A Genetic Epidemic of Down Syndrome?

Over at Left Brian Right Brian, Sullivan (aka Matt Carey) is attempting to make an argument that there can be a genetic epidemic.  He says -

It’s a phrase that is heard a great deal in online discussions about autism: there are no genetic epidemics. Genes don’t change quickly enough for a genetic condition to see an increased prevalence over a single generation, right? Well, yes and no.

Before going on to suggest Down Syndrome might be a model for how a genetic epidemic might work.  But while it is true that Down Syndrome is caused by an underlying genetic mutation and that the number of people with Downs is increasing, it does not classify as a "genetic epidemic".

The main reason is that for the most part¹ Down Syndrome is not an inherited condition.  A child does not have Downs because their parents passed along any sort of mutation.  Downs is caused by a mistake in cell division during the development of the egg, sperm, or embryo.  So while there is a genetic mutation responsible for the condition, the mutation is caused by some non-genetic factor.

Down syndrome (again, for the most part) doesn't run in families.  Having one child with Downs only slightly increases your chance of having a second child with Downs.

And a person with Downs in unlikely to pass the condition along to another generation because most people with Downs have problems with fertility.  Men with Downs are thought to be sterile and women are only fertile about a third of the time.  Although, if a person with Downs has a child, there is a significant chance that their child will inherit their condition.

So while Downs might be caused by genetic mutations (aka is "genetic"), it is not a "genetic epidemic" because the increasing number of people with Downs is not due to the mutations becoming more common in the gene pool.  What is causing the increasing number of people with Downs is due to a few social changes.

First, there is an ongoing trend for women to have children later in life and maternal age is a demonstrated risk factor for Downs (although, as a side note, I have never read a good explanation as to why exactly that is the case).  The risk of having a child with Downs increases with age, going from 1 in 1,600 when a mother is twenty to 1 in 1,000 at thirty and 1 in 90 by forty.

Second, people with Downs are surviving longer.  Down syndrome comes with a variety of related health problems and these problems can shorten a persons life.  About eighty years ago, people with Downs often didn't survive past their tenth birthday.  Over time that age has gradually increased so that today people with Downs can expect to live to 50 and beyond.

Finally, people today seem to be more willing to have a child with Downs.  Abortions of children with Downs is a contentious subject, but as life expectancy and quality of life for a person with Downs has been improving, there seem to be fewer abortions.

These three factors can explain the modest increase in the number of people with Downs that has been seen over the past several decades.

So, back to Sullivan's original argument, there is no "genetic epidemic" of Down Syndrome.  There isn't really even an "epidemic" of Down Syndrome - the growth has been modest and is what would be expected given the social trends listed above.

More importantly, there aren't that many similarities between Downs and autism -

In autism, there is no single mutation that can account for more than a small fraction of cases and yet having one child with autism does significantly increase the risks of having another child with autism.  Down Syndrome centers on a mutation on one chromosome while autism has been tentatively linked to many places across our genes.  So, from the little we know about the genetics of autism, the underlying genetics of autism and Down Syndrome are quite different.

Autism does not automatically come with medical conditions that shorten life expectancy.  And there is no prenatal test for autism² and so there are no widespread abortion trend that is reversing.

Autism is growing much, much faster than Down Syndrome.  Over a twenty four year period (1979 to 2003), the number of people born with Down Syndrome grew by 31.1%.  Over the same time period, the number of case of autism grew from about 2-4 per 10,000 to 66 per 10,000 or about 1500%³.  If you take a more conservative time period (2000 to 2008), autism has grown about 160%.  Anyway you slice it, autism is growing significantly faster than Downs.

The only real link between Downs and autism is parental age, and that is a nebulous one at best.  Some studies of autism involving parental age show a small relationship with maternal age but not paternal age, some show a relationship with paternal age and not maternal, some show a relationship with both, and some show no relationship to either.

At the present time, the best that we can say is that there might be some relationship between maternal or paternal age and risk of autism.  Compare that to Down Syndrome where there is a clear link between maternal age and the risk of having a child with Downs.

So, not only is Sullivan wrong to suggest that there is a "genetic epidemic" of Downs Syndrome, he is wrong to suggest that there is any epidemic of Downs at all or that this tells us anything about what is going on with autism.

1. There are actually three different known types or styles of genetic mutations that can cause Down Syndrome. The most common one, trisomy 21, accounts for more than 90 percent of cases and is almost never inherited.  Another form, which is almost not normally inherited, involves a mutation of chromosome 21 where only some of the cells have the mutation and others don't.  The final form, called translocation Down Syndrome, can be passed from carrier parents to children.  However, only about 4% of Downs cases are these type and only about half of these cases involve inheritance.  Source

2. Well, technically, there are prenatal tests for some known genetic forms of autism, such as Rett Syndrome and Fragile X.

3. A large part of the growth in autism from the 1980 to 2000 is probably due to the condition being added t the DSM III and then its changes in the DSM IV.