Dear Neil

I have searched a few chapters and a couple of books for you to read and understand schizophrenia. I hope it will help you to find your answers.

Article one

Article number two is a chapter from a book

Third one is a handbook.

fourth one is abook on Molecular Biology of Neuropsychiatric Disorders.

Thanks & best regards

Thanks, Sibtain. I appreciate your input to our discussion.

Before we get into detail over this, could you please explain what the uses of this statistical formula will be? Surely it is never going to be accurate enough for prediction in an individual case, even if if were thought desirable to tell an adolescent say that he had a 50% chance of developing schizophrenia in the next 30 years.

Anthony,

Excellent question. I had in mind that the formula could be used potentially in several ways: 1) prodromal stage - likelihood of individuals developing full-blown sz, and creating treatment plans accordingly, 2) active phase -measuring the severity of symptoms on several scales (eg, positive, negative symptoms), again informing research and treatment, and 3) chronic phase - likelihood of recovery and/or participation in social and vocational activities.

Any model created would be, as you pointed out, probabilistic and only a guide with regard to its application to any particular individual. My hope would be finding a way to move away from a purely subjective assessment of acuity and prognosis toward a more empirial approach.

"Is there a multifactorial (multiple genes, epigenetic events) model to explain the onset of schizophrenia"

If I have understood your model, it could explain the liklihood of getting schizophrenia, but would not vary with the stage of the disease, thus not predict its onset nor prognosis.

Anthony,

I think you are correct - as stated in my original question, some of the factors (eg, genetic) may not change much over time, although the epigenetic ones might well change over time as the genetic code becomes activated/suppressed by negative or positive environmental experiences. So if we had the means to create a dynamic, interactive model, then it could serve in a probalistic role for all stages of the disease.

But I'm not a geneticist, so getting a bit out of my depth here. I'm hoping that some others with specialized knowledge in this area can contribute to the discussion.

Schizophrenia tends to come in acute episodes, sometimes with complete recovery, so I don't see how this variation can be explained by epigenetic events, ie How to explain schizophrenia offset. Some genes for schizophrenia have been identified, but I think it will be a long time before the epigenetic mechanisms are identified.

Anthony,

We've discussed the issue of diagnostic reliability among the psychotic disorders and this certainly applies to sz. I know that there are acute episodes in sz patients, but the ultimate causes of these exacerbations are probably quite varied - everything from medication non-adherence to environmental stressors to comorbid substance use. All of these factors (plus probably a dozen more I didn't mention) can cause or increase the probability of an acute episode.

And making the picture even cloudier is that different patients have higher or lower thresholds for handling stressors before decompensation occurs. This resistance or lack of resistance is the interaction of genetics and environment, the epigenetic component. I have not done a lot of research searches lately on the genetic-environment interaction, but I believe progress is being made linking certain genetic markers to psychiatric symptoms and behavior. Do you agree?

"I believe progress is being made linking certain genetic markers to psychiatric symptoms and behavior. Do you agree?"

Rapid progress is certainly being made, but I suspect that there are thousands of genes that can lead to hallucinations, for example.

"All of these factors (plus probably a dozen more I didn't mention) can cause or increase the probability of an acute episode."

I fear it may turn out that chance, or some undetectable environmental changes, are major triggers of attacks (as opposed to the illness itself).

Many epidemiological studies have linked diverse infections, both prenatally and in adulthood with schizophrenia http://www.ncbi.nlm.nih.gov/pubmed/?term=infection+schizophrenia. Recent data from the Danish register showed that about 45% of recently diagnosed schizophrenics had a hospital contact with some kind of infection prior to diagnosis - a large effect that did not seem to related to any specific pathogen. Nielsen et al, 2013 http://www.ncbi.nlm.nih.gov/pubmed/24379444

Many SZ genes (particularly from the MHC region in GWAS) Corvin and Morris 2014 http://www.ncbi.nlm.nih.gov/pubmed/24199664 are related to the immune system and many are also related to pathogen life cycles (toxoplasmosis/ herpes and others) http://www.ncbi.nlm.nih.gov/pubmed/23913659,23533776,18552348 suggesting a gene/environment interdependence

Infections are impossible to predict, so I doubt whether any statistical formula would be able to factor this in. Some are however preventable or treatable and perhaps the answer lies in identification and prior detection of the major culprits.

Chris,

Very interesting set of studies about the role infection may play in the etiology of sz. What do you think of the possibility of identifying particular alleles related to the operation of the immune system that might contribute to the probability of the onset or acuity of sz?

Neil

I guess this needs a combined GWAS/SZ/seroprevalence study preferably with detection methods for multiple pathogens. My guess is that SZ risk genes would be slightly different for each pathogen, but impacting on the same pathways. It would be useful to take larger blood samples to try and correlate seroprevalence and inflammatory and other biomarkers with the genetics, but this is difficult as each different study currently has to have ethical consent and can't be done post hoc with freezer samples. Many of the pathogenic insults may also have been prenatal, and genetics would have to be correlated with maternal history. Not easy given the many objections to access to medical records (except in Denmark !). Although most agree that gene/environment interactions are important, this hasn't really been addressed much in large scale studies, although this is probably in progress.

There have been a few studies on the microbiome in autism and multiple sclerosis and perhaps someone will try this in schizophrenia and related disorders.

"My guess is that SZ risk genes would be slightly different for each pathogen"

My guess is that schizophrenia is not related to any pathogen, nor to pathogens in general, but to infection in a specific part of the body (eg middle ear), or to some non-specific concomitant of infection, like fever.

I also agree Chris Carter's hypothesis because I gathered publications which make believe that schizophrenia can be due to a pathogen, especially a bacteria.

And Indeed an import problem is that if the pathogen is different for different kind of schizophrenia, the way to find a treatment for them will be very difficult.

here is all the document I created about pathogen and schizophrenia:

http://www.wiki-theory.com/index.php5?title=Bacteria_infectious_hypothesis_of_schizophrenia

I guess different pathway are impaired in schizophrenia: like reduced autophagy, increased kynurenic acid, and eventually different gene methylation. I guess in my document that all this change are mainly made because of toxins released by the bacteria.

In your attached document, you say you included all medical publications consistent with your hypothesis that a bacterium is involved in schizophrenia.  But you should mainly be looking for, and including, all studies that refute your theory.  I suspect there are enough of these epidemiological studies to make it not worth while to directly test your theory.   For example, why is concordance no greater for dizygotic twins than for siblings?

Yes indeed, I included all the medical publication I found consistent with my hypothesis of a bacteria involved in schizophrenia.

I'm also interested in publication that refute my theory, even if I don't known lot of facts that refute my theory or only facts that are difficult to explain with my theory. It would be an interesting work to gather also this facts that refute my theory but I find it more complicated than gathering facts that supports my theory, I don't known what keyword I should use to find such publications in pubmed.

you say "why is concordance no greater for dizygotic twins than for siblings", I don't clearly see why the concordance would be higher for dizygotic twins than for sibling guessing my hypothesis is true? did you mean monozygotic twins?

"", I don't clearly see why the concordance would be higher for dizygotic twins than for sibling guessing my hypothesis is true? did you mean monozygotic twins?"

Twins are more likely to catch the same infection than siblings, especially if occurring in the womb.  I think someone used this argument in the past to disprove their own infectious theory.

Thanks for your answer.

In the document I created, and if my theory is true, there is not enough publication which would permit to known when happen the infection:

Whether the infection happens in nearly everyone, and in non schizophrenic patient the toxin is blocked.(As monozygotic twins concordance is not 100% but about 50%, if the infection happens in nearly everyone, we can assume that sometimes for the same genome the immune system succeed to fight the toxin and sometime no)

Or if the infection happen during pregnancy, childhood, or adulthood

Or sometime during pregnancy, sometime during childhood, sometime during adulthood.

An other thing whom there is not enough publication, is that the same schizophrenic patient can hold a soup of pathogen (I mean lot of different pathogens) or maybe it's not the case.

To come back to The Neil's question:

I found few month ago this document which I only succeed to access to the abstract, It tells there would be 8 different kinds of schizophrenia, based on other genome wide association study:

http://www.ncbi.nlm.nih.gov/pubmed/?term=25219520

You can also find some genome wide association study about schizophrenia on pubmed telling mainly (if I well undestood them) that the most significant association is found in the MHC part of the genome, and secondly in genes related to some neurotransmitters.

There is also this study:

http://www.ncbi.nlm.nih.gov/pubmed/?term=24402055

Which seems to have find the differences in methylation between schizophrenia and controls.

And finally:

http://www.ncbi.nlm.nih.gov/pubmed/?term=25113377

This document tells that the transcriptome is changed in schizophrenia patients compared to control, I cite the abstract "Pathway analyses of these genes indicated that they are involved in the lysosome, Fc gamma receptor-mediated phagocytosis, regulation of actin cytoskeleton pathways, along with several cancer pathways"

But to my knowledge there is no model which permits to tell depending on methylation, on genome, on the environment,  what is the probability to become schizophrenic. But it comes closer, even if we are far from that, If I remember well, in some Genome wide association study, they succeed to tell a probabilty to get the disease during the whole life depending on the genes you have, and some people would have up to 70% to get the disease.