Chiropractic Safety in Older Adults: What the 2026 Updated Review Found

What the Authors Did

Daniels and colleagues, working under Clinical Compass and partial NCMIC Foundation funding, registered their protocol prospectively with PROSPERO and audited the search with PRESS methodology. They searched PubMed, Cochrane CENTRAL, CINAHL, AMED, and the Index to Chiropractic Literature on December 12, 2025. Inclusion criteria targeted adverse events attributed to chiropractor-provided treatment in patients over 55, drawn from RCTs, observational studies, and case reports or case series. Pilot studies, surveys, narrative reviews, populations under 55, and papers lacking adverse event detail were excluded. Risk of bias was assessed in duplicate using Scottish Intercollegiate Guideline Network checklists.

Of 2,295 titles screened, 125 went to full-text review and 25 papers were included. The pool: 6 randomized clinical trials, 4 observational studies, and 15 case reports or case series. One RCT was rated low risk of bias and five were acceptable. Three observational cohorts were low risk and one acceptable. The authors counted 412 study-related adverse events. Nine were severe. None were catastrophic.

The Findings, in Order of Clinical Weight

What is common: mild-to-moderate

The bulk of adverse events were predictable and self-limiting. Changes in pain quality, transient muscle or joint soreness, post-treatment stiffness, numbness, weakness, fatigue, headache, dizziness, and lightheadedness. None of those will surprise a clinician who treats older spines. None of those should surprise a patient who has been told what to expect after a first visit.

What is reassuring is the consistency. Across 6 RCTs, 4 cohorts, and 15 case reports, the dominant adverse-event pattern looks the same. Older adults experience the same mild-to-moderate transient effects that younger adults do. Age does not appear, in this dataset, to be a multiplier on garden-variety side effects.

What is rare: severe events from case reports

Six severe adverse events appeared in case reports. Four of those were vascular in nature, consistent with the long-running concern about cervical manipulation and arterial events. Twenty of the case reports involved spinal manipulation specifically. One followed TENS application. One followed repetitive shoulder abduction. Notably, no case reports described severe adverse events following soft tissue work, exercise, or other modalities.

That last point matters. The signal in the case reports is concentrated on one modality, spinal manipulation, especially in the cervical region. It is not a general "chiropractic is risky in older patients" signal. It is a "specific intervention, specific population, specific risk" signal that has been on the literature's radar for two decades, most prominently in the population-based work of Cassidy and colleagues showing no excess risk above primary-care visits for the same population.

What is missing: standardization

The honest weakness in the dataset is methodological. Several studies lacked explicit definitions of an adverse event. Several lacked structured data-collection processes. The authors flag this directly. Without standardized definitions and severity classifications, the certainty of the "no catastrophic events" headline carries less weight than it could.

Translation: the field is not measuring adverse events the same way across studies, so we cannot be sure we are counting them all. The conclusion is not "we have proven safety in this population." The conclusion is "in studies that bothered to track adverse events, no catastrophic ones turned up, and the common ones were mild and transient."

How This Sits With the Larger Evidence Base

The Daniels review does not exist in isolation. The largest population-based study on cervical manipulation and stroke, by Cassidy and colleagues in Spine (2008), used a case-control and case-crossover design across an Ontario population and found no excess risk of vertebrobasilar stroke after a chiropractic visit compared with the same patient visiting a primary care provider. The interpretation has held up for 18 years: patients in the early stages of vertebrobasilar dissection tend to seek care for the neck pain or headache the dissection is producing, and they sometimes see a chiropractor first because the symptom looks musculoskeletal. The provider does not cause the dissection. The dissection causes the visit. I take this apart in detail in my Forward Review of the 2026 STOP-CAD subanalysis, including the valid critiques of the paper and where the chiropractic-side rebuttal overreaches.

Whedon and colleagues looked at Medicare-aged populations using administrative-data analyses with similar conclusions for the most common chiropractic events. Carnes and colleagues' systematic review of adverse events across manual therapies broadly placed serious adverse events at a rate consistent with what we would expect from a clinic-based intervention used at scale.

The Daniels 2026 review fits inside that body of work. It is not the first word on safety, and it is not the last. It is an updated synthesis with the explicit honesty about its own methodology that the prior work has occasionally lacked.

Limitations and Caveats

The methodological problems above are not unique to this review. They are characteristic of the entire body of clinical safety evidence on manual therapy. The Daniels team's contribution is to pool the available studies in the over-55 population and to honestly flag the definition problem. That second part matters as much as the first.

What I Take From This for Practice

I see older adults every week at Vanguard Chiropractic in Tulsa. The most common reason an older patient hesitates to start care is the safety question. A family member said something. They read something online. Their primary care provider made a face when they brought it up.

The Daniels review gives me a better answer than I had before. I can tell that patient: in 25 studies summed across the field, with about 400 adverse events tracked, none were catastrophic, and the common ones look exactly like what we expect in younger patients. I can also tell them, honestly, that the field is not great at standardizing how adverse events get measured, so I want to know about anything they feel after a visit so I can document it properly.

The case-report concentration on spinal manipulation, especially cervical, is also useful. It tells me that for an older patient with cervical-spine concerns, vascular history, or anticoagulant use, my technique selection matters. Mobilization-based work, soft-tissue protocols, and gentle setup positions are not "lesser" care. They are appropriate care for a population where the rare-but-real events cluster around one specific technique. That clinical judgment is the part the literature alone cannot make for you.

One: Catastrophic events from chiropractic care in adults over 55 are rare enough that 25 studies and 412 tracked events did not produce one.

That is not "zero risk." It is "low enough that the modern literature is not turning up the worst-case scenarios as a regular event." A patient or referring clinician who asks "is the evidence strong" should hear "the certainty is mixed, the definitions vary, and the case reports are the wobbly piece." A patient who asks "did the worst-case scenarios show up across 25 studies" should hear "no, they did not."

Two: Mild-to-moderate effects, especially after the first or second visit, are normal and short-lived.

Soreness, mild stiffness, occasional brief lightheadedness. If anything is more than that, or lasts more than 48 hours, the patient should call. The expected effects are part of the standard pre-treatment conversation, not a surprise we hide from older patients in the name of optimism.

Three: For patients over 55 with vascular history, clotting disorder, or anticoagulant use, the conversation about technique selection is not optional.

Mobilization, soft-tissue work, and exercise-based approaches are appropriate alternatives to high-velocity cervical manipulation in those populations. A good chiropractor will raise this before the patient has to. The Daniels signal supports that conversation. It does not change its content.

Where the Field Should Tighten

Three asks for the next iteration of this work.

1. Adopt a standardized adverse-event definition across studies. The CARM (Chiropractic Adverse Reporting Methodology) framework or its successor would let the next systematic review pool data more confidently.
2. Include administrative-database studies in the search. The Daniels authors flagged this as a limitation. Rare events show up in big data, not in small RCTs.
3. Stratify reporting by technique class. The case-report signal in this review concentrated on one specific intervention. Future analyses should preserve that detail rather than rolling it into a single "chiropractic" bucket.