| Didn't read the whole paper, but the summary is idiotic. GAO made undercover calls to the companies seeking health advice. ... GAO made undercover contact ... and asked about supplement sales ... So it's not that the companies are pushing such things, it's that GAO prodded it out of them. one donor who had a pacemaker implanted 13 years ago to treat an irregular heartbeat was told that he was at decreased risk for developing such a condition. It's almost as if the person who wrote this doesn't understand that statistical likelihoods are not the same as ironclad facts. The above anecdote, seemingly included to shock us, seems like a perfectly reasonable outcome to me. There are plenty of people who aren't at risk genetically, who still end up with problems. One donor was told that he was at below-average, average, and above-average risk for prostate cancer and hypertension. Again, perfectly reasonable. It would seem that each company is testing for different SNPs. A given person's DNA can certainly read different ways when we're not reading the entire genotype, but just a limited (and potentially differing) set of sequences from within it. (However, I would hope that each of these tests included a description of which SNP are tested for, so that it's at least possible to interpret properly with research) Also, none of the companies could provide GAO's fictitious African American and Asian donors with complete test results, but did not explicitly disclose this limitation prior to purchase. I've only ever looked at 23andMe, but they certainly disclose this [1]. AFAIK, 23andMe is the best known of such companies (indeed, the only one I can name), and it has this disclosure. I think that, again, they're sensationalizing. At this point I got disgusted and stopped reading. [1] https://www.23andme.com/health/ethnicity/ ADDED as a follow-up to my comment about misunderstanding statistics. The proposal would have the consumer hearing this information through a filter of their MD. How many MDs really understand physics? How many of them are going to be able to explain the Bayes concepts behind "if your father and your uncle both have prostate cancer, then it doesn't double your risk...". Is making the MD the gatekeeper to this actually going to help? |