Research and Evidence

In Plain Terms

If you read nothing else on this page, read this. Five honest sentences about what the research behind structural correction does and does not support.

• • Posture and spinal curves can be measured reliably. Two 2026 reviews confirmed that different examiners measuring the same X-ray get consistent results. This is the best-established part of the evidence.
• • Traction can change those measurements. A review of nine controlled trials found that neck extension traction increased the neck curve more than the comparison care did.
• • Several trials found the symptom improvements lasted longer. Both the treated and comparison groups usually improved at around ten weeks. The difference showed up at six months to two years, when the comparison groups slid back.
• • Most of this literature is weak by design. 167 studies sounds like a lot. Only 34 are controlled trials or reviews of them; most of the rest describe single patients, which cannot show that treatment caused the change.
• • The evidence is concentrated, not independently confirmed. Most randomised trials come from one research group, and the authors disclose financial interests. We set that out in full below rather than leaving you to find it.

None of this predicts your result. It describes groups of people in published studies. What applies to you is a question your examination answers.

Why We Publish This

Our brand promise is that we measure, we correct and we prove it. “Prove it” means documenting change through reassessment — and being honest about what the published literature does and does not establish.

This page lists the 167 published studies we are willing to cite about structural correction and Chiropractic BioPhysics® (CBP), the method we practise. Of those, 34 are controlled trials or reviews of controlled trials. The remainder are case reports, case series and association studies, which carry much less weight. We say which is which next to every study.

We also keep a further 64 studies off this website entirely: 40 that we could only verify by title and never read, and 24 single-patient reports about conditions unrelated to the spine. We do not consider either group publishable, and we do not cite them.

The Strongest Evidence

These are the controlled trials and the reviews of controlled trials — the only tier of this literature that compares treated patients against a comparison group. Read the limitations with each one: they are short, and they matter.

• Systematic review · 299 traction patients vs 315 controls · 2021

  A 2021 systematic review of nine controlled trials found that neck extension traction increased the neck curve by 12–18°, and patients kept their improvements for up to 1.5 years.

  Limitations: Most RCTs from one CBP-affiliated group; 3 trials non-randomized; no meta-analysis.

  Oakley 2021
• Systematic review · 80 patients · 2020

  A 2020 systematic review of three controlled trials found that lumbar extension traction increased the low-back curve by 7–11°, and that the pain relief lasted longer than with rehab alone.

  Limitations: Only 3 trials; both RCTs from one group. CBP authors. No meta-analysis.

  Oakley 2020
• Randomised controlled trial · n=120 (60/60; 102 completed 1 year) · 2018

  A 1-year randomized controlled trial of 120 patients with neck trigger-point pain found that adding Denneroll traction to standard trigger-point therapy gave similar relief at 10 weeks, but at 1 year the traction group had significantly less pain and disability.

  Limitations: Alignment change was measured only on posture photos, not X-ray. No short-term pain advantage. A single research group with a CBP-affiliated co-author. The 1-year figures above come from the first reader's full-text check; an independent checker could only confirm them against the abstract, so verify them before quoting the numbers.

  Moustafa 2018
• Randomised controlled trial · n=66 (33/33) · 2023

  A randomized controlled trial of 66 older adults with neck pain and forward head posture found that CBP correction (mirror image exercises plus a neck traction orthotic) improved head posture by about 13° and lowered pain from 4.7 to 0.5 out of 10 at 3 months, outperforming a standard exercise program whose results faded.

  Limitations: Alignment was measured only by CVA (a posture angle); no X-ray ARA was reported in the text reviewed. Follow-up was only 3 months. Patients and therapists were not blinded (assessors were; PEDro 8/10). The printed 6-week Berg p-value (0.48) conflicts with its own confidence interval (2.2 to 4.7), so do not cite a 6-week balance result either way. The authors are CBP-affiliated.

  Suwaidi 2023
• Randomised controlled trial · n=64 (32 traction / 32 control) · 2013

  A randomized controlled trial of 64 patients with sciatica from an L5–S1 disc herniation found that adding lumbar extension traction to standard therapy improved the low-back curve, back and leg pain, disability and nerve function more than standard therapy alone, and the advantage held at 6 months.

  Limitations: Abstract only, with no effect sizes available. A single research group. The comparison was passive therapy, not exercise. Follow-up was 6 months. PEDro score 7/10.

  Moustafa 2013

These summaries describe published research. They are not a promise of results, and individual results vary.

Evidence by Condition

Each condition page carries the strongest few studies relevant to that presentation, with the same design-and-limitations treatment used here.

• • Neck pain and loss of the neck curve (16 studies)
• • Anterior Head Syndrome (forward head posture) (15 studies)
• • Spondylolisthesis and disc degeneration (15 studies)
• • Low back pain, sciatica and disc herniation (25 studies)
• • Headaches and migraines (11 studies)
• • Pinched nerve and arm pain (9 studies)
• • Whiplash and auto injury (10 studies)
• • Scoliosis and lateral shifts (11 studies)
• • Hyperkyphosis and hunched upper back (14 studies)
• • TMJ and jaw pain (4 studies)

Some groupings in this library have no page of their own, because the evidence is not strong enough to build one: post-surgical care, paediatric spine and posture, fibromyalgia, straight back syndrome, and measurements of nerve conduction. Those studies appear in the index below and nowhere else.

What This Research Does Not Show

• • It does not show that spinal alignment treats disease. Studies on asthma, ADHD, infertility, diabetes, reflux and similar conditions are single case reports. We do not cite them as evidence of anything, and we make no claim about them.
• • It does not show that measurement changes predict how you will feel. Several studies report improved alignment alongside improved symptoms; that is an association within an uncontrolled group, not a mechanism.
• • It does not show that any correction is permanent. The longest follow-ups run one to three and a half years, and most involved continued periodic care.
• • It does not show that reliable measurement means effective treatment. The measurement-reliability reviews below show the X-ray measurements can be reproduced between examiners. That is all they show.
• • It does not establish that care prevents surgery or future degeneration. We make no such claim.

Who Produced This Research, and What They Disclose

Most of the randomised trials in this literature come from one research group working in Egypt and the United Arab Emirates — principally Ibrahim Moustafa and Aliaa Diab — usually publishing with Deed Harrison, DC, president of CBP NonProfit, Inc. Several of the reviews are led by Paul Oakley, DC. Research funding for a number of these papers came from CBP NonProfit.

These authors declare their financial interests in the papers themselves, which is what researchers are expected to do. We reproduce two of those declarations here, word for word, so you can read them as published rather than as our summary of them:

“PAO is a paid consultant to CBP NonProfit Inc.; DEH teaches and sells products to doctors for spinal rehabilitation; There are no other conflicts of interest declared.”Conflict of interest statement, Oakley, Harrison & Haas 2021 — the systematic review of nine controlled trials cited above. Oakley 2021

“Dr. Paul Oakley (PAO) is a paid consultant for CBP NonProfit, Inc.; Dr. Deed Harrison (DEH) teaches chiropractic rehabilitation methods and sells products to physicians for patient care as used in this manuscript.”Conflict of interest statement, Oakley, Kallan & Harrison 2022 — the review of 60 published case reports. Oakley 2022

Read plainly: an author of several of these papers teaches the method and sells equipment used in the care being studied, and another is a paid consultant to the organisation that funds some of the research. Declaring this is normal and proper scientific practice, and none of it suggests the findings were falsified or that anyone acted improperly.

What it does mean is that the evidence base is concentrated rather than independently replicated. In medicine, a finding becomes dependable when separate groups with no stake in the outcome reproduce it. That has happened only to a limited extent here — one of the trials we cite on neck pain was run independently in Egypt, and we point that out in its summary. This is the most important caveat on this page.

We use this method, so we have an interest in it being effective. That is exactly why we publish the declarations alongside the findings instead of only the findings.

Can These Measurements Be Trusted?

A separate question from whether treatment works is whether the measurements themselves are reproducible. Two 2026 systematic reviews examined that for the neck and the mid-back.

• Systematic review · 2026

  A 2026 systematic review of 51 studies found that the main X-ray measurements of neck curve and head posture show good-to-excellent reliability, supporting their use to track changes in alignment.

  Limitations: Shows measurement reliability only, not treatment benefit; methods and statistics varied a lot across studies; CBP-affiliated authors. Numbers (818 screened, 51 included, 15/27/9 quality split) confirmed against the published abstract.

  Lightstone 2026
• Systematic review · 2026

  A 2026 systematic review of 66 studies found that measurements of the mid-back curve on X-ray generally show good-to-excellent reliability.

  Limitations: Measurement review, not treatment evidence; 15 of 66 studies were low quality; CBP-affiliated authors.

  Betz 2026

These summaries describe published research. They are not a promise of results, and individual results vary.

Complete Study Index

All 167 studies, grouped by what kind of study they are. Strongest tier first. Each links to the published paper.

Of these, 68 were summarised from the full paper and 99 from the abstract or a secondary source. Where a study’s specific figures have not yet been checked against the full text, our summary omits those figures rather than repeating them.

How We Decide What to Cite

• • We name the study design and the number of participants every time.
• • We state the limitations next to the finding, not in a footnote.
• • A single patient is described as a case report, never as “patients”.
• • We do not cite a study we have not read at least in abstract, and we say which.
• • We match the claim to the intervention actually tested. These trials tested extension traction and postural orthotics; where that differs from mechanical spinal decompression, we say so rather than blurring the two.
• • We do not present association as cause, or improvement in a measurement as a promise about symptoms.
• • Where we were wrong, we correct the page and log the change.

The full editorial standard we hold ourselves to is published in our repository alongside the data, and these principles are the same ones described in why we measure rather than assume and corrective care versus symptom relief.