Manual Therapy Combinations Reduced Headache Frequency and Severity. The Trial Quality Holds the Field Back.

The Source Paper

Rehman and colleagues published their work in the Journal of Osteopathic Medicine in late 2025 (with an eCollection date of May 1, 2026). The team searched 11 databases through September 2023, screened 11,657 records, full-text reviewed 167, and included 18 randomized controlled trials. The included population was adults with headaches associated with musculoskeletal dysfunction. The interventions were OMT techniques classified into three families: articulatory (ART), high-velocity low-amplitude (HVLA), and soft-tissue (ST). Trials compared OMT to other treatment, to a different OMT technique, or to combined OMT modalities.

Outcomes pooled for analysis included headache severity, headache frequency, headache-related disability, quality of life, and return to work. Harm outcomes included all-cause dropout, dropout due to inefficacy, and adverse events. Risk of bias was rated using the Cochrane RoB 2 tool. Quality of evidence was rated using the GRADE framework.

The relevance for a chiropractic readership is direct. The technique vocabulary in osteopathic medicine and chiropractic medicine overlaps substantially. HVLA in OMT is the same biomechanical event that chiropractic calls a diversified adjustment or spinal manipulation. ST work in OMT is what chiropractic calls myofascial release or pin-and-stretch soft-tissue work. ART in OMT corresponds to the lower-velocity articulatory mobilization used in chiropractic care for patients who do not tolerate or do not need the higher-velocity thrust. The pooled evidence about OMT for headache is, for practical clinical purposes, also pooled evidence about chiropractic for headache when the underlying technique is the same.

What They Found

The pattern is consistent across the moderate-quality findings: a combination of techniques outperforms single-technique comparisons for the outcomes that patients actually care about, which are how often the headache happens, how bad it is when it does, and how much the quality of life improves over the course of care. The effect sizes (SMD around 0.4 to 0.6) sit in the small-to-moderate range. They are not the kind of dramatic effect sizes that change a field overnight; they are the kind of effect sizes that make a clinically meaningful difference for the patients who respond.

The disability and adverse-event findings are worth reading carefully. The lack of significant association on disability is not the same as a finding of no effect. With the trial quality at the level the authors describe, the meta-analysis simply cannot detect a real-world disability change with confidence, and that is honestly reported. The lack of significant adverse events is the more reassuring half of that pair: across 18 RCTs, the harm signal did not separate from comparison treatments, which is consistent with what every other systematic review of manual therapy for headache has found.

Why the Trial Quality Holds the Field Back

None of the 18 trials met every Cochrane risk-of-bias criterion. That is the same finding that the chiropractic-headache-guideline team reported in their 2026 multi-institutional Delphi consensus, and it is the same finding that turned up in the de Zoete Cochrane review of SMT for chronic low back pain. Manual-therapy trials, across conditions and across professions, tend to share the same set of methodological problems: difficulty blinding the practitioner, difficulty constructing a credible sham control, small sample sizes, heterogeneous technique definitions, short follow-up windows, and outcome measures that are not standardized across studies.

Two specific structural issues are visible in this paper. First, the technique categories (ART, HVLA, ST) are themselves loosely defined across trials. Two trials labeled as testing HVLA may use different specific manipulation techniques, applied to different vertebral levels, with different practitioner training backgrounds. The pooled estimate is correct as a pooled estimate, but the heterogeneity of what is actually being delivered limits how directly the findings translate into clinical recommendations.

Second, the comparison arms vary widely. A trial of HVLA versus sham produces a different effect estimate than HVLA versus active mobilization, and both are different from HVLA versus pharmacologic therapy. The pooled meta-analysis works because the outcomes are standardized in SMD units, but the clinical interpretation has to account for what the comparator was. The effect size of "manipulation versus sham" answers a different question than "manipulation versus standard care."

Why this matters for headache patients in Tulsa

Most headache patients I see in walk-in care at Vanguard Chiropractic fall into two categories. The first is tension-type headache: a band-like pressure across the forehead or temples, often related to sustained postural load, screen time, or jaw clenching. The second is cervicogenic headache: pain that originates from upper cervical dysfunction (typically C0 to C3) and refers into the head along trigeminocervical pathways, recognized as a distinct headache subtype in the International Classification of Headache Disorders, 3rd Edition. Both presentations have a clear musculoskeletal driver and both are within the population the Rehman meta-analysis pools.

The clinical translation of this paper for a patient choosing chiropractic care for either of those headache types is: combine manipulation with soft-tissue work. That means cervical HVLA (the diversified adjustment most patients experience as the popping or releasing sensation at the upper neck and base of skull), paired with sub-occipital and upper trapezius release, and where appropriate articulatory mobilization at segments that do not tolerate HVLA. The Rehman pooled data supports this combined-technique approach more strongly than it supports any single technique used in isolation.

What the paper does not support, and what no honest review of the literature would currently support, is a guarantee of headache resolution. The effect sizes are moderate. A meaningful subset of patients will see substantive headache reduction. Another subset will see partial improvement. A smaller subset will see little change, and those patients are best served by an early referral to neurology if a primary headache disorder (migraine with or without aura, cluster headache, medication-overuse headache) becomes the more likely explanation. Cervicogenic headache is the diagnosis manual therapy is best positioned to address; primary migraine is not, regardless of how well manipulation is delivered.

How this fits with the Trager 2026 guideline

The 2026 Trager chiropractic headache guideline (covered in this issue's prior commentary) recommended spinal manipulation as a first-line option for tension-type and cervicogenic headache, with multimodal care as the preferred clinical approach. The Rehman meta-analysis provides additional evidence for that recommendation, specifically for the combined-technique angle. The two papers tell a consistent story: the case for manipulation is strongest when manipulation is part of a multimodal manual-therapy program, not when it is delivered as a standalone intervention. This is consistent with the broader pattern documented in the Cochrane review of spinal manipulation across musculoskeletal conditions.

For Vanguard Chiropractic patients with headache complaints, the practical implication is the care model already in place. A typical visit pairs cervical adjustment with sub-occipital and upper trapezius soft-tissue release plus posture and movement coaching for what to do between visits. The Rehman pooled data and the Trager consensus guideline both support that approach, and the body of evidence is now mature enough that the clinical recommendation is stable even as the field continues to push for higher-quality individual trials.