Advanced Maternal Age Associated with AIS?

[Pooka1]

How do they define concordance in mz twins? Do they both have to be affected and that's all, or do they have to share the same curve pattern/severity as well? I know some strange things can happen in mz twinning, but, at least in females you have to consider X-inactivation. If any part of the combination of genes that contribute to the scoliosis is carried on the X chromosome, that could account for the <100% concordance in females, correct? However, straight X-linked diseases would primarily affect males (as was not made clear in my statement about my family hx). Does anyone know if they separate the males from the females in conducting mz twin studies? I think that could be useful information.

Excellent point. Lyonization (activation of different x-chromosomes in each twin) has been documented to account for the presence of a genetic disease in one identical twin and not the other.

I am not so sure how common it is, though and may not be common enough to necessitate throwing out all female identical twin data.

Ogilvie's suggestion about twins studies possibly being inappropriate for scoliosis studies (or any midline disorder) is a point well taken in my opinion. All these twins studies may be dead on arrival. Most published research results are false.

[rohrer01]

I don't think female studies should be thrown out but just compared to the males. It may give us some indicator if there is a prevalent mutation carried on the x-chromosome. I'm sure they've already thought of that as a component. I have a hard time keeping up with the big dogs. ;-)

I'm finding it difficult to wrap my head around why most affected individuals are female.

[hdugger]

I'm finding it difficult to wrap my head around why most affected individuals are female.

I *think* what's going on is that males and females may have the same tendency to curve. But males, because of some masculine characteristic like greater muscle mass or denser bones, may curve a little but don't tend to progress.

I think that's somewhat supported by the numbers. Little curves show up in roughly equal numbers in males and females, but females tend to progress.

If that's true (and it may well not be) one would expect that increasing those characteristics (muscle mass and bone density) in females would also protect against progression. But that's entirely a thought process, unsupported by any research.

[Pooka1]

If anyone has that Danish Twins study, we might cut to the chase the GIGO level on zygocity at least (not incidence of scoliosis though) if the ratio of identical to fraternal is far off 0.3 (assuming most of these twins were not born in the IVF error in which case the ratio goes down because mono- is constant but di- goes up).

[Pooka1]

It actually makes me wonder whether or not the one percent of AIS cases they've identified as not being effective with any type of treatment is actually some other form of scoliosis such as juvenile, which seems to have a much more determined progression towards surgical levels.

That's a good point. As a percentage, is JIS about 1% of AIS?

But also recall Dr. Hey's comment about bracing success with JIS. Plus the majority of bracing successes in this small sandbox are JIS. By far.

[skevimc]

But that is the reason the self-reporting is flawed. Many people have scoliosis (Cobb > 10*) and never know it. It is clearly GIGO.

...

I believe I can convince you they are contradictory. Consider this article:

http://www.ncbi.nlm.nih.gov/pubmed/9306532

So all these twins presumably have severe scoliosis because they are in the clinic per your suggestion. And yet the ratio of monozygotic to dizogotic is greater than one when out in the general population the "natural" ratio before IVF was ~0.3 and would only go DOWN from there in response to IVF. And yet it is greater than one among the patients here.

It's a small patient population but this is actually more proof beyond the GIGO argument (which is the strongest argument by far) because the ratio of these test subjects is way off what would be expected by their presence in the general population.

I'm assuming GIGO means garbage in garbage out?

You're correct that self-reporting has flaws. But so does strictly clinical selection.

It's not my suggestion that the cases in clinic are more severe. I'm stating that they ARE more severe. The meta-analysis you posted confirms that. So the patients being counted in these studies, not only have scoliosis, but are being treated for it. Only 4 pairs of the monozygotic twins both had curves <20° (none of the dizygotic). The remainder had at least one, and most were both, curves >25°. So this is a decidedly biased group toward more severe curves. So the concordance rate being higher for more severe curves makes sense to me in light of the scoliscore data. The more severe progressive curves have a stronger genetic influence.

The Danish twin registry is one of the largest in the world. They screened >22k pairs of twins. Mono:di is 0.48. (Although I’m not entirely sure what your point here is.) They screened out respondents where only one twin responded, twins older than 50 (to eliminate adult onset) and twins who said they were diagnosed <10 years old. Without a doubt there were some twins that were screened out because both twins didn’t respond and/or they had a curve small enough to never be diagnosed and so never counted.

So it seems to me that the clinical data represents a worst case scenario by representing only those who need treatment. While the Danish study represents the best case scenario of the general population of curves found in twins and probably under-reports the number of cases. I’d agree that the Danish study is severely under-representing the dizygotic twins.

In light of the scoliscore data, both studies seem to be in agreement, whether they agree with that or not. More severe curves are under heavy genetic influence and less severe curves potentially have less genetic influence. So over sampling the severe curves decidedly biases the results toward concordance. And the less severe curves or curves small enough to go unnoticed, which are potentially less genetically influenced and thus never reported, heavily biases the data toward discordance.

[Pooka1]

So the concordance rate being higher for more severe curves makes sense to me in light of the scoliscore data. The more severe progressive curves have a stronger genetic influence.

Why do you say this? What is the evidence for this?

The Danish twin registry is one of the largest in the world. They screened >22k pairs of twins. Mono:di is 0.48. (Although I’m not entirely sure what your point here is.)

I'm very surprised it is high not low. I think there might be some reporting issue there per se. The ratio ~0.3 is the natural ratio as far as I know prior to IVF. In my experience running into and talking to other parents of twins, it is my non-expert opinion that there is a significant amount of identical denialism for reasons I will mention only if asked. I will say that it is my impression that more people deny identicalness than accept it in my personal experience.

But getting back to the issue at hand, any deviation from ~0.3 in that large a group is going to directly demonstrate GIGO in terms of self-reporting zygosity. Or maybe ~0.3 is wrong. But I will say it was in every twin book and reference I have read. They all might be referring back to the same study maybe but I wonder.

I’d agree that the Danish study is severely under-representing the dizygotic twins.

This is very strange in my opinion and needs an explanation. The natural ratio is NOT half... fraternal twining is mush more common even before IVF. So if some of the fraternal twins are erroneously reporting themselves as identical which seems to necessarily be the case, then the concordance rate in fraternal goes up. SO that might bring the ratio right but there is still the scoliosis incidence under-reportage issue.

More severe curves are under heavy genetic influence and less severe curves potentially have less genetic influence.

Why do you say this? What is the evidence that all curves aren't under equally high genetic control?

[Pooka1]

And by the way...

If the concordance rate for mono- is 13% and the concordance rat for di- is 0% then the concordance rate approaches infinity in the limit. Or it becomes progressing undefined, take your pick. (smiley face)

Either way, "It's all genetic."

[Pooka1]

Oh actually I think I can explain why the mono-/di- ratio is too high in the study and why that affected their results...

I think if both twins had scoliosis and they looked at least somewhat alike that they simply assumed they were identical. That would explain the wrong ratio and the 0% percent concordance in di- when no other study has ever come close to replicating that.

The other studies group together and they group AWAY from this Danish Study. When that happens, it is often a result of different methods which in this case you actually mentioned (severe bias in clinical versus mild bias in self reporting/GIGO). In my field when there is one aberrant report it is usually mentioned and then dismissed as I have seen researchers do with this Danish study. It's very annoying when a method flaw gets by the reviewers and then all comers have to constantly bring the paper up only to shoot it down. That appears to be what is happening with the Danish Twins study as far as I can tell and that may not be far given how far afield this is from my field. That ratio is wrong and thus there is likely a math error that appears to have carried through to the conclusions.

[skevimc]

Why do you say this? What is the evidence for this?





Why do you say this? What is the evidence that all curves aren't under equally high genetic control?

I'll try to get to your other points but wanted to comment on this one before I left work.

The evidence I'm using is the Ward paper (scolioscore). Patients with fewer genetic traits, i.e. low scoliscore, are seen as having, and in fact seem to, have a low risk for progression beyond 40°. Those with more genetic traits, the opposite is true.

[Pooka1]

I'll try to get to your other points but wanted to comment on this one before I left work.

The evidence I'm using is the Ward paper (scolioscore). Patients with fewer genetic traits, i.e. low scoliscore, are seen as having, and in fact seem to, have a low risk for progression beyond 40°. Those with more genetic traits, the opposite is true.

Wait a minute. There is a difference between being controlled by more or less genes VERSUS various degrees of the condition being more or less genetic. Do you see that distinction? You can have ONE genetic trait to use your terminology and that is COMPLETELY genetic as in Down Syndrome.

And there may be "protective" genes in the case of the non-progressors that come along with the package like endogenous retroviral insertions come along so well with certain genes that they a constitute a knock down argument for evolution of humans and other apes from a common ancestor all by itself.

Or Scoliscore could be measuring genes that coincidentially go along with the actual controlling genes. They wouldn't know if that was or wasn't the case as I understand it and maybe I don't.

Do the Scoliscore folks state what you did in those terms? I have never come across a statement of theirs saying the more progressive cases are more genetic. It's all genetic.

[Pooka1]

Plus out of the markers, maybe most have a small control and a few have most of the control in both progressors and non-progressors. Did they rule that out? How?

[skevimc]

Wait a minute. There is a difference between being controlled by more or less genes VERSUS various degrees of the condition being more or less genetic. Do you see that distinction? You can have ONE genetic trait to use your terminology and that is COMPLETELY genetic as in Down Syndrome.

And there may be "protective" genes in the case of the non-progressors that come along with the package like endogenous retroviral insertions come along so well with certain genes that they a constitute a knock down argument for evolution of humans and other apes from a common ancestor all by itself.

Or Scoliscore could be measuring genes that coincidentially go along with the actual controlling genes. They wouldn't know if that was or wasn't the case as I understand it and maybe I don't.

Do the Scoliscore folks state what you did in those terms? I have never come across a statement of theirs saying the more progressive cases are more genetic. It's all genetic.

OK. I see your point. As far as actual function of the genes are concerned, no, nothing is known about the role they play in scoliosis progression. It's based solely on a regression formula. So in a sense, it's not all genetics, it's really just statistics. Because correlation does not equal causation. But serisously... In the list of genes they used for the formula they do have an indication of what they say is protective and non-protective (I forget the actual words they used). I couldn't find anything written as to how they determined that. Since everything was based mostly on the regression equation, I'd imagine the values were based on the coeficient. So if something was negatively correlated they termed it 'protective' and if something was positively correlated it was causitive or whatever they used. (I'll look it up later).

The idea that the genes postively AND negatively regulate the curve is certainly a hypothesis. Although I can't imagine being able to actually test that. To test the complex interplay of 53 genes and 6 or 8 gene to gene interactions with enough power to confirm the positive and negative regulators would be exceedingly difficult (to put it mildly). Especially when adding in the extra layer of which genes are actually being coded into usable proteins. Regardless, I do see how what you're suggesting is possible but I think it's probably one of those "how many birds are in flight at any given moment" kind of things.


Let me ask, what role do you see the environment, defined however, playing in eitiology and progression?

[skevimc]

Actually I'm also interested in what the people on the environment side of the argument see the role that genetics plays in the eitiology and progression.

[hdugger]

I'm not sure there is a straight "environment" group - at least not that I've heard. Most people assume there's some genetic component and some environmental component.

i'm half-and-half. I assume that there has to be a genetic role, just because there are so many physical similarities between these kids. So, they have to share some gene for body type. And then some gene, in varying degrees, that controls joint hypermobility.

Beyond that, I can't tell how much is written into the genes and how much is environmental.

Do the genes tell the spine to start curving? Based on the scoliscore test, I'd say no, because even the low scoring kids get that message. And there are lots of high scoring kids who don't start to curve at all.

So, I think, the genes must control the things that play a role in the spine collapsing. Maybe they make the bones weaker, or the muscles imbalanced, or some other part of the process.

That puts the environment in control of the initial trigger - the thing that makes the spine go wrong in the first place.

And, I assume, one can intervene environmentally at any point and try to counteract the activity of the genes (i.e., make the bones stronger, balance the muscles, etc)

I always think of big curves as a perfect storm. You have to have the tendency, plus maybe some environmental condition, plus some kind of asymmetric physical activity. The more you have of one thing, the less you need of the other. So, kids who score 200 probably need very very little to push them over the edge, as do kids with odd environmental conditions (rickets, for example), or I assume kids with some very imbalanced muscular condition.