Supporters of "integrative" or "holistic" treatments often assert that physical manipulations or dietary adjustments can improve problems of behavior or mood, and it would be silly to deny those possibilities. It would be equally, if not more, silly to presume that this claim is correct without some clear evidence to support it.
There is some evidence, it's true. But is it clear evidence? Is it the kind of evidence we should demand as support for the effectiveness of any treatment of any type? Let's look at one recent publication and see whether it's up to standard.
The paper I have in mind is about the use of chiropractic treatment with children who have Attention Deficit/Hyperactivity Disorder (ADHD), a problem that deserves much research attention because of the way it can interfere with school achievement and with social relationships. This article (Alcantara, J., & Davis, J., The chiropractic care of children with Attention Deficit/Hyperactivity Disorder: A retrospective series) was published in 2010 in Explore: The Journal of Science and Healing (pp. 173-182). The authors examined 4 cases from their records in which children and adolescents had received spinal manipulations and advice about dietary supplements from chiropractors over several months. A 15-item checklist of ADHD symptoms showed that the symptoms improved from the pre-treatment to the post-treatment period. Alcantara and Davis concluded that the treatment was an effective way to alleviate ADHD.
Well. Let's have a look at what we've got here. Need I repeat that 4 cases is not enough to draw clear conclusions from? Think about it this way. What if you had a dresser drawer full of socks, half of them black and half of them white? And what if you took 4 socks out of the drawer without looking? It wouldn't be too surprising if all the socks you picked up were black, or if all of them were white. But what if you pulled out 20 socks without looking? It would be surprising (i.e., the probability is small) if all 20 socks were black, or all 20 socks were white. And if you pulled out 100 socks, and they were all the same color? We'd certainly say that could not happen by chance alone.
Similarly, if we take 4 cases from thousands of possibilities, it's not unlikely that we'll get a set of cases that really don't represent the nature of all the possibilities.That means that the conclusions we draw from our 4 cases may be true about those cases-- but they aren't necessarily true about all the other cases. That's why we demand that evidence about the effectiveness of treatment has to be based on a large number of cases.
In addition, the Alcantara and Davis study is essentially a "before and after" design. It looks for differences between the original condition of the participants and their later condition. But there's no way to know that any such differences resulted from the treatment, rather than from other factors-- like getting older, receiving attention from others, believing that improvement is possible, and so on. In the case of children and adolescents, improvement may come about because parents and teachers, cheered up by their belief that the treatment will be helpful, behave differently toward the kids. They may be less angry and thus create less stress for the children, or their increased hopefulness may lead them to evaluate the children's behavior more positively and to rate them as better on those checklists.
To find out whether a treatment is effective, a comparison or control condition is needed. That is, the children given the chiropractic treatment need to be compared not just to themselves at an earlier date, but to similar children who did not receive the same treatment. If this is not done, there's no way to assess whether change occurred spontaneously; that could occur with increased age, or simply because most problems fluctuate in seriousness, and children are likely to have been brought into treatment when at a particularly low ebb, with nowhere to go but up.
There are several other points we'd want to see in a well-designed study. One is that individuals need to be randomly assigned to treatments. When the child or the parent chooses a particular treatment, many factors of belief or expectation come along with the choice, and chances are much increased that moods, behaviors, and checklists will be affected by those factors. Another important point is that such investigations should be "blinded"-- in other words, the children, the parents, and the teachers do not know which child is getting a real treatment and which one is getting a sham treatment that closely resembles the real one. Finally, a good study would separate the dietary supplements and the spinal manipulation; the Alcantara and Davis study made it impossible to know which of those very different factors was having any effect, if there was an effect at all.
Once again, it's "buyer beware" in the holistic treatment market. A published study (no doubt "peer-reviewed" as well) is not necessarily good evidence for the claims it makes.
By the way, when did the adjective "chiropractic" become a noun?