7 Reasons why I oppose genetic testing as an indicator of lifestyle/supplement/food choices

Imagine 12 jaws of 12 geneticists drop to the ground in horror as one of us read the line “using direct to consumer genetic testing results to determine lifestyle decisions like diets, exercises and supplements.”

That happened during my human genetic class last year where we discussed different studies of human genetics, the state of technologies, how genes were discovered and etc. Most of it was dry science. We were assigned key research papers to read, then we discussed how it added to human knowledge and criticized how the studies could be better or more convincing. Let me tell you, it’s either you find this really boring or you are really crazy about it.

Not until we got to commercial human genetic testing that it got extremely political. We debated the pros and the cons and whether direct-to-consumer human genetic testing should be openly available.

As a geneticist and a consumer, I strongly oppose human genetic testing at our current state of technology and knowledge. Here are my 7 reasons.

1. Human genetics methods are very different from the studies of genetics in other systems, and so our knowledge of human genetics is very incomplete. Remember Gregor Mendel, the father of genetics? He studied Pea genetics by mating them and then analyzing the patterns of segregation of the phenotypes (or observable traits) such as flower color, flower position, stem length etc. Years later, his discovery and methodologies were picked up and geneticists start studying genetics by mating model organisms (plants, yeasts, flies, worms, mice, frogs, fish), and then analyzing patterns of segregation and phenotypes. The reasons these organisms were selected as model organisms include that they can be mated easily, easy to examine physiology, their short generation times, and that they produce large numbers of progenies; allowing for simple genetic analyses.

You would know by now that it’s not possible to pair up humans based on traits of interest and mate them, wouldn’t you? Therefore, a canonical way of studying human genes responsible for a certain trait or disease was to analyze the traits that are available in the family history (or pedigree), which is also often incomplete.

I got very used to being very critical of research studies after having worked with 4 different genetic model organisms because it was much easier to get answers from these organisms than in humans. Even then, in 2012, there is no organism whose genetics is completely understood.

The human genetics studies frustrated me as if I got my hands tied in the dark as a geneticist. The ethical aspects of things prevented us from conducting certain experiments in humans, but since the findings were in enough demands that the papers could be published in high-impact journals with inconclusive data.

Technology is quickly evolving and Moore’s law often applies to cutting edge life science research. It is widely known that the typical amount of time requires to identify and confirm a responsible gene for a single gene human disease is 20 years. An example is the Cystic Fibrosis gene where 1 single change in the DNA is responsible for the disease.

2. The human genome is extremely complex and repetitive. Lots of our genome is made of repetitive DNA called transposons. The significance of this is that it gives the computers a lot of hard time while trying to assemble the first drafts of genomic DNA. Humans have 46 chromosomes in  23 pairs. The first time scientists attempt to assemble this genome draft is by the method called “chromosome walking.” They had to shred the DNA into pieces, clone it into bacteria so that they could be amplified in large amounts (because sequencing technology can only sequence 500 base pairs at a time). Then, we sequence the next piece and look for overlaps. So, if there are a massive amount of the same sequences in more than 90% of the human genome, it is very hard to assemble an accurate copy. Think trying to put together a jigsaw with the same pattern throughout.

Anyway, computer geeks figured it out and then we got a “reference” copy of the human genome. But as technology evolved (or sometimes they discover mistakes), they go back and *correct* the sequence. So if you work with DNA sequences and genes, once in a while you see a gene that used to exist in 1995 and no longer exists in 2000. You get the idea.

I visited a genetic testing booth at a product show, and the “expert” guy told me that the results are 100% accurate. That’s probably what you want to hear if you send a blood sample to test if you have HIV. But heck, even HIV tests are not 100% accurate and it’s been around for decades. Scientists know that there’s no such a thing as a 100% accurate experiment. Big lie.

3. Lots of what determine your genetic outcome is NOT in your DNA sequences. LOTS.Epigenetics” is what we call whatever that determine genetic information not in the DNA sequence. Did you know adding a chemical structure to your DNA can either turn the gene ON or OFF? Our DNA is wrapped around globular proteins called histones, and the density of histones on our the DNA strand affects how much a gene functions. After the gene is turned ON, the gene product can be chemically modified at many layers.

It’s so complicated that, in conclusion, knowing just your DNA sequence alone will barely be informative.

4. Our knowledge of genetic variation within and among populations is extremely limited. We know a reference human sequence of 1 or a few people. There are also studies that attempt to study what difference from one person to another does by profiling many people. However, there are genetic variants that are different from a Chinese person to a Japanese person to an Italian  Person to a Jamaican person, and we don’t completely understand the whole picture. Can you imagine how much money it would take to have all these data?

For example, there is a genetic variant commonly known for lactose intolerance among Europeans and North Americans. There is also a group of Africans that don’t have this variant but they are lactose intolerant. This puzzled human geneticists for years, until someone finally decided to analyze hundreds of these Africans who were or were not lactose intolerant and found a different genetic variant that may cause lactose intolerance among them (Nature Genet 2007; 39:31-40). Because the Europeans/North Americans and Africans don’t share genes (i.e. they don’t get married and have kids together), these variants are specific to each population. Not only that, our genes do change and evolve, and mutations arise as times go.

Indeed, if you already did your commercial genetic testing and you are different enough from the subjects that were recruited for human genetic studies in the past 50 years (i.e. you are not a white North American), the results may indicate that it’s “good enough” to be useful to you. That’s because these genetic tests are based on the same scientific data that have been generated over the past 50 years.

5. Most traits involve multiple genes, like hundreds or thousands, and we are very resilient to mutations.

A few years back I was enthralled by the field called “Systems Biology.” It involves studying cells and organisms like super-robust electrical circuits, so much that electrical engineers start looking at cells and wonder why living things are so hard to screw up. Systems biologists analyze how genes function to compensate for one another, or before one another in a biological process. One of my summer internships in systems biology involved generating graph images of genes (dots) and relationships between genes (lines) that looked like they could be in an abstract art museum.

For example, if you are tested to have a mutation that affect a gene involved in vitamin B1 processing, another gene of yours may be compensating for that mutation you tested for. That’s why you are alive and still walking. If you really lack that gene and absolutely need to supplement, you would not survive as a human being.

5. The tests you get at 23andMe and Navigenics for the same trait are different. There are so many new next-generation sequencing technologies coming out, each getting more accurate and cheaper per the same volume of sequencing. Next-gen sequencing often involves shredding the DNA samples into very small sizes (25 – 75 base pairs) to determine the sequences before mapping to the reference human genomes. Now, remember that you are not the same person as the person used to generate the reference genome and that our genomes are very repetitive (see point #2).

The gold standard in testing for DNA sequence is called Sanger sequencing. It’s the most accurate and conclusive of all tests. Many companies use next-generation sequencing technologies. I’ve seen many times that different machine types and analysis methods gives different conclusions. It’s also possible that different companies look at different parts of your genome to determine one same conclusion.

6. We still don’t fully understand the behavioral responses to these test results. One major point in the debate is that we don’t fully understand how the general population will respond to our accessibility to direct-to-consumer genetic testing. The general population are very poorly educated about direct-to-consumer genetic testing as to the state of technology, the informativeness of the data and how it should be interpreted. Even most physicians don’t fully understand points 1 – 5 above.

We don’t know placebo effect/emotional impact/cultural impact we will have towards knowing some of our genetic information or having that used as a marketing pitch towards us.

7. Your genetic information can and will tell things about your family members. An issue that arose that never arose before in medical history is the medical dilemma between the duty to warn and confidentiality. If you work in a hospital you will be trained not to give away any medical or identity information of the patients without consent. However, if you know that someone may be harmed, the hospital staff may have the duty to warn them.

Examples exist where knowing the genetic information of the child tells something about the parents or other family members. Oftentimes, it is simple, like some families that discover that they have stomach cancer genes and decide to have their stomach removed. Other times, it’s not.

As a fitness professional and a geneticist, I do not believe that genetic testing can give you the information you need to live a healthful life. What I believe is systematically examining your phenotypes or traits with respect to how your body response to certain things. If you do take the tests, do know it for exactly what it is (e.g. mutation at base pair number 453 in the peptide dismutase gene as tested by the Ilumina Sequencer model123 with 8x coverage and analyzed by so and so software), but don’t believe the implications that come with it. Most interpretations and implications attached to these genetic tests are VERY POWERFUL marketing.

Talking about marketing, how many people do you know are taking weight gainers and doing nothing else to gain weight just because the weight gainers say “weight gainer” on the label?


Leave a Reply

Your email address will not be published. Required fields are marked *