No, I’m Not Irish

So where did that red hair come from?

by Kasia Bryc, 23andMe Population Geneticist

Most people who see my unruly wild red hair often assume I must be Irish.Kasia1
I’m not.

As a first generation immigrant from Poland, I don’t have any ancestors from Ireland.

Yes, really, just take a look at my 23andMe Ancestry Composition results.

Yet it is true that red hair is strongly tied to Ireland. Over 10 percent of 23andMe customers of Irish heritage, whose four grandparents were born in Ireland, have red hair. In most other parts of continental Europe, red hair is less common, and its frequency tends to hover between one and three percent.

For me, my red hair holds special significance. Both my parents had brown hair, yet both my brother and I have red hair, thanks to the joys of recessive inheritance (No, we’re not adopted. And yes, our family relationships have been confirmed genetically through 23andMe).

Red hair acts, for the most part, like a recessive trait, meaning that you need two copies of the red-hair genetic variant to see red hair. Which explains why my parents, each carrying one copy of the variant, didn’t have red hair, but could pass it on to their children by chance, which they did, twice.
Kasia's Ancestry Composition

As a child I puzzled over my hair color – it seemed rather poor camouflage, and surely ill-adapted – how could a trait as odd as red hair have come to be?

That early fascination with human traits and evolution helped lead me to a career as a population geneticist. Nowadays I love wrangling with just such questions as I study the genetics of people across the globe.

The red hair gene

To dive into the genetics of red hair, we first have to understand the function of a gene called the Melanocortin 1 Receptor (MC1R). MC1R acts as a lever that toggles the cells to produce a brown or black pigment instead of producing a reddish yellow pigment. The color of the pigment, and how much is produced, leads to much of the diversity in appearance of humans around the world.

Genetic mutations can change the normal function of MC1R, leading to a loss of function. That means that the protein the gene makes is changed so much that it doesn’t work as it should. So the pigment-producing cells don’t get their signal to make black-brown pigment that is so protective against the sun’s harmful UV rays,and instead produce more reddish yellow pigment, which, in your hair, gives it a red color.

About 90 percent of 23andMe customers who report having red hair carry at least one of three of the most common genetic variants associated with red hair variants. Though there are other known and unknown variants involved in determining human pigmentation.

 

Do you carry one of these variants associated with red hair?

Three variants highly associated with red hair among 23andMe customers. Not yet a customer? Visit our store!

The evolution of red hair

For a long time, red hair was thought to be an adaptive trait. Not the hair color itself, but rather, red hair and lighter skin pigmentation are a bit of a package deal, and this lighter pigmentation was thought to be adaptive in northern Europe. Lighter skin color allows the skin to produce sufficient vitamin D when levels of light are low, when you don’t get sufficient levels of the vitamin through your food (such as from fish). So if you were a farmer in, say, Ireland, lighter skin would help you keep your vitamin D levels sufficiently high.

However, there has been a surprising dearth of evidence for positive selection of red hair in Europeans, and there seems to be no association between red hair and the evolution of light skin in European populations. In contrast, the gene SLC24A5, which is involved with skin pigmentation, has a strong signal of selection. So it seems that for red hair, the vitamin D hypothesis isn’t really substantiated by science.

Lack of selection

Instead, the current theory most evolutionary scientists believe suggests that selective pressure to maintain proper protective black-brown pigmentation is highly constrained in regions closer to the equator that have strong sun exposure. The production of black-brown pigment, necessitating a functioning MC1R gene, helps shield cells from ultraviolet radiation that can increase the risk of skin cancer.

In the far north, it seems, that selection may be relaxed. Comparing the human MC1R gene sequence to a chimpanzee’s genome suggests there was a lot of functional constraint on this gene… right up until humans migrated out of Africa. Mutations in MC1R that take away production of black-brown pigment may matter less where the sunlight isn’t as powerful. As a result, we observe these mutations, and the accompanying red hair. Sometimes, the mutations drift up in frequency by chance, and red hair becomes (more) common.

For an even more nuanced contemplation of skin pigmentation and adaptation, and its possible relationship to the switch from hunter-gatherer lifestyles to farming, check out this read.

Red_hair_map

Where are 23andMe customers with red hair from?

A closer look at the three most common red hair variants among 23andMe customers shows that red hair allele frequency in Europe does correlate with distance from the equator. The variant that is most common (and is involved in a third of red hair among 23andMe customers) is R151C. Unsurprisingly, this variant is most common in individuals whose grandparents were born in Ireland and in the UK.

Over 50 percent of 23andMe redheads carry at least one copy of this variant. Are you one of them?

The variant D294H is also most frequent in Ireland, but overall it is less prevalent, and found in a limited range: mostly Ireland, the UK, and France, Switzerland, and Belgium.

Unlike the first two, the variant that I carry, R160W, isn’t most common in Ireland. Instead, it is found at its highest frequencies in Sweden and Lithuania, as well as other parts of Scandinavia, Central and Eastern Europe.

So what do these maps of variants tell us about why red hair might have come to be?

Well, there does seem to be a correlation between where the variants lie – each variant is most frequent above the 35th parallel, which suggests that there may really be a relaxed selective constraint on MC1R function this far north of the equator.

redhair worldIn researching traits among 23andMe customers, I learned a lot about red hair. I was amazed at how the genetics of just one trait are so complex. For example, not everyone with two MC1R variants reports having red hair. Isn’t that how recessive traits should work?

But instead, red hair variants aren’t completely penetrant, meaning they don’t always cause the trait. The variant that I carry, R160W, is especially illustrative: only 50 percent of individuals that carry two copies of this variant have red hair.

On a personal level, I felt reassured that what I knew about my family’s history was consistent with what I found to be true about red hair. But even more, I realized just how widespread red hair variants are across Europe, especially its northern reaches.

 

  • John

    Where’s your picture? When you write an article like this we need to see a picture of you with your red hair. 🙂

  • John

    Where’s your picture? You can’t write an article like this and not show your red hair. 🙂

  • Ponto Hardbottle

    I do not agree with the generalization about the color of the two types of melanin. Eumelanin is reddish brown, and Phaeomelanin is brownish red. There is no yellow, that is just a Eurocentric bias thinking it has something to do with blond hair which is due to Eumelanin not Phaeomelanin.

    I am of the Maltese ethnic group and carry R150W and if I had Eastern European ancestry I would shoot myself. I have four red haired children. It is a curse rather than a blessing in Australia, it turns red haired people into leathery purple faced people here.

    • Janine O’Flaherty

      Eumelanin is brown-black and pheomelanin is red-yellow. I have no idea where you got your information but it is incorrect.

      Blond hair is not caused by pheomelanin because blond hair is not actually yellow, it is shades of beige, indicating low amounts of eumelanin but not the presence of pheomelanin. If blond hair had some pheomelanin going on, it would present as strawberry-blond, i.e. a sandy/peachy pale orange.

  • JodyM

    I have red hair but don’t carry any of the three variants. Any idea why the .6% who don’t have these variants end up with red hair? I am pretty much a mutt with family from all over northern and southern Europe.

    • 23blog

      Hi Jody,
      There are definitely other MC1R variants that cause red hair – these three are just the most common and associated with red hair in 60 percent of redheads. Obviously that doesn’t cover everyone. It is interesting that you have none of those variants something that is rare among redheads who are 23andMe customers. So you can take that to mean that you are special.

      • JodyM

        Lol. Thank you! 🙂

      • Berneice

        I just got my 23andme results back last week, and was VERY surprised to find out I did not have any of the three common redhead variants. According to my results, I should have dark brown hair, but my hair is very much red, and has been my whole life. The mystery continues. Glad to know there are other extra special redheads out there with me! =)

  • Colyn Ward

    Kasia, are there studies of men who are not redheaded but have red in their beards? How does different beard color fit into the statistics? (I’m not even sure if the question is asked within 23andMe.) I had a friend who had dark brown hair and tanned very well but had a very bright red beard if he grew it. Quite a startling thing to see! Thanks for the article — it’s all very interesting.

    • 23blog

      Hi Colyn,
      That’s a great question, but we do not have data on beard color. We do ask questions about hair thickness and whether someone’s hair is curly or straight and how men’s facial hair fills in. But nothing yet on color. Maybe we should start.
      Thanks.

  • happykt

    My grandmother had bright red hair, the color of a copper penny and she was born in what is now Western Ukraine to Polish and Ukrainian parents. Her son (my dad) had red brown hair and pink freckled skin and burned to a crisp. I also have the most freckles on skin but do not have pink skin or red hair but am highly allergic to direct exposure to the sun and used to get boils and sores on my skin and try to stay out of the sun as much as possible.

  • 23blog

    Hi Jim11,
    Kasia said that according to our estimates based on customers who reported their hair color, 86% of individuals with two copies of *R151C reported having red hair, and everyone we saw with two copies of *D294H reported red hair. On average, 72% of individuals with any two MC1R alleles report to have red hair.

    Unfortunately, our data doesn’t have very subtle shade information. As you point out, it is common for hair color to change over time, but I’m unaware of the mechanism behind it.

  • Linda Hara

    3 percent of your DNA is Northern European. Ireland is in Northern Europe.

  • Linda Hara

    3 percent of your DNA is Northern European. Ireland is in Northern Europe.

  • KSnowQueen

    Do tests detect any red hair inducing gene mutation other than the three mentioned above if someone has it,or does it just say hat you don’t have the three?just wondering

    • 23blog

      In reporting out the likelihood of having red hair we look at three specific variants, rs1805007, rs1805008 and i3002507. These three variants are all in the MC1R gene.

  • Jen Bur

    I got my red hair from two separate genes. R151C rs1805007ct and a separate R160W rs1805008ct. Its the t at the end of each that causes red hair. A t on two separete genes. Since it
    was caused by two separate genes, could I have recessively inherited from one parent?

  • Janine O’Flaherty

    Actually, people with two copies of R160W have red hair 74% of the time according to a study (Beaumont et al. 2007), and 73% of the time according to self-reported 23andMe research… not 50% of the time as you stated.

    I too am homozygous for R160W. Here is a well-lit picture of me so you can see my natural hair and skin colors: http://i65.tinypic.com/2lj59mw.jpg

    I’m curious to know what shade of red the same mutation produced in you.

    • 23blog

      Hi Janine,
      This response from Kasia, who wrote the post:

      Thanks for the comment! Yes, my hair (http://kasiabryc.com/assets/static/kasia1.jpg) is fairly similar to yours (although it used to be a little darker when I was younger). According to Table 1 from the Beaumont et al. 2007 study (http://hmg.oxfordjournals.org/content/16/18/2249/T1.expansion.html) I found, it looks like 4 of 7 individuals (or 57% of individuals) who were homozygous for R160W had red hair, pretty similar to the number I reported here based on 23andMe customers who consented to participate in research. The sample size used to generate our 50% statistic is over 300 individuals, so I think that 50% is quite reliable. The 23andMe self-reported research number you quote corresponds to 73% of individuals with any combination of all the different red hair alleles have red hair on average, not just the R160W mutation. So which particular red hair alleles you inherited lead to quite different chances of having red hair!