Why are we motivated to test?
- to satisfy a deep curiosity, a need to understand the genetic mechanisms that link genotype to phenotype.
- to become more informed and hopefully make better life choices, with particular attention to identified health risk areas.
- to find near cousins, and more information about our deep ancestral heritage.
For us, it was all of the above. Here, let’s concentrate on the first two motivators, curiosity that seeks out information to enable better life choices.
DNA testing has been offered for over a decade to the public. After having tested my non-recombinant DNA over the last several years for genealogy purposes, I decided to purchase autosomal genetic testing (the other 97% of my DNA) and convinced Debby to be tested also. The most cost-effective source at the time was 23andMe, a privately-owned direct-to-consumer (DTC) testing company founded in 2006. See my related discussion of my 23andMe testing experience.
Following are a summary of our experiences, reflections on value, and interpretations of the state of this brave, new DTC medical information industry. There are many unanswered questions about the present and future of this industry and of the place for DTC in the scheme. There are many naysayers who are uncomfortable with genetic health testing in general, and DTC in particular. In the USA, there is a major regulatory agency, their major lobbyist, and their Institutional Review Boards (IRB) who oppose the DTC concept unless it is constrained beyond the point of usefulness. There’s a lot to talk about.
The Nature Of DNA Health Analytics
By analytics we mean here a mathematical and scientific approach that allows researchers to derive meaning and guidance from data. Data is the foundation of DNA-based health research, which largely follows a process:
- data collection and analytics
- research into the fundamental disease process and pathways, following leads provided by analytics
- research into diagnostic methods, disease intervention, and interventional drugs, once process and pathways are identified.
Health research often begins with academic research grants for studies whose results are published in peer-reviewed journals. These typically test one or more hypotheses and publish study data. Analytics then aggregates the data from one or more related studies and analyzes the data statistically, to determine patterns in the data resulting from underlying associations between genotypes and health outcomes.
It is hoped such associations may suggest which research directions can be most profitably pursued. The discussion here addresses the first phase of the research process, data collection and analytics. These typically involve two types of information, a genotype (DNA signature), and a phenotype (list of known traits) associated with such a genotype.
Genome-wide Association Studies (GWAS) attempt to statistically match traits to specific DNA mutations by identifying genetic patterns in populations. This usually involves studying a population whose members express the trait under investigation, and a control population, looking for commonality of SNP alleles in the trait group that are absent in the control, or vice versa. Further, the study population should be large enough and random enough to suppress unintended common factors of the group (such as age, sex, weight, ethnicity, etc.) that might also correlate with the trait. The desired result is a statistical imputation of SNP association with a trait, although the converse knowledge, lack of association, is also useful.
Ultimately, the goal is to understand genetic linkages of disease and to eventually devise methods to interrupt disease process. This means additional studies moving to identifying and understanding the anomalous gene function(s) that produce the trait. This is where research organizations will commence to monetize their efforts and transition to profitability as commercial diagnostic and treatment entities. This is also where the big jump in progress occurs, from mere association to causation.
What To Expect From Genetic Health Screening
Enlightenment regarding health’s direct relationship to genome is the major benefit of my genetic screening. The bulk of the derived benefit is information only, derived from the explanatory power of many of the results and the studies behind them. Enhanced understandings help us put things in proper perspective. We can learn things about ourselves with the potential to sweep away mysteries and answer lifelong questions, even perhaps regarding our personality traits, .
American medical technology is maintaining stride with the new genetic results, but the practice of medicine is likely a decade behind, awaiting better data, the establishment of protocols to guide actions, and the enlightenment of the medical industry regulators.
I now experience a refined view of what it means to be in a ‘normal’ state of health. At the level of phenotype, I had thought of a world of sick and well people, where well was an expression of the normal (disease-free) state. Now I realize there is no comprehensive concept of normal at the level of the genotype, and hence phenotype normalcy is also a suspect concept.
At the DNA level, the wild genotype has been the researcher’s term for the biological state of a natural population. Mutations were defined as variations from wild type. But ‘wild type’ is a population-specific, gene- or location-specific concept. Wild is assumed to be the majority state (allele) in a population, but different populations may have different majority alleles. All organisms are a combination of billions of wild and mutated alleles. This renders the concept of normalcy of genotype unworkable. We are all unique.
One’s test results will provide a lot of hints regarding connections from specific alleles to risks for clinical outcomes. Some of these hints will be assigned greater importance. One will receive corroboration of the utility of one’s testing by recognizing personal health outcomes that resonate with the provided hints. This will provide one a sense that the test report really is describing you.
Some of the important hints will not have manifested in one’s self. But they will incite awareness and a desire to know more. As we study about their potential, we may learn life choices that increase or decrease potential risk. While a physician could offer the same advice in the abstract, having the genetic evidence in front of us provides a more powerful instinct to toe the line in making good choices.
In rare cases, there will arise a marker of very strong significance that predicts a very bad outcome. It is this eventuality in the back of each test subject’s mind that probably provides the most primal urge to be tested. If the outcome has an available medical intervention, then one experiences testing nirvana. Else, one will still have positive life choices which would not be available without fore-knowledge.
Some specific information returned in my tests is further described below. While possibly these specific results are unique to 23andme, in general one could expect similar information from other testing enterprises.
My enlightenment is augmented by appreciation for complexity. Some parts of one’s genetic signature are responsible for each phenotypic trait one expresses. But the process by which the genome codes for phenotype will involve incomprehensively many individual chemical interactions. Researchers have their work cut out for them.
What Not To Expect From Genetic Health Screening
Do not expect directly actionable medical prediction or advice. The medical industry is not yet ready to participate in utilization of genetic testing information, and the information itself in its current state of refinement is too vague and internally self-contradictory to be useful.
In our experience to date, there are not yet established medical protocols for addressing most of these genetic findings. If we pass along notable findings to our medical practitioners, the information hopefully will be recorded in our charts for future reference.
Do not expect simple answers from genetic testing. Statistical association is not causation, and those seeking causation relationships from the test data will be disappointed or misled.
Do not expect the test results will mean what they seem to say. The extreme complexity of our biological process currently defies our attempts to assign health-predictive value to any given allele. Test results seem to imply prediction when they note an X% increase or decrease in relative risk for a disease outcome for a given allele. Such predictions are orders of magnitude less significant to outcome than they seem from the statistics. One senses this disconnect when observing an allele, with a majority frequency in the population, being associated with significant risk increase for a rare disease.
Then why test? The lesson here is that most of the value of these individual statistical observations will be derived by the researchers looking to further develop understanding of cause and effect. It remains to be determined how any given set of alleles combines in effect to produce a clinical symptom. But by joining a genetic screening effort and contributing freely with our phenotype information, we help advance the research toward unleashing the power of individual medicine.
Arguments Against Testing
Many parties, including the American FDA/AMA, remain critical of DTC genetic tests purporting medical relevance. They assert that even small interventions based on inferred benefits can be dangerous to one’s health in absence of professional guidance. But being an informed consumer of health services is one of the most important steps we can take toward the direction of improved health. And to be informed, we need the data.
The naysayers are correct that relevance of most results to individual health decisions is low, although a few pertinent findings do surface. They are also correct that when that rare relevance is found, it is best assessed by a licensed physician. Only then can it be actionable, but only after protocols are established in the future. Nobody but the naysayers’ strawmen would question such observations.
Naysayers argue there is no real added value in this testing, for if we followed our health care providers’ instructions to the letter, paid attention to our family history, ate right, exercised right, and got plenty of rest, we would be as disease-free as it was meant for us to be. I would agree with that 80%. It’s the other 20% that I am after here, plus the informational value of all the data, whose exploration keeps our minds energized. The naysayers seem governed by a Catholic philosophy that too much knowing is harmful to the members of the flock, that such knowing is best the province of clergy.
Naysayers claim that the data is of no value to someone without a PhD in statistical bioinformatics and a close friend who is a PhD geneticist. I couldn’t disagree more. It takes a well-educated person to get the most from it, but the innate intelligence of any curious individual will capture meaning even from the raw data I encountered.
I expect most of the ~500K people who have signed up for 23andMe testing are like me, curious, well-educated, comfortable with statistical concepts and basic genetics. But curiosity is the only necessity to get one’s money’s worth. Let’s not sell the customer short.
Naysayers are shocked there is potential knowledge being communicated by the SNP data that may indicate life-threatening conditions. While that is the pot of gold if there is a successful intervention available, what if there is not? This is a dilemma. But whether there is a physician in the loop or not, the customer needs to decide before pushing the test button whether she is a know or don’t-know type. If the answer is ‘don’t want to know’, then getting tested is a bad idea. You can’t only test for the good stuff. Only test if you want to know everything.
At surface value, the FDA’s argument against DTC medical genetics is that health-related genetic testing constitutes a medical device subject to certification and IRB regulation. There is some basis for this argument and a requirement for a regulatory role. The health industry has been a hotbed for charlatans over the centuries. But to throw a research company involved in serious medical research under that bus is not sensible or fair.
Under the surface, the AMA’s elitism continues to shine through this argument, promulgating the idea that the public isn’t well-enough informed to make appropriate use of their detailed health data. Hence data could be dangerous to their health. Unless, of course, a middle man, one or them, is inserted to interpret it. This is self-serving, patent nonsense, and the actual dangers they cite to ‘prove’ their point are valid subjects for ridicule. No, I wouldn’t excise my body part because genetic testing data suggested a remote possibility of a life-threatening disease someday invading that part.
We see in these arguments the usual risks to continued free availability of our data. The government has consistently caved in to business lobbying, lobbying by those that would insert themselves between our publicly-funded or personally-funded data and the general public, charging exorbitant fees and imposing restrictions for subsequent access. Let’s not allow that to happen with our genetic data. DTC is the right model, and an informed public requires such a model.
Until the AMA becomes our partner and not our master, expect the obstructionist status quo from the FDA. The people will now have to be more energetic and resourceful to become informed about their health. But being informed is still a possibility so long as the raw data is freely available to us.
We all owe it to ourselves to address the naysayer’s concerns, then move past them to learn all that can be known about our potential health factors. Testing can inform our decisions on diet and lifestyle, and make us more sensitive to signs of potential health trouble. While many of us know, in the abstract, what constitutes a healthy lifestyle, real data, emphasizing specific dangers, is often the impetus we need to toe the line.
By understanding the nature of results related to brain function and its influence our personality traits, we can become better learners and life partners. By shining our own light on the mysteries of our bodies, we can ask more informed questions of our health providers, making us better patients.
The Road Ahead
Genotyping and analytic association of genotype and phenotype are in their infancy. The medical science behind personalized medicine is about to become much richer and more useful, and I am pleased to be in on the ground floor. In order to fulfill the promise, however, in utero genetic testing may have to be widely available and encouraged.
Before we condemn the AMA for their off-stage obstructionism, we should see the germ 0f truth in such objections and reflect. They perhaps argue the raw data of genetics will have virtually no utility or appeal to uninformed people lacking exposure to genetics, those who cannot tell the difference between association and causation, who have no understanding of the published studies linked to the research. It seems unlikely that such people would spring for the cost of the testing, but somebody should worry about what harm comes if they do.
OK, I’ve thought about it; other than the confused client generating random questions for their medical providers to the point of nuisance, it is hard to imagine serious consequences. As for non-serious consequences, yes, the establishment has a closet full of them. You don’t even have to ask, they trot them out regularly.
The AMA is likely more comfortable with a testing model that collaborates with established medical centers affiliated with genetic research institutions. For example, the Coriell Institute for Medical Research, their affiliated Coriell Personalized Medicine Collaborative, and their spin-off, Coriell Life Sciences, provide different models for genetic information dispersal that works through established medical providers and seeks IRB approvals. To get past the AMA hurdle, DTC genetic testing companies such as 23andMe may have to re-orient their business model and collaborate with medical providers also.
However, it will be a long while before such a model, involving a sanctioned provider of personalized medicine, is available to me. After all, these are dinosaurs we are talking about, the big heavy ones that move slowly. I likely won’t live so long. And then in all probability, by current models, I would need a diagnosis to be able to benefit from information related to my medical need.
But the whole idea is to get the information before the diagnosis, to avoid the diagnosis. This is where the DTC model is superior. It is available to anyone, no gates to entry. It is available now, not after I get sick or die. It is available to me, not to an intermediary who doesn’t know me from a few markings on a chart. It is affordable and delivers true value to those who understand how to use it. These people, my people, should not be denied access to their genetic birth right by bureaucrats. This is America, for goodness sake! (That felt good.)