Sep 01 2009
Three Quick Hits: Science and Mental Illness
Here are three quick hits about advances in psychological science as it pertains to mental illnesses. I’ll share a summarizing paragraph from the news release and comment on what I find must noteworthy about it.
1) ADHD and the Reach of the Placebo Effect
Now a recent review of research by University at Buffalo pediatric psychologists suggests that such medication, or the assumption of medication, may produce a placebo effect — not in the children, but in their teachers, parents or other adults who evaluate them.
[emphasis mine - source]
This is a great example why the tightest research is not only blinded (the subjects unaware as to whether they are receiving the true treatment or the placebo), but double blinded. Double blinding adds this layer of error prevention: those doing the measuring/evaluating are also unaware of whether individual subjects are receiving the real or sham treatment — and so their own expectations do not infiltrate the methodology and corrupt the results.
2) Schizophrenia and the Importance of Null Results
Research co-led by an academic at the University of Hertfordshire, concludes that cognitive behavioural therapy (CBT) is of no value in schizophrenia and has limited effect on depression.
[emphasis mine - source]
Positive results are noteworthy. But so are negative. Unfortunately, “our variable-manipulation generated results” is perceived as somehow more scientifically relevant than “our variable-manipulation resulted in no change.” But both are equally important. Yet the second class goes unreported/unacknowledged far too often.
Somewhat tangentially, this result makes sense to me. Primary characteristics of schizophrenia include delusions and disorganized thought. Cognitive therapy, in a sense, relies upon re-organizing a person’s thoughts and emphasizing a more functional perception of reality: things that people with schizophrenia struggle mightily with. IN using it you play to their weaknesses.
3) Depression and Missing Scientific Precision
[I]n acute therapy patients respond better to both drugs than to a sham drug. There is greater alleviation of symptoms, and in some cases they recede to such an extent that some patients no longer fulfill the criteria for a depression diagnosis. As far as relapse prevention is concerned, there is at least an indication that patients benefit more from duloxetine and venlafaxine than from a sham drug. In contrast to duloxetine, there is also proof in the case of venlafaxine that the drug provides more effective protection than placebo against a renewed occurrence of depressive symptoms (recurrence prevention). In the direct comparison of venlafaxine and duloxetine, neither drug displays superiority over the other with regard to alleviating depressive symptoms.
[emphasis mine - source]
While the above research-finding summary is specific — a good thing — it lacks precision. Good science and good science writing is precise. So okay, the drug treatments outperform placebo, but by how much? Are these essential numbers (and what is more precise than numbers) left out because the general public is math-averse? Or because findings such as this one seem so much more hum-drum when the actual effect-sizes are disclosed?
