by Dr Geoffrey Modest
The FDA just approved direct-to-consumer marketing for genetic risk information (23andMe Personal Genome Service Genetic Health Risk) for 10 conditions, though noting that “the tests cannot determine a person’s overall risk of developing a disease or condition … there are many factors that contribute to the development of a health condition, including environmental and lifestyle factors.” This approved test involves saliva samples, assessing more than 500,000 genetic variants associated with increased risk of: Parkinson’s disease, late-onset Alzhemer’s, Celiac disease, Alpha-1 antitrypsin deficiency, Early-onset primary dystonia, Factor XI deficiency, Gaucher disease type 1, Glucose-6-phosphate dehydrogenase deficiency, Hereditary hemochromatosis, Hereditary thrombophilia. see https://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm551185.htm
The FDA reviewed the data for 23andMe through a premarket review pathway for low-to-moderate risk devices, with expectations about assuring test accuracy, reliability and clinical relevance, and also to make sure the results be clearly understandable by consumers. But the FDA now intends to exempt further tests added on by 23andMe from further premarket review, and may well exempt other genetic testing companies after submitting their first premarket notification. These exemptions “would allow other, similar tests to enter the market as quickly as possible and in the least burdensome way”. [and, I might add, this is before confirmation of Trump’s pro-industry FDA nominee Scott Gottlieb, who has “received millions of dollars from various investment and pharmaceutical firms” per Bloomberg Technology…..]
Statnews had a really impressive review of the 23andMe test, at the cost of $199, revealing many of its limitations (see https://www.statnews.com/2017/04/07/genetic-analysis-need-to-know/ ). For example, they note that those having the specific variant for Parkinson’s disease tested increases their risk 3-fold, from a baseline of 0.3% to 1%…. Or, that those with Apo ℇ4 alleles may not get Alzheimer’s, and those without it may (the frequency of the Apo ℇ4 allele varies by ethnicity, 15% in Caucasian, 25% African-Americans; the presence of one allele increases the risk of Alzheimer’s by 2-3 fold, and two alleles by 8-12 fold). So, the presence of a genetic variant, either for the Parkinson’s gene or if only 1 allele of Apo ℇ4, still makes the development of the disease unlikely (and actually rare, in the Parkinson’s case). And still about 10% or so of those who are homozygous for Apo ℇ4 do not get dementia.
–the big issues here, to me, are:
–these tests may well have pretty low sensitivity and specificity, as well as low positive predictive value
–patients may have trouble understanding the wording: 3x higher incidence of Parkinson’s sounds like a lot, but the actual 1% incidence not so much. Can be very confusing
–and, there are real concerns about the psychological effects of finding out one has a somewhat higher likelihood of a bad disease for which there is no current treatment. Will there be more depression, anxiety, decreased social cohesion/more isolation, hopelessness/even suicide?
–focusing on the genes undercuts the very important role of environmental/lifestyle factors: it really reinforces the conceptual deterministic framework that one’s future is set by one’s genes, undercutting the oftentimes dominant message that our environment and lifestyle are really important
–and it reinforces the conception that technology is the answer to our ills…
Some recent article on dementia are tangentially related to the above.
–The WHO reported that dementia deaths have increased, unseating AIDS as one of the top killers in the world (see http://edition.pagesuite.com/popovers/article_popover.aspx?guid=30673b53-29ff-49a5-a387-796e55c1aa5e ), and taking over the number 7 slot of the top 10 causes of death. And, as per this article in Bloomberg News, about 100 experimental treatments for dementia have failed to make matters better. Part of the issue causing the “elevation” of dementia is the aging population and probably that it is more often diagnosed now. But, so far, drugs do not seem to be the answer
–in this light, and complementing the above point that genes often do not play a decisive role, there was a recent study finding that lower adherence to a Mediterranean diet was associated with more significant loss of brain volume (see Luciano M. Neurology 2017;88:1)..
–Background: increased adherence to Mediterranean diet (lots of fruits, veges, legumes, cereals, olive oil as primary fat, moderate consumption of fish, low to moderate intake of dairy and wine, and low intake of red meat and poultry) is associated with less inflammation, better cognitive function, and lower risk of Parkinson’s and Alzheimer’s, as well as cardiovascular and cancer mortality. And cross-sectional studies have found higher consumption of components of the Mediterranean diet are associated with larger MRI-based brain volumes and cortical thickness. Higher fish and lower meat intake seemed to be the most important players.
–The current study was a prospective one of 562 Scottish men and women, assessing diet and brain structural changes from age 73 to 76
–50% female, 30% Apo ℇ4 positive, 4% diabetic/38% hypertensive/22% cardiovascular disease/BMI 28
–baseline cognitive ability: Mini-Mental Status Exam 29 (30=max, so no significant baseline dementia), and they assessed reading ability and general cognitive ability which relates to IQ (no comment on the scales they used or their validity). Diet was assessed only at baseline, age 70.
–change in brain structure from age 73 to 76:
–total brain volume: decreased 19 ml (from 990), gray matter volume decreased 9 ml (from 465), mean cortical thickness decreased 0.05 mm (from 3.11 mm)
–the group with highest adherence to Mediterranean diet had more carriers of Apo ℇ4 alleles (reason for this unclear in this healthy population who did not have underlying dementia), yet had greater total brain volume and gray matter volume at age 76
–in the fully adjusted model (controlling for those factors found in prior studies related to Mediterranean diet and brain MRI measures: age, sex, education, BMI, diabetes, general cognitive ability, MMSE), there was significant association between Mediterranean diet components and total brain volume change between age 73 to 76 (p=0.04), and presence of Apo ℇ4 genotype did not change this. Fish and meat consumption were not found to be the drivers of this association. [perhaps it is a different combination, or even the all of the components together: parsing out specific components may be a tad reductionist and undercut potential interactions between the individual components. Better to eat well overall]
–so, there was a significant association between the diet and brain volume changes over this 3-year period
–and, the effect size of the Mediterranean diet on brain volume was substantial: half the size of that due to normal aging
–of course, this was not a randomized controlled trial, so there could well be confounders (do those choosing to adhere to a more Mediterranean-type diet do other, unmeasured healthful things that may really be the ones that decrease cognitive decline, such as exercise??)
but, all in all, this study supports the concept of environmental/lifestyle factors being really important in the development of Alzheimer’s/cognitive decline, that this appeared to be independent of the known genetic risk factor of Apo ℇ4, and adds to the argument against a genetic-determinant view of the development of this important condition (as is conceptually promoted by 23andMe etc)