Posted by Larry Hoover on March 22, 2007, at 14:05:53 [reposted on March 25, 2007, at 4:34:59 | original URL]
In reply to Re: Orothomolecular treatment for bipolar disorder, posted by Squiggles on March 22, 2007, at 13:06:32
> So, even if preliminary clinical trials on drugs, nutrients, and minerals are too early to yield valid results, eventually, after some years, it should become obvious that vitamins and amino acids, for example ARE or ARE NOT efficient in *correctly* diagnosed psychoses. I think the major mental illnesses are too hard-wired to confuse with the placebo effect, unlike say, situational anxiety and stress.
>
> So, we should know by now whether bipolar disorder can be cured by vitamin B or amino acids. If someone claims to be able to cure orthomolecularly, then either the subject is not bipolar, or not enough time has elapsed to see the illness recur. This is assuming that we know that vitamin B or amino acid X do not work.I disagree with your basic premise. You assume homogeneity of diagnosis can be determined a priori. In fact, the currently observed clinical/statistical failure of any specific intervention to cure all subjects pretty convincingly argues for broad heterogeneity with convergent symptomotology/phenotypy, within any currently defined psychiatric diagnosis.
Let's assume, for argument's sake, that bipolar really exists as five different dysfunctions (genotypy), each expressing in a similar way (phenotypy). However, we cannot yet tell one from the other. We diagnose by symptoms alone. We may have found the perfect treatment for one of the dysfunctions, according to our biochemical investigations, but when we gather a collection of subjects together and blindly treat them, our success falls well short of our expectations, statistically. Yet, for individuals, some are "cured". Others might be helped a bit. And others yet, are cursing the cost/lost time/side effects, absent any benefit.
Some treatments work for some people, but not for others. Orthomolecular treatments do work. But not predictably so. You still have to do the experiments. The only scientific way to demonstrate absence of benefit from orthomolecular treatment is to never try. All the rest is statistics, with all attendent fallacies (type 1/type 2 error, sampling bias, blah blah blah).
In the end, the only thing that matters are the experiments with N=1, i.e. those done on you.
Lar
poster:Larry Hoover
thread:744029
URL: http://www.dr-bob.org/babble/alter/20070320/msgs/744033.html