In response to the previous article discussing OCD and Involuntary Psychiatric Care, S. P. has decided to post a reply on his blog. He also used different proxies to comment on this blog, despite the fact that his comment privileges were removed several weeks ago because of his constant abusive behavior and unwillingness to address arguments. Unsurprisingly, S. P. engages in multiple personal attacks (calling me a zealot, claiming that I must be mentally retarded, a fascist and a pathetic useful idiot) and rehashing of the same old assertions that have been debunked before. Unlike S. P. I am going to be the bigger person and just address the arguments.
Details are not irrelevant!
It is clear that the personal story of S.P. was an important influence for the origin of his animosity towards psychiatry. In my previous post, I made a provisional conclusion that the story lacked crucial details. The major details S. P. gives us is that he has an exaggerated fear of contracting HIV through usual contact, that he was diagnosed with OCD and that he underwent involuntary psychiatric care in an unnamed European country. However, to investigate whether the decision to commit S. P. for involuntary psychiatric care was justified (an by extension the emotional underpinnings of his position), one needs to know the precise details of the conditions, such as severity and the level of functional impairment, as well as the identity of the European country (to be able to check the laws regulating it).
My general argument was that it seemed implausible that he was subjected to involuntary psychiatric care just because an exaggerated fear. I assume most people have more or less irrational fears: wasps, dogs, clowns, heights, spiders, snakes, lightning and thunder, flying and of course germs. But obviously the majority of people with irrational fears are not subjected to involuntary psychiatric care. The conclusion I drew was that there is more to the story that S. P. has shared. Now, I certainly realize that anxiety issues (as well as other issues related to mental health) are sensitive things and I obviously cannot force S. P. to share if S. P. does not want to. However, this does not change the fact that these questions are very relevant. How severe is the condition? How much functional impairment was there? What are the laws regulating involuntary psychiatric care in the unnamed European country?
Individual freedoms and laws regarding involuntary psychiatric care (Sweden v.s. the U. S.)
S. P. dislikes the Swedish laws that regulates involuntary psychiatric care. To summarize, Sweden require that the patient has a severe psychiatric condition, refuses voluntary care and cannot be taken care of in any other way that around-the-cloak psychiatric care. According to S. P., the U. S. requires that the life of the patient or a third person needs to be threatened. What S. P. fails to grasp is that the third condition in the Swedish law is related to whether or not the life of the patient is threatened, but is more specific and focuses on not being able to take care of him or her self, or by other people. This highlights an important difference between Sweden and the U. S. Sweden has a generous welfare system, but the U. S. generally does not. It boils down to how much do we think that the government should have to care about people.
I find it peculiar that S. P. considers the U. S. to value freedom. Is S. P. aware that the FBI regularly abuses the Patriot Act?
Accuracy of medical testing
I confronted S. P. on his belief that medical tests are generally 100% accurate. S. P. then says that he clearly did not mean 100%, but used a metaphor. I find this hard to believe because the argument is often used by opponents of science-based medicine and metaphor is really a form of analogy, so S. P. is using the term incorrectly. The general message of my argument was that medical tests that are not 100% accurate does not undermine the existence of the condition, or the biological basis of the condition.
Misunderstanding consensus
S. P. clearly misunderstood my transition from medical testing to HIV/AIDS. What I said was that “Since the second blog post tells me he accepts the consensus position that HIV causes AIDS, I will pick an example from this area.” When I used the phrase “consensus position” I am making the descriptive statement, that “HIV causes AIDS” is generally accepted by the scientific community (much like the existence of biological factors influencing the development of mental conditions). I did not imply that S. P. accepts that HIV causes AIDS because it is a consensus position. I am just referring to the current consensus position as “HIV causes AIDS”.
Biological markers, again
S. P. makes a subtle shift when it comes to the existence of biomarkers. In a previous entry, he quoted the American Psychiatric Association out of context saying that “brain science has not advanced to the point where scientists or clinicians can point to readily discernible pathologic lesions or genetic abnormalities that in and of themselves serve as reliable or predictive biomarkers of a given mental disorder or mental disorders as a group”. I explained that this is because the mental conditions are heterogeneous (just like many other diseases) and that they are strongly multifactorial, so we should not except the existence of a particular biomarker that can be found in all cancer patients and cannot be found in individuals without the mental condition. This is true for other conditions as well, such as type-I diabetes. The genetic risk factors for type-I diabetes are different in different subgroups with the disease and people with certain genetic risk factors that lack other risk factors may never develop the disease. Does this mean there is no genetic influence for type-I diabetes? Of course not.
Remember, the entire quote from the American Psychiatric Association was:
Research has shown that serious neurobiological disorders such as schizophrenia reveal reproducible abnormalities of brain structure (such as ventricular enlargement) and function. Compelling evidence exists that disorders including schizophrenia, bipolar disorder, and autism to name a few have a strong genetic component. Still, brain science has not advanced to the point where scientists or clinicians can point to readily discernible pathologic lesions or genetic abnormalities that in and of themselves serve as reliable or predictive biomarkers of a given mental disorder or mental disorders as a group. Ultimately, no gross anatomical lesion such as a tumor may ever be found; rather, mental disorders will likely be proven to represent disorders of intercellular communication; or of disrupted neural circuitry. Research already has elucidated some of the mechanisms of action of medications that are effective for depression, schizophrenia, anxiety, attention deficit, and cognitive disorders such as Alzheimer’s disease. These medications clearly exert influence on specific neurotransmitters, naturally occurring brain chemicals that effect, or regulate, communication between neurons in regions of the brain that control mood, complex reasoning, anxiety, and cognition. In 1970, The Nobel Prize was awarded to Julius Axelrod, Ph.D., of the National Institute of Mental Health, for his discovery of how anti-depressant medications regulate the availability of neurotransmitters such as norepinephrine in the synapses, or gaps, between nerve cells.
Now, S. P. has shifted towards denying the existence of biological influences on mental conditions completely, just because none of the risk factors are unique to the conditions. So not only is the quote out of context, but S. P. has shifted the usage of the term biomarkers to support his anti-psychiatry position. In other words, S. P. confuses the existence of a unique and reliable biomarker for a given condition with the existence of a biological basis for the same condition. There is no contradiction in saying that there is nothing that categorically and uniquely distinguish people with ADHD from all other people without it, but that there are biological influences for the origin of ADHD. S. P. is ignoring the multifactorial origin of mental conditions.
Cancer, psychological questionnaires and migraine
I pointed out that there is no unique biomarker on cancer cells that uniquely distinguish them from healthy cells. Guess what S. P. does then? He points out that there are imaging techniques that can be used. This is true, but the same can be done for e.g. ADHD and schizophrenia. S. P. cannot escape it.
S. P. then makes the straw man assertion that I think psychological questionnaires are as reliable as diagnostic tests for HIV. Psychological questionnaires have been demonstrated to have a high reliability, but of course tests like PCR are more reliable. But that is not the issue. The issue is that there are reliable tests for many mental conditions.
S. P. rejects my migraine analogy by saying that there is physical pain associated with migraines. But there are physical symptoms associated with a depressive episode or an anxiety attack as well. So what is the qualitative difference? If you accept the existence migraines, you must accept the existence of anxiety disorders and many other mental conditions with physical symptoms.
SSRIs and suicidality
I linked to a study showing that SSRIs do not increase suicides in the reference section. I can even link to a review in the Lancet (Hall, 2006), showing that treatment with SSRIs actually reduce suicide risk and the supposed increase in suicidality early in treatment is due to not controlling for base line level of suicidal thoughts.
SSRIs are more effective than placebo and the studies S. P. references demonstrates it!
I found this to be endlessly entertaining. S. P. thinks that SSRIs are no more effective than placebo, and links two studies that he thinks demonstrates this. However, the results of both studies contradict his claim! The golden rule in discussion scientific studies is to actually read the study. This rookie mistake shows that S. P. is probably not being completely honest when he says that he has a high level scientific degree. Let’s take them chronologically, starting with Turner.
Turner et. al. (2008) looked at studies on SSRIs submitted to the FDA and found that, like all other medications, studies that show a larger effect are more likely to be published. Taking this into account, he calculates an overall effect size for both published and unpublished studies and concludes that:
For each of the 12 drugs, the effect size derived from the journal articles exceeded the effect size derived from the FDA reviews (sign test, P<0.001) (Figure 3B). The magnitude of the increases in effect size between the FDA reviews and the published reports ranged from 11 to 69%, with a median increase of 32%. A 32% increase was also observed in the weighted mean effect size for all drugs combined, from 0.31 (95% CI, 0.27 to 0.35) to 0.41 (95% CI, 0.36 to 0.45).
In other words, the effect size of published and unpublished studies was 0.31 compared to placebo. This lies between a small effect and a moderate effect. So SSRIs are more effective than placebo. S. P. quotes the article out of context, emphasizing the publication bias, but ignoring the fact that even when the bias is taken into account, all the SSRI drugs tested were still superior to placebo.
What about the Kirsch et. al. (2008) study? This one actually concludes that SSRIs are no more effective than placebo, but the results do not support this conclusion. The results show that SSRIs have an effect size of 0.32 (looking at both published and unpublished studies):
Represented as the standardized mean difference, d, mean change for drug groups was 1.24 and that for placebo 0.92, both of extremely large magnitude according to conventional standards. Thus, the difference between improvement in the drug groups and improvement in the placebo groups was 0.32, which falls below the 0.50 standardized mean difference criterion that NICE suggested.
In fact, Kirsch studied pretty much the same data as Turner et. al. So how could they reach the opposite conclusion? It is because Kirsch used an arbitrary standard for clinical significance that is no longer being used by National Institute for Clinical Excellence (NICE). As we see, the effect size from Kirsch is even bigger than that from Turner. Even psychotherapy has an effect size of 0.22, so if you reject antidepressants, you have to reject psychotherapy as well. A bold move for any proponents of anti-psychiatry to make, save perhaps the lunatic fringe.
So when we look at the actual data for effect size of antidepressants such as SSRIs, controlled for publication bias, it is around 0.3 compared with placebo. This shows that SSRIs have a practically significant better effect than placebo.
There is really nothing else of substance in the rebuttal S. P. posted so I think I will stop here for now, but i suspect the rock will roll down soon enough again.
References and further reading
Hall, W. D. (2006). How have the SSRI antidepressants affected suicide risk? The Lancet, 367(9527), 1959-1962.
Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, et al. (2008). Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration. PLoS Med 5(2): e45.
Turner, E. H., Matthews, A. M., Linardatos, E., Tell, R. A., & Rosenthal, R. (2008). Selective Publication of Antidepressant Trials and Its Influence on Apparent Efficacy. New England Journal of Medicine, 358(3), 252-260.