research.  For example, if people who have a more serious problem tend to drop out at a higher rate, then even a completely ineffective treatment may appear to be providing benefits.  This would occur if one merely compares the condition before and after the treatment for only those who finish the treatment (ignoring those who were enrolled originally, but have since been excluded or dropped out).  For the purposes of ITT analysis, everyone who begins the treatment is considered to be part of the trial, whether he or she finishes it or not.  Such an analysis creates a higher standard of study by eliminating non-treatment effects.  It is expected in good-quality trials and was performed in the LBP study.

The numbers of participants who fell out of the normal range for each blood marker were analysed. We used a threshold analysis of normal pathological ranges because this is the criterion that is adhered to when diagnosis is given in general practice. These were analysed using standard chi-squared statistical analysis, taking into account the attrition rate percentages of both groups. Thus all randomised participants were analysed, including dropouts, as such eliminating any intention-to-treat issues.

The next section will describe the results obtained with the NET treatment.  In this study we compared the treatment groups (who received true NET) with a control group (who received a “sham” NET treatment).

NET reduced pain and disability associated with chronic low-back pain. After six months, the treatment produced significant reductions in:

Pain (MPQ) p<0.001
Disability (ODQ) p<0.001
Anxiety (DRAM- MPSQ) p<0.001
Depression (DRAM-Zung) p<0.001
Health (SF-36)

• vitality     p<0.001
• physical functioning     p<0.001
• bodily pain     p<0.001
• general health perceptions     p<0.001
• physical role functioning     p<0.001
• emotional role functioning     p<0.001
• social role functioning     p<0.001
• mental health     p<0.001

1. TNF alpha p<0.001
2. Interleukin-1 p<0.01
3. Interleukin-6 p<0.001
4. C-reactive protein (CRP) p<0.001

1. Interleukin-10 p<0.004

For the first time, a stress-relieving technique was able to demonstrate universal changes in all measures of pain, disability, emotion, and health in both subjective paper-based outcomes as well as objective blood-based assessments of inflammation in a group of chronic LBP sufferers.

NET was effective in normalising the inflammatory and anti-inflammatory markers associated with chronic LBP.  Based on this protocol, NET is effective in the short  and medium term (one, three, and six months) for reducing chronic LBP as measured by the MPQ, ODQ, DRAM, SF-36, as well as IL-1,IL-6, IL-10, TNF-alpha, and CRP.

The changes in the blood markers noted in this study are significant because their reduction may be associated with a reduced risk of serious illnesses, especially cardiovascular disease risk (Aubert 2008).

Recent research suggests that patients with elevated levels of CRP are at an increased risk of diabetes, hypertension, and cardiovascular disease.  Our research significantly reduced the levels of CRP with NET treatment, but not with sham NET treatment.

The role of inflammation in cancer is not well known.  Some organs of the body show greater risk of cancer when they are chronically inflamed.  Some studies have shown that anti-inflammatory drugs could lower colon cancer risk.  Therefore any treatment capable of producing similar reductions in CRP may be able to reduce the risk also.  Only further study will determine this definitively.

Further study of various disease states (diabetes, chronic bowel disease, cancer, and cardiovascular disease) with NET intervention is encouraged. Of particular importance are studies that investigate the long-term (greater than one year) effect of NET therapy, as well as studies that investigate at-risk populations of diseases such as chronic cardiovascular, bowel, cancer, and diabetic disease who receive NET intervention.

• There is significant association between anxiety and chronic heart disease risk as measured by CRP (Bankier et al 2009)
• Anxiety is associated with inflammatory disease and CRP is associated with the increase in inflammation (O’Donovan et al 2010)
• Inflammation (IL-6) might link post traumatic stress disorder (PTSD) to cardiovascular disease (CVD) risk using CRP as the indicator of increased risk (von Kanel et al 2010).

• Increases in psychosocial variable intensity are associated with increased expression of inflammation cytokines (IL-1, IL-6, and TNF alpha) (Amati et al 2010, O’Connor et al 2009)
• Treatment of psychosocial variables is associated with a decrease in inflammatory cytokines (particularly IL-6) (Miller et al 2005, 2009).

• Psychosocial factors are known to affect recovery from acute low-back pain.  They consist of thoughts, emotions, behaviors, social context, and interactions (Fritz et al 2002)
• Psychosocial variables are associated with increased inflammatory blood markers (van Kleef et al 2010, Maletic & Raison 2009, Maes et al 2009).
• Increased psychosocial variables are associated with increased stress (Dixon et al 2009, Pace et al 2006, Glaser et al 2003, Kiecolt-Glaser et al 2003).
• Increased stress elevates CRP (Ramond et al 2010, McDade et al 2006, Miller et al 2009, Miller et al 2005)
• Early childhood stress is associated with an increase in CRP (Danese et al 2008)
• Increased level of depression is associated with increased CRP (Pikhart et al 2009)
• Psychosocial variables are associated with strong emotional states (Aubert 2008)

Inflammation is part of the complex biological response of vascular tissues to harmful stimuli.  Inflammation is a protective attempt by the body to remove the injurious stimuli and to initiate the healing process.  Without inflammation, injuries would never heal.  However, chronic inflammation can also lead to a host of diseases, including cancer.  It is for that reason that inflammation is normally closely regulated by the body.

Inflammation can be classified as either acute or chronic.  Acute inflammation is the body’s initial response to a harmful stimulus.  A cascade of biochemical events propagates and matures the inflammatory response.  Prolonged inflammation (by definition a state that lasts longer than three months), known as chronic inflammation, leads to a progressive shift in the type of cells present at the site of inflammation.  It is characterized by simultaneous destruction and healing of the tissue caused by the chronic inflammatory process.

The inflammatory response must be actively terminated when no longer needed, to prevent unnecessary “bystander” damage to tissues.  This occurs with great damage in conditions such as rheumatoid arthritis.  Failure to do so results in chronic inflammation, and cellular destruction.  Resolution of inflammation occurs by different mechanisms in different tissues.

For those who would like a description of the statistics used in this study the following will be of interest.  There were two main statistics used for the different components of the study (the subjective and objective measures).  The subjective paper-based and objective blood-based tests were investigated with the REML method (similar to an ANOVA with correction for unequal groups).  The objective blood tests used a chi-square analysis for categorical data (there were two categories: normal and abnormal range values of the various outcome measures).

The REML analysis is identical to the ANOVA analysis when the assumptions for an ANOVA hold, however the REML method allows many more scenarios to be analyzed.  An ANOVA is an analysis of the variation present in an experiment.  It is a test of the hypothesis that the variation in an experiment is no greater than that due to normal variation of individuals’ characteristics and error in their measurement.  (Date accessed 05/15/11: http://www.tfrec.wsu.edu/ANOVA/basic.html).

With the advent of modern computers it is now possible to model the statistics more meaningfully.  The REML algorithm allows that to occur, but it also allows the variance to change over time, which (generally) no ANOVA-based test will do.  Note that our study was conducted over six months, so the assumption that nothing changes over time is inappropriate.  This is why the REML method was chosen.

Variance is used as a measure of how far a set of numbers are spread out from each other.  It is used to describe how far a (treatment) value falls from the mean (treatment) value of the group.  Take the extreme example of a control group, and a group under a 100% successful treatment (for pain relief, say).  The variance of the pain levels for the control group at the start should be the same as at the finish.  The variance of the treatment group should be like that for the control at the start, but collapse to 0 (no pain across all subjects) at the finish.  Of course this extreme is unlikely to be reflected in practice, but a REML algorithm allows the separate variances to be estimated.

In our study we measured outcomes at baseline, and at one, three, and six months.  Therefore there were baseline and three time points in this study.  When there are only two time points, the split-plot ANOVA is equivalent to the REML (because with two time points there is only a single correlation).  Only in a long-term trial will a REML really improve the precision of the statistic.  However, with three time points the REML accounts for the changing variance with time.

Chi-square is a statistical test commonly used to compare observed data with data we would expect to obtain according to a specific hypothesis.  So here we measure what is expected versus what is observed.  What we need to know is whether the deviations (differences between observed and expected) were the result of chance, or were they due to treatment.  In the LBP study we tested whether the blood tests were “normal” or “abnormal.”

The World Federation of Chiropractic has its biennial conference at which chiropractic researchers from all over the world come together to present the latest research in chiropractic related fields.  We chose to present the findings of our low-back pain study at the WFC for its obvious relevance to chiropractors.  The biennial congress was held April 7-9, 2011, at the Rio Intercontinental Hotel, Rio de Janeiro, Brazil.  Approximately 120 papers were presented: 32 platform and approximately 90 poster.  Note: platform presentations are deemed to be of higher value than poster presentations.

1.    Basic Sciences, including the use of animals and computer-related research or investigation on cadavers, in fields such as anatomy, physiology, biomechanics, biochemistry, immunology, etc.
2.    Diagnostic Sciences, including the evaluation of various diagnostic or analytical methods or instruments.  When new concepts are presented, there should be accompanying data collection on normal and abnormal populations of patients.  Clinical examination techniques such as palpation or diagnostic imaging and inter-examiner reliability studies are encouraged.
3.    Clinical Management, including clinical trials, retrospective studies, and multiple case studies.  The presentation of new adjusting/manipulation methods are encouraged, but such presentations should include some clinical data collection on why the techniques should be considered.
4.    Special Interest, in which papers on anthropology, epidemiology, cost of care, standards of care, and educational methods are presented.  Philosophical papers will also be reviewed, provided they follow formal philosophical argument with references, rather than the presentation of personal opinion.

A team of qualified reviewers was selected prior to the conference to review research submissions.  These reviews were led by Scott Haldeman DC, MD, PhD, and Eduardo Bracher DC, MD, PhD.  The exact membership of the scientific review committee was anonymous.  The scientific committee reviewed all submitted works and then selected those deemed of high enough quality to be showcased at the conference.

Peer review is a generic term for a process of self-regulation by a profession, or a process of evaluation involving qualified individuals (typically researchers who hold a DC and PhD or equivalent) within the chiropractic profession.  Peer review methods are employed to maintain standards, improve performance, and provide credibility.

Professional peer review focuses on the performance of professionals, with a view to improving quality, upholding standards, or providing certification.  Professional peer review activity is widespread in the field of health care, where it is best termed clinical peer review.  In academia, peer review is common in decisions related to faculty advancement and tenure.  Scholarly peer review (also known as refereeing) is the process of subjecting an author’s scholarly work, research, or ideas to the scrutiny of others who are experts in the same field, before a paper describing this work is published in a journal.  The work may be accepted, considered acceptable with revisions, or rejected.  Peer review requires a community of experts in a given (and often narrowly defined) field, who are qualified and able to perform impartial review.  Impartial review, especially of work in less narrowly defined or inter-disciplinary fields, may be difficult to accomplish; and the significance (good or bad) of an idea may never be widely appreciated among its contemporaries

• Amati M, Tomasetti M, Ciuccarelli M, Mariotti L, Tarquini LM, Bracci M, Baldassari M, Balducci C, Alleva R, Borghi B, Mocchegiani E, Copertaro A, Santarelli L. Relationship of job satisfaction, psychological distress and stress-related biological parameters among healthy nurses: a longitudinal study.  J Occup Health.  2010;52(1):31-8.
• Cuellar JM, Golish SR, Reuter MW, Cuellar VG, Angst MS, Carragee EJ, Yeomans DC, Scuderi GJ.  Cytokine evaluation in individuals with low back pain using discographic lavage.  Spine J. 2010;10(3):212-8.
• Christiansen D, Larsen K, Jensen OK, Nielsen CV. Pain response classification does not predict long-term outcome in patients with low back pain who are sick-listed.  J Orthop Sports Phys Ther.  2010;40(10):606-15.
• Danese A, Moffitt TE, Pariante CM, Ambler A, Poulton R, Caspi A. Elevated inflammation levels in depressed adults with a history of childhood maltreatment.  Arch Gen Psychiatry.  2008;65(4):409-15.
• Dixon D, Meng H, Goldberg R, Schneiderman N, Delamater A. J Pediatr Stress and body mass index each contributes independently to tumor necrosis factor-alpha production in prepubescent Latino children. Nurs.  2009;24(5):378-88.
• Du Bois M, Szpalski M, Donceel P. Spine J. 2009 May;9(5):350-9.  C, Barz T, Rolli Salathé C, Tamcan O, Mueller U, Theis JC.  Patients at risk for long-term sick leave because of low back pain.  Int Orthop.  2009;33(2):301-13.
• Fritz JM, George SZ.  Identifying psychosocial variables in patients with acute work-related low back pain: the importance of fear-avoidance beliefs.  Phys Ther.  2002;82(10):973-83.
• Heymans MW, van Buuren S, Knol DL, Anema JR, van Mechelen W, de Vet HC.  The prognosis of chronic low back pain is determined by changes in pain and disability in the initial period.  Spine J. 2010;10(10):847-56.
• Kallewaard JW, Terheggen MA, Groen GJ, Sluijter ME, Derby R, Kapural L, Mekhail N, van Kleef M. Discogenic low back pain.  Pain Pract.  2010;10(6):560-79.
• Li FC, Zhang N, Chen WS, Chen QX. Endplate degeneration may be the origination of the vacuum phenomenon in intervertebral discs. Med Hypotheses.  2010;75(2):169-71.
• Glaser R, Robles TF, Sheridan J, Malarkey WB, Kiecolt-Glaser JK. Mild depressive symptoms are associated with amplified and prolonged inflammatory responses after influenza virus vaccination in older adults. Arch Gen Psychiatry.  2003;60(10):1009-14.
• Igarashi A, Kikuchi S, Konno S, Olmarker K. Inflammatory cytokines released from the facet joint tissue in degenerative lumbar spinal disorders. Spine (Phila Pa 1976).  2004;29(19):2091-5.
• Kiecolt-Glaser JK, McGuire L, Robles TF, Glaser R. Psychoneuroimmunology and psychosomatic medicine: back to the future.  Psychosom Med. 2002;64(1):15-28.
• Maletic V, Raison CL. Front Biosci. Neurobiology of depression, fibromyalgia and neuropathic pain. 2009 Jun 1;14:5291-338
• McDade TW, Hawkley LC, Cacioppo JT. Psychosocial and behavioral predictors of inflammation in middle-aged and older adults: the Chicago health, aging, and social relations study.  Psychosom Med. 2006;68(3):376-81.
• Melloh M, Elfering A, Egli Presland C, Roeder C, Barz T, Rolli Salathé C, Tamcan O, Mueller U, Theis JC.  Identification of prognostic factors for chronicity in patients with low back pain: a review of screening instruments. Int Orthop.  2009;33(2):301-13.
• Miller GE, Rohleder N, Cole SW. Chronic interpersonal stress predicts activation of pro- and anti-inflammatory signaling pathways 6 months later.  Psychosom Med. 2009;71(1):57-62
• Miller GE, Rohleder N, Stetler C, Kirschbaum C. Clinical depression and regulation of the inflammatory response during acute stress.  Psychosom Med. 2005;67(5):679-87.
• O’Connor MF, Bower JE, Cho HJ, Creswell JD, Dimitrov S, Hamby ME, Hoyt MA, Martin JL, Robles TF, Sloan EK, Thomas KS, Irwin MR. To assess, to control, to exclude: effects of biobehavioral factors on circulating inflammatory markers. Brain Behav Immun. 2009;23(7):887-97.
• O’Donovan A, Hughes BM, Slavich GM, Lynch L, Cronin MT, O’Farrelly C, Malone KM. Clinical anxiety, cortisol and interleukin-6: evidence for specificity in emotion-biology relationships.  Brain Behav Immun. 2010;24(7):1074-7.
• Pace TW, Mletzko TC, Alagbe O, Musselman DL, Nemeroff CB, Miller AH, Heim CM. Increased stress-induced inflammatory responses in male patients with major depression and increased early life stress.  Am J Psychiatry.  2006;163(9):1630-3.
• Pikhart H, Hubacek JA, Kubinova R, Nicholson A, Peasey A, Capkova N, Poledne R, Bobak M. Depressive symptoms and levels of C-reactive protein: a population-based study.  Soc Psychiatry Psychiatr Epidemiol.  2009 Mar;44(3):217-22.
• Ramond A, Bouton C, Richard I, Roquelaure Y, Baufreton C, Legrand E, Huez JF.  Psychosocial risk factors for chronic low back pain in primary care–a systematic review.  Fam Pract.  2011;28(1):12-21.
• Reuben DB, Cheh AI, Harris TB, Ferrucci L, Rowe JW, Tracy RP, Seeman TE.
• Peripheral blood markers of inflammation predict mortality and functional decline in high-functioning community-dwelling older persons.  J Am Geriatr Soc. 2002;50(4):638-44.
• Rumayor A, Januzzi JL.  Association between anxiety and C-reactive protein levels in stable coronary heart disease patients.  Bankier B, Barajas J, Martinez- Psychosomatics.  2009;50(4):347-53.
• van Kleef M, Vanelderen P, Cohen SP, Lataster A, Van Zundert J, Mekhail N. 12.  Pain originating from the lumbar facet joints. Pain Pract.  2010;10(5):459-69
• Van Nieuwenhuyse A, Crombez G, Burdorf A, Verbeke G, Masschelein R, Moens G, Mairiaux P; BelCoBack Study Group.  Physical characteristics of the back are not predictive of low back pain in healthy workers: a prospective study.  BMC Musculoskelet Disord.  2009;10:2.
• van Vuuren B, Zinzen E, van Heerden HJ, Becker PJ, Meeusen R.Work and family support systems and the prevalence of lower back problems in a South African steel industry.  J Occup Rehabil.  2007;17(3):409-21.
• von Känel R, Begré S, Abbas CC, Saner H, Gander ML, Schmid JP.  Inflammatory biomarkers in patients with posttraumatic stress disorder caused by myocardial infarction and the role of depressive symptoms. Neuroimmunomodulation.  2010;17(1):39-46.