This is an exploratory Factor Analysis. It is a model: a simplification. It is not reality. As Sweden has both a full clinical database and full criminal database and family relationships are known, the familial linkage can be ascertained for each disorder.
The simplest (one factor) model would be Shared Factor plus unique and Shared environment plus unique, but the authors suggest a six factor solution: three genetic factors, one environmental, and two unique factors.
From the results:
As hypothesized, all disorders and violent criminal convictions loaded in the same direction on the general genetic factor (range 0.31–0.60), indicating that all conditions partly shared the same genetic origin. Aside from the general genetic factor, schizoaffective disorder (loading=0.67), schizophrenia (loading=0.56) and bipolar disorder (loading=0.40) loaded together, indicating that psychotic problems shared a genetic pathway independent of the general factor. The second genetic subfactor included loadings on drug abuse (loading=0.65), alcohol abuse (loading=0.51), violent criminal convictions (loading=0.47), ADHD (loading=0.46) and anxiety (loading=0.39). Although the strongest loadings were on typical externalizing problems, we interpreted this factor as non-psychotic problems because anxiety also loaded on it. The non-shared environment factor, which we interpreted as mood problems, included loadings on major depression (loading=0.86), bipolar (loading=0.72) and anxiety (loading=0.46).
I have some difficulties with this. Currently, we have over a hundred candidate genes for schizophrenia from increasing large Genome wide association studies (GWAS). As a colleague pointed out to me today, if you need studies involving hundreds of thousands of participants (which is the size of the GWAS studies that have results) then you are probably wrong. The clustering of disorders — separating psychoses from others and having a neurosis component is not unusual, and the signficant variation in rates of depression from one nation to another would fit with a suggestion that there is more environmental contribution than genetic for these disorders.
But this is a model. It is not a mechanism, less so a classification. With mathematics, particularly exploratory factor analysis, all results have to be taken with a skepticism. It may, however, point us away from overly complicated classification, and back to the ancient suggestions that we are mad, acedic, choleric or felonious.
The diagnostic categories of psychosis are so big, fuzzy and overlapping that research has been inconclusive for decades – in private the experts on schizophrenia c1980 would all say it was *many* conditions, not unitary – but in public, when it came to grants and publishing, it was a different story.
The two most interesting papers on schiophrenia in recent years that I have found most interesting are:
The failure of the schizophrenia concept and the argument for its replacement by hebephrenia: applying the medical model for disease recognition. Michael Alan Taylor, Edward Shorter, Nutan Atre Vaidya and Max Fink
http://onlinelibrary.wiley.com/doi/10.1111/j.1600-0447.2010.01589.x/full
Which suggests that Kallbaum’s original hebephreia was a much more coherent category than Kraepelin’s Dementia Praecox (the damage done by Kraepelin’s ‘lumping togther’ tendency, amplified by the Neo-Kraepelinians of the late 20th century, would be hard to exaggerate!)
and
The incidence of admissions for schizophrenia and related psychoses in two cohorts: 1875–1924 and 1994–2010. David Healy, Joanna Le Noury, Stefanie Caroline Linden, Margaret Harris, Chris Whitaker, David Linden, Darren Baker, Anthony P Roberts
http://bmjopen.bmj.com/content/2/1/e000447.full
Which suggests that schizophrenia (?hebephrenia) may have been a disease arising with the industrial revolution, which is now dwindling – this may give clues to its aetiology.
Hope you enjoy!
I have a book somewhere (by Torrey) which notes that the number of madmen increased rapidly during the industrial revolution. The wards, built bigger than the planners thought they would need, became overcrowded.
Catatonia is now very, very rare. I think I see one or two cases every year, at most. Much more common is hallucinations and delusions, particularly of persecution, with less organization, and often an affective component.
Which I currently call “psychosis NOS”.
I don’t think any one is that happy with the current DSM: the hypomania/mania split, the issue of affective elements in psychosis, and the lack of acknowledgement that we don’t know at first presentation what we are dealing with are all problems.
However, we have not a replacement, nor will we, until we understand the biology of these things. Current theories relating to synaptic networks have more face validity than the dopamine/receptor hypothesis.
But we have been wrong before. My skepticism remains.