23andMe: The Consequences of Personal Genetic Information in the Hands of Consumers

23andMe: The Consequences of Personal Genetic Information in the Hands of Consumers

Tags: , , , , , , , , , ,

By Kasey Hemington
Image courtesy of Andrew Tubelli

The stories on the company blog are sensational and alluring – a PR dream come true. Long searching adoptees learn of their true ethnic roots and visit native countries that they never imagined they belonged to. Upon discovering a relative’s cancer diagnosis, a woman demands further screening at her own check-up when the doctor finds something but wants to wait. An older gentleman is alerted that given his genetic profile, his medications are likely the cause of his newly developed muscle weakness, and he fiercely petitions his doctor for alternatives.

The company, 23andMe, is selling a Personal Genome Service (PGS): customers send in personal information, saliva samples, and 99USD in return for a slice of their raw genetic data – a set of single-nucleotide polymorphisms selected by researchers – and accompanying ancestral and health interpretations. Ethically, legally – it’s a nightmare. The Silicon Valley startup’s privacy statement is a rosy promise to protect personal information except where ‘required by law’, but that doesn’t mean customers are getting the whole picture. For example, most customers probably aren’t aware that while the recent United States Genetic Information Nondiscrimation Act (GINA) protects Americans against genetic discrimination by health care providers and employers, there is no protection against discrimination from long-term care insurance providers. Here in Canada, there is no act like the GINA at all, meaning employers and health insurance providers of Canadians purchasing the kit could request results and use them as basis for discrimination.(1) A failure to provide them may be treated as any other failure to provide health information. The Parkinson Society Canada even lists a warning for the 23andMe Parkinson Community Research project, stating that “for Canadians, there is potential risk to their privacy as well as insurance or employment implications”.(2) What’s more, while accounts can be deactivated, it isn’t clear if all clients’ personal information can be permanently erased. One can’t help but be wary that endless Facebook-esque privacy battles will follow, except with consequences bigger than photo and wall post regrets.

Ultimately, the goal of the project itself is information exploitation – not just selling trendy saliva kits with potential health implications for individuals. With founder Anne Wojcicki’s love of big data and her former marriage to Google co-founder and now 23andMe investor Sergey Brin, her long-term plan certainly includes the mining of your data. Ms. Wojcicki herself recently interviewed with Fast Company, saying “I want 25 million people. Once you get 25 million people, there’s just a huge power of what types of discoveries you can make. Big data is going to make us all healthier”.(3) The discoveries she refers to also imply company profits. Pharmaceutical companies will pay to access genetic or behavioural data that predict risk of developing cancer, diabetes and Parkinson’s disease. Hospitals will pay for personalized drug metabolism profiles to optimize patient dosage. Academic institutions will want to access the massive data sets they could only dream of collecting with traditional methods.

At this point, if you’re thinking of buying one of these fascinating spit-collecting contraptions despite the aforementioned complications, you might be out of luck. Technically, the kits are still for sale, but due to a recent warning letter from the United States Food and Drug Administration (FDA) on November 22nd, 2013, 23andMe has suspended the provision of its neatly packaged interpretations. Customers now only receive raw data – useless unless they are genetics experts or are willing to spend months reading scientific articles and trusting their personal interpretations.(4)

The FDA’s communication was two-fold: The main issue is that, as the PGS is marketed as a tool in diagnosis, treatment and prevention of disease, it is classified as a medical device, and 23andMe has failed to provide the data necessary to demonstrate device accuracy. Highlighted concerns include persons taking drastic and potentially harmful prevention measures against disease or failing to adequately protect themselves when PGS reports do not accurately reflect true risk. 23andMe has informed the public that it is complying with the warning by including only genetic results without interpretation for the time being, and is working to complete the necessary studies.

A second issue, for which the FDA did not request any specific actions, pertained to self-management of medications. 23andMe isn’t just about the risk of disease – the health report portion of the PGS is divided into disease risks, carrier status for conditions, specific traits, and drug responses. The fear of customers taking it upon themselves to manage medications without a complete understanding of the drugs – even if the results are 100% accurate – is real. From the letter, though, it appears the FDA will consider offering full approval even if this issue is not addressed. FDA territory may only include accuracy of the PGS as a medical device, but other risks such as the self-management of prescriptions as mentioned by the FDA, also need to be considered in the provision of genetic services. Failure to closely monitor how people will interpret the information is just as important as the scientific accuracy of the tests.

23andMe’s blog directs readers to a series of studies that appear to answer a related question: Will newfound knowledge of high genetic risk for disease cause long-term stress and anxiety?(5) According to these studies, the answer is no, though some reviews and blogs by individuals do express the opposite. But even if we move past this burning question and assume that the decision to get tested is that of the individual, there are other fundamental flaws in the information as 23andMe presents it.

Let’s get one thing straight: Humans don’t intuitively understand numbers, and we’re incredibly vulnerable to interpreting them differently based on their presentation. This is why the PGS costs $99 instead of $100 and why companies decide to split stocks – making the company appear more affordable although the underlying value has not changed. This is why people buy lottery tickets and why casino’s stay in business. Here’s what it means to impressionable humans interpreting the information in a PGS: In the health risks section (according to online reviews), red and green ink are used to indicate a higher or lower than average risk for developing a disease, respectively. This puts unnecessary emphasis on positively or negatively comparing oneself to an arbitrary average comprised of different ethnicities, ages and lifestyles. For example, Jimmy finds out he has a 3.4% chance of developing restless leg syndrome (in red) compared to a hypothetical average of 2% and only an 18% chance of getting cancer, compared to 23% on average (green). Jimmy might run to the doctor the minute he has a poor night’s sleep. If he does have restless leg syndrome, he might be diagnosed more quickly, but no one has taken Jimmy’s environment into account when considering his risks; he is 68, a smoker, and morbidly obese. With his true risk of developing cancer likely higher than 18%, will Jimmy take positive steps toward cancer prevention and discuss cancer screening with his doctor as a result of his 23andMe report? Will a genetics company communicate to him that environmental factors are far more important than genetics in this case, or that 18% is still a risky die to roll?

University of Toronto Professor of Biochemistry Dr. Laurence Moran recently described a viable solution (at least for Canadians) to this potential public health crisis waiting to be ignited by the fiscal motives or ignorance of private industry. He suggests that the Canadian government should provide free genetic counselling. His theory is that since government counsellors, unlike private providers, would be trusted to provide unbiased advice, the private competition would simply die out.(6) This is reasonable in theory, but the Canadian government has no plans to revolutionize healthcare through genetic testing in the immediate future, and so we can at least commend 23andMe for moving things forward at lightning speed, despite the road bumps.

A feasible, more immediate addition to Prof. Moran’s long term suggestion might be the requirement of a closer working relationship between government bodies like the FDA, and genetics firms. Fully disclosing the methods used in determining disease risks to medical personnel would ensure the information is useful amongst the medical community and promote doctor-patient discussion. Meanwhile, the provision of direct web links to local health authority disease screening recommendations would strengthen an individual’s sense of responsibility for their own health outcomes. Newfound genetics knowledge does have the potential to revolutionize healthcare, as long as there is a plan in place to ensure accurate genetic information is presented without bias, and that it empowers the public to take constructive steps towards improved personal health.


  1. Office of the Privacy Commissioner of Canada. (2012). Genetic Information, the Life and Health Insurance Industry and the Protection of Personal Information: Framing the Debate. Retrieved December 28th from http://www.priv.gc.ca/information/research-recherche/2012/gi_intro_e.asp
  2. Parkinson Society Canada (2013). 23andMe Parkinson Community Research. Retrieved December 28th from http://www.parkinson.ca/site/c.kgLNIWODKpF/b.5743161/k.1E3F/23andMe_Parkinson_Community_Research.htm
  3. Murphy, E. (2013). Inside 23andMe Founder Anne Wokcicki’s $99 DNA Revolution. Retrieved December 28th from http://www.fastcompany.com/3018598/for-99-this-ceo-can-tell-you-what-might-kill-you-inside-23andMe-founder-anne-wojcickis-dna-r
  4. Food and Drug Administration (2013). 23andMe Inc. Retrieved on December 28th from http://www.fda.gov/ICECI/EnforcementActions/WarningLetters/2013/ucm376296.htm.
  5. 23andMe (2013). Few Worries about Genetic Testing. Retrieved on December 22nd from http://blog.23andMe.com/23andMe-research/few-worry-about-genetic-testing/.
  6. Moran, L. (2013). Michael Eisen defends 23andMe against the FDA. Retrieved on December 28th from http://sandwalk.blogspot.ca/2013/12/michael-eisen-defends-23andMe-against.html.