Integrative Medicine and Wellness: Research Interpretation in Natural Medicine

One of the consistent controversies in the complementary and alternative medicine (CAM) community is how to interpret the vast quantity of research that comes out on a weekly basis. Market forces both supportive of and against CAM interventions often help to distort findings in the public eye.

Perhaps the biggest difficulty in interpretation of CAM research is that studies of unconventional approaches are almost uniformly of lower quality than studies in modern surgery or pharmacology. This lower quality is generally more a function of resource limitations than any intent, and is reflected in smaller numbers of subjects, problems with randomization or blinding of studies, or difficult to measure outcomes.

Still, lower quality research is still science, and is important in advancing our knowledge, as long as our interpretations are valid. Here are the different levels of research study, and some idea of the level of evidence they represent:

Double-blind, placebo-controlled trial: Usually the highest standard of research study. Here, researchers can demonstrate cause and effect most clearly. Even though this method can yield flawed results, this level of research is often sufficient to make a conclusion about a new treatment.

Unblinded clinical trial: Here, subjects in the study get a treatment, but they may not be compared to a similar group not receiving treatment, and are aware of the treatment they are getting. While this type of study is not definitive, it can help to identify potentially useful treatments worthy of further study.

Epidemiology or survey study: These are studies that survey a large group of people and attempt to find correlations between disease and health habits. These studies do not demonstrate cause and effect, but when they yield consistent results, can still sway medical opinion. Much of what doctors know about diet and exercise comes from this type of research.

Anecdote or case report: These describe the outcome of a single patient after a specific intervention. While patients and consumers often find this level of evidence compelling, I do not. Published case reports go through a peer-review process, which makes them more credible than other anecdotes. Interesting and important treatments often are first reported as a single case, but further research is virtually always necessary to confirm results.

Preclinical research: Animal and test-tube studies help to show the mechanisms by which treatments may work. They should not be the basis for general health recommendations, but when results are consistent with higher level studies, they can help to fill in knowledge gaps.

Meta-analysis: As their name suggests, this type of study gather together multiple research studies, sometimes of different levels of evidence. A meta-analysis is only as good as the studies analyzed, but can often highlight bias inherent in individual research studies.

Complicated and subtle interventions like diet and lifestyle changes are simultaneously thought to be some of the most important predictors of human health outcomes, and among the most difficult treatments to effectively study. For this reason, diet and lifestyle recommendations are often made based on the results of survey research, supported by consistent animal and test-tube findings. If you have been puzzled by how frequently healthy diet recommendations seem to change, this lower level of evidence may help to explain the inconsistency.

There are several questions I tend to ask about new treatments before I consider bringing them into my clinical practice. Here are a few: