Genetic denialism and claims about locked wards and suicide from non-RCT designs

Some things I’m saying in social media in the week ending July 31, 2016 and will probably follow up on.

This post is a bit of an experiment. I share some of things that I’m talking about on Facebook and Twitter that have a high likelihood that they will be reflected in later blog posts. Comments welcomed, as well as encouragement or discouragement about pursuing further in later blog posts.

 On Facebook: Genetic Denialism

I had  recently blogged about a US neurologist writing in The Conversation that clinicians should consider the possibility that migraine headaches are due to patients having been abused as children. The neurologist arrived at this recommendation with very superficial attention to the available evidence.I would’ve thought that tickling with a complex and poorly understood condition like migraine headaches, the neurologist would recommend exploring comorbidities and idiosyncratic reactions to medications, rather than making a psychological issue of something that is not obviously psychosomatic.

A focus on childhood adversity and a rejection of genetic influences and migraines reflects a recurring theme in a much larger literature. For a while, child adversity has been being pitted against hereditary in discussions where increasingly sophisticated research about genes and genetic expression are given short shrift. Oliver James is a key example of someone who considers the choice between early adversity and genetic influences settled on the basis of politics.

I posted a link to this article on Facebook:

The cruel grandmother, shamed mother and psychotic half-brother who shaped David Bowie’s life and work.

I said

More nonsense from the generic denialist Oliver James. He has a lot in common with Richard Bentall  and this piece could’ve been written by Bentall. James basically takes considerable evidence of familial transmission of psychosis in the family of David Bowie and uses it to deny a role of genetics.

David Bowie Passes away at 69

David Bowie photographed by Chris Walter in 1974. © RTNWalter / MediaPunch/IPX

It’s worth pondering this piece, not only because it’s wild inferences about influences on David Bowie’s music, but because the article reveals the form of some thinking of generic denialism takes. When genetic influences are expressed through adverse family circumstances, genetic denialists like Oliver James take them as evidence that genetic influences do not exist.

beans_on_toast430x300Behind all this is a lot of entrenched politics, not science. There used to be a group in Manchester, UK, Psychologists against Nazi Psychiatrists. While it appears to have dispersed, a number of members remain very influential at Liverpool University. Some of the faculty there like David Pilgrim still attack mental professionals as Nazis who are willing to entertain the idea that that there are genetic influences on abnormal behavior [see Pilgrim’s bizarre views in the comment section here]. British Psychological Society President Peter Kinderman (along with Bentall, also at Liverpool) drew heavily on Pilgrim’s comments without attribution and took the “Think child adversity, not genetics or else you are a Nazi” to a whole new level. I know, you have to see this bias before believing it is so strong

I also commented on Facebook:

I think that a lot of childhood abuse reflects environmental transmission of parental vulnerability to severe mental disorder. It’s not simply environmental or genetic, but an interaction between genes and environment.

Stay tuned

As I keep saying, there are so many dubious papers to critique, but so little time to blog. But I’ll likely be following up on the issue of childhood adversity. Likely prospects are a consideration of some new claims in the literature that the influence of childhood adversity on the brain has been traced traced. Another prospect is been on the back burner too long into consideration of the role of childhood adversity versus genes of development of psychosis and schizophrenia. This is extremely politicized literature were too much is being made of mediocre data. And what everybody seems to be ignoring is that even with the mediocre, highly confounded data that are available, the risk for psychosis associated with childhood adversity is about the same as smoking tobacco or marijuana.

On Twitter

There’s been ongoing discussion going about an article in Lancet Psychiatry article.

Huber CG et al. Suicide risk and absconding in psychiatric hospitals with and without open door policies: a 15 year, observational study The Lancet Psychiatry, 380 (2016) . doi:10.1016/S2215-0366(16)30168-7 1

With tweets such as Suicide Risk No Different on Closed and Open Psychiatric Wards

James C.Coyne@CoyneoftheRealm tweeted

Headline ignores that suicidal persons are not randomly assigned to locked and unlocked wards. Stop distorting.

 Not sure what your point is, but evidence for means restriction in reducing suicidality is quite strong.

Acutely suicidal persons left in unlocked wards is stupidly irresponsible. Low quality data don’t convince otherwise.

Unfortunately, the article is behind a pay wall and Lancet Psychiatry is still not listed in PubMed, restricting access. But I do have access to the abstract and the confidence intervals for death by suicide are huge, (OR 1·326, 95% CI 0·803–2·113; p=0·24).

So why am I offering an opinion on an article that I haven’t yet seen?

A fair question, to which I respond – I’m taking the risk of relying on prior probabilities.

I have studied suicide in Germany, where the study was conducted, including some intervention studies. The low frequency of deaths by suicide means it’s difficult to show anything influences suicide. This study is observational, not an RCT. Everybody tweeting about it seems to assume that propensity scores can overcome the problems of not having RCT data. Propensity scores are attractive, but poorly understood by most people who use them. They are vulnerable to on measured confounds. They require dropping data for patients for whom a match can be found, in this case patients hospitalized in locked versus unlocked wards.

I definitely will be blogging about this, after putting out a request for the article on Twitter.

 Other things I’m saying on Twitter James C.Coyne@CoyneoftheRealm

#2016APAPREZ #APA2016 Will you apologize to innocent victims of psychologists’ involvement in #torture?

We need to recognize that communicating about evidence impossible in some conversations – or terribly inefficient. pic.twitter.com/VtRdihYqyc

Remarkable how academic institutions remain silent about alleged scientific misconduct by some of their researchers. http://health.jotwell.com/restoring-the-integrity-of-the-pharmaceutical-science-record-two-tales-of-transparency/ …

Imagine testing yoga vs acupuncture (AC) in cancer pts receiving chemo, no control group, concluding yoga not < AC. twitter.com/drs_reynolds/s…

US law allows 72 hour hold. Can only be extended if qualified professional observes active risk to self or others.

Szasz had libertarian views about not locking up acutely suicidal people. Only treated neurotic, nonsuicidal patients in long-term therapy

Suicide ideation poor surrogate outcome for services research. Too nonspecfic, easily changed w/o outcomes affected.