Nutrigenomics has been one of the biggest trends in Nutrition and Functional Medicine in the last decade. Here are what I consider the most compelling reasons why:
- The understanding that nutritional factors influence gene transcription and expression
- The background work of Bruce Ames, and myriad papers since, showing that high dose cofactor nutrient supplementation can push deficient cycles forward
- The drive towards personalized medicine, identifying the unique dietary and lifestyle factors each individual needs to optimize to create desired outcomes
- Greater availability of genetic SNP (single nucleotide polymorphism – unique gene types) testing via companies such as 23andme
What is Nutrigenomics? Definitions vary, especially since the field is still so new. Common threads are the ability of dietary components to up- or down-regulate specific gene expression (epigenetic and transcription factor modulation), and the impact of certain genotypes on individual responses to different dietary components (e.g. PKU, lactose intolerance). In short, nutrigenomics looks to match different dietary recommendations to individuals based on their genotype.
Our Own Internal Universe
Nutrigenomics is a vast topic. So vast, in fact that we still can’t really conceive of its scope. The detail and rich complexity of the multilayered structures of the human genome that we are merely starting to visualize was entirely unanticipated at the outset of the human genome project. There may only be an estimated 19,000-20,000 linear human protein-coding genes (a little more than a chicken, but less than a grape), but that represents around 1% of the entire human genome. The remainder is a far-reaching landscape, previously (and erroneously) referred to as ‘junk’ DNA that is actually a refined and complex array of overlapping “regulatory landmarks” and “transcription maps,” acting within the tight yet dynamic folds of DNA and DNA-bound proteins (who themselves have central roles to play in gene expression and configuration).
Genes also don’t act alone – the final output in terms of cellular function is dependent on the expression of many different genes, whose coded proteins act in concert via myriad metabolic pathways. Genes even interact with other genes to influence their own expression (a process called epistasis). These challenges complicate a scientist’s search for single genes responsible for a complex disease, such as cardiovascular disease, obesity, Alzheimer’s, ALS and suddenly it becomes painfully clear why research efforts in this vein have proved elusive. Here are some examples:
- Major Heart Disease Genes Prove Elusive
- The Elusive ALS Genes
- Genes and Autoimmune Disease – A Complex Inheritance
- Epistasis: Gene Interaction and the Phenotypic Expression of Complex Diseases Like Alzheimer’s
With complex, chronic conditions we’re in an entirely different territory than known single-gene defect diseases like sickle cell disease or cystic fibrosis, which are caused by mutations in single genes.
Just to throw another vast overlapping landscape into the mix, there is the genome of our own microbiome. Boom! – that just opens up a whole new can of worms since our commensal microbes contain genes that produce nutrients, signaling molecules, immune modulators, and even hormones that our bodies utilize.
It’s against this backdrop that nutrigenomics attempts to determine clinically-relevant interventions using diet and lifestyle factors. That’s a tough challenge.
More Human Reductionist Tendencies at Work
I have an admittedly geeky background. My post-doc training in nutritional biochemistry and laboratory science instilled an understanding (and love) of biochemical pathways, and their related nutritional inputs (both as substrates and products (amino acids and organic acids) and cofactors/coenzymes (minerals and vitamins). It’s a complex topic and it was deeply rewarding to be able to learn it in depth and share that knowledge with others. Indeed, if I lived by a spiritual precept, it was that “God is in the mechanism”: identifying imbalances in pathways through laboratory analysis (my lab was an early “omics” adopter, before the advent of high throughput screening, we looked at hundreds, but not thousands, of data points) should be correctable with pin-point precision nutritional interventions given in the right doses Wow. It was the brilliant Dr. Linus Pauling’s Orthomolecular Medicine come to life. What clinical condition couldn’t we correct with this careful analysis of what lies under the “metabolic hood” of an individual? They were heady, thrilling times for me. When our laboratory later turned its attention to DNA analysis of the microbiome, I felt invincible, if a little naïve…God is there, no doubt, but as I was to learn, the picture is far more vast and interactive than a smattering of data points would suggest.
Clinical practice is the great leveler, is it not? As my time at the lab drew to a close, and I turned my attention more broadly towards patient care, I was quickly faced with the realization that identifying and correcting said imbalances didn’t often yield the outcome I anticipated. Occasionally I saw brilliant turn-around of a condition that fit tightly with what the data suggested. But more often than not, this deep analysis was often proven to be reductionistic. I still value and rely on broad laboratory investigations. But I understand and appreciate that attempting to reduce an individual’s physiology to a handful of inputs and outputs will rarely yield the clinical outcomes we expect.
(This awareness doesn’t make the practice of functional medicine any less rewarding. It is still the most deeply satisfying professional endeavor I could undertake.)
We are in danger of applying reductionist principles to nutrigenomics too. The draw of being able to ‘master’ nutrient-gene interactions is too compelling. We want to be able to know what to do for each specific gene variant. Even for genes such as MTHFR and APOE4, where there may be more research, my patients don’t respond uniformly to methyl donor (or methyl-related) supplementation or dietary fat adjustments. Simplistic interpretation and treatment of nutrient-gene interactions , while exciting in the abstract, are ultimately insufficient to address the human sitting before us looking for a way out of their predicament. And furthermore, only rarely does the SNP investigation fit neatly with what one would expect to see with regard to the associated gene products.
All said, I just don’t see that nutrigenomics will ultimately play out in the way that current popular thinking suggests.
That doesn’t mean all is lost though. We will still be able to move forward with approaches that are just as satisfying intellectually, and even more so ultimately for their efficacy.
Patterns Across Multiple Genes – A Step Forward for Clinicians
Researchers recognizing the issues with looking for single gene-disease connections are now investigating combinations of genes in connection with a percentage increased risk for certain diseases. This is more useful, in my opinion, though still with challenges.
Here are some examples:
- Combinations of SNPs in CYP1B1/COMT/GSTP1 and CYP1B1/COMT/MnSOD increase breast cancer risk by 2.7 and 12.2-fold respectively
- 63 genetic markers associated with 5.7 times the risk for prostate cancer
- 95 genetic variations increase risk for specific type of obesity that can be mitigated by exercise
These ones are worth a smile:
- 939 genes relate to intelligence
- 599 genes implicated in neurotic behavior. I wonder how many of those genes you have to have before you can be technically labelled ‘neurotic’?
As a hard-working physician myself, I can tell you that the clinician sitting in front of you won’t be able to identify most of these gene pattern associations in their head. OK, maybe the breast cancer SNPs above, but not the 95 obesity/exercise SNPs. Certainly not hundreds of genes related to neuroticism. No matter how many of those intelligence genes they have. Further, there must be a weight given to significance of an individual genes; not all are equally impactful. And beyond that, different patterns of gene variants alter outcome. The clinical application of this kind of research is going to require computing power. It’s “big data.”
And of course the big wild card of “environment” (exposome) must be considered for each individual: Big data must embrace the “N of 1” investigation. Yes, the research might have concluded that the average increased risk is significant, but an average is just that – an average. How do you know who has less risk, and who might even have more risk? Here’s an illustration, using data from that gene combination and 12.2-fold increased breast cancer risk above:
Take a closer look at the 95% confidence interval next to the 12.2 odds ratio (OR). 1.4 to 102.3. While one person with this gene combination might have only a small 1.4 times the risk, someone else with this gene combination might have over 100 times the risk! Which one of those is in your consulting room? (although with larger data sets, this odds ratio would tighten up)
Back to our earlier picture of the ‘internal universe’ of our cell nuclei, though: these studies are still only able to look at the linear, two-dimensional attributes of the genome. They entirely miss the complexity of the three-dimensional structural interaction, the layer of epigenetic regulation, and gene-to-gene interactions. It’s in these spaces that we can start to look for the answer to why having a certain genotype doesn’t always translate into disease. And why we need to think about more than SNP=nutrient need.
Swim Out of the Pond and Head Upstream
While we wait, watch, and judiciously apply some of the nutrigenomics research that continues at a clip in this still-nascent field, where I feel most comfortable taking my patients is much more upstream. It’s not about avoiding complexity, it’s about being realizing that we need to approach this amazing, inspiring complexity in a different way. That’s what the Methylation Diet & Lifestyle is all about. Our ebook arose from conversations and collaboration with my nutrition program director, Romilly Hodges, as we unpacked the research on epigenetic methylation in the context of our functional medicine and patient care. (New to methylation and epigenetics? Read more here.)
With this in mind, I think about:
- Providing optimal and safe levels of nutrient precursors and cofactors so that DNA regulation and metabolism can function appropriately. Diet is foremost here, gut support, plus judicious supplement use.
- Including factors that create optimal genetic regulation; direct modulators of epigenetic and three-dimensional DNA structures. These include diet, lifestyle, microbiome and environment as what we call “DNA methylation adaptogens”.
- Removing factors that get in the way of appropriate gene expression or that damage DNA strands or structures. Diet, lifestyle, microbiome and environment are involved here too, as well as stemming methyl donor drain in areas of metabolic dysregulation.
If you’re as excited as I am about applying these principles in practice, I encourage you to use our Methylation Diet & Lifestyle resources to guide you.
As we challenge the current status quo around how to interpret and use nutrigenomics data, and posit a more upstream approach, we need to put our thinking to the test, also.
Check out the epigenetics research study we are conducting using our Methylation Diet and Lifestyle program through the Helfgott Institute at the National University of Natural Medicine in Portland, Oregon. Our study is actively underway, and we anticipate early findings in 2019. I’ll be presenting on our study at PLMI’s Thought Leader Consortium in October this year.