German Studies

 That Can't be Replicated Outside of Germany

The development of the National Center for Complementary and Alternative Medicine (NCCAM) at the National Institute of Health (NIH)
 led to a number of large, multicenter trials with rigorous methodology for the assessment of herbs, vitamins and supplements for a number of conditions.
 These studies failed to replicate a number of prior studies which supposedly showed efficacy for these herbs in a variety of areas.
 The NCCAM studies involved much larger numbers of patients, and employed the placebo controlled randomized study designs that the earlier
 studies had employed. This failure to replicate is not a one time occurrence; rather it seems to be a repeated pattern. So what gives?
 Many of the earlier so-called "placebo controlled" studies came from Germany, where there is a strong bent towards herbal medicine.
 Given the fact that there data on the efficacy of herbs is far and few between, the fact that German doctors rely so heavily on herbal
 may explain why their health care (in terms of outcomes) is in last place amongst industrialized countries (thank goodness for Germany, otherwise we Americans
 would be in last place).

 So what gives? Why is there such a systematic disconnect between the earlier trials and the more recent ones? For one thing, the "placebo controlled"
 trials from Germany may not be as placebo controlled as you might think. Most of these studies did not assess to what degree the patients
 believed they were getting study medication. This is important because there are subtle ways patients (and treating doctors) can guess
 what they are on. For instance, the investigators of the NCCAM funded trial of Saw Palmetto for prostate hypertrophy explained in their paper
 that Saw Palmetto had a pungent odor, and they had taken great care to mask the scent. I did not see any similar efforts in other trials of
 Saw Palmetto. Similar concerns apply to Valerian root, which is used for insomnia. In addition, many of the initial trials were sponsored
 by the manufacturers of the herbs or supplements. Since non-replication of these initial trials by NCCAM funded trials seems to be the rule rather
 than the exception, where does that leave us?

 I think that many of these earlier positive trials are driven by bias and fundings sources. I read one study of Artichoke Leaf Extract
 where rather than measure the outcome at the end of the trial, they added up the treatment score at each week and used the summed score
 as the outcome measure. There was no baseline score so you didn't know if the two groups were different at baseline. And the score at week one
 was elevated in the artichoke group and stayed elevated at the same level throughout the trial.

 A more extreme example is the case of Ranjit Chandra, M.D., who published studies, one in The Lancet, that claimed that
 multivitamins reduced the length of colds. He published an article in his journal Nutrition Researchby a mysterious author named
 "Dr. A.L. Jain" that "replicated" his findings. Unfortunately it was not possible to prove that Dr Jain was an actual human being
 and the studies of Dr. Chandra were apparently never conducted.