Why 23andMe Genetic Testing Is A Waste Of Time And Money

Remember last year when our brothers in the US offered a discount to Gizmodo readers for genetic testing service 23andMe? Yeah well I totally did it, and I kinda wish I hadn't.

Access to your genetic data used to cost thousands of dollars, but with the rise of online genetic testing services, like 23andMe and Navigenics, both of which are available to Australians, it's become more accessible to people like you and me.

23andMe is financially backed by Google co-founder Sergey Brin. It's at the forefront of technology that makes genetic data accessible for the average person by bringing the cost down from what would be thousands of dollars to just a couple of hundred. I paid just $US99 with an offer for Gizmodo readers, plus a mandatory $US5 per month for a 12-month personal genome service subscription.

I tweeted about my plans and found that people were equally divided over whether I should or shouldn't have signed up for it. At no point did I expect this to become a moral debate. It was my decision, and mine alone, but I found myself trying to justify it to people I didn't really know, as well as to family and friends.

Some said that there are certain things in life that are best left unknown. Others said that knowing where my risk factors are could improve my quality of life. My boyfriend was one of those people who thought it was a bad idea. "F**k that!" he said. "Do you want to live the rest of your life worrying about something you may or may not get?"

He had a point, and if anyone's opinion mattered, it was his. I spat in the tube and put it aside for more than a month, deliberating over whether I should or shouldn't ship it off to California for analysis. In the end, I decided that once my results were in, I could choose to not view them if I didn't want to. After all, I had already paid for the service.


I'm a twentysomething woman with mostly Asian ancestry. My paternal grandfather was of mixed heritage, but other than that I'm pretty much all Korean. I sometimes wish I was a couple of inches taller, or that my boobs were bigger, or that I was better at maths, but those complaints are trivial when I know that I have my health and expect it to stay that way for the foreseeable future. Or so I thought.

23andMe kind of ruined that illusion for me. Apparently I have a 5.1 per cent risk of getting colorectal cancer, and my body doesn't do a very efficient job of metabolising caffeine. I don't really drink coffee, but I do like my Red Bulls from time to time! Yes, 5 per cent is a small figure to worry about, but you can't help but be alarmed when a hard number is put on something you'd think should not be so easily predictable.

The Law

According to the Private Health Insurance Act (2007), health funds cannot refuse to insure you based on age, sex, health risk profile or genetic profiling. My own health insurer, nib, promises on its website that a genetic testing service will not impact my health premiums.

However, life insurance is a different story. According to the Insurance Contracts Act (1984):

Subject to this Act, an insured has a duty to disclose to the insurer, before the relevant contract of insurance is entered into, every matter that is known to the insured, being a matter that:

(a) the insured knows to be a matter relevant to the decision of the insurer whether to accept the risk and, if so, on what terms; or

(b) a reasonable person in the circumstances could be expected to know to be a matter so relevant.

As part of my investigation, I spoke with Keith Henry, a broker specialising in risk insurance. He confirmed that applicants are required to disclose information that could impact on insurability, but he reassured me that insurance companies generally do not request that genetic testing services be done.

"No application forms and no underwriting that I've been involved in over the last 32 years have requested a genetic test," Henry said.

I'm pissed off that 23andMe didn't make this information clearer prior to the signing-up process. All they say on their website is that the Genetic Information Nondiscrimination Act (GINA) "protects Americans from discrimination (in health insurance and employment decisions) on the basis of genetic information". This is obviously unhelpful to me as an Australian.

If fact, they say nothing about duty of disclosure in matters involving life insurance other than this sneaky little statement: "GINA does not cover life or disability insurance providers", which, again, is irrelevant to anyone who isn't American.

It's dodgy behaviour, at least in my eyes. They know that if we know the legal implications of undergoing genetic testing, fewer people would sign up, which would obviously affect 23andMe's commercial interests.

The Report

That's not all that 23andMe failed to mention. It's not until after you pay the money, send off your sample and get the results back that you realise that most of the data is irrelevant to you.

The first time you log in to see your results, a dialog box pops up with the option to not see your health results. Of course, by now, I was dying to find out what the results were. I clicked through to my "disease risk" profile and saw this:

I almost had a heart attack right there. This made no sense. I was 1.21 times more likely than the average person to suffer a heart attack, despite having no family history of anything remotely related to heart disease.

But then I clicked on the link that took me to a detailed report and saw this:

The data is based on the assumption that I am of European ancestry, and between the ages of 40 and 79. I rolled my eyes. There was no option for other ethnicities, because there are "comparatively few studies looking for genetic associations in populations that form a minority in the countries where much of the latest research takes place".

My eyes glazed over as I scanned through the rest of the detailed "heart attack" report, including ways to reduce risk, a chart showing "marker effects" that I still don't understand, and a highly technical explanation of chromosomal region 9p21 that is, again, only applicable to people of European descent.

23andMe fact: The heritability of death from a heart attack is estimated to be 38 per cent for women and 57 per cent for men. This means that genetic factors contribute slightly less to this condition than environmental factors in women, but contribute slightly more in men.

Next on the disease risk report was one of the few on the list that actually was somewhat relevant to me. Apparently, I have an elevated risk of colorectal cancer, assuming that I'm Asian and between the ages of 25 and 79. I was 23 at the time of taking the test, and that's about as relevant as the whole report gets to my ethnic/age group.

Again, despite having no family history of colorectal cancer and being outside the assumed age range, this struck a chord with me. Heritability is estimated to be at 35 per cent, which means environmental factors play a bigger part in my odds of getting it or not. The report recommends that I get regular screening after the age of 50, exercise, don't get fat, don't drink too much alcohol and limit my intake of red meat. Doesn't sound too bad -- there's not much I have to change already, although I could probably do with more exercise.

Most of the health results are based on studies that compare different "markers". For example, my risk of keloid -- the formation of dense, painful and persistent raised scars -- varies depending on the genetic variation. According to the quoted study involving people of Japanese ancestry, one gene marker is at "moderately higher odds of keloid if skin injury occurs", while another is at "typical odds".

It appears that I also have a gene that puts me at "slightly higher odds" of schizophrenia, according to a study of 4200 Japanese people. This worries me a little bit, as I currently take medication for anxiety and have a family history of depression.

The report also revealed that I have a reduced ability to break down a toxic byproduct of alcohol and cigarette smoke called acetaldehyde, which may or may not explain why I can only have a few cigarettes or a few drinks before I start feeling ill. Apparently, I also have a gene that puts me at "substantially higher odds" of heroin addiction, which may or may not explain why I've always had a gut feeling about staying the hell away from it in a way that I don't feel about other illicit drugs.

If you can garner any good news out of 23andMe's genetic testing service, it's the "decreased risk" section of the results. Again, a lot of the results are irrelevant to me as a non-European person under the age of 30. Interestingly, I have a gene that puts me at lower odds of stomach cancer. Considering that my anxiety usually manifests itself in stomach pains, this can only be a good thing.

This section of the report says that I'm less likely to experience irregular heartbeats, fertility problems, lumbar disc disease, and a condition called tardive dyskinesia that causes tics in people who take or have taken antidepressants. Also of note, my genetic data puts me "typical odds" of asthma based on a study comparing 487 Koreans with asthma to 286 people without asthma. I found this interesting as a sufferer of childhood asthma, but other than what I mentioned above, that's all that the health report had to offer.

The "carrier status" part of the report revealed that I didn't have any genetic predispositions to a list of 24 diseases, and the "drug response" section showed nothing alarming, other than reduced odds of responding to Hepatitis C treatment and that I'm likely to be a slow metaboliser of caffeine. The report also says that I can taste certain bitter flavours (you mean some people can't?), have dry earwax, am likely to be lactose intolerant, share a genotype with "many world class sprinters", reduced sensitivity to the smell of BO, decreased sensitivity to pain and less efficient at avoiding errors. I guess you win some, you lose some.


23andMe also failed to adequately inform me that there's no way to trace my paternal line because it's "determined by the genetics of the Y-chromosome". Maybe this is common sense to some, but it isn't to me. If I really want to know my paternal ancestry, I would have to get my dad, brother or paternal uncle to do the test and cough up more money.

My maternal haplogroup was, however, revealed in detailed fashion. 23andMe delves into the history of my maternal haplogroup, and even provides a chart full of numbers and letters that don't explain themselves to ordinary people like myself. I'm not able to work out what any of it means, except that I might have distant relatives in South America and Siberia.

The Result

So what did I learn? That I can't drink more than one glass of milk a day without my stomach exploding; that I really should never touch heroin; that if I think my boyfriend smells, he needs to be put in the shower right away; and that if I get fired for being inefficient at avoiding errors I could maybe seek career advice from Usain Bolt.

The report, despite being overwhelmingly positive, did nothing to make me feel better. If anything, it's made me feel worse. If I do happen to contract Hepatitis C one day, I'll probably freak out to the point of being inconsolable. Every time I have a drink now, I'll worry about my body's "reduced ability" to break it down. That's no way to have a beer.

My suggestion is that unless you have a specific reason behind your motivation, don't bother with genetic testing services, especially if you're not of European ancestry and you're under the age of 30. I can see how it may be useful as a preemptive measure for people with a family history of a debilitating disease, or for people who are seeking out long-lost relatives. But for the rest of us, the money is probably better spent towards a gym membership, more fruits and vegetables and health/life insurance. And not having to worry for the rest of your life about something that may or may not happen to you? That's priceless.

WATCH MORE: Science & Health News


    Genetic Testing has always felt a little like Professor Farnsworth's Death Clock. Interesting in theory but as soon as you use it your friends start bagsing (baggsing?) your stuff.

    I have no moral objection to Genetic Testing, I just know that if I got it done it would be all I could think about for the rest of forever. For example there is a history of debilitating dementia in my family and if I saw my chances represented as a number it would cause me to go crazy much faster.

      As far as I'm concerned it's like standing in front of a fire exit at a party
      You're never blocking it as you have legs and are flammable, hell if anything if a fire goes down you're in the best position to gtfo

      In other words, you find out you're susceptible to dementia then you can do something about it as opposed to slowly finding out you're already in the throes of dementia

    Environmental and behavioural factors are far too powerful to ignore, or make light of. IMHO analysis like this is only useful when it's intimately personalised and/or only used in a very general fashion, such as looking for certain genetic disorders and metabolic indicators - things about which we can have some certainty.

    As your example illustrates there are plenty of demographics for which relevant health statistics are thin on the ground, making the prediction of environmental/behavioural-linked effects totally unreliable.

    I'm getting flashbacks to that movie gattaca.

    Good article!

      Same; Gattaca is a brilliant movie. Really interesting article, Elly.

      "I'm sorry, the wind caught it."

      Seriously, this is scarily close to the plot of Gattaca.

        >"I'm sorry, the wind caught it."
        >Seriously, this is scarily close to the plot of Gattaca.

        I don't get it. Why aren't they fined for littering? After all, we know whose DNA it was now!

    Very interesting report! :-) Although, it must weigh on your mind heavily. I think I would rather not know, I like surprises.
    But I wouldn't worry too much, those statistics look like only slightly elevated risks, but don't forget your more likely to be killed by a toppling vending machine than be eaten by a shark. So, live it up!! :D

      >I think I would rather not know, I like surprises.

      If you could know if a burglar was going to sneak up on you at night or not, wouldn't you want to know? Wouldn't you want to be able to prepare/arm yourself? Do you really think there's anything that happens after death? There is not. This is all we have. Make the most of it. That includes using all information available to you to make the best decisions in life.

    Good article Elly :)
    Do 23andMe say what they do with all this info? Is your email bout to be flooded by heart and colorectal medication spam??

    I don't see any particular reason to be worried there, Elly. Those sample groups are so small that the information provided is statistically unsound. 4200 Japanese people? Less than 800 people for the asthma conjecture? The most worrisome thing there is that you actually payed money for this :|

    Those numbers only have meaning if you give them meaning. Enjoy life. Kismet.

    If you were so apprehensive about knowing this info, why did you bother? You have to take responsibility for your own decisions.

    Blaming the service provider for not giving you more info when you could have easily researched it more on your own is your fault.

    Plenty of people have had 23andme tests and blogged about their results online. You could have investigated it more thoroughly.

    I hope you cancelled your subscription so you don't keep getting annoyed every time you log in to check the updates.

      I am taking responsibility for my decision. I was fully prepared to hear what my genetic data had in store for me. I just think it's dirty play by these companies to not be fully upfront with the ramifications of undergoing genetic testing. Perhaps I shouldn't be so disappointed since they're a business and profit is their number one motive. It's just not socially responsible.

      The subscription is a mandatory part of the service. I have to be subscribed for a minimum of 12 months.

        Great article but its pretty well known that once you find out this shit you have to advise your insurer about it. haven't you seen the michael moor doco that discusses how shody the insurance companies are about this type of stuff are in the US ? Why would you expect it to be any different here ?
        OK and this company saying they wont ever share this information with Insurers ? What about when the insurance companies take them to court and a judge tells them to hand it over. They have no option to.

          I think the nastiest possible scenario is that insurance companies are already doing their own tests without your consent.

        Lifetime subscriptions to the service have been available almost from launch, at a cost of $400 per account.

        to Elly re 23andMe

        quote you "

        The data is based on the assumption that I am of European ancestry, and between the ages of 40 and 79. I rolled my eyes. There was no option for other ethnicities, because there are "comparatively few studies looking for genetic associations in populations that form a minority in the countries where much of the latest research takes place".

        Actually, they wouldn't need to see you. If they knew what they were doing, they'd ID your genetics as Asian. I sense a con. It's own by a tech guru, after all, Sergay Brin

      Here here!

    I guess they aren't sequencing your entire genome are they? and they don't provide it on a bluray disk or anything?

      Nope, but they should!

        Re: Data on disk.

        You can download your raw data by going to the "account" drop down tab and left clicking "Browse Raw Data". There is then a tab on the top right labeled "download raw data".

        Regarding the data being not as useful because of insufficient data re. your demography, they're doing ongoing correlations using their "Google algorithms" 24/7. So your data will become more useful as time goes on.

        Many of your points are very accurate, but some are way off. The site very clearly has always stated that the paternal line is unavailable to females respondents. I think you must have either just missed that or ignored it or didn't understand it before your testing. I was tested around the same time as you and discussed this very fact with my coworker when she wanted to do the testing at the same time.

        I was disappointed in the ancestry side, but every health study you quote from the site was available to you to read before you even buy the test. You have to take responsibility to understand what you are signing up for. Especially when it's as sensitive and important as this. It's disingenuous of you to suggest that because you didn't read the provided information that it is somehow their fault for not spoon-feeding it to you more aggressively.

        Also Elly you clearly wrote this article without understanding the science around human genome sequencing. Sequencing an entire human genome as you say 23andme should provide is technically difficult, unreliable, not useful, and prohibitively expensive. For instance, when you wrote this the cost to do that was around $40,000 dollars US. One firm says they offer it now for $1000, but is known to be unreliable. Steve Jobs had entire genome sequenced for $100,000.

    i think this was a really brave article to write and it was very interesting to read and see how these resuts can affect you even when you actually get pretty good results.

    It's also pretty good of you to reply to the inane "well why did you do it then?" attacks, the providers of these tests offer this mysterious product and don't fully disclose the kind of information you might or might not be able to find just based on your gender or ethnicity. Without your article we wouldn't know that.

    With a brain like yours, you've got a great life ahead, don't let genetics get you down!

      Thanks. If this article can unveil some of the mystery surrounding genetic testing, then my job is done here :)

        Pretty unbalanced article. 23andMe possible are on a slippery slope depending on how they present the information but you also seemed to make every effort to pretend to be as ignorant and uneducated as possible so you could have a rant. Genetic testing isn't that complicated but it also needs further development because maybe the customer isn't happy with a simple percentage of risk either way. Leaving aside the geneticist comments which I am sure are technically accurate I thought the website did give sufficient information for me to take a balanced evaluation of my results. I didn't freak out either. Perhaps it would be better if Elly didn't go near such products which obviously irritate her so much.

    A rather honest and personal thing to share with the interwebz Ms Hart - kudos.

    The life insurance point is interesting - no insurer requires people to get it done (yet) but if you do get it done, you would potentially need to disclose the results.

      I used to work for an insurance adviser and whilst I am by no means qualified to talk about the subject in depth, I'd wager that any half decent adviser would be able to convince the underwriter to ignore this information.

    I wouldn't be too concerned with the results:
    1. These studies, like you have mentioned are fairly useless - although the number of studies screening different populations for variations which may infer a particular characteristic are growing, the sample sizes are simply too small to speculate on odds ratios on an individual basis.
    For example, the HapMap consortium which screens several different populations for variations does have a representative "Asian" population - however it is composed of Chinese people living in Utah!

    2. It is widely known that people from Asian descent have higher tendancies towards addiction (it is thought to have lead to the opium addiction among Chinese). From memory it's due to changes in an enzyme needed to metabolise the drugs (It was in a seminar I attended several years ago).

    3. Yes, there are particular markers, whether they be repeat regions used just for screening (like STRs) or SNPs which may be a location marker for screening or the actual contributing factor, which may suggest an increased risk for a particular characteristic, it is still just an increased risk over the general population. Genetics is generally only about 50% of the story and environmental influences make up the other 50%... so really, the ball is still in your court!

    Hope that helped!

      Interesting! Especially point #2. I had no idea.

      Re: #1, the Chinese data in HapMap is actually from Beijing -- it's the European-ancestry data that is from Utah. Whether these can still be considered "representative" of the broader group of Asians or Europeans, or Asians living in the US or Europeans in Europe, is a reasonable question though.

      Also, 4200 is plenty large enough to establish an odds ratio. However, Ania is right that it is still tricky to apply an odds ratio to an individual, given that they are based on populations and are mostly predictive of populations (like all probabilities).

      Well number 2 only relates to heroin or all drugs? I know for a fact that I am not going to be addicted to alcohol as I do not metabolize it well. As a person who gets sick from it right away I doubt I can get addicted to something that never gets me a good feeling in the first place.

      Does Point 2 have anything to do with addiction to gambling? I was in Crown Casino the other day and it was about 99% Chinese (assumption) and I was struck by the intensity of the gambling I saw. I did wonder if it was a cultural difference or genetic difference that produced so many gamblers. Surely everyone knows you can't beat the system?

      Elly, did you get your Neanderthal ancestry? I did 23andMe and I'm 2.8% Neanderthal and knowing this has made me look at things quite differently. Homo Sapiens aren't evolved from Neanderthals but are cousins, so I'm kind of like a hybrid and that's the coolest thing ever in my opinion.

    Yes and yes, a fine journey indeed and, most importantly, well told. But, at the end of all things, you seemed displeased with the results; i can see a certain sense of remorse here.
    i guess, so to speak, that's the point you make: life is about the journey and the destination is nothing more than an imagined, unreachable mirage beyond a never ending horizon into eternity and forever.
    Like the time i walked in the desert for three days looking for salvation; i found nothing until i found rescue from a kindly man in orange overalls.
    Destination is but cruel illusion; something we can never find. The journey is all we ever have at this very moment, right now. The present is everything, the past and the future just ideas outside of what is real at this point in time as it stands here. i never consider the past or the future, they are unimportant to me.
    it has been said, "Colours do not exceed five, yet all their variations cannot be seen. Tastes do not exceed five, yet all there variations cannot be tasted. The shih of battle do not exceed the extraordinary and the orthodox, yet all their variations cannot be exhausted."
    i guess life is too complicated to be rationalised and contained into the parameters of such a simplistic scientific quantification or analysis atop an inspired business strategy. i guess the world and all things is just - with the big picture in mind - bigger than that in all ways and always.
    Never consider tomorrow or its infinite possibilities; it is a dangerous brand of speculation and fear. Be present; right now, nothing else matters.
    Your old friend,
    Plainview (still looking for salvation in all the wrong places).

    Of course then theres the issue of the reliability of the tests overall. New Scientist has a disturbing article about these labs getting results majorly wrong, even when it comes to important stuff like Paternity testing...


    I checked the 23andme webpage and they have a page dedicated to education and they even provide a free demo where you can create an account and look around and see exactly what info you get:

    Too late now, but unfortunately it would have saved you time, money, and peace of mind if you had explored a little more.

    Hopefully others reading your blog will check out the demo so they can decide first if this is something they think is cool.

    I think they did more than enough to help people make an informed decision. To call it dirty play is a bit extreme. You weren't tricked into it. It was your choice.

    Elly, your genetics have not failed you. If that's you in the picture, you're very pretty. Coupled with your geek points, your boyfriend is a lucky guy.

    As DWH mentions, 23andMe makes a ton of information available on their website. You don't even need to create a demo account to see all of the reports, what genetic markers they use to calculate the risks, details about the studies referenced, and how applicable they may or may not be to you:




    Whether the information is useful in a clinical sense or to you personally is still fine to opine about but you can't really argue that 23andMe doesn't disclose these things.

    You should have sent in dog spit instead

      THIS! Can Gizmodo please redo one of these tests? Just to see if they pick it up?? Scientific purposes of course... :)

        How are you going to make a dog spit?!

          Mark Serrels, Award-Winning IT Journo - Always asking the tough questions.

      Yeah, if someone is willing to shell out the money and send in some animal drool and see if they get human results lol

    DWH and SW remind me of those scientologists whose sole purpose in life is to reply to publicly posted articles bullocking the practice...

    dudes, you're wasted at 23andme...give Tom Cruise a call, he'll love 'ya!

    Elly, nice effort to write the article. The useful part of the article is to point out that in USA they have the GINA act which prevents discrimination on the basis of DNA results. Australia is so backward in this respect; most people here have not even heard of such a thing as GINA - thus it holds us all back from embracing this new technology. Sometimes the USA really does get things right.
    I agree that people under the age of 30, with an Asian background, will struggle to get value out of 23-and-me reports. On the other hand, the analysis seems a lot cheaper than health club membership! I used DeCodeMe which is a bit fancier. The results are on my website. I am open to comments about my genomic data - and I doubt that Australians will ever have to worry about insurance to the degree Americans need to. On the other hand, we really do need a GINA legislation here as well to reassure people who would otherwise like to know.

      Hello Prof. Barry Marshall if that is indeed you! DeCodeMe has discontinued interestingly.

      Elly - very nice article. Good to know about some of the consequences. I'm glad that you - of Asian decent has written this.

    I thought 23andMe wasn't designed to make money, as being underwritten by google?

    I found that the article was a very unbalanced view of genetic testing.
    I have done genetic testing with both FTDNA and 23 & Me. But unlike what you are reporting I did my research before I did my first test.
    There are many, many sources on the internet, both by companies offering these services and by other groups and people offering information about the companies and the services. The type of testing offered by the companies I mention are vastly different and you can achieve different results.
    One type of testing is genealogically based and does not go into the health field at all. Most people who are serious genealogist would use this style of service.
    The other is primaruly a health based testing service that leads to research into diseases and hopefully cures. The genealogical portion is more of a broader results than the other so it does a different thing.
    Both types of service offer the third alternative of autosomal testing which again is useful in genealogy to go around an unknown - that is find cousins instead of direct blood line ancestors.
    I thought that the information provided on both these websites was very good and enabled me to make a balanced judgement between what they can provide and what I wanted to find out.
    If I recall correctly it was some 18 to 24 months of research before I did the first DNA test and it was onother couple of years research before I did a 23 & me test.
    The other point I would point out is that the information is on the web to answer most of the question that the report you received gave you. Again it is research. And the companies I have found are very good at answering question that reports raise.

    23andMe is largely a lot of show, and very little 'go', for anyone.. let alone asian females under 30. It sounds cool in the MARKETING CLAIMS, BUT like most marketing claims falls way short of it advertising.
    They recently dropped the cost temporarily to a flat $109 USD for a limited time, from a earlier cost of $400-500+ USD, and even at $109, I do not see much of value.

    The medical results in most cases are preposterous when you assess the mini-studies that are the basis for their assertions, and that fact that the conditions they concern may have many other causes or related factors than the SNP they are evaluating as potentially causative.

    Delta23/CCR5 (HIV resistance if you are of northern European heritage) being one of the very FEW exceptions of some value (or sickle-cell, if you are african orgin).

    Most of the medical conclusions are close to totally worthless and come from tiny studies of anecdotal participants that apply to very tiny populations, limited medical scenarios, and only in very vague assumptions. The same conditions can still be developed/contracted by other combinations of causes or environmental factors- not strictly by the one lone SNP 23andMe bases their predictions upon.

    For this reason, a '23andMe' prediction of ones medical susceptibility to condition(s) that are not solely related to the SNP that the prediction is based on, more likely gives false worry AND/OR false assurance, than any valuable information.

    What 23andMe does NOT offer, SNP testing, (esp for males who can get Paternal Y-line and Maternal dna line SNP results) is available in much greater detail WITH STR markers from a standard MTdna / Y-dna test(which 23andMe does not provide).

    Getting SNP confirmation is a part of most STR packages, and you will find that if you are like me, you dont even remotely match anyone who has the same SNP Haplogroup that I do.

    This means that if you are going strictly by the ''family/cousin'' matches that 23andMe is telling you that you have, and linking you to as ''family'',
    you are extremely likely to be getting a lot of really bad and false matches to your new found ''cousins''.

    I have Y-line STR markers, and also the SNP confirmation- NOT from 23andMe - but from a traditional Y-dna test..
    If I were to accept ALONE the results from the 23andMe 'matches' to others who share the same SNP's that I do, without being able to double check their match-accuracy from my Y-dna STR numbers, I would be falsely matched as 'Cousins' to a lot of people who are in no way related to within thousands of years,
    since when your matches are compared via STR similarities (which 23andMe does NOT test for) my STR markers are not predicted to intersect with some of these ''3rd cousins'' anytime in the past 1500 years- contrary to the 23andMe conclusions based on less pertinent info.

    I am not saying 23andMe is utterly worthless, but I am saying its not worth much at all in my estimation, beyond the SNP ancestry of your Y/MTdna paternal lines, which are available with much more data from the longstanding STR testing.

    The vast majority of 23andMe medical predictions / results are based on a voodoo science that is not much better than waving a chicken-foot as you jump over a broomstick.

    Oh boy... This is the pitfall of having a third party company take on requests for SNP analysis from random individuals - the average person doesn't have the background to put these results into context (no offense to you, truly). The platform 23andme is used by researchers, yes (it used to be Illumina's 550 array, though I've heard they're moving to Illumina's omni). But researchers who use the platform aren't focused on just a smattering of SNPs and assigning clinical stats to different variants. It's a limited approach, necessarily guided by the fact that most people don't give a crap about other details. I can understand why - even my eyes glaze over sometimes when I'm working with huge data sets. The biggest complaint I've heard in this debate is that GWAS studies fail in terms of complex mental disorders and obesity, etc. But that's such a small amount of research compared to what SNP arrays are most often used for.

    23andme doesn't run the arrays, it sends them to Illumina. Illumina sends back the raw data. 23andme gives you its analysis, which does leave much to be desired. I'll definitely admit that. That's why I would look at the raw data myself. Of course, most people aren't interested in combing through raw data. 23andme is trying to give people just the data points an average person would be interested in along with clinical statistics. But linking up basic science with clinical implications is a slippery slope.

    They should make certain limitations more clear but insinuating that they're out to rip you off isn't true. It's a US-based company, are they really responsible for putting insurance practices in other countries on their website, particularly when that information is already available online elsewhere? They should warn that statistics vary among ethnic groups and that most studies are done with Europeans, but the fact that most studies are European-centric isn't a fault of theirs. There are a lot of bottom-feeding companies following this trend who are really trying to rip customers off. 23andme, Pathway Genomics and deCODEme are decent companies with people in charge who actually understand the science. I'd be interested in doing this just for the raw data, but I'm more interested in seeing what non-pathogenic CNVs I have and not so interested in risk-associated SNPs. Because I'm a nerd like that. I would say only do this if you're a science nerd or if you have the medical knowledge to put the results in context. Or if you know you won't freak out over generic statistics. I guess my post is really addressing the needless pearl-clutching I've heard across the board. Genetics testing is useful and not the problem, statistics taken out of context is the problem. Please rest a little easier with the knowledge that your personal habits aren't taken into account in this kind of testing and that most often a single "risky" SNP implicated in x disease isn't enough to put you at real risk. Especially for something like schizophrenia. I'm kind of surprised that they're including mental disorders without explaining it better, that's the part I think is irresponsible.

    Correction- 23andme gets the arrays from Illumina and runs them in-house these days I think. I was thinking of what an Illumina employee told me a few years back, probably not the case now. The point was that 23andme's intellectual contribution is in the analysis, not the arrays themselves. I just hate to see the technology itself being needlessly criticized.

    I've just sent off for the 23andMe test kit. There is a massive amount of information on the site about the testing and post test results, including mock results so you can get a feel for 23andMe. BUT you have to put the time into exploring all this. I spent a few hours checking out the site first so I feel I was able to make an educated choice about the testing. From reading through the online community at 23andMe you can also see that the information provided by 23andMe is really just an essential springboard for further ancestral research which I'm excited about doing - its not really intended as an instant handbook for everything about your ancestors. So basically the moral of the story is you will get as much as you put in. BTW they do state VERY CLEARLY on the site that you can't test Y-DNA in females - given we don't have any. This is not a money gathering "trick" it's basic biology - females are XX males XY so how can they test for Y in women? In regards to the health genetics - yes if you don't have any background knowledge in health genetics the findings may be daunting (I have had genetic vulnerability education via my psychology degree). For individuals wanting carrier testing a service like this will save them a fortune as the Cystic Fibrosis test done by 23andMe is the standard 32 mutation panel you would gain from a genetic counselor. So if you are serious about getting genetic DNA testing for ancestral research or medical reasons 23andMe is a cheap way to do this. In regards to the ethnicity issue 23andMe provides information of all the research used in regards to testing so you can check out its relevancy first.

Join the discussion!