Many genetic tests just uncover one small bit of
your disease risk formula.
Image courtesy of Tungsten

Since 23andMe is offering data on Alzheimer’s risk but only with an upgrade, I decided to get that upgrade. I got the new results last week and opened it right up. There was a new Alzheimer’s link with a little lock on it.

The lock is to keep me out until I feel ready to deal with the news. Well, I’m ready…

When I opened the lock, I was given information about one gene, APOE. This was pretty disappointing. In my case, I was able to guess my APOE status with the old information.

I understand that for people who couldn’t guess their APOE status before that these new results might be useful. But I expected more. Lots of genes all impact our risk of getting Alzheimer’s and scientists are making real progress figuring these out.

What I was hoping to see was a compilation of all the information out there. See, my APOE data suggests that I am at a lower risk. But what if all my other genes are of the wrong type and put me at a higher risk? Then I might be at a higher risk than I think I am.

And my situation is better than people who have APOE gene versions that put them at a higher risk. What if all of their other genes put them at a lower risk? Now they are worried for no reason.

This points to one of the big issues with genetic tests and complex genetic diseases. Right now we know a bit about some of the genes involved but we don’t yet know the whole story.

It is kind of like having a whiteboard with a vast, awful, Jimmy Neutronesque formula scrawled across it. This is our genetic risk for getting something like Alzheimer’s.

Unfortunately, a blanket of ignorance hides most of the formula from us. We can only get a glimpse of one small part of it. What we then try to do is figure out our risk from that glimpse.

So let’s say we manage to uncover a bit of the blanket in the corner and we see a 0.5X. Does that mean that I am half as likely to get Alzheimer’s? Hardly. We have managed to see a piece of the whole formula but the rest remains hidden.

My APOE status is really just a bit of the formula for my Alzheimer’s risk peeking out from under the blanket. I don’t know the rest of the formula and so don’t know the rest of the risk.

APOE status is a well-studied risk that we know a lot about. We have completely uncovered it on our whiteboard. Some studies are staring to reveal other parts of the board but they aren’t completely clear yet. This is probably why 23andMe left them out.

I decided to look at the hazy parts of my equation using good old SNPedia. Remember, they have lots of information about DNA variants and disease that isn’t as filtered as is 23andMe’s.

Here is what I found when I put my data from 23andMe through SNPedia. I have listed the SNP, what I have at the SNP (the allele), and my risk for Alzheimer’s for those that give me a higher or lower chance for Alzheimer’s. I haven’t put in the ones that have normal risk.

I came out very well here. I am at a much lower risk for getting Alzheimer’s when I look at these SNPs. Yay!

But what if I just had s4934 data? I would come out thinking I was at a higher risk when I am actually at a lower one. Lots of needless worry over nothing…

Now I shouldn’t get complacent, this is still only a bit of my risk for Alzheimer’s. There may be some killer SNP lurking in my DNA, ready to spring Alzheimer’s on me. Or a long list of small, not-yet-discovered SNPs that will eat away at my supposed lower risk for Alzheimer’s. We just don’t know yet.

And I haven’t even begun to talk about environmental impacts. We know so little about that one with Alzheimer’s that we haven’t even found the whiteboard yet!

37.7749295 -122.4194155

I might be better off NOT knowing some things about my genes.

Last blog I talked about how I was able to wrest information about my APOE gene from my 23andMe data. I wanted to know because of this gene’s link to late onset Alzheimer’s disease.

APOE comes in three versions: e2, e3, and e4. People with two e4 versions are around 15 times more likely to end up with Alzheimer’s than are people with two e3 versions. And if these e4 folks do get the disease, it tends to come at an earlier age.

I found out that I actually have two e2 versions which protects me somewhat from getting late onset Alzheimer’s. Good news? I guess…

In all of this, I didn’t ask whether this information is worth knowing. Being e2/e2 doesn’t mean I won’t get Alzheimer’s…it just means I am at a lower risk. But at least I didn’t find out I was e4/e4. Then I’d know I was more likely to get Alzheimer’s but not be able to do anything medically useful with that information.

Being e4/e4 would not have meant that I would for sure end up with Alzheimer’s. So it wouldn’t be like having two copies of the delta-508 marker of the CFTR gene. In that case, I would almost certainly have developed cystic fibrosis. No, two copies of e4 would just mean that I was at a higher risk.

And knowing this wouldn’t be able to help me medically at all. There aren’t any good preventative measures I could take to stave off Alzheimer’s.

Now this isn’t always true with these kinds of increased risk genetic markers—some are definitely worth knowing. Women who have certain BRCA1/BRCA2 markers are at an increased risk for getting breast or ovarian cancer. They can choose to screen early (and often) in the hope of catching the cancer early when it is more treatable. Or, more drastically, they can choose to have their breasts and/or ovaries removed. Neither is really an option for Alzheimer’s.

So knowing my APOE status isn’t really that useful medically (at least not yet). I can’t do anything useful with the information other than wait and see if I end up with Alzheimer’s. Which is pretty much what I would have done without the test.

I’m bringing all of this up because 23andMe is now offering people their APOE information (with an upgrade to their new chip, of course). People can now find out their particular combination of e2, e3, and e4 markers*.

In the past, the question was whether or not direct to consumer (DTC) genetic testing companies should offer such a test. That ship has either sailed or is getting ready to leave the harbor. The test will be made available to people who really want the information.

So now the key question is whether knowing your APOE status is worth it. The answer to this question will be different for different people. Given this, the most important thing is for people to have the information they need to make the right choice about whether they want to know their APOE status or not.

One way to figure this out is with a genetic counselor but most DTC tests don’t mandate that you need to talk to one before you are tested. This means it is absolutely critical that the online information provided by DTC companies are presented in an easy to understand way that does not oversell the genetic test. The DTC companies need to be upfront in the fact that this test is not predictive and that there are no proven preventative measures that can keep Alzheimer’s at bay.

I’ll let you all judge how well 23andMe has done at letting people know about what you can learn from the APOE test and what you can do with that knowledge. Click here to read what they have to say.

* You have always been able to get your APOE status with a more expensive genetic test from deCODEme.

Learn more about how genes and the environment work together to cause Alzheimer's.

I put this video in the last blog but it probably should have gone in this one. It shows the unintended consequences of finding out you are e4/e4.

37.7749295 -122.4194155

I was having lunch with a colleague the other day and we got to talking about genetic testing (yes, we are that geeky). I told him about how my 23andMe test had missed my being prediabetic. This says a lot about what 23andMe’s test can tell me about my risk for diabetes. Not much.

Now as I have said before, the fact that 23andMe can’t tell me that I’m likely to become diabetic from their DNA test isn’t their fault (other than, perhaps, some overselling on their part). Diabetes research just isn’t far enough along to be able to accurately predict whether someone will get diabetes or not. Or to make any really meaningful prediction about diabetes at all.

We then got to talking about Alzheimer’s. Turns out his family tree is littered with Alzheimer’s on both his mother’s and father’s side. He ran into a similar problem–23andMe doesn’t report anything for Alzheimer’s.

But here the situation is a bit different than for diabetes. There are a couple of DNA markers that can tell us a lot about our future risk for late onset Alzheimer’s. It just so happens that 23andMe doesn’t report on them.

These markers deal with the APOE gene. This gene comes in three common versions: e2, e3, and e4. People with two copies of the e4 version are 15 times more likely to develop Alzheimer’s and if they do end up with the disease, it tends to come earlier. (One copy of e4 increases your risk about 3 times.)

With a little help from my colleague, I decided to dig a bit deeper into Alzheimer’s and see what I could figure out from what 23andMe did provide. Turns out I can figure out a whole lot about my APOE status. And that I can help other people figure out more from their results too.

The two markers (also called SNPs) that deal with APOE and Alzheimer's are rs429358 and rs7412. Here, according to SNPedia, are the combinations of these markers that tell you your APOE status:

To my surprise, 23andMe gives information on one of them, rs7412. Even though this isn’t usually enough to tell whether you have the dreaded e4 version or not, in my case it was. I almost certainly do not have any versions of e4 (yay!).

See, my results at rs7412 are TT*. If I am interpreting these results correctly, this means I almost certainly have two copies of the e2 version of the APOE gene. In terms of my Alzheimer's risk this is great news as being e2/e2 actually lowers your risk for getting the disease. (There is a chance I could be e1 but this doesn’t seem to be a very common version at all.)

Of course this doesn’t mean I won’t get Alzheimer’s…I still might. After all, the risk for someone over 85 getting Alzheimer’s is almost 50%. Even though this number lumps e4 and e2 folks together, e2 people do not have a 0% risk for Alzheimer's nor do e4 people have a 100% chance. People with e4 are just more likely to have Alzheimer's than e2 people.

It might help to think about getting Alzheimer's like drawing an inside straight in poker. Everyone might do it but people who are playing with wildcards are more likely to get it. These are the e4 folks. But it isn’t a for sure thing…you still may not draw the straight even with deuces wild.

Same thing with not having e4. Now I am less likely to draw that inside straight because there are no wild cards but it can still happen.

Because I was TT at rs7412, my case was pretty easy to figure out. Next blog I’ll try to help people out who are TC or CC at rs7412.

*Remember, we have two copies of each of our genes. So I have a T at rs7412 in one copy and a T at rs7412 in my other copy.

A video showing why knowing your APOE status isn't necessarily a good thing.

Learn more about current Bay Area genetic research from KQED's Forum.

37.7749295 -122.4194155

If you’re an adult and you can drink milk, then you are a mutant. Image courtesy of Collectie Willem van de Poll, via Nationaal Archief.

Over the last couple of weeks it has become obvious that my daughter is lactose intolerant. In most of the world, that wouldn't be a big deal. One study I saw claimed that at least 3 in 4 adults worldwide are lactose intolerant. And in some countries (like Thailand), over 99% of adults can’t drink milk as an adult.

This makes sense since lactose intolerance is the norm for mammals. Humans are one of the few animals where a sizable minority of adults are lactose tolerant. Scientists have noted this and have renamed the Eurocentric "lactose intolerance," lactase persistence.

Sigh, scientists never make things easy do they? Why not call it lactose tolerance so everyone understands?

They named it lactase persistence because of how lactose is digested in our bodies. Lactose is digested by the enzyme lactase and lactase persists longer in people that are lactose tolerant. Hence, lactase persistence.

Our cells make lactase by reading the lactase gene. In most mammals, this gene gets shut off in adulthood.

The programming for turning the gene off later in life is found in the DNA around the lactase gene. People with lactase persistence have a DNA change that messes with the programming. Now the lactase gene stays on so these folks can keep drinking milk and eating ice cream.

The “on” version is actually dominant over the normal one. In other words, lactose intolerant people need to get the normal lactase gene from both parents. So my wife and I are at least carriers—we can drink milk but carry one normal lactase gene.

I actually know that I am more than a carrier. I have two normal lactase genes even though I can still drink milk. This means that my gene hasn’t shut off yet but that it probably will at some point. Scientists don’t yet know why it shuts off early in some people like my daughter and later for others like me.

So how did some people end up able to drink milk as adults? Mutations and natural selection of course.

Most likely there is always a low level of lactase persistence in any mammalian population. Mutations (or new DNA changes) can and do happen and changes like this in the lactase gene are bound to be pretty neutral. In most situations there won’t be any advantage or disadvantage to having it and so it will stay rare.

This can all change if adults suddenly have to start drinking milk as happened in certain cultures in Europe and Africa. In these places, the few people with the right lactase mutation had an advantage and so did better than the lactose intolerant. Eventually, most people in these places could drink milk as an adult.

What is really interesting to me is the fact that European and African milk drinkers don’t share the same DNA difference. Each population had a distinct lactase mutation that became the norm. This is called convergent evolution—two populations arrive at a similar trait with different DNA changes.

However it happened, my daughter is now adjusting nicely to the milk-soaked culture she finds herself in. There are pills that let her drink milk as well as lactase-treated milk and ice cream. This means she can still have her favorite dessert in the world, chocolate ice cream.

37.7749295 -122.4194155

A few minor tweaks to genetic testing companies' websites could make their offerings more transparent to the public and the FDA.

The last couple of blogs I have been talking about direct to consumer (DTC) genetic tests. I talked about how the FDA has begun looking into them and why the FDA isn’t happy with what it sees.

In this blog I thought I'd propose a couple of different ways these DTC companies can present their data that might mollify the FDA. These changes will also let consumers know what they're really getting and whether they want it at all.

Before starting, I want to say that I will focus on 23andMe, a Bay Area company. I'm not picking on them. They are just the company I know best and one of the few that is good enough to survive the FDA's scrutiny. I also know a lot about them because I have taken their test.

23andMe has a very good website. They present complicated data in an understandable and easily searchable way. Their major weakness, though, is that they implicitly promise more than they can actually deliver. In essence, even though they are pretty good about disclaimers, they aren't good enough.

One of the first things the company should probably do is to reorganize the first page that potential customers see. They need to make sure that potential customers have a good idea about what they can and can't get from these sorts of genetic tests.

For example, right now a prominent feature is a box that lets the viewer search for the diseases 23andMe “covers” along with a list of popular topics. People may come away thinking 23andMe has useful tests for most of the diseases listed. They don't.

They have some useful tests for a few, rare genetic diseases. But the bulk of their tests are not at all useful yet in figuring out someone’s risk of getting a certain disease. What they have for the more complicated diseases is a way for people to compare their DNA to various studies in the scientific literature.

Maybe a study was done that found a DNA difference involved in diabetes. Customers can see whether or not they have this difference too but this tells them nothing about their risk for diabetes. It gives just one piece of a giant puzzle. They are not getting any meaningful results that can predict their risk for diabetes. This box should probably be heavily modified or even eliminated.

In fact, the website really should be organized into different sections that are labeled by how medically useful they are to the customer rather than by how strong the DNA study was scientifically. Maybe they could split their tests into three sections.

The first would be carrier testing. These tests can tell you if you have a hidden genetic disease that you could pass down to your child if your partner has it as well. This would get high marks for reliability, scientific validity, and usefulness.

The next section would be more fun related stuff. This would have ancestry and some of the traits testing. It would be able to tell you what your earwax is like, where your mother's, mother's, mother's, etc. mother came from, the odds that your child might have blue eyes, etc.

The final section would include the bulk of what is tested. These are the tests that compare your results to results in the scientific literature for complex diseases. Many of these tests would score high in scientific validity but get no points for usefulness. As I said before, most if not all of these tests will not give you an accurate risk assessment for the diseases they look at. Period.

There isn’t any reason these results shouldn’t be included, though. Maybe people enjoy seeing the results or want to use them to watch progress in the field or whatever. But the companies need to say upfront that these tests are not that useful for determining risk. This needs to be obvious enough that someone wouldn't buy the product just for that test.

As a last point, 23andMe (and all genetic testing companies) need to be much more upfront about what their tests can offer based on race. The carrier tests are probably pretty good for most everyone (although they may miss any nonwhite versions of many diseases). The fun section might be pretty useful to the nonwhite world for ancestry but probably less so for traits as they have mostly been determined for people of European descent.

Most of the rest of the tests they offer that deal with more complex diseases have only been validated for white people. This needs to be explicit on their website so nonwhite people know they aren't getting as much bang for their buck. Buyer of color beware!

These kinds of changes will go a long way towards making these sites more transparent to potential customers. And they may even keep the FDA at bay.

37.7749295 -122.4194155

Maybe the FDA could set up a sort of Consumer Reports for genetic testing.

Over the last few years, a bunch of companies have sprouted up that offer genetic testing over the Internet. The most controversial of these are the ones that offer consumers the chance to predict their future health risks.

These particular direct to consumer (DTC) tests took a real kick in the teeth from Congress this month. It was a big enough wallop that I don’t think the companies will be able to survive in their current form. The real question now is what they will look like when all of this shakes out.

There has been some low level concern by the FDA about this sort of testing for quite a while now. But what really focused their attention was when a small company from San Diego made plans to offer their test through Walgreens. What the FDA saw was a test that promised more than it could deliver.

To be honest, this is true in one way or another about most if not all of these genetic testing companies. Few of them are able to deliver everything that many consumers think they can.

So the FDA has set out to right this wrong. Unfortunately, because of the industry’s lack of organization, the FDA lumps all these testing companies together. This is not fair.

Some companies like 23andMe, deCODEme and Navigeneics are basically good companies that need some FDA guidance. Other companies are rip-offs that should be shut down. One of the genetic testing industry’s tasks now is to provide the FDA with the tools to distinguish between these two groups of companies.

What might help is some sort of Better Business Bureau for genetic testing. Or maybe some sort of self policing umbrella organization that only the best companies can join. Or some outside group that evaluates each company. Any of these would help both Congress and consumers have a better idea about which companies they should use for their testing.

These companies can already be CLIA certified but that isn’t necessarily that helpful from a results perspective. It is important in that it tells a consumer that the company runs the tests very well according to professional standards. What it doesn’t do is tell consumers which claims are believable, which ones are iffy, and which ones are straight out lies.

Even though it will be very tricky to police the results, these companies need to figure out a way to do this or they’ll continue to get lumped together. And consumers will lose the chance to get testing through some of the better companies.

Maybe the FDA could set up a sort of Consumer Reports for genetic testing. It would be staffed by a few genetic scientists who would look at the scientific merit of these tests and determine whether the claims are true or not. This would then be published on a website and companies that were up to snuff would get a seal of approval.

Right now I don’t think any company except maybe DNA Direct would get that seal. So maybe we need some sort of rating system instead. One DNA helix means stay away, five means a perfectly legitimate company, etc. This could let consumers know which are the better companies without the FDA having to straightjacket the industry into irrelevance.

I'll tackle the problem of interpreting and presenting data in my next blog. What I’ll try to do is propose some ways to present the data so that any potential customers won’t be bamboozled (or at least minimize the risk).

Before I leave, I do want to voice a word of caution. A working assumption in the discussion here is that the better companies are pretty good. A sting operation by the GAO (no, really) may cast some doubt on this.

As you listen to this YouTube video, think about the fact that the first example is from Navigenics, one of the good companies. The good ones need to be more careful with their communications to the public or they may be wrongfully lumped in with the bad ones. Or with too many examples like this, maybe I should say rightfully…

37.7749295 -122.4194155

Results like these get a company noticed by the FDA. (Click the image for a larger version.)

A month or so ago a small company in San Diego struck a deal with Walgreens. Together they were going to sell genetic tests directly to consumers from drug store shelves. Sort of like how some paternity tests are now sold.

The resulting furor has caused Walgreens to back off. It has also awakened the slumbering beast—the FDA.

Last week the FDA sent a letter out to five different direct to consumer genetic testing companies about the genetic tests they offer. The FDA is arguing that these genetic tests are medical devices and therefore it has the authority to regulate them.

These tests actually consist of thousands of individual tests. If the FDA wants to regulate each one separately, there is no way that these companies will be able to make a profit. They’ll have to raise prices beyond what most people are willing to pay or simply fold up shop.

This is probably why companies like 23andMe and deCODEme shied away from medical sorts of tests before and instead focused on recreational genomics. They offered tests to learn about the DNA behind your earwax, lactose intolerance and lots of other little, fun traits. But people would learn very little about their disease-causing DNA.

The companies must have found that they had trouble finding enough people willing to plunk down 400 or 500 dollars for a fun genetic test. A new TV, iPhone, Wii are all more fun and cheaper.

So to scare up some customers, the companies began to offer more medical information in these tests. Undoubtedly they thought that people would be willing to pony up 400 dollars to learn about their risks for diabetes, heart disease, lots of rare genetic diseases and much more. A bargain by any standard!

Leaving aside the fact that most of these individual tests aren’t as useful as you might think, including medical tests introduces another problem. Including medical tests puts these companies square in the sights of the FDA which is not known for its liberal attitude towards medical devices.

To get enough customers to stay profitable, these companies need to offer medical tests. But if the FDA regulates these tests then the profits will go away. Or the companies will have to charge a lot more money which will drive customers away. A Catch-22 if I’ve ever heard one!

These companies may need to provide detailed evidence for each individual genetic test. Many of them probably won’t stand up to the FDA’s scrutiny and will have to be removed. Let’s hope for their sake that the remaining ones are still interesting enough to consumers. If the companies can survive the cost of complying with the FDA that is…

37.7749295 -122.4194155

It may not look like much, but the two bits of shared DNA shown in blue show that I am at least this person's fifth cousin. (Click the image for a larger version).

My latest dive into my DNA began with an email. Someone contacted me via 23andMe and said that we may be fifth cousins and asked if I would like to compare genomes.

Ok I thought, why not? And so I contacted them. We compared our genomes and this is how much DNA we had in common. Not exactly awe-inspiring is it?

Even though it looks pretty wimpy, it is actually a couple of good sized chunks of DNA we share. Odds are that I am indeed related to this person. To understand why, we need to step back a bit and remember how DNA is passed from generation to generation.

As you undoubtedly know, we get half our DNA from our moms and half from our dads. This DNA is passed down in the form of chromosomes.

What you may not know is that except for the Y chromosome, we don’t inherit an exact copy of our parents’ chromosome. Instead we inherit a mix of each of their pair of chromosomes. Clear as mud as usual, Dr. Starr…

We all have two copies of each of our chromosomes, one from mom and one from dad. So we have two copies of chromosome 1, two copies of chromosome 2 and so on up to chromosome 22. The chromosome 1 we get from mom is actually a mix of her two chromosome 1’s. Same thing with our chromosome 2 and so on.

This mixing happens with each generation. I have tried a couple of times to explain this in words but I got tired of writing chromosome over and over. So instead here is an image that I think explains things a bit better than words:

The image focuses on a single pair of chromosomes. It starts out with the chromosomes of your four grandparents (G1-G4). G1 and G2 pass a chromosome down to mom and G3 and G4 pass one down to dad.

As you can see, the chromosome that was passed down didn’t exist before. It is a unique mix of each grandparent’s pair.

The same process then happens with mom and dad. More mixing and matching happens leading to your unique pair of chromosomes.

Note that the chunks of DNA from your grandparents get smaller as they head down the family tree. Your children would inherit even smaller chunks.

Over the generations, some chunks would become so small that they could no longer be easily linked to the originals. And eventually, some ancestors’ DNA would disappear entirely from the record.

This is why the fact that this woman and I share a couple of chunks of DNA is so significant. The DNA of fifth cousins needs to travel down two different branches of the family tree from great, great, great, great grandparents. The only reason we have any chance of seeing the relationship at all after all this traveling is because we have 23 pairs of chromosomes.

This kind of analysis is brand spanking new and wouldn’t have been possible even a few years ago. Back then, we could only look at little snippets of our DNA. These old school tests have trouble telling if two people are cousins let alone fifth cousins!

So now we can find a bunch of folks related to us. We each have 64 great, great, great, great grandparents which means we have tons of fifth cousins.

And now I run into a common problem I have with genetic tests like this…so what? I can find lots of relatives but is there anything I can do with that information?

I certainly don’t know who any of my 64 great, great, great, great grandparents were. If my grandfather was born in 1906, my great, great, great, great grandparents were born around 1800 or so. I don’t know anything about my family tree from way back when and I am not sure knowing I am related to Jane Doe from Poughkeepsie helps me figure it out.

I suppose that as more and more of us plumb the depths of our DNA we’ll be able to better trace these sorts of relationships. But in the meantime, this is probably only useful for the real genealogy hounds out there. Since I’m not one of these folks, I wonder how useful it is even to them…

37.7749295 -122.4194155

23andMe's DNA testing was always fun. Now it is more useful as well.

As anyone who follows this blog knows, I had my DNA tested awhile back by a company called 23andMe.  I wrote about what I learned and didn’t learn from their testing in a bunch of blog entries.

In my mind 23andMe has always been a sort of recreational genetics testing company.  You can find out about your earwax, whether you are likely to have blue eyes or be lactose intolerant and lots of other minor sorts of traits.  This is stuff you probably already know but for geeks it is pretty cool to see them written out in their DNA.

The company always offered some health data too but it wasn’t that strong.  For example, they could tell you if you carried the most common DNA difference that could lead to cystic fibrosis (CF) but not about the less common ones.  In fact, I gave them an incomplete for their carrier testing a few months back.

Since then, the company has gone away from being a place where you get your DNA tested for coolness’ sake to one with a focus on health and/or ancestry.  With this change has come a much-improved product for people interested in what their DNA tells them about their carrier status for a variety of genetic diseases.

Carrier status is important if you are considering having a child.  If you and your partner both carry the broken versions of a gene that could lead to a disease, then your child would be at an increased risk for getting that disease.  For example, if both you and your partner have a nonworking copy of the CFTR gene, then, depending on the exact DNA you each have, your child could have up to a 25% chance of ending up with CF.

This is why the first iteration of 23andMe carrier testing wasn’t as useful as I would have liked.  They tested only one of the 100’s of different DNA variants in the CFTR gene that can lead to CF. Since this DNA variant only accounts for about half the cases of CF, there was a good chance that something would get missed.  This is no longer true.

As part of the refocusing, 23andMe looks for 31 different variants in the CFTR gene that are known to cause CF. Now this isn’t hundreds but is more than the 23 recommended by the American College of Medical Genetics.  And in fact 23andMe includes these 23 in the 31 it tests.

Of course the testing still isn’t perfect but no testing is.  Some of the tests are only useful for certain ethnic groups.  And there is no upfront genetic counseling to help you decide whether or not genetic testing would be useful in your situation anyway.

But the bottom line is that 23andMe’s testing for genetic diseases that you might be carrying is much stronger than it was before.  So much so that it can even give you some piece of mind for many of these diseases.

In some ways I’ll miss the more whimsical look at DNA that 23andMe used to represent.  But this obviously wasn’t a good business model for anyone except those enamored of DNA.  And 23andMe does need to make a profit…

37.7749295 -122.4194155

Current DNA tests could not have predicted that George W. Bush would be our 43rd President of the United States.

Time recently had a great article on helicopter parents. These are the parents who hover around their kids, protecting them from any harm. They are undoubtedly doing this to ensure their kids’ success in life.

I don’t want to get into the plusses and minuses of this parenting style…to each his own. What I do want to do is to warn them away from a new genetic testing company that seems designed to target them.

This testing company, called My Gene Profile, claims to be able to use genetics to help parents figure out where their child’s inborn talents lie. The idea, then, is for parents to point their children towards interests or careers that match up with what the genetic test says.

The talents the company is looking at are not simple ones like tongue rolling or bending your thumb back (neither of which we can yet determine genetically). They claim to be able to tell you if your child will be smart, creative, good at sports, and near as I can tell, five other similarly broad traits.

This is impossible given our current knowledge of genetics. And frankly, I am not sure we’ll ever be able to figure any of this out with a simple genetic test. Most of these traits are more than just the DNA we inherit.

Let’s take IQ as an example. Most of the studies I have seen point to about half of someone’s IQ coming from genes and the other half from the environment. Any test done right now won’t look at how the environment affected a child’s DNA. And they certainly won’t look at how the environment influenced the growth and development of the brain or how it affected synapse connections or about a million other things to do with intelligence and the brain.

Still, 50% from genes is a lot. If we could get a complete readout of how our genes influence our IQ that might be at the very least interesting. But we can’t.

Scientists believe there are at least 100 genes that contribute to IQ. So far they’ve only identified a handful and none of them have been shown to have reproducibly significant effects on IQ.

For example, scientists have found that having certain versions of the CHRM2 gene affects your ability to organize things in a logical way. The effects aren’t huge though. 23andMe (a company that I have tested with) reports that the variations that they look at in this gene can lead to a 6 point swing in IQ. Woopty doo.

If you drill down a bit farther, some scientists claim that you can get much larger differences. For example, at the furthest extremes, people with one set of variations in this gene averaged an IQ of 85 while people with a second set averaged an IQ of 103. Sounds impressive.

Except that a larger follow up study was not able to see the same effect. In this study, scientists weren’t able to find any connection between variations in the CHRM2 gene and IQ. And CHRM2 is by far the best characterized IQ gene.

Most likely the way that genetics contributes to IQ is that each of the 100 or so genes tweaks IQ a bit higher or lower. So to get a complete readout on IQ you’d need to look at all of these genes. This is difficult to do right now since we only know about a few of them.

And to make things even more complicated, the genetic contribution to IQ probably isn’t a simple summing up of these different variations. They don’t exist in a vacuum—these gene variations all interact with each other too.

What this means is that we may never be able to get an accurate prediction about genetic IQ from our genes. There are lots of possible combinations all with different outcomes. In other words, everyone’s IQ genetics may be unique which would make predictions impossible.

The bottom line is that there is not nearly enough data out there to figure out someone’s IQ or intellectual potential. And this goes for athletic ability, creativity and any other similarly complicated trait. We can’t even predict eye color yet very well from our DNA!

So consumers be aware of what a genetic test can and can’t deliver based on what scientists know. Testing for cystic fibrosis, pretty good. Testing for intelligence, not so much.

A final example. Imagine that Einstein’s parents had tested his genes for IQ with a company that looked at four or five IQ genes. And let’s say that he happened to have versions of these genes that lead to a lower IQ. Of course, since it is Einstein his other 95 or 96 or so IQ genes swamp out the effects of these few genes. But the testing company misses this and recommends that he not take an academic career. A little knowledge is a dangerous thing.

*This is common with genetic studies. There is a promising result with a small group that disappears when scientists look at a larger group.

37.7749295 -122.4194155