The challenges of molecular nutrition in the diet and health relationship

Published: 31-01-2014

What are biomarkers of health?
Twentieth century nutrition research has focused on optimizing nutritional requirements from a public health perspective. Over the last 15 years, the concept of functional foods has emerged and matured. This coincided with the obesity wave and related pathologies. Thus, nutrition research, in large part, is aimed at unraveling the relationship between diet and these issues (primarily weight maintenance, type 2 diabetes, cardiovascular diseases, as well as gut health, immune status, etc.).

Consequently, the biomarkers used in this research were derived from the disease processes. Yet, the processes on which dietary compounds act are often unrelated to disease progression, but are involved in maintaining optimal health: metabolism, innate immunity (“inflammation control”), maintenance of optimal resilience in oxidative stress, neurohormonal control of regulatory set points in metabolic and inflammatory homeostasis, etc. Quantification of these processes cannot easily be done in a “resting” condition, since this would result in “normal values.” So a paradox evolved: how can we demonstrate health optimization if all relevant biomarkers already are “normal” or “optimal”? This led to the reintroduction of the concept of “challenging homeostasis.”  By imposing a perturbation of the homeostatic situation, we can quantify the efficacy of the related processes to return to their homeostatic state. A classic example related to nutrition is the oral glucose challenge test. Many other challenge tests are now being developed or revitalized, with stressors in the area of inflammation, exercise, oxidation, mental activity, etc.

The impact of nutrigenomics on biomarkers of health
In the implementation of these new biomarkers, it became apparent that the amount of relevant information that can be obtained from these “challenge tests” now far exceeds the original scope. For example, the oral glucose tolerance test was originally designed to quantify the plasma glucose curve after glucose absorption. Modern analytical technologies allow simultaneous quantification of hundreds of other plasma metabolites and proteins, thus providing a comprehensive molecular physiological overview. These include organ-specific metabolic responses on insulin, the resulting oxidative stress and inflammatory reaction, the lipid and fatty acid response between liver and adipose, and the muscle absorption of selective amino acids, etc.

Likewise, an inflammation challenge test (for example induced by LPS) not only reports on C-Reactive Protein, a classical biomarker of inflammation, but also demonstrates the 24-hour “inflammatory wave” as demonstrated by a panel of 120 plasma cytokines and related proteins, and whole genome transcriptome analysis in white blood cells (PBMCs) — revealing the intricate complexity of the triggered innate immune system. The power of “genomics” technology is thus shifting from “hypothesis generation” to complete assessment of all relevant parameters, pathways and processes related to a biological function. This development has a major impact on nutrition and health research, as often, effects of nutrients cannot be pinpointed to a single receptor interaction (as is the case for drugs) or to a single biomarker. In our efforts to understand the complexities of these systems, we see indications that we might need to switch to “process biomarkers” or even “process descriptions.” Many molecular parameters that, together, describe the process can now be quantified, and all information can be evaluated together to accurately describe diet-induced changes in physiological processes.

Nutrigenomics and personalized nutrition
A fundamental problem in nutritional intervention studies is the large interindividual variation in phenotypical changes induced by the intervention. This is counteracted by rigid stratification of the intervention study groups and by increasing the size of the study cohort, but still many studies produce disappointing results. There are multiple reasons for this diversity, related to genetics, life stage and lifestyle. In fact, now that we are able to very accurately quantify the “nutritional phenotype” (i.e. all molecular and physiological parameters involved in the diet and health relationship) with the omics technologies, we see subgroups emerging that respond differently to a nutritional intervention.

The relation between diet and blood pressure is a striking example. Many intervention studies have shown marginal effects of, for example, milk polypeptides on lowering blood pressure, but the interindividual variation mostly prevented strong conclusions or claims. A recent publication (Horrigan et al. 2010) demonstrated that the micronutrient riboflavin lowers blood pressure in cardiovascular disease patients, but only in those subjects with a specific genetic variation in a folate-metabolizing enzyme MTHFR.

Other examples of the impact of genetic variation exist in the areas of fatty acid, lipid and cholesterol metabolism related to cardiovascular health outcomes. But also independent of genetic variation, large phenotypic variations in response to dietary interventions are observed and mechanistically understood. Several early phases on the metabolic syndrome may evolve from one lifestyle (“caloric excess”). This raises the issue of whether or not we should have a better look at establishing diet and health relationships, and possibly also health claims, for specific subpopulations. From a purely scientific point of view, this may be the optimal way forward: select a study population which, based on mechanistic observations, reacts in a homogeneous manner to a dietary intervention. From a practical viewpoint, this is not easy since it may require a phenotyping or diagnosis of the consumer related to dietary advice, possibly using invasive methods like plasma biochemistry. Also, this is unattractive from the view of the current business model of food companies.

Yet, examples already exist and are successful in the food product market. Cholesterol-lowering margarines based on plant sterols are targeted to the subpopulation with moderately elevated plasma cholesterol (invasive diagnostics involved!). Coffee is sold in two variations (with and without caffeine), based on the phenotypic response of the consumer (“awareness” at the wrong moment). Although all consumers know about this differential impact of coffee, few realize that this phenotype is partly determined by a genetic variation in the caffeine-metabolizing enzyme in the liver, and probably, none of them have taken or even considered a genetic test to determine their choice for “regular” versus “decaf.”

These are obvious and accepted examples, but could be seen as “low hanging fruits,” with many other fruit still a bit hidden between the leaves of the tree. Many early-adopter companies have currently emerged that link a diagnosis (genotypic and/or phenotypic) to dietary advice. Others go a step further and include these services in lifestyle advice.

Nutrigenomics, in its broad definition (the application of a systems approach in the nutrition and health relationship by quantification of all relevant parameters and processes), is changing the face of nutrition and health research. This extends far beyond the first accusation of a “fishing expedition” and explores a new generation of biomarkers, adding a new meaning to “molecular physiology” and facilitating the (market) introduction of personalized nutrition.

European Nutrigenomics Organisation (NuGO) and TNO
Many of the above developments are the result of the activities of the European Nutrigenomics Organisation (NuGO) and TNO. NuGO is an EU-funded Network of Excellence, where 23 European universities and research organizations have joined forces to develop nutrigenomics technologies, research and concepts. Upon the completion of the EU contract, NuGO will continue its activities as a new legal Association, with a focus in nutrigenomics infrastructure activities and research networking. The Annual NuGOweeks are highlights in global nutrigenomics research. (see

At TNO, all the above-mentioned activities are operational and fully integrated in the practice of human intervention studies, analytical facilities and bioinformatics.

Ben van Ommen is Director of NuGO and Senior Research Fellow of TNO.

Contact details:
Ben van Ommen, PhD
TNO Quality of Life

References and further reading:
Horigan G et al. – Riboflavin lowers blood pressure in cardiovascular disease patients homozygous for the 677C T polymorphism in MTHFR – J Hypertension 28:478–486 (2010)

Williams CM et al. – The challenges for molecular nutrition research 1 – Linking genotype to healthy nutrition. Genes & Nutrition 3: 41-9 (2008)

Van Ommen B et al. The challenges for molecular nutrition research 2: quantification of the nutritional phenotype. Genes & Nutrition 3: 51-59 (2008)

Daniel H et al. The challenges for molecular nutrition research 3: comparative nutrigenomics research as a basis for entering the systems level. Genes & Nutrition 3: 101-106 (2008)

Van Ommen B et al. The challenges for molecular nutrition research 4: the “nutritional systems biology level.” Genes & Nutrition 3: 107-113 (2008)

Van Ommen B et al. Challenging homeostasis to define biomarkers for nutrition related health. Mol Nutr Food Res. 53:795-804 2009)