Chiropractic adjustments can heal your DNA?

The headline above should be good for a laugh, but believe it or not, chiropractors around the world are claiming that they can help your body repair its DNA. All of them cite the same 2005 article as evidence, so I read the article to find out what it was all about.

The article is titled "Surrogate Indication of DNA Repair in Serum After Long Term Chiropractic Intervention – A Retrospective Study," written by Clayton Campbell, Christopher Kent, Arthur Banne, Amir Amiri, and Ronald W. Pero. They published it in 2005 in a chiropractic journal called the Journal of Vertebral Subluxation Research. This journal has many of the trappings of a scientific journal, but it's really all make-believe: it does not appear to be properly peer-reviewed, it is not indexed by standard biomedical databases, and (most damning of all) it is based on a concept, "subluxation," that does not exist. That's right, even the UK's General Chiropractic Council admitted in 2010 that subluxation was a mirage, saying:

"The chiropractic vertebral subluxation complex is an historical concept but it remains a theoretical model. It is not supported by any clinical research evidence that would allow claims to be made that it is the cause of disease or health concerns."

So about that journal article. The claim that chiropractic treatments could somehow improve your body's ability to heal its own DNA seems wildly implausible, but that's what Campbell and colleagues claim. Their press release, which was reproduced verbatim on many chiropractor's websites, said:

"In a landmark study published in the Journal of Vertebral Subluxation Research, chiropractors collaborating with researchers at the University of Lund found that chiropractic care could influence basic physiological processes affecting oxidative stress and DNA repair."

(I can't help remarking that authors don't usually boast that their own work is a "landmark study," but let's move on.)

Unfortunately for Campbell and colleagues, their study has fundamental flaws that completely undermine their claims, as we'll see below. Nonetheless, many chiropractors' websites are touting this amazing "benefit" today, including sites that were updated as recently as a few weeks ago, such as: this one (updated Oct 31 2011), this one, this one, this one in Australia (updated Oct 2011), this one in Australia, and many more.

So what did Campbell et al. actually study? First, they didn't measure DNA repair at all. They measured serum thiol levels, which at best are a very indirect indicator of DNA repair. And they ran a very small study, with just 76 patients, all who came to chiropractic clinics with back pain, whom they divided into 3 groups. The three groups were:

1. No chiropractic treatment, 30 patients
2. 2-12 months of chiropractic, 21 patients
3. 1-6 years of chiropractic, 25 patients

It was not placebo-controlled, blinded, or randomized, which presents major methodological problems regardless of what happened. Before I tell you the results, which group do you think the chiropractors would want to do the best? Bingo! The group that saw chiropractors for many years did the best, as measured by plasma thiols. At least that's what Campbell reported.

But the results were very odd: first, they saw a drop in plasma thiol levels (a drop is a bad outcome, for this study) in patients treated for 2-12 months, from 124 down to 105. But hang on: in the long-term chiropractic treatment group, the average level was 146. So are we supposed to believe that chiropractic is bad for you in the first year, but good for you after that? The problem gets worse, though, when you look at their claim that "there were statistically significant differences in the serum thiol levels of the three groups." None of the serum thiol levels were significantly different: their claim is simply wrong.

[Note: skip the next paragraph if you don't care about the statistics. But the statistics matter.]

Yes, that's right - Campbell et al. got their statistics wrong. Oops! They reported that the 2-12 month group had signficantly lower serum thiol levels, and the 1-6 year group was significantly higher levels, with a p-value of 0.001. From the numbers in their own tables, I was able to compute the true significance values, to determine if their reported value of 146 (plus or minus 60) was significantly higher than the control group's average of 124 (plus or minus 48). It turns out that this difference isn't significant at any level, and certainly not at a p-value of 0.001. A decent journal would never have published this painfully bad analysis.

There are other problems, but this huge error in their central result is devastating. And not surprisingly, no one has replicated these non-results since.

This hasn't deterred Chad Mathey, a chiropractor in Colorado, from posting this comment on his blog just a few weeks ago:

"This [the Campbell et al. study] is an incredible article! This talks about one of the many reasons people do and should stay under regular Chiropractic care. It’s not just for pain and people are starting to finally understand this."

Incredible indeed. As in "not believable" and "not even close to true."

This is another illustration of how pseudoscientists use the trappings of science to do make-believe science, and then advertise their "findings" to the world, just as Dr. Oz did in his recent apple juice and arsenic experiment. Dr. Oz didn't even publish his findings - he just announced them on his show. Campbell and colleagues used a pseudoscience journal. After all, who's gonna know?