What are We Eating?

NUTRITION We are what we eat Studies linking how genes and diet interact are helping food companies design products capable of protecting people prone to certain diseases Sep 4th 2003 |From the print edition

Eat what your genes likeAnthony Blake

SOME people eat three-egg omelettes topped with slivers of bacon and show no sign of a spike in cholesterol. Others indulge in one chocolate bar after another and stay as thin as a rake. Many, however, are less fortunate. Current research suggests that the culprit may be found in one's genes. Differences in genetic make-up may not only determine the body's ability to metabolise certain nutrients, such as fats and lactose, but also its susceptibility to disease.

The good news is that, within five years or so, researchers should learn how to modify people's diets to fit their genes and thereby prevent or delay the onset of a possible illness. At least, that is the goal of nutritional genomics—a new field that studies how genes and diet interact.

Projects in nutritional genomics are sprouting around the globe. Europe is merging its efforts in the field by launching NuGO early next year, a network that aims to integrate and develop the new branch of research. In America, the National Institutes of Health recently granted the University of California at Davis $6.5m to establish a Centre of Excellence for Nutritional Genomics. Technology Quarterly

Expect the unexpected Supercharging the grid Falling, flailing, virtual doll Voltage ahoy Solar cells come down to earth Ups and downs of chip design Nanomaterials move up a notch Soul of a newer machine A web address for every car? Fountain of truth? Who watches the watchers? Uncrackable beams of light More light than heat Banking on the technology cycle Battle for the big screen Out of the ether We are what we eat Reinventing Europe Embryonic man

Reprints R

elated topics

Genetic testing Health care Cardiovascular medicine Health and fitness Diet and nutrition

In addition, there are international projects under way, such as HapMap, that focus on studying the pattern of inheritance of single nucleotide polymorphisms or SNPs (pronounced “snips”). These are places where the message encoded in the genome may vary by a single genetic “letter” between individuals. SNPs may determine differences in appearance, such as hair and eye colour, predisposition to illnesses, and how people respond to foods and drugs. Nutrition will be an important part of the new paradigm of “personalised medicine” and preventive health care, says Craig Venter, who spearheaded a private effort to sequence the human genome, and is now collaborating with Duke University Medical Centre to include genomic information in health-care planning.

Only a few diseases are based on mutations in single genes—as is the case with, say, cystic fibrosis or Huntington's chorea. As scientists have learned more about the human genome, they have found that many illnesses, including cancer and type II diabetes, are the result of an interaction between a number of genes and their environment. More than 100 genes have been implicated in the development of coronary artery disease (though carrying only one such gene is still a risk factor). But to express those genes, there needs to be a trigger—such as diet. Researchers estimate that diet may be the defining factor in a third of all cancers.

A gene is a recipe for making a protein. At the molecular level, various nutrients interact with genes by binding to DNA transcription factors—which regulate gene expression. Thus, the amount and type of food consumed affects the production of proteins directly.

Similar to “pharmacogenomics”, which studies the effects of common genetic variants on drug response, nutritional genomics investigates the effect of diets on different individuals, groups and populations. But unlike drugs—which come as refined compounds, are administered in specific doses and have relatively short-acting effects—foods, like genes, act in concert. “They're like a great big symphony,” explains Wasyl Malyj, who directs the laboratory for high performance computing and informatics at Davis's new centre. Moreover, the effects of foods are slow moving—and often take years before becoming visible.

As a result, no one meal is ever going to be detrimental for most people. The trick will be to find out what constitutes the best balance of nutrients over long periods. For example, a recent paper in the New England Journal of Medicine describes the effects of a traditional Mediterranean diet—lots of olive oil, fresh fruits, vegetables, legumes and nuts, moderate amounts of fish and wine, and little meat and dairy products. Those who adhered closely to the diet lived longer and were less likely to die of heart disease or cancer. There was, however, no association between longevity and any individual food component of the diet.

The complexity of human nutrition poses challenges. Clinical trials require a large group of people to be followed over many years. That makes studies costly and hard to conduct. Moreover, people often forget what they have eaten.

But progress is being made thanks to “systems biology”, which uses the tools of genomics, molecular biology and bioinformatics to study the complex interactions of genes, proteins and nutrients at the cellular level. DNA chips can now look at thousands of genes at once, allowing complex gene-expression profiles to be created quickly. In addition, studies of human cell cultures and “knockout mice”, which lack the ability to make specific proteins, provide other means of collecting data. The goal is to find early molecular profiles (“biomarkers”), which may be a useful step on the way to identifying diseases before they pose health problems.

Changes in DNA have occurred in human populations as evolutionary responses to changes in diet. For example, a change in a single gene 10,000 years ago allowed a group of northern Europeans to become lactose-tolerant and continue to consume milk products into adulthood. But such evolutionary responses take time. They also depend on the weeding out of inappropriate genes by the early deaths of carriers. In other words, they only happen because some genes cause diseases when exposed to the new diet.

For example, roughly half the adult population of the Pima tribe in Arizona have developed diabetes. This seems to be in response to a modern American diet, and its associated way of life. Another group of Pima, which lives in Mexico and still follows a more traditional diet and a less sedentary lifestyle, shows no significant increase in the disease. Singapore, with residents of Chinese, Indian and Malay descent who share similar prosperity and eating habits, serves as another interesting example. A portion of this ethnically-mixed population is now affected by the formerly uncommon coronary heart disease, but in unequal proportions—a result of a difference in genetic predisposition.

What is acceptable for one individual may actually harm another, says Ronald Krauss, director of atherosclerosis research at Children's Hospital Oakland, in California. Dr Krauss, who also chaired the American Heart Association's dietary guidelines committee in 2000, says its current recommendations benefit most people. However, it may be useful for some to adjust their diet based upon their genetic profile. “Some healthy people may increase their risk of heart disease with an extremely low fat/high carbohydrate diet”

Dr Krauss, who has been researching the effects of various types of low-density lipoprotein (LDL), or “bad cholesterol”, suggests that some healthy people may actually increase their risk of heart disease by embarking on an extremely low fat/high carbohydrate diet—something that is sometimes recommended as a way of lowering cholesterol. He has found that people with large LDL particles in their blood, a state of affairs that is relatively safe, can develop an abundance of small LDL particles, which is more dangerous, after a few weeks on such a diet.

One of the first projects of Davis's new centre will be to compare tissue samples from white and black American men. For reasons not yet understood, black American men have a 60% higher incidence of prostate cancer than their white compatriots. Analysing the degree of genetic variations between the two groups may provide clues to the importance of environmental factors—such as diet and access to good health care—which could play a part in the disparity. Eventually, hopes Raymond Rodriguez, the director of the Davis centre, nutritional genomics will usher in a new era of consumer genetics that will translate into practical lifestyle changes and dietary choices.

Food as medicine

Businesses certainly hope so. The current epidemic of metabolic diseases, such as obesity, diabetes and heart disease—in which diet is an important risk factor as well as a preventive agent—provides an opportunity for companies to develop foods that help populations at risk. Add to that the clinical proof of diet/gene interactions, and the traditional distinction between medicine and food becomes increasingly blurred. As a result, food companies may begin to think more like drug firms. Pharmaceutical spread-AP

While they have only a fraction of the R&D budgets that pharmaceutical companies make available, some food companies are already marketing “functional foods” that address the needs of specific populations and have gone through independent clinical trials. One example is Unilever's Flora pro.activ—a bread spread that reduces cholesterol by 10-15%. Unilever, an Anglo-Dutch giant supplying foods and other consumer products globally, has formed a partnership with Perlegen, a genetics company based in Mountain View, California, which has catalogued no fewer than 1.5m common genetic variations in humans to date. The aim is to develop new foods that match the needs of specific populations. Nestlé, the largest food company in the world, has similar agreements in place.

As more interactions between diet and genes become known, simple genetic tests via “cheek sweeps” (DNA testing based on swabbing the inside of the cheek) may determine what people need to eat. A handful of companies, such as Sciona in Britain, has already begun to sell products (available online and through some health-care providers) that test for genes that influence the metabolism of alcohol, folate and other nutrients. Based on the results, Sciona makes suggestions for diet modifications. But experts agree that, thus far, such tests have little meaning. “Telling somebody what to eat based on a few genes is not appropriate,” says Jim Kaput, founder of NutraGenomics, a start-up that plans to offer genetic tests once the knowledge is more advanced.

However, before such tests become commonplace, a number of ethical, medical and legal issues will need to be addressed. One fear is that insurance companies might gain access to the information and charge higher premiums, or even exclude people from their plans, on the basis of their genes. And, because of the complexity of diet/gene interactions, it will be essential to train physicians to interpret the test results. Despite these hurdles, most researchers agree, the potential benefits outweigh the possible harm. As Dr Ordovas, professor nutrition and genetics at Tufts University in Massachusetts, says, predicting illnesses 40 or 50 years before they may arise is not only a convenient way to minimise the need for expensive drugs—but also a powerful tool to improve the quality of a person's life.