This week Dr. Phil Wood’s article, Nutrigenetic testing: Is it ready for prime-time?, looks at the one goal of modern genetics and medicine is to provide “individualized medicine,” or medical care based on an individual’s genotype. Nutrigenetics is one example of the developing potential for meeting that goal.
Nutrigenetic testing: Is it ready for prime-time?
Philip A. Wood
One goal of modern genetics and medicine is to provide “individualized medicine,” or medical care based on an individual’s genotype. Nutrigenetics is one example of the developing potential for meeting that goal. The idea behind nutrigenetics is that if we know our genotypes, then we might be able to make dietary changes based on that genotype and, as a result, prevent the development of disease processes in type 2 diabetes and metabolic syndrome, such as insulin resistance, pro-inflammatory state, or heart disease. Some direct-to-consumer genetic testing companies are already selling individualized services based on this concept. But what can this information do for an individual, and is nutrigenetics really ready to go?
The general procedure is this: A customer buys a kit from a company. The kit contains a swab, which the customer rubs on their inner cheek and sends back to the company. From the swab, the company extracts the customer’s DNA and analyzes it for variations known as single nucleotide polymorphisms (SNPs) at the sites of various genes. These are genes that may influence diabetes, obesity, cardiovascular disease, and other disease processes. The rationale for the company’s analysis is that some SNPs at these various loci have been associated with increased risk for factors important in type 2 diabetes, such as insulin sensitivity, heart health, and inflammation. Therefore, the analysis may identify genetic loci which indicate that the customer has a genotype for increased risk for the conditions of interest and that, based on this genotype, some dietary change might reduce the risk of developing that disease. This all sounds very useful. However, as it currently exists, the approach has two major problems. One is the different ways genes are involved in disease, and the other is an apparent lack of disease-specific dietary recommendations. Although companies that offer these services are usually capable of providing the genotype data, I am not convinced that those data can help the companies’ customers.
First, let’s consider some genetic concepts. Many of us have a strong concept of Mendelian genetic diseases, in which disease traits are recessive or dominant based on mutations at a single gene locus. Such diseases are said to be monogenic. For example, someone who inherits two mutated copies of the cystic fibrosis gene is most certain to develop cystic fibrosis. This same pattern of inheritance is true for many monogenic disorders such as sickle cell disease and phenylketonuria (PKU). In many such cases, not even the environment or the rest of the individual’s genetic background will prevent the development of the disease. However, in one unusual example of PKU , drastic reduction in the patient’s dietary intake of phenylalanine greatly improves the patient’s condition and prognosis, and this has been shown in many research studies (1). In this case, there is definitely a nutrigenetic benefit because knowing the genotype provides specific information about changing the diet to improve the inherited condition of the patient.
Next, let’s look at genetics and type 2 diabetes. In almost all situations, type 2 diabetes is not monogenic but multigenic, or based on sequence variations at more than one gene locus. Also, it is influenced by the environment and behavior, so its development is not nearly as straightforward genetically. To understand diabetes, obesity, and other common diseases, we need to change our thinking to consider multiple genes, the environment, behavior, and other influences on disease development. The genetic basis of diabetes comprises multiple genes scattered around the genome. Each of these genes can be represented by one of multiple variations, and each variation interacts with a host of other potent influences such as diet, gestational and neonatal environments, and many other factors we are just starting to understand. Yes, genetics is important. But for multigenic diseases, there is very limited value in knowing your genotype at a few specific loci that may or may not influence your disease risk. In contrast, for monogenic diseases like PKU, knowing your genotype at a single locus can be extremely valuable.
Whether the variant is monogenic or multigenic, our knowledge is very limited when it comes to the genetic bases of disease. Most variants are known to researchers only as an association between a particular genetic variation and a disease or process such as diabetes, insulin resistance, or inflammation. Although such associations between genetic variant and disease may reach statistical significance, the impact of that association is minimal because most of the gene variants detected so far are associations with small physiological effects (2, 3). A very few of such associations have been replicated in multiple studies, and even fewer suggest relationships as significant as those between specific monogenic variants and diseases like PKU or cystic fibrosis. Most of the time, it is not known whether a genetic variant and the risk of a disease have any cause-and-effect relationship.
In the nutrigenetic paradigm, the next step is to develop a diet that will complement the customer’s genotype. I have no direct experience with the companies offering these services, but I have heard directly from two people who have submitted their own DNA samples. They received their genotype results and the following recommendation for their diet: “Eat more fruits and vegetables. Reduce intake of sugar, starch, red meat, and trans and saturated fats.” Well, we don’t need a genotype analysis for that diet recommendation! Furthermore, almost no clinical studies have been done to ascertain the efficacy of diet recommendations based on genetic variants. For example, we do not know whether increasing our vitamin intake by such-and-such dose will overcome such-and-such genetic variant and reduce disease risk.
In summary, we hear almost weekly that a new “diabetes” gene has been discovered. What this most likely means is that geneticists have detected another genetic variant that has a significant association with—but likely a small effect on—the development or management of diabetes (2, 3). As a consequence of this limited knowledge, only very limited value can come from direct-to-the-consumer nutrigenetic testing (where the genetic effects are small) that is then used for a nutritional consultation (without clinical trial data to back up its efficacy). Diabetes is a complex trait consisting of multiple gene-gene and gene-nutrient interactions (4). More research is needed before we can fully realize individualized approaches to such complex disease traits. For now, go with what you know: Eat sensibly, exercise, do not smoke, and monitor your blood pressure. Those actions are likely to be just as effective as today’s nutrigenetics, but they save you over $300 on genetic testing, and even more on dietary supplements.
1. Scriver CR, Kaufman S. Hyperphenylalaninemia: Phenylalanine hydroxylase deficiency in Metabolic and Molecular Basis of Inherited Disease, McGraw-Hill, Baltimore, 2001, pp. 1667-1724.
2. Das SK, Elbein SC. The genetic basis of type 2 diabetes. Cellscience 2:100-131, 2006.
3. Janssens ACJW, Gwinn M, Valdez R, Narayan KMV, Khoury MJ. Predictive genetic testing for type 2 diabetes – May raise unrealistic expectations. BMJ 333:509-510, 2006
4. Wood PA. Multifactorial genetic diseases of lipid metabolism in How Fat Works, Harvard University Press, Cambridge, Massachusetts, 2006, pp. 110-128.