In recent years, I too have started down the genealogy path, discovering my industrial and military roots going back over a hundred years or so. I don't know whether age had anything to do with my desire to undertake this search; that and the realisation that we all end up as old photographs and historical facts and figures in a shoebox eventually, but nonetheless the detective work began.
I wish I was able to tell you that Royal blood courses through my regal veins and I am something like the 1,023 person in line to the British throne. But alas, at least for the moment, all I know is that my relations were apparently all hard-working manual labourers or military types and not in anyway Landed gentry. Some of them actually seem to have met rather unfortunate ends it has to be said.
Why the focus on the family history?
Well, today's post concerns an interesting piece of research by Patrick Sullivan and colleagues* (full-text) looking at familial history of bipolar disorder and/or schizophrenia as a risk factor for autism. The authorship group includes a name, I believe, which has already featured on this blog, Dr Cecilia Magnusson and her very interesting research on migration and autism (see here).
I've talked before about the possibility of various 'intersections' between conditions like autism and schizophrenia. So for example, the appearance of autism or autistic features in cases of schizophrenia (see here) and similarities in some cases with regards to gastrointestinal and immune-related findings (see here). Indeed specifically with the relationship between autism and schizophrenia in mind, the work of the late Curt Dohan seems always to crop up, on a potential role for diet, gluten and casein components of diet, in cases of the two conditions. I will however, for now, just put that to one side, apart from this letter which I will come back to at a later date.
The paper by Sullivan and colleagues is open-access so no real need for me to go through it with a fine-toothed comb. A summary however:
- This was really three studies all rolled into one based on the examination of various registries held in Sweden (registry #1 and registry #2) and Israel (registry #3).
- Cases of autism spectrum disorder (ASD) were initially picked out from the registries based on the ICD criteria. ASD cases were matched with various asymptomatic controls based on several parameters including age and year of birth; from the Swedish registries at a 1:10 ratio in a case-control fashion. The actual total number of people with ASD included for the entire study numbered above 30,000 cases.
- The rates of schizophrenia and bipolar disorder were inspected among parents of ASD cases (registries #1 & #2) and siblings (registries #1 & #3).
- Results: based on the Swedish registry data (#1 and #2), the "exposure of schizophrenia in parents" was associated with odds ratios (ORs) of 2.9 for a diagnosis of autism.
- Similar 'exposure' of schizophrenia in siblings (#1 and #3) was also linked to an increased risk of autism; the ORs ranging from 2.6 to a staggering 12.1 (#3). Indeed bearing in mind the smaller ASD group included for study based on the Israeli registry (#3) (n=386), the 95% confidence intervals were reported as 4.5 - 32.5.
- When analysing for any potential effect from comorbid learning difficulties (LD) in ASD cases, the results suggested exposure to schizophrenia in parents and/or siblings seemed to be more powerful in cases where no LD was present.
- The data on familial bipolar disorder, whether parent or sibling, also suggested some increased risk of autism; although not to the same extent as with schizophrenia.
The authors seem to be pretty confident in their findings. To quote: "The findings were clear. The presence of schizophrenia or bipolar disorder in first-degree relatives was a consistent and significant risk factor for ASD in all 3 samples". I have to say that allowing for potential issues with the diagnostic criteria used and data collection methods, I'm inclined to agree with their findings.
Bear in mind however some important details reported in the study with for example, regards to the effect of learning disability and dare I say, a degree of protection that it seemed to confer with regards to familial psychiatric history. Does this mean that those cases of autism and comorbid learning disability might have a slightly different pattern of aetiology from the autism cases without LD? I must also link to this press release associated with the study and quote a sentence: "Most people with a family history of one of these disorders actually get nothing - the vast majority in fact".
Then to the question of why, why in some cases there may be a link? As much as I like to tug on the loose threads of the genetics research on autism and schizophrenia as a function of the 'genes are everything' mentality that some people hold, one has to allow for the fact that genes must show some involvement in the current study. I will perhaps temper that last statement by pointing out that the growth in the gene*environment viewpoint and the increasing interest in epigenetics are potentially as important as the 100% heritability argument. I assume for example, that most parents and siblings lived in the same household with the participants identified with autism and so their environmental exposure patterns would be considered similar through things like household environment, drinking water, food and diet, pathogen exposure, etc. Bearing in mind also the possibility of other factors being potentially related to schizophrenia/psychosis such as T.gondii exposure, anti-gliadin antibodies, etc.... too much?
To end some Top of the Pops 1980s Gold. Level 42 and Running in the Family. No pun intended.
* Sullivan PF. et al. (2012). Family history of schizophrenia and bipolar disorder as risk factors for autism Archives of General Psychiatry DOI: 10.1001/archgenpsychiatry.2012.730