Genetic Testing

I read the posting on how it is possible for 2 non-tongue rolling parents to have a child that can roll his tongue. Prior to reading your web site posting, I stumbled across an article in the April '07 edition of Men's Health. The article was discussing Paternal Discrepancy and stated that one of the tale-tell signs was tongue rolling.

I immediately started wondering about my own child. I have no other reason to doubt that I am not his biological father. Neither my wife nor I can roll our tongues, yet my child can. Most of my child's other genetic traits are the same as mine, and my wife shares a lot of those as well, i.e. Blonde hair, blue eyes, etc...

The article in Men's Health said that 4% of fathers in the U.S. are unknowingly raising a child that is not theirs. I have not mentioned this to my wife, but it is keeping me up at night. Should I pay $100 for a paternity DNA test, or should I just not worry with it at all?

-A curious adult from Texas

May 18, 2007

Thanks for this question and for bringing the article to my attention. The article is an interesting one and I can't really comment on the 4% number (although even if true it does mean that 96% of us are raising our own children). But I can comment that the "tale-tell" signs are not as clear cut as they say in the article.

The article brings up six different traits: tongue rolling, cleft chin, blood type, ear lobes, finger length and blonde hair. Some of these are downright useless for predicting who your daddy is.

Really, the only one that would raise a red flag for me is blood type. And even that isn't foolproof (as I'll show later).

So please do not be concerned if your child can roll his or her tongue. This is no indication of paternity. Nor are most traits really.

This is because nothing in genetics is absolute. All you have is what is more or less likely.

The article compares traits like these to tells in a game of poker. A tell is when a player does something that gives away the player's hand.

For example, let's say you notice that whenever Sally has a bad hand, she itches her nose. Now when she tries to bluff but itches her nose, you know she's bluffing.

The idea is that you can tell if you are not the father of your child if these certain traits don't line up right...if she itches her nose, it isn't your child. It is a very clever and enjoyable analogy. But the language they use make the traits sound much more clear cut than they are.

Look, even a tell in poker is not absolute. Maybe Sally just happened to itch her nose a couple of times during the game and they happened to be when she had a bad hand.

Now when you try to use the information about her tell against her, it doesn't work. You lose. Just like your own child might lose if you take these traits too seriously.

What I want to do is go over each of the traits they listed and talk about exceptions to these supposedly iron clad rules. Some of the traits they mention are not signs of anything. They are more like urban legends.

Others are more likely to raise a red flag. But even these are not for sure. So please be more careful than the language of the article would seem to indicate.

The Basics

Before going into each trait, let's go over some basics first. That way we will all be on the same page for when we start talking about each specific trait.

First off, we have two different copies of each of our genes -- one from mom and one from dad. Second, each of these copies can come in different forms (called alleles).

This is why two people without red hair can have a redheaded child. Or why two parents with brown eyes can have a blue eyed child.

In each case, the parents have two different copies of a gene for the trait. For example, the parents of the redhead each have a copy of the MC1R gene that leads to red hair. And one that does not.

Each parent does not have red hair. But they carry the red hair gene (they are carriers). Which means that each child has a 1 in 4 chance to get a red MC1R gene from each parent. And end up a redhead.

The same sort of thing works for eye color. The tells that the article talks about are situations that work in reverse. Two redheads that have a child without red hair. Or two blue eyed parents having a child with brown eyes.

These things are impossible in a simple black and white world. But genetics is anything but black and white. It is all shades of gray.

The gray comes from at least two other things we need to remember:

1) Our genes do not work in a vacuum. They are affected by the environment and all the other 25,000 or so genes in our cells.

The environment and all of these other genes can affect whether our genes work right or not. And sometimes, we pass the gene to our kid in a different environment. And/or with different genes. So now the gene works.

Let's take our example of two redheads. A person ends up with red hair when both copies of their MC1R gene don't work.

Imagine one copy of one parent's MC1R gene doesn't work because of another gene. For example, maybe there is some other gene whose job it is to shut off the MC1R gene. Then that parent would have red hair even though his or her MC1R gene by itself would say no red hair.

Now that parent passes the working MC1R gene down to his or her kids without the other shut off gene. And guess what? These parents now have a kid without red hair. Even though it is "impossible."

2) Our genes are not set in stone. They can and do change. This is where all the wonderful differences we see around us come from. And these changes are not confined to our past.

For example, 7 out of 8 cases of dwarfism happen because of a new change in the person's DNA. The parents carry no sign of the trait. But their child does.

For our red hair example, this would mean that one of the parents had a change in their MC1R gene that changed it from red to not red. Not likely, but it could certainly happen. Like our dwarfism example.

OK. Now we have a solid foundation to work from. Remember, the article mentions six traits: tongue rolling, cleft chin, blood type, ear lobes, finger length and blonde hair. There are certainly exceptions to each of these. And some of them are a totally useless tell.

The six tells

Tongue rolling: This is the worst of the lot. A study done way back when concluded that tongue rolling was a dominant trait (like brown eyes or not-red hair).

What this would mean is that two parents who couldn't roll their tongue could not have a tongue roller for a child. This is wrong.

As we have talked about before, further studies have shown that tongue rolling doesn't really follow this pattern. For example, identical twins, who share all of the same DNA, can't always both roll their tongues. In fact, one study showed they weren't any more likely to each be able to roll their tongue than any brother or sister.

If tongue rolling were a simple dominant trait, then if one twin could roll their tongue, then the other should be able to as well. But they can't always both roll their tongues. So something else is going on here.

So if you have a tongue rolling child and you can't roll your tongue, this is not a cause for concern. Relax and enjoy your kid.

Cleft chin: Another bad example (although the article was a bit more careful in its language in this case). The article states:

Cleft Chin - While not as definitive as the other tells, "it's extremely rare for a child's chin to be cleft if both parents lack the trait," says Pond.

I am not sure where the extremely rare comes from. But as we have talked about before, cleft chin is a classic example of something called variable penetrance.

What this awful phrase means is that sometimes someone can have the right gene but not have the trait. It might be that a lot of folks have a gene that shuts the cleft gene off. So they have an off copy that can get turned back on in a child.

Or maybe something environmental can affect the trait. Like sometimes injury in the womb can keep a child's eye blue. Even though the child has the genes for brown eyes. And so could have a brown eyed child themselves.

This tell is a bit better than tongue rolling (which is worse than worthless). But cleft chin still wouldn't worry me too much.

Blood type: This one would actually raise a red flag for me even though there are plenty of exceptions. This is because the exceptions are much less common than the previous two tells. But exceptions certainly happen!

The most interesting case for me happened in Japan. An AB dad had a child with an O wife and had an AB child. This is impossible according to straight up genetics. And is listed as a tell in the article.

But this guy was the dad. What had happened was that his A and B alleles got stuck together. So he passed it on as a single AB. This also means he could have had an O child. Another "sure tell."

In fact, as we talk about here, AB parents having O children happens more often than you might think in certain Asian countries because of something called cis-AB.  And as we talk about here, people from India can have an O blood type and have an AB child more often than you might think because of the Bombay blood group.

Here are three other discussions about ways that blood types can sometimes go against the rules:

An AB child with an O parent

An A or B child with O parents

An Rh+ child with Rh- parents

Blonde hair: According to the article, two blonde haired parents can have only blonde haired children. The red hair example we used before shows that there is at least one case where this isn't a particularly good tell -- if they have a redhead.

All we need is for both parents to be carriers of the red hair gene. Now two blondes have a redhead.

There are also cases of blondes having darker haired children. This probably comes from the fact that hair color is due to many genes and not one (click here for more details).

In fact we don't understand hair color all that well. All sorts of things can affect hair color.

The hair color of children can darken over time. A child can be born with dark hair that turns light again. And then darkens again.

And the form of a skin color gene that makes Caucasians so pale may also influence hair color. So someone carrying the darker version of that gene might end up with a child with darker hair.

So hair color is not a particularly good tell either. It might raise a red flag but not nearly as big of one as blood type might.

Ear lobes, finger length: I am not sure what to do with these. The tells are when parents with attached earlobes have a child with unattached earlobes. Or when both parents have a ring finger longer than the index and the child doesn't.

We know, for example, that finger length is definitely influenced by testosterone in the womb. As we have talked about before, women tend to have an index and ring finger that are about the same length. Men tend to have a longer ring finger.

This difference is thought to happen because of the extra testosterone that men get in the womb. What in the world does that have to do with paternity?

Imagine a man has a child that gets less testosterone in the womb. Dad has a long ring finger and his son doesn't. Cause for worry? No. The environment affected his finger length.

And attached and/or unattached earlobes is a really terrible tell.  As we talk about here, the original studies have not held up over time.  Parents with attached earlobes can have a child with unattached earlobes.  In fact, some people can have one of each earlobe! I would not sweat my kid's ear lobes.

Only DNA is a good tell

So, what have we learned? Of all these tells, the one that would concern me the most is blood type. Even though there are exceptions, they aren't all that common. At least not compared to the other cases. For which there are many exceptions.

The only really reliable tell is a DNA test. And even that can fail sometimes (click here for the details).

So don't freak out if your child can roll his or her tongue. This is not a good reason to have a paternity test done!

By Dr. Barry Starr

There are no foolproof traits that can determine paternity.

A DNA test is the only good tell and even it isn't perfect.