Join 3,415 readers in helping fund MetaFilter (Hide)

99% chance of Alzheimer's Disease
April 21, 2014 6:01 AM   Subscribe

Chatting casually with a bioinformatician yesterday, I was told that I have 99% chance of developing Alzheimer’s disease, based only on my mother and father’s nationalities. How should I take this claim?

Yesterday I was talking to a friend of a friend, who is an academic bio-informatician. Early in the conversation I happened to mention that I have an English mother and a father from the Caribbean island state of St Vincent and the Grenadines.
Later in the conversation he told me, on the basis of the above information that I have a 99% chance of developing Alzheimer’s disease in later life. (I wasn’t asking.)

My question is, does he really have a sound basis for making that claim, given the information I gave him? The estimate of the probability of me developing Alzheimer’s disease is 99% conditional only on my parental nationalities?

Obviously I would like the answer to be no. But this is this guy’s academic speciality.

I did not tell him this: but, as far as I know, all of my grandparents lived to their 70s and 80s and did not develop Alzheimer’s disease. Also, I did not mention to him that my maternal grandmother was not English but Austro-Hungarian. (Presumably, he was making certain assumptions about my parental heritage.)

Perhaps there is a more nuanced interpretation which he skipped over. Like, “conditional upon not dying of something else first you have a 99% chance of developing Alzheimer’s disease from age X” ?
Could there be a wide confidence interval for this probability?

(The guy also said he himself had a 95% chance of developing the disease –but he had analysed his own genetics.)

So, how seriously should I take what he said, and how should I interpret it? (I should have dug deeper in the conversation but I was kind of shocked.)
posted by mister_kaupungister to Science & Nature (27 answers total)
I would assume that he was fucking with you. That sounds totally dubious, and also like a majorly unethical thing to tell someone if it were true.
posted by ArbitraryAndCapricious at 6:03 AM on April 21 [74 favorites]

Yeah, I'd assume he was bullshitting you for whatever reason. The next time you see him you might mention that that's a pretty cruel and stupid thing to say to someone whose genes you haven't actually analyzed and - most importantly - who hasn't asked.
posted by rtha at 6:09 AM on April 21 [9 favorites]

Sounds like this guy is a grade A asshole.

He doesn't know enough about you to be able to tell you such a thing and I agree with ArbitraryAnd Capricious that it's WILDLY unethical to tell you in an informal setting.

As you mentioned, he knows nothing of your actual genetic makeup, just some random stuff.

We did 23andMe. My Mom has been saying that she's half-Hungarian her whole life. Imagine her shock when she discovered that while her people may have settled in Hungary, that she has NO Hungarian DNA at all. We're pretty much Ashkenazi Jewish all the way back to Adam and Eve.

Discount this entirely and I'd think about calling this guy's institution and reporting his ass. Because this is pretty heinous behavior.
posted by Ruthless Bunny at 6:09 AM on April 21 [13 favorites]

Worth checking out, but do it in a professional setting with a medical professional, not a math professor. What this person did, in the setting they did it in, automatically discounts by 99% the accuracy of their 99% prediction.

Anybody who, in a casual setting, would make such a claim, is not to be trusted other than the have a penchant for upsetting people without cause.

christ, what an asshole
posted by lampshade at 6:14 AM on April 21 [4 favorites]

That guy is a monumental turgid DICK. And even if he had a printout of your genetic information in hand - which he did not - saying you have a 99% chance of ANYTHING is, 1. Inaccurate, 2. Unscientific, 3. Immoral, and, 4. Unethical.

My credentials: I have a not-good genetic mutation. Every time a geneticist or associated professional has discussed it with me, they've used ranges of risk ("40% - 60% by age 70", NOT point estimates, e.g. "99%", which are less accurate), and they've been gentle and measured in their delivery.

I would seriously contact his institution and let them know. You seem to have responded to his statement well (with skepticism), but not everyone necessarily would.
posted by julthumbscrew at 6:19 AM on April 21 [3 favorites]

Another vote for someone who is totally off base and WILDLY unethical. First off, there is no way anyone could give you that specific an estimate based on nationality of your background alone. NO WAY. More importantly, there are a raft of guidelines for ethical clinical genetic work and part of that is meeting with a counselor who will walk you through the kinds of tests they are proposing doing, what can and cannot be uncovered in that process and who will give you the option of choosing when and how to procede for testing. So, for example, you might be getting a work up for X syndrome and the kind of test may have a chance of turning up something else. In that case you would always be allowed to choose whether you want to pursue knowing about the additional diagnosis at all and what the benefits or drawbacks are.

I'd definitely tell this guy what he did was awful and totally agree that you're within your rights to contact his institution to let them know that he's doing this kind of stuff.
posted by goggie at 6:22 AM on April 21 [3 favorites]

There does appear to be some increased risk of Alzheimer's among Carribean Blacks and Hispanics. I have no idea if that includes you, and a casual read does not suggest that the estimate of 99% is anything like accurate.

I would seriously contact his institution and let them know. You seem to have responded to his statement well (with skepticism), but not everyone necessarily would.

He was not representing "his institution," and was engaged in a casual conversation. I'd just steer clear of the guy.
posted by JohnLewis at 6:28 AM on April 21 [6 favorites]

This man is a jerk, and you should pay no attention at all to what he said.

The most generous interpretation we can make here (and I'm not sure it's worth being generous!) is that Alzheimer's is one of those diseases, like prostate cancer, with fairly high incidences in those of us who live long enough. Ballpark figures for uk residents are something like 1 in 6 people over 80 showing some signs of dementia, with about half of those cases caused, probably, by Alzheimer's specifically (so that's still less than 10%, but pretty high odds compared to most medical conditions.) that's not to say that even 10% 'suffer' from the disease, just that there are some symptoms in life, or signs of the disease in death.

The genetics of late onset (ie after age 60) Alzheimer's are still vague enough, and the disease still clearly environmental enough that it seems very unlikely for any test he claims to have had to have given him a 95% chance of developing it (see the NIH fact sheet)

Tl;dr ignore this idiot.
posted by AFII at 6:34 AM on April 21 [3 favorites]

I actually have been looking into probabilities and dementia - sadly, a close relative has an early-onset form of Alzheimer's. It seems like family history is a huge determinant. In certain small populations, I suppose nationality might serve as a proxy for family history, but I have not seen anything in the research I've read to suggest that someone with no family history of AD would be at a 99% risk...what would that require? Check out this CDC reference page on AD and genetics - it's clearly pretty basic (and you can use that info to read more seriously) but there's nothing about simple direct inheritance on there. For you to have this 99% chance of AD based on no family history and two particular parents, it would have to be something really simple, where each parent always has "half" of the genetic conditions needed for AD and those genes getting passed down in a super-simple way to the children. (Like, every person from each population would have to be a "carrier", but not develop the disease themselves.) That's just not how AD works - early onset AD is the variant that is autosomal dominant, and you would absolutely know if you had early onset-AD in your family, because it is an awful, tragic thing and even a sporadic case is a huge situation. They seem to have some ideas about how late-onset AD works, but it is clearly a lot more complicated.

Whatever this person meant (as far as I can tell from researches that have been, like, a big deal to me personally) what they ended up conveying to you is not accurate.

IANAG, of course.
posted by Frowner at 6:35 AM on April 21 [1 favorite]

It sounds to me like he's a guy who has a fatalistic view of things and exaggerates conversationally. That's not a clinical opinion, it's a conversational opening to "Hey, we're all kind of screwed and in the same boat, right?"

I'd expect the same type of individual to sarcastically point out that life has a 100% mortality rate. This wasn't a medical opinion.
posted by mikeh at 6:36 AM on April 21

I tend to agree with the position that making the decision to convey this to you in the way he did casts such doubt on his judgment and good sense that the relevance of his information drops to a very, very, very low range.

But if you're nervous about it, ask someone who will discuss it with you in a professional way. That's not what this is.
posted by Linda_Holmes at 6:49 AM on April 21 [1 favorite]

Did he say when? There are some that view dementia in the same way geneticists view cancer: It's inevitable provided you live long enough. Perhaps it was a trick, designed to test whether you noticed what wasn't being said.
posted by kisch mokusch at 6:51 AM on April 21 [7 favorites]

He was bullshitting you. There's no level of risk specificity that could be accurate to within 4% (the difference between your supposed 99% risk and his 95% risk) that operates purely on the basis of parent's reported country of origin.
posted by gerryblog at 7:35 AM on April 21

Another thing that occurs to me - this dude might very well be an academic bioinformatician whose specialty is something that is not at all related to predicting AD. Academic specialties are....pretty specialized. And sometimes people who are experts in one thing assume that they are also experts in a related thing just because they're smart people who are fairly well-informed.
posted by Frowner at 7:42 AM on April 21 [2 favorites]

This man is an idiot. The idea that you can predict who will get Alzheimer's based on nationality is complete and utter bollocks. He is either just flat out lying to you to get a reaction, or if I am being very generous, possibly doesn't understand that there is a difference between senile plaque formation (which is pretty common if you live long enough, and means nothing in itself) and clinically detectable Alzheimer's. At best, "if you live to be 100 there's a 99% risk of you having microscopically detectable amyloid plaques at post-mortem", which is pretty fucking different to a 99% chance of getting Alzheimers, and even given that proviso 99% seems really high to me (plaques detectable in 60% of over 80s at post mortem is the figure I've heard quoted, and only 8% of over 80s actually have Alzheimer's). I'm not sure if he doesn't actually understand this, or if he does but has dumbed this down to the point of meaninglessness, or if he's being a dick, but either way it's not something that you need to worry about.

I mean, it is flat out untrue that 99% of people in the UK get Alzheimers. Risk increases with age, but only 30-35% of over 95yr olds have some kind of cognitive impairment, of which about half (ie 15-18%) will be Alzheimer's. Similarly it is not true that 99% of people in St Vincent will get Alzheimers. I'm not aware of any country where 99% of people develop Alzheimer's. In fact the incidence of dementia may have gone down recently, which lends weight to the fact that there is a strong environmental component to the disease.

Hopefully this factsheet will be interesting, it's for a UK audience and certainly does not support the idea that having UK parentage guarantees dementia.

In terms of this being his academic specialty, bioinformatics is not clinical medicine, clinical genetics, neuroscience or epidemiology, so no it is not. I have worked with bioinformaticians who were great at bioinformatics (and were far too lovely to have done this kind of shit) but who could not tell you any more about the specifics of the disease they were working on beyond the name, because that isn't what they do.
posted by tinkletown at 7:50 AM on April 21 [5 favorites]

I've read a lot about Alzheimer's disease the past few years since my grandmother was diagnosed with it. What he told you doesn't seem consistent with anything that I've read. Based on what I've read, my impression is that why some people get the disease and others don't is still not well understood, and that the chance of any individual getting it is the result of many many factors both genetic and non-genetic (and interactions between the two), most of which probably haven't been identified yet.

Some of the genetic susceptibility factors might be more common in certain ethnic groups but (1) you still wouldn't be able to tell if any one individual had any of them just based on nationality or ethnicity, and (2) none of them as far as I know give you anything like a 99% risk.

Of anything that's been identified so far the genetic factor that apparently has the biggest effect is the ApoE allele. But even if a person has the highest risk version of that (two copies of the E4 version), it's still not a 99% chance, and furthermore, it wouldn't be possible to know that you had that combination based just on what you told him about your parents' nationalities.
posted by treese at 8:11 AM on April 21 [1 favorite]

It's absolutely unethical for someone to make a statement like this. Obviously, from your own observations, there is no dementia in your family. So it's patently BS. But it's really scary to hear such a thing. You can increase your chances of having a healthy brain (and body), so live a healthy life, which has a myriad of benefits. Also, avoid assholes like this guy.
posted by theora55 at 8:18 AM on April 21

99% of the time that someone claims a 99% chance of something, it is pulled directly out of their ass. It's a red flag number, and it usually just means that someone subscribes to a notion really hard.

Plus, statistical data is NOT presented like that. If there really were some data showing that your specific racial makeup were at an increased risk of disease, it would be something like "X% of subjects who reach a specific age, in these specific circumstances (current location or something) are diagnosed with this disease by some different age, with this degree of confidence" or something like that. 99% is crap. About 96% of people with Alzheimers develop it after age 65, and I cannot figure out what percentage of people live to age 65, but I'll betcha it's already less than 99%, so there would have to be another really wild, really specific statistical anomaly with mass early onset Alzheimers for that to even be possible.

So all of that isn't 100% impossible (only 99%), but if something that pronounced and specific were actually happening in reality, it just seems like it'd be huge news that everyone would know about, and someone probably would have contacted you already asking to study you.

Tell your friend's friend to stop messing with you and being a jerk. He's either a bad comedian or a bad scientist.
posted by ernielundquist at 8:22 AM on April 21 [3 favorites]

If that were true, 99% of all children with both English and Caribbean parents would develop Alzheimer's Disease. That obviously isn't the case.
posted by the jam at 9:05 AM on April 21 [4 favorites]

This guy was wrong.

Um...not even people with extreme early-onset AD genetic markers (very rare markers) have a 99% chance of AD.

The funny thing about genetic disease markers is that if you have the disease, then the marker tells you a cause. But if you have the marker, it is not certain of the outcome. There are exceptions to this, but no AD marker is one of them.

I am most certainly not a bioinformatics expert, but I do work in genetics and genomics for a living.
posted by BearClaw6 at 9:38 AM on April 21 [2 favorites]

"you're within your rights to contact his institution to let them know that he's doing this kind of stuff."
Are people really suggesting that you inform on this guy to his employers because he said some weird comment in a private conversation? That's ridiculous. If you're really that worried about this, get a genetic test done and investigate your family's medical records. Otherwise, forget about it and don't talk to the guy.
posted by Ideefixe at 9:39 AM on April 21

Bullshit. Alzheimer's has a heritable component, but it's nowhere close to 99% heritable and even if it were there's no way that the trait would be fixed in both British and Caribbean populations. Genetics is almost never that clear-cut about anything, and if it were that clear-cut about Alzheimer's then it would be common knowledge and the disease would be much easier to understand and treat than it is.

If predicting who was going to get Alzheimer's were as simple as knowing where their families came from then the state of research and clinical practice around the disease would look very different than it does. Molecular genetic pathology is not my specialty (I work in evolutionary genetics, and in frogs rather than in humans, and I'm only a grad student) but I know enough to tell you with certainty that this guy was being stupid. We know that Alzheimer's is caused by a complex suite of genetic and environmental factors, some of which we understand and some of which we don't. This is how most things work – simple determinants are very much the exception in genetics, not the rule. It's a lot more complex than just knowing where your folks are from.

The guy you were talking to was shooting his mouth off and should've known better. Scientists of all people should know how to hedge their words on a complex and incompletely-understood subject; it should be second nature, really. Was he drunk at the time? Because he sounds like he was just waving around what I like to call his "science cock". That is to say, he was giving in to the temptation to use his technical knowledge of an esoteric subject and his perceived authority as a scientist to make himself appear wise and powerful.

Telling people how they're going to die is something that people have claimed to be able to do since forever. Scientists should know better than to pull that crap, but we're human and that means that some of us are assholes and/or insecure people who need to feel superior to those around us in order to feel good about ourselves. It's ugly, but it happens. Scientists can be as conceited as anyone else; perhaps it's even more common in our community than in the general population.

Nobody can know whether you're going to get Alzheimer's or not. A qualified genetic consultant or gerontologist might be able to give you a picture of your level of risk and what if anything you could do about it if they had access to enough information, but nobody can tell you anything just from a casual conversation. They'd want to interview you about your family history, your personal medical history, and your lifestyle. They'd also want to run some tests. Even then, what they would give you would be a qualified estimate that would be considerably more complex than just saying "you have an x% chance of getting Alzheimer's Disease". It would be complex enough that they'd probably want to sit you down and chat with you about the results to make sure you understood what they were and were not able to tell you.

Anyway, rest easy. You don't know anything more about your chances of getting Alzheimer's than you did before you met this bioinformatics douche. I'm sorry that he threw you for a bit of a loop; I can totally understand how that would happen. He was being stupid, and unscientific, and an asshole, and he should've known better. A very charitable interpretation would be that he was being carelessly hyperbolic and incorrectly assumed that you would pick up on that, but I think there was a general helping of callous, condescending conceit in there as well. If you're genuinely concerned about your risk of the disease then you should talk to your doctor about getting a referral to an appropriate specialist who can give you real (if complicated and fuzzy) information instead of bold and inappropriate assertions.

We (society as a whole, I mean) do know some things about who is more and less likely to get Alzheimer's, and if you really want to know then your doctor could probably put you in touch with somebody who could give you the best picture that contemporary medical science can provide. It wouldn't be anything like "You are 99% likely to get Alzheimer's" though; it would instead be more like "Based on what we know about your personal and family history, and the results of the tests we've run, and assuming that we're not missing any important information (which we might be, because our understanding of this disease is not great and our knowledge of you is also imperfect) we can say with reasonable confidence that your chances of developing Alzheimer's (assuming you live long enough and don't die for some other reason instead) are somewhat higher than the overall population average, meaning that you have about 2-5% chance of developing it by age 75, a 20-30% chance of developing it by age 85, and a 50-65% chance of developing it by age 95." I just made all that up by the way, those figures were just pulled out of the air to illustrate my point and bear no real relation to anything. It'd probably actually be a lot more complicated even than that, and would involve you being given a fat information packet and a 30-60 minute consultation session with your specialist as they explained the results to you.

I hope that eases your mind somewhat. Unfortunately, nobody can tell you for certain whether or not you'll get Alzheimer's. Fortunately, nobody can tell you for certain whether or not you'll get Alzheimer's. That's science for you. ;-)
posted by Scientist at 10:30 AM on April 21 [4 favorites]

Oh, and feel free to contact this guy's institution if you like. I guarantee though that they won't care even a little bit and that your complaint will go nowhere. Academic researchers are not held to any kind of standard of ethical or professional rigor in terms of using their technical knowledge to make assertions in casual conversation.

If he'd said something like that in a journal article then nobody would publish it. If he said it at a presentation then his peers would lose respect for him. If he said it in front of a class then someone could complain and if he made a habit of it it might come up at a tenure review. However, saying it to a random person in casual conversation just means that he was being kind of a dick.

If you call his school to tell them essentially "hey, one of your professors was kind of a dick to me at a party a while ago" then they will just think you are weird.
posted by Scientist at 10:34 AM on April 21 [1 favorite]

Ok so the 99% (and 95%) figures are baseless. The background to this conversation was him telling me about how useful and forward moving his field is, so I put this down to a weird attempt to impress this upon me. I'm not going to call his institution as some of you suggested, we were just chatting at a bar.

But thanks everyone for clarifying things and some interesting links.
posted by mister_kaupungister at 10:45 AM on April 21 [1 favorite]

Ha, so definitely just "waving his science cock around" then (great phrase!).
posted by tinkletown at 1:10 PM on April 21 [4 favorites]

Just chiming in to add that I'm an academic bioinformatician (it says "bioinformatics" on my business cards, so it must be true!) and while I don't work in human genetics, I think this guy is full of it. First, ancestry is not usually a good proxy for genotype at an arbitrary locus to begin with, unless you were born to parents from an extremely isolated population (not true for either of your parents). Second, predicting diseases with non-Mendelian inheritance and incomplete penetrance is very hard -- genetic associations can currently only explain 23-33% of the heritability of AD [ref 1, 2]. Third, ApoE4/E4 homozygotes, who have the absolute highest risk of late-onset AD, are only 14% of the population in Northern Europe, and their risk is still less than 90% [ref]. Fourth, you're male, which means your risk is lower than it would be if you were female.

From the way you're describing it, it sounds like he may have gotten defensive and insecure about the utility of bioinformatics and overcompensated as a result. As background, it's not uncommon for wet-lab scientists to be particularly skeptical of bioinformatics, and as a result, bioinformaticians often get put on the spot and made to justify their existence. But throwing out inflammatory, hurtful, and inaccurate statements is not only being a terrible advocate for the field but also a total dick move. He owes you a retraction and an apology.
posted by en forme de poire at 4:42 PM on April 21 [5 favorites]

Ideefixe: "you're within your rights to contact his institution to let them know that he's doing this kind of stuff."
Are people really suggesting that you inform on this guy to his employers because he said some weird comment in a private conversation? That's ridiculous. If you're really that worried about this, get a genetic test done and investigate your family's medical records. Otherwise, forget about it and don't talk to the guy.

If we were talking about a plumber, painter, or lawyer, I'd agree.

This man is a professional in a particular field, and made a professional-sounding statement within that field's context. If I were his boss, and knew he was playing fast & loose with his professional reputation (and by implication our company's), I'd care.

There's a reason lawyers don't walk around sayng, "Do it, man! As a lawyer, I'm telling you it's totally legal!" to everybody. There might be professional repercussions.

His statement was unethical, unprofessional, and damaging. I have no problems with him facing professional consequences for it.
posted by IAmBroom at 11:54 AM on April 22 [3 favorites]

« Older I want to upcycle this glitter...   |  I went all feminist hulk and t... Newer »

You are not logged in, either login or create an account to post comments