Have scientists discovered a gene for binge-drinking? No, but the research is still important
Yesterday a load of headlines about genes being linked to binge-drinking appeared on various news sites. Only the BBC (so far) seem to have fallen in to the ‘gene for x’ trap, the first paragraph of their article being particularly poor:
Scientists believe some people have a gene that hard-wires them for binge drinking by boosting levels of a happy brain chemical triggered by alcohol.
Not quite, but what did the study* actually look at?
Stacey and colleagues are interested in the effect of dopamine (the ‘happy brain chemical’ referred to above) on the way people respond to drugs, and the reinforcement people get from them; in other words, addiction. In particular, there is a gene which has previously been associated with alcohol intake in a genome wide association study (although that study did not provide definitive evidence of an association, only a suggestion, and only in a subpopulation). The gene in question (RASGRF2 for any gene spotters out there) codes for a protein which is involved in dopamine transmission (read more about dopamine here), which is why the researchers were interested in its effects on alcohol consumption.
The paper which has triggered all the articles is a combination of research in to mice who have been bred with or without the gene of interest, and humans who underwent fMRI brain scans.
What did they find?
- Mice who didn’t have any copies of the gene in question (mutant mice) drank less ethanol infused water than their ‘wildtype’ littermates, when given free reign to drink either water or ethanol and water.
- When looking at the brains of the mice, there was a difference in levels of dopamine in the spaces between neurons, with mutant mice not showing the more usual increase in dopamine after an injection of ethanol.
- The neurons of mutant mice, in a certain brain area (the ventral tegmental area), fired less frequently than their littermates’ neurons, suggesting what the authors call ‘reduced cell excitability’.
- When they sliced the brains of the mice, they found differences in the potassium channels in to the cells, which they say could be the mechanism for the difference in how often the cells fire, as these differences could affect how dopamine enters the cells and allows neurons to fire.
And what about the humans? The scale of this study is very impressive, as they have scanned the brains of 663 fourteen year old boys (they only used males as the original study only showed an effect in males – subsample analyses like these increase the risk that the findings are spurious).
- Using a task in the brain scanner designed to elicit reward anticipation (alas the details are in the supplementary material, so I don’t have them - see *), they found brain activation in certain brain areas consistent with dopamine, including the ventral striatum.
- When they split the brain scans by genetic differences (thought to be equivalent to the genes in the mice, although there are issues with directly comparing animal and human genetics - and sometimes the media isn't even clear what they're reporting), they found differences in the activation they saw in the ventral striatum by genotype, suggesting that perhaps the people with this genotype respond differently to reward anticipation.
- When looking at number of times the same boys had used alcohol by age 16, they found a difference between this same genetic split, with those who showed a higher activation being more likely to have drunk more than 6 times, than those who didn’t.
BUT, they did not find:
- A ‘gene’ associated with binge drinking (in fact, paragraph 6 of the BBC article states ‘this is not proof that the gene causes binge drinking’). Firstly, the study did not look at a single gene, it used ‘haplotypes’ which are a set of polymorphisms on a single chromosome which are all statistically related. Maybe a technicality, but still important when looking at dodgy headlines. More info about the haplotype is unfortunately in the supplementary material, so apologies for the lack of clarity here.
- At no point do the authors make it clear how many people they are comparing in each genetic group. They mention that one ‘haplotype block’ (with 613 individuals) is associated with what they define as ‘lifetime drinking behaviour’, but at no point do they state how many people they are comparing to each other, so it’s very hard to judge the meaningfulness of the findings. It could be that there are only very few people with a certain genotype, and they are driving the difference. They mention a frequency of 22.7%, so this may be the split, but it’s not wholly clear.
- They don’t assess binge drinking in humans AT ANY POINT. Their measure of alcohol use is a binary measure, comparing people who had drunk less than 6 times in their lifetime to people who had drunk more than 6 times. Now, when I was 16 I had definitely drank on more than 6 occasions, but I wouldn’t call myself a binge drinker! It isn’t quite clear whether they are talking about binge drinking 6 or more times, or any form of drinking, but they certainly don’t look at problematic drinking in teenagers, which would be considerably more than 6 binge drinking episodes in a lifetime, if indeed that is what they have assessed.
The authors are clear in their quotes that they don’t believe this study shows a ‘binge drinking gene’, and indeed it doesn’t. It is certainly possible that likelihood to consume alcohol is affected by differences in genetics (for example, I’ve written before about the genes which code for proteins that metabolise alcohol; variation in these affects likelihood to drink). It’s also possible that how rewarding alcohol is found to be could affect likelihood to drink, and to drink more heavily. However, this is opposed to the ‘low responders’ I wrote about recently; behaviourally it seems that those who are likely to go on to develop alcohol problems in later life are people who don’t get an initial big response from alcohol, rather than those that do.
*NB I’d love to put a link to the original article, but for some reason it’s still not accessible on the PNAS website. Now, articles that appear in the press before the paper is in the public domain are a worry, but that’s another story. I was able to get hold of the paper from the press office, but at the time of writing I still don’t have the supplementary material, so apologies to the authors if anything I’ve said isn’t mentioned in the paper appears in there, but I can’t check! Will happily update.