Deep divefrom Episode 10Longevity8 November 202614 min

Your Genes, Your Habits and How Old Your Brain Looks

A deep dive into one paper: what was known before, what the researchers did, what the numbers say and what they don't.

The paper: Association of genetic risk, lifestyle, and their interaction with MRI-based brain aging (Translational Psychiatry)

MP3

Chapters

Transcript

The script the two AI voices read from. The times are worked out from the script, not heard, so a line can start a few seconds early or late.

Read the whole episode

Cold open

Host 1You heard the trailer — this is the whole story. Welcome to the Hastrup Review of Science, where artificial intelligence reads real science, so you don't have to pretend you understood the paper.

Host 2Today is one paper, published in October in the journal Translational Psychiatry. It asks how your genes and your lifestyle work together to shape how old your brain looks on a scan.

Host 1Before anything else, what does it even mean for a brain to look old, as opposed to the person being old?

Host 2It means a computer looks at a brain scan and guesses your age from the picture alone. If it guesses seventy and you are sixty, your brain is said to look ten years older than you are, and that difference is the thing everyone in this paper cares about.

How a scan learned to guess your age

Host 1Where did the idea of guessing someone's age from a picture of their brain come from in the first place?

Host 2Looking at the wider field, it grew up over roughly the last fifteen years, as machine learning met large collections of brain scans. Researchers trained programs on thousands of healthy people of known age, so the program learned what a typical forty year old brain or a typical eighty year old brain looks like.

Host 1And why would that be useful, beyond being a neat party trick for neuroscientists?

Host 2Because the gap between brain age and real age turned out to mean something. In earlier studies, people whose brains looked older than expected tended to do worse on memory tests, and the gap was larger on average in people with dementia, schizophrenia or after a stroke.

Host 1So the gap works a bit like a warning light on a car dashboard, telling you something might be wearing out early?

Host 2That is a fair picture, with one correction. It is a warning light for groups of people, on average, and it is not yet a reliable reading for one single person in a doctor's office.

What was known before

Host 1What did people already know about what pushes that gap up or down, before this paper came along?

Host 2Two broad things, from the wider literature. Brain age is partly heritable, meaning it runs in families and is linked to many small genetic differences, and it is also linked to habits like smoking, heavy drinking, poor sleep and inactivity.

Host 1If both of those were already known, what was the piece nobody had worked out yet?

Host 2How the two combine. Genes and habits had mostly been studied in separate papers, and the authors say their joint influence remains poorly understood.

Host 1When you say how they combine, do you mean whether a healthy lifestyle can cancel out bad genes?

Host 2That is exactly the question underneath it. Scientists call it a gene by environment interaction: whether the effect of your habits is bigger or smaller depending on which version of a gene you carry.

What the team did

Host 1So who did they study, and how many people are we talking about?

Host 2Forty six thousand, eight hundred and forty three people from the UK Biobank. That is a large British research project in which volunteers, mostly in middle and older age, gave blood for genetic testing, answered questions about their lives, and many had brain scans.

Host 1You mentioned scans that look at the brain in more than one way. What does that actually involve?

Host 2The abstract calls it multimodal magnetic resonance imaging, which means several kinds of scan from the same machine. Typically, one kind shows the size and shape of brain regions, another shows the wiring between them, and another shows activity, and the model got to learn from all of them together.

Host 1How do they know their age guessing program is any good, rather than just making confident wrong guesses?

Host 2They report a number called R squared, between zero point seven two and zero point seven five. In plain words, the program's guesses account for roughly three quarters of the differences in real age between people.

Host 1Is three quarters good, or is that the kind of score that would get a student a polite frown?

Host 2It is solid for this kind of work, though not exceptional. The remaining quarter is partly what they actually want to study, the brain looking older or younger than expected, and partly just measurement noise, and the two cannot be cleanly pulled apart.

Host 1Once they had a brain age for every person, how did they go hunting for genes connected to it?

Host 2They ran what is called a genome wide association study. That means checking millions of single letter spelling differences in the genetic code, one by one, to see which ones are more common in people whose brains look older than their age.

Host 1And how did they measure lifestyle, did they follow people around with clipboards?

Host 2No, lifestyle came from what people reported, things like smoking, drinking, exercise and sleep, combined into an overall score. The abstract does not list the exact ingredients, so I will not guess at them.

Host 1You said earlier that some people had more than one scan. Why does that matter?

Host 2Because it lets them do two kinds of analysis. A cross sectional one compares different people at a single moment, and a longitudinal one follows the same people over time to see whose brains aged faster between scans.

What they found

Host 1Then what did they find in the genes?

Host 2Nineteen independent places in the genome linked to brain aging, which mapped onto one hundred and twenty two genes. Independent means each place is a separate signal, not just a neighbour riding along with another one.

Host 1Does nineteen spots mean there are nineteen genes for an old brain?

Host 2No, and that is a common misreading. Each spot nudges brain age by a tiny amount, and in the wider field these traits are understood as the sum of thousands of such nudges, of which nineteen are simply the ones strong enough to stand out in this sample.

Host 1And what about the people with more of these risky versions, or with worse habits?

Host 2Both groups showed faster brain aging, in the snapshot and over time. People with higher genetic liability, meaning they carry more of the risk versions added up, and people with an unhealthy lifestyle, both tended to have brains that looked older and aged faster.

Host 1How large was that effect, in years of brain age?

Host 2The abstract does not give the size in years, and I would rather say that than invent one. What it does say is that both predicted acceleration, so the direction is clear but the size is not something we can report from it.

When habits share genes with the brain

Host 1There was something about a genetic link between lifestyle itself and brain aging, which sounds odd. How can a habit have genes?

Host 2Habits are partly heritable too, through things like taste, temperament and how easily people get hooked on things. The team used a method called linkage disequilibrium score regression, which compares the genetic fingerprints of two traits across the whole genome to see if they overlap.

Host 1So if I have this right, some of the same genes that nudge people toward healthier habits also go with younger looking brains?

Host 2That is the most natural reading of their negative genetic correlation, yes. It also means a warning: some of the link between habits and brain age may be shared genetics rather than the habits themselves doing the work.

A word from our sponsor

Host 2This episode is brought to you by Lost Tuesday, the scented candle that quietly burns away one Tuesday you did not need.

Host 1Unlike some jars that sell you an extra day, Lost Tuesday is a properly controlled subtraction.

Host 2Tested on forty six thousand Tuesdays, with an R squared of zero point seven three against the Wednesday control group.

Host 1I lit one last month, and I still have no idea what I did on the twenty first.

Host 2Side effects may include a missing dentist appointment. Lost Tuesday: every hour it burns comes back to your brain as exactly one year younger, fully preregistered.

One spot that listens to lifestyle

Host 1Let's get to the single gene that changed with lifestyle. What did they find there?

Host 2They scanned the whole genome a second time, now asking where the effect of lifestyle depends on which version you carry. One spot passed their threshold, a variant with the catalogue number six two two eight seven zero nine six, sitting just upstream of a gene called HRASLS.

Host 1What does that gene do, in words a non-biologist could follow?

Host 2Looking beyond this paper, that gene is known to make an enzyme involved in handling the fats that build cell membranes. The paper itself does not prove why it matters for the brain, so treat that as background, not a finding.

Host 1And what was special about the people who carry the risky version of that spot?

Host 2Their brains seemed more sensitive to lifestyle in both directions. With poor habits, carriers showed bigger brain aging than non carriers, but with healthy habits, carriers showed less brain aging.

Host 1So if I have this right, that version isn't simply bad, it just makes your brain react more strongly to how you live?

Host 2Yes, that is precisely the pattern they describe. Some researchers call this a sensitivity variant rather than a risk variant, because whether it helps or hurts depends on the surroundings.

Host 1How do they know this spot is doing anything at all, rather than being a statistical coincidence?

Host 2They took it to the lab. Using a dual luciferase reporter assay, they showed the stretch of code around the variant acts like an enhancer, a dimmer switch that turns nearby genes up, and the two versions turned it up by different amounts.

Host 1What on earth is a luciferase reporter assay, and does it involve fireflies?

Host 2It does, in a way. Luciferase is the enzyme that makes fireflies glow, so researchers attach a piece of human DNA to the glow gene inside cells in a dish, and the brighter the cells, the more that piece of DNA is switching genes on.

Host 1Cells in a dish are a long way from a person's head, though. What does that experiment actually prove?

Host 2It shows the variant can change gene activity, which makes it more than a random statistical blip. It does not show that this happens in human brain cells, or that it changes how a real brain ages.

Why it matters

Host 1Pulling back, what's the big idea someone should walk away with from all of this?

Host 2That genetic risk for an older looking brain is not a fixed sentence. In this data, lifestyle was linked to brain aging on top of genetics, and for at least one spot, the people with the riskier version seemed to gain the most from healthy habits.

Host 1For an ordinary person listening, does this change anything about what they should do tomorrow?

Host 2Honestly, very little. The usual advice, not smoking, moving more, sleeping well, drinking less, was already supported by the wider field, and this paper adds a modest extra reason, not a new instruction.

Host 1Could I get a genetic test now and find out whether I'm one of the sensitive carriers?

Host 2Not in any useful way. One variant with a small effect, found in one study, is nowhere near ready to guide individual advice, whatever the abstract's phrase about precision strategies might suggest.

Host 1Who might actually change what they do because of this paper, then?

Host 2Mainly other researchers. It gives them a specific spot to test in other groups of people and in brain cells, and a reason to study genes and habits together rather than in separate boxes.

The honest caveats

Host 1What are the biggest reasons to be careful with these results?

Host 2The first is the volunteers. The UK Biobank is known to be healthier and wealthier than the general population, and mostly of white European ancestry, so the findings may not carry over to everyone.

Host 1And lifestyle was self reported, so couldn't people simply have been flattering themselves?

Host 2Yes, that is a real weakness. People tend to underreport drinking and overreport exercise, and that kind of error usually blurs results rather than creating them, but it adds noise.

Host 1Does the shared genetics you mentioned mean we can't say that habits cause the brain to age?

Host 2Correct, this is an observational study, so it shows links, not cause. People with healthier habits may differ in many other ways, from income to education, and the genetic overlap is one more tangle in that.

Host 1You said gene by lifestyle interactions are hard. How often do these findings hold up when someone checks again?

Host 2In the wider literature, not often, which is why the field treats single interaction hits with caution until they are repeated in an independent group. The abstract does not mention an outside replication, so this one should be seen as a promising lead, not settled.

Host 1Is there anything tricky about the brain age measure itself?

Host 2There is a known quirk. Age guessing programs tend to overestimate the age of young people and underestimate it in older people, and studies have to correct for that, so how well that was done matters a great deal for the final numbers.

What to watch next

Host 1So what would you be looking for next, to know whether this holds up?

Host 2Three things. Whether that spot near HRASLS shows the same pattern in other large scanned groups, especially beyond European ancestry, and whether it changes gene activity in actual brain tissue.

Host 1You said three things, so what is the last one you would watch for?

Host 2Trials. The real test would be whether changing habits, not just having them, slows brain aging on a scan, and whether carriers of that version benefit more than everyone else.

Host 1If you had to sum the whole paper up in one honest sentence, what would it be?

Host 2A large British study found that genes and lifestyle both line up with how old a brain looks, and one genetic spot may make some people more sensitive to their habits, a careful first step that now needs repeating.

Sign-off

Host 1That's the deep dive for this week. Every weekend this same feed carries the roundup, the week's new papers in one short episode. Thanks for listening to the Hastrup Review of Science.