Foods and food products have been a major part of a revolution in consumer trends, with increasingly personalized values of convenience, cost, packaging, and taste all emerging as distinguishing variables in the food marketplace. Now food is poised to take a major step to increase the value to the consumer by personalizing diet and health.

Consumers can readily perceive and appreciate many of the inherent values of foods, such as taste, texture, and convenience, yet improved health is not immediately perceivable as being associated with particular foods. Similarly, the metabolic differences among humans are not obvious without assessment technologies to measure them.

There is now an imperative to understand these variations in human metabolism. Metabolic dysregulation is increasingly epidemic in the human population, leading to health problems such as obesity, atherosclerosis, diabetes, and hypertension (Alberti, 2001; Puska, 2002). The questions therefore become “How will science and technology find solutions to the causes of metabolic disease?” and “How will the food marketplace deliver these solutions to consumers?”

The causes and consequences of metabolic diseases are different for different people. Metabolism is closely linked to diet, and even with drugs and lifestyle changes in place, food choices will be a vital key to resolving these metabolic health issues. Therefore, food choices will provide an important component to the overall goal of improving metabolic health. Understanding the basis of choices provides the means to deliver nutritional value and health through products consumers prefer. The keys to foods for metabolic health and delight will be to understand the differences among consumers, discover technologies that distinguish these differences, and apply metabolic knowledge to match differences to specific foods and diets.

Although personalizing foods for health may sound like a new idea and the marketing of food products matched to individual health issues seems decades away, the truth is quite the opposite. The concept is centuries old—food and products that have been designed for specific metabolic issues of individual consumers have been in the modern food marketplace for decades. It is not a question as to whether personalized foods will become a part of the food marketplace, but simply when they will become the rule rather than the exception.

Differences Among Individuals
There are basic differences among individuals that relate diet and health.

• Phenotypic Variation in Humans. The term phenotype refers to all of the discernible and measurable properties of each individual. It is the sum of all functional attributes related to health and is the observable physical or biochemical characteristics of an organism:
Phenotype = Genotype + Environment + (Genotype × Environment)

Each person’s phenotype is either slightly or dramatically different from another’s due to many factors (Table 1). In terms of diet and health, if we are to personalize our food choices according to our needs and aspirations for health, it is important to understand how these differences arise.Table 1—Sources of biological variation

• Genetics. There are significant variations in the population beginning at the level of the basic genetic sequence, the genotype. From the most obvious of genetic differences, gender, to subtler differences down to the level of single nucleotides, humans differ.

--- PAGE BREAK ---

Many of the differences in genotypes are deducible by the phenotype or obvious physiological responses to diet that are sensible to the individuals themselves. For example, lactose intolerance, which is a genetically based variation among humans, can be deduced from the intestinal consequences of consuming lactose. Lactose-free dairy products allow those individuals who cannot metabolize lactose to choose dairy products, even though it was not necessary to determine their genetic profile to provide them with these product options.

The human genomics initiative has undertaken the mammoth task of assigning the naturally occurring genetic variations in humans to their metabolic consequences (Muller and Kursten, 2002; Chagnon et al., 2003; van Ommen and Stiernum, 2002). As the understanding of human genetics and diet—nutrigenetics—grows, more and more products will emerge to deliver health value according to those differences.

• Environment. Throughout their lives, individuals are exposed to highly varying environments, including their diets, that affect health. Family traditions in cuisines, food preferences, cost, convenience, age, and lifestyle all have important impacts on the dietary choices each of us makes throughout our life.

Many of the differences in human health are due to these environmental influences and are independent of genetics. For example, identical twins, who are genetically the same, will enjoy different health status if they eat different diets. Different diets, exercise, and lifestyle choices all influence overall health. There are several examples where food products are designed to recognize the health effects of particular environments on individuals. From the perspective of lifestyle, sports beverages are designed to provide water to rehydrate athletes during and after exercise, and sugars and salts to restore energy and electrolytes.

• Genetics × Environment. The same environment can have different effects on different individuals because of their genetic makeup. Genetic predispositions determine to some extent how environment influences health. For example, consuming a high-fat diet increases HDL cholesterol in some individuals, but decreases it in others (Tai et al., 2003).

• Metabolic Memory. Humans have a remarkable plasticity in the ways that they can adapt to different environmental conditions. The nutrition research community has begun to study this persistence of phenotypic traits induced by diet, which is termed nutritional “imprinting” (Levin, 2000) and nutritional “programming” (Singhal et al., 2003). Biological organisms possess myriad mechanisms to “learn” from the environment to which they are exposed, including diet, and to adapt a variety of structural, biochemical, and regulatory processes to improve their responses to this environment (Reddy and Hashimoto, 2001). Such adaptations are not always detrimental as implied by terms like imprinting, so the processes of acquired metabolic phenotype are more appropriately referred to as metabolic memories, in keeping with their variety, complexity, and value. Metabolic memorization can be a positive aspect of life and the basis of such disparate but positive attributes as athletic training and olfactory preference (Decombaz et al., 2002; Zhang et al., 2003).

• Phenotypic Aspirations of Consumers. Each of us has aspirations for a lifestyle that we would like to achieve. These aspirations can be quite different, and achieving them may require different dietary choices. The most obvious examples of phenotypic aspirations are those of athletes, whose goals of strength, speed, endurance, and flexibility are already linked to certain dietary selections. Many desirable attributes may in the future be improved by diet, including intellectual, musical, and artistic performance. Even sensory or behavioral traits, e.g., organoleptic preferences and moods, can be altered and their experience heightened by diet (Hudson and Distel, 1999).

There are simple examples of dietary choices that one makes now that can influence future health and enjoyment of life. As a simple example, if someone who is genetically predisposed to become lactose intolerant as they age would like to continue to consume lactose-containing foods, maintaining a diet that contains lactose will encourage the development of microfloral bacteria that metabolize lactose in normal quantities in the diet (Hertzler et al., 1997).

As the science of personalized health builds knowledge of human variation, metabolism, and diet, more and more choices should be available to individuals to guide their own goals for health and performance. The goal of this science is to build a body of knowledge to enable each human not to be more constrained in lifestyle choices but to achieve whatever optimal health status and lifestyle they aspire to.

--- PAGE BREAK ---

Distinguishing Differences
There are philosophical as well as technical aspects to health assessment.

• Personal Assessment. Traditionally, health was assessed by analyzing biofluids, such as blood and urine, for the presence of biomarkers. These biomarkers are used to diagnose potential patients for the presence of specific diseases or problems in health. In contrast, for health assessment, the assumption is that the individual is healthy and there is no disease. Hence, diagnostic biomarkers to detect disease are inappropriate to measure health. Instead, an individual must be assessed for overall metabolic status. Cholesterol provides a model. Cholesterol levels in blood do not represent a disease per se, but the abundance of cholesterol—a normal metabolite—reflects an aspect of one’s metabolic state. When cholesterol levels are high, this measurement reflects a metabolic condition that if unchanged can develop into a long-term threat to health.

Technologies to measure cholesterol were developed years ago and are routinely available. Now technologies are being developed to measure many more metabolites quantitatively. Science must then use the abundance of these metabolites to build an overall pattern of health that can guide each consumer in appropriate dietary choices.

• Technology Platforms. Three technology platforms are emerging with sufficient analytical power to measure many metabolites routinely: mass spectrometry, nuclear magnetic resonance (NMR), and high-resolution chromatography. On the horizon, immunoassays, immobilized antibodies and immobilized enzymes may prove to be even faster and more cost-effective (Table 2).Table 2—Technologies to measure biofluid metabolites

The important principle of all of these platforms is the ability to measure hundreds of molecules at a time, rather than single analytes, and to transfer analytical data into high-capacity-database computer systems.

Modern mass spectrometers measure molecular weight with astonishing accuracy. Mass spectrometry as a metabolic profiling technology cannot yet measure metabolites with quantitative accuracy equal to that of NMR.

NMR-based technologies are capable of resolving hundreds of magnetic resonances from all of the metabolites in biofluids. NMR is highly quantitative. The challenge to NMR as a metabolic profiling technology is to assign all of the disparate resonances in a sample into identified metabolites.

Chromatographic separation of hundreds of metabolites is now possible because of the efficiency and resolving power of microbore columns, in liquid or gas phases, using pressure or electrophoretic flow. The challenge to this technology platform is that biofluids must be pre-processed prior to introduction into columns.

The ultimate goal of using technologies is to measure all of the metabolites in a human biofluid. This goal will not be attained until some time in the future, yet each of the technologies is capable of measuring literally hundreds of times the number of metabolites being measured currently in biomarker diagnostics. Therefore, technologies have capabilities of metabolite analysis now that could move health assessment forward immediately.

As the value of measuring metabolism increases, providing this knowledge to consumers will drive industry to deliver newer, faster, and cheaper technologies into the personal assessment marketplace. The computational tools that are necessary to manage this information are already being put into place for other aspects of health (Kunze et al., 2002).

--- PAGE BREAK ---

Using Knowledge of Metabolism
Scientific knowledge of metabolism can be used as a blueprint for acting on individual differences.

• Linking Metabolites to Phenotypes. The technologies to measure metabolites in humans would be of little value if the variations in metabolic profiles were not understood in terms of discernible health. Knowledge of metabolism needs to relate the distribution of metabolites to existing health status and to predict the trajectory of health that such a profile will develop over time.

The scientific field of nutrition and metabolism is still building the knowledge base to link health to metabolite assessment, but it is not necessary to know all aspects of health to be able to provide health improvements immediately. The consequences of variations in many metabolites are known and can be used to predict many aspects of health, including obesity, diabetes and atherosclerosis. Examples of the relations between metabolism and phenotype are summarized in Table 3.Table 3—Diagnostic results, health implications, and potential diet solutions

Cholesterol levels in blood have been studied for decades because of the association between high cholesterol and heart disease. The absolute quantity of cholesterol in blood provides a quantitative estimate of cardiovascular disease risk for an individual. Those whose cholesterol levels are high have been empowered to act on their health by taking various steps to improve their metabolism to reduce cholesterol, including diet, exercise, and drugs.

Cholesterol also illustrates how more systematic measurement of metabolites will make cholesterol information more actionable. Measuring cholesterol in blood provides an indication of long-term risk, yet the single measurement does not say why it is high in that individual. Knowing why one person’s cholesterol level is high can lead to a more specific and appropriate means to resolve the problem.

Several biochemical mechanisms lead to elevated serum cholesterol in blood, and three examples illustrate the value of a greater understanding of personal metabolism: in some individuals, cholesterol absorption in the intestine is inordinately high; in others, endogenous biosynthesis of cholesterol is too high; and in still others, cholesterol is poorly metabolized to bile acids.

Measuring only cholesterol in blood does not distinguish these three mechanisms, but measuring additional metabolites does. High concentrations of phytosterols in plasma reflect a high absorption of sterols in the intestine, and for individuals with this problem, adding foods to their diets that inhibit intestinal absorption of cholesterol is effective (Mussner et al., 2003). The concentration of mevalonic acid in blood reflects the rates of sterol biosynthesis; for individuals who synthesize too much cholesterol, foods that inhibit absorption are not as effective, but drugs and foods that inhibit cholesterol biosynthesis are (Yoshida et al., 1993). Measuring the concentration of the bile acid metabolite 7-alpha-hydroxy-4-cholesten-3-one in plasma reflects the rate of bile acid synthesis, and those who produce few bile acids can enhance this pathway for cholesterol excretion by consuming foods that stimulate bile acid metabolism and avoiding foods that inhibit bile acid synthesis (Shoda et al., 1997).

Homocysteine is another example of a metabolite linking metabolism and health. High homocysteine is associated with increased risk of heart disease, stroke, and neurodegenerative diseases. Assessing homocysteine plus other related metabolites provides the information necessary to guide food choices to lower homocysteine.

The metabolic problems of overweight, diabetes, dyslipidemias, osteoporosis, and hypertension are health issues that develop over time, but in most cases those affected are unaware of the nutritional problems before they are severe and difficult to reverse. Routine metabolic assessment could readily identify such health issues. For example, high levels of triglycerides in blood are an early indication of metabolic dysregulation of energy, which in many cases can be improved rapidly by decreasing foods rich in rapidly absorbed glucose equivalents and increasing foods rich in polyunsaturated lipids. High circulating concentrations of calcitriol (1,25-dihydroxyvitamin D3) in blood reflect poor calcium status, a threat to bone health, hypertension, and obesity (Weaver and Boushy, 2003). This state of inadequate calcium status could be usually resolved by consuming foods rich in absorbable calcium and vitamin D, if only the affected individuals knew that this was a problem.

--- PAGE BREAK ---

• Linking Food Composition to Metabolites. The public health problem of avoiding deficiencies of essential nutrients was resolved by analyzing various food commodities for their content of essential nutrients. This information was the basis for the food recommendations, enrichments, fortifications, and food processing designs that were needed to match food choices within an overall diet to adequate intakes of all essential nutrients. Personalizing foods for metabolic health will likely take a similar path initially.

Because metabolic health depends on more than just the essential nutrients, food composition databases will need to be more complete for optimizing health. Nonessential nutrients, secondary plant metabolites, proteins, lipids, saccharides, nucleotides, the presence and activity of enzymes, the structures of macromolecules and tissues—all of which are recognized to affect the health value of food products—will be the information basis for redesigning foods. Databases are being updated by agencies such as the U.S. Dept. of Agriculture to include more detailed information about each agricultural commodity. These data will be the palette with which future foods and diets are designed.

The values that particular compositions provide to food product designers and consumers will also dictate, in part, the evolution of value in agricultural commodities. Food materials with unusually active components will become increasingly more valuable as the individuals who can take most health advantage of those components are identified.

Providing Health and Delight
The food marketplace is one of the most creative and dynamic systems known. It is thus impossible to predict exactly how personalized foods will evolve over the next decade. It is only certain that they will. The necessary inputs to solving health issues are knowledge of human variations and knowledge of foods that match individual needs. As the technologies of assessment evolve, the opportunities to provide greater personal guidance will emerge as well.

The process of personalizing foods is underway and it will continue at a speed driven primarily by self-knowledge of consumers or their care providers. Consumers who know their metabolic status will be increasingly guided to food products that improve their health.

It is probable that the process of personalizing foods will develop as a series of stages, each of which is driven by greater knowledge content. The first stage will be nothing more sophisticated than better guidance of food choices. The food supply is already quite diverse in the composition of all of the macronutrients, and the first stage of personalizing foods for health will likely take advantage of this diversity to simply guide each consumer to appropriate choices in the existing food marketplace. However, the approach of guiding food choices for health is flawed because consumers do not want to lose their freedom of choice for foods that they prefer by other criteria.

• Designing Foods. In a marketplace populated with consumers armed with self-knowledge of their health needs, the food industry and its supporting agricultural enterprise can begin to deliver foods that apply their own knowledge of food to improve the health of consumers. The food industry will be propelled to a new level as a knowledge-based enterprise.

Providing healthy foods is relatively easy. Providing healthy and delicious foods is the real challenge. Foods are an intimate part of everyone’s daily life, and for centuries, the pleasures of foods have defined much of life’s delights. During the last generation, immediate choice of foods has become a major added value of the food marketplace. Consumers who have increasingly been able to choose from a wide range of foods will not wish to abandon all of the values of taste, flavor, texture, convenience, etc., simply for metabolic health, nor should they. Foods must be simultaneously safe, delicious, convenient, affordable, etc. Thus, if foods with greater health are to succeed in the marketplace, the values of personal health must be added to all of the existing values of the same or similar foods.

Food that is healthy for an individual must also be acceptable. This will mean that foods of the future must be able to maintain existing extrinsic values while at the same time altering their composition to match specific intrinsic health goals. Is this possible? Of course—reformulating existing foods to the metabolic needs of consumers is an old idea, in some cases a very old idea.

--- PAGE BREAK ---

The food industry has assembled considerable information about the composition, structure, and physical properties of edible agricultural commodities, the biochemistry of these biological systems, and the principles of process engineering of biomaterials necessary to formulate, assemble, and distribute foods. The focus of much of the food industry’s research and development over the past century has been to manipulate specific food properties or “functionalities” while maintaining flexible control over the basic food compositions (Table 4).Table 4—Processes that impart organoleptic properties of foods

This trend in research and application has been driven by various factors, including cost, safety, ingredient availability, and, increasingly, health. An example is the value of sweetness. Sweetness was a major driver of food technology development starting centuries ago with the crystallization of natural sugars (sucrose) from sugar cane, to production of intensely sweet sugars (fructose) by enzymatic hydrolysis and isomerization of starches, to the discovery and chemical synthesis of intensely sweet, noncaloric sweeteners (e.g., aspartame). The latter sweetener has obvious value to diabetics for whom the routine consumption of sugar can be deleterious to their health.

Another example is the shortening properties of plastic fats. The functionality of “shortening” as an explicit food ingredient began with the rendering or creaming of animal fats, was broadened industrially with the hydrogenation of refined vegetable oils, and is now being extended agriculturally by selective breeding of plants to modify the fatty acid composition of seed oils (Ursin, 2003).

Process engineering has also dramatically increased our abilities to manipulate various ingredient compositions into specific desired physical properties of final foods. Traditional frozen dairy dessert (ice cream) is one of our most delightful foods. Ice creams were co-invented at various points in culinary history by simultaneously joining cream-based foams with the phase changes of water and fat to form stabilizing crystals. Newer processing technologies take advantage of extrusion and gas injection to produce frozen foams with widely varying compositions while achieving the similar desired physical properties.

An even more flamboyant example of the ability of food process design to produce a desired result is imitation caviar. Judicious understanding of the gelling properties of proteins in mixed hydrocolloid suspensions and simple flavoring techniques led to the production of one of cuisine’s most expensive items using simple, inexpensive raw materials (Tolstoguzov, 2002).

The number of food products with organoleptic properties that are achievable with a wide range of final compositions of macromolecules is very wide. All of these, and many more, food products illustrate the scientific knowledge that the food industry has amassed to manipulate food composition widely while achieving product properties that are recognizable as pleasurable foods by discerning consumers. This level of understanding of the chemical and biochemical events that underlie food processing has provided the knowledge base necessary to begin personalizing foods that are simultaneously healthy and delightful.

Personalized Food Marketplace
How will the system of agriculture and the food industry change as food choices become personalized? The simplest answer is, no one knows. Initially, all that is required to add a significant degree of personalized health is just to steer individuals to choices from existing products.

However, with a more personalized and knowledge-based food marketplace, the agricultural enterprise itself will begin to change. More-knowledgeable consumers will exert increasing demand for desired properties all the way to the agricultural producer. What to produce, how, and when will be driven less by cost drivers for the food industry and increasingly by more individual consumer values. In effect, agriculture will shift from a “push system” driven by the producer’s ability to generate commodities, to a “pull system” driven by the consumer’s needs for health and desires for delight.

Such trends are already apparent in different sectors of agriculture, including those already demonstrating a health-related value. For example, the red wine industry observed a significant increase in market sales and overall value when scientific research linked red wine to a reduced incidence of heart disease (Frankel et al., 1993). This value of reduced heart disease risk did not lead to a standardization of wine about a lowest-cost “healthy” wine relying solely on its perceived health value. Instead, the wine industry continued to expand into increasingly diverse varieties, each coveted by distinct consumers, in which the health feature was just one part of an overall value package.

--- PAGE BREAK ---

Over the longer term, the gradual evolution of the food industry into a knowledge-driven provider of personal health will enlist the power of biology to expand value throughout agriculture (Watkins et al., 2001). This change in agriculture and food is illustrated in Fig. 1. In the current food production system, the food industry realizes profit by leveraging economies of scale and minimizing costs throughout the food chain. In such a technological model, the biological properties of input commodities are in many instances liabilities, as the industrial practice seeks to unify the chemical properties of ingredients and processes. In most cases, values are added in as discrete ingredient functionalities, much as traditional cuisines add ingredients. Biological diversity compromises the industrialization economies of scale as processing and formulations are combined to neutralize the effects of natural variation.The food process chain from agricultural raw materials to the final health effects of products. In more traditional processing, engineering is dominated by added functionalities for sensory and acceptance criteria and renovations designed to minimize cost throughout the production chain. Nutrition and health are properties added on as an isolated part of quality and safety to meet mandated general product compositions. As food products take on a more intimate role in the management of personal health, the wellness of the consumer becomes a key value driver, and various features of the health of the consumer will “bioguide” all of the stages of food processing, from softer refining to point of sale production and individualized benefits to the consumer.

However, biological diversity in agriculture will become an asset as the biological diversity of consumers is recognized. The current industrial process systems that are based on simple chemical and bulk technologies will change to take advantage of this diversity. As a result, as the biological variation of food components become associated with specific health values appropriate to particular consumers, the food production process will adopt technologies to capture them. Processing will be guided by the endogenous values of agricultural inputs. Again, to refer to red wine as an example, fermentation of must in the presence of the skins and seeds of red grapes selectively captures the endogenous flavonoids and anthocyanins that are natural to the grape. Furthermore, the unique attributes of pinot noir vs cabernet sauvignon grapes are not liabilities to be processed into homogeneity, but assets adding to the value of high-cost varietal wines.

The genomics era will produce a new level of understanding of life sciences, and this will lead to more professional use of living organisms and their properties. Food processing will take increasing advantage of strategies to produce assorted foods in which biological principles will guide processing rather than the other way around. Not surprisingly, the industries that are “paying” for genomics science will use bioguided processes first. But, as biological tools become more and more a part of the pharmaceutical and fine chemicals industries, the food industry will apply them as well, as food process design becomes more and more bioguided (Ward et al., 2004).

With power comes responsibility. Consumers must recognize that they will become more empowered to improve their existing health and expand their prospects for future health. But they will also need to take responsibility for pursuing self-knowledge and using it wisely. Producers and the food industry will be empowered to connect more intimately with their consumers and markets but will need to take responsibility to deliver what they promise and not promise what they can’t deliver.

by J. Bruce German, Chahan Yeretzian, and Heribert J. Watzke
Author German is Professor, Dept. of Food Science and Technology, University of California, 1 Shields Ave., Davis, CA 95616 and Senior Scientific Advisor for the Nestlé Research Center, P.O. Box 44, CH-1000 Lausanne, Switzerland. Author Watzke is Dept. Head and author Yeretzian is Group Leader, Nestlé Research Center. Send reprint requests to author German ([email protected]).


Alberti, G. 2001. Noncommunicable diseases: Tomorrow’s pandemics. Bull. World Health Org. 79: 907.

Chagnon, Y.C., Rankinen, T., Snyder, E.E., Weisnagel, S.J., Perusse, L., and Bouchard, C. 2003. The human obesity gene map: The 2002 update. Obesity Res. 11(3): 313–367.

Decombaz, J., Schmitt, B., Ith, M., Decarli, B., Diem, P., Kreis, R., Hoppeler, H., and Boesch, C. 2001. Postexercise fat intake repletes intramyocellular lipids but no faster in trained than in sedentary subjects. Am. J. Physiol. Regul. Integr. Comp. Physiol. 281: R760–769.

Frankel, E.N., Kanner, J., German, J.B., Parks, E., and Kinsella, J.E. 1993. Inhibition of oxidation of human low-density lipoprotein by phenolic substances in red wine. Lancet 341(8843): 454–457.

Hertzler, S.R., Savaiano, D.A., and Levitt, M.D. 1997. Fecal hydrogen production and consumption measurements. Response to daily lactose ingestion by lactose maldigesters. Dig. Dis. Sci. 42: 348–353.

Hudson, R. and Distel, H. 1999. The flavor of life: Perinatal development of odor and taste preferences. Schweiz. Med. Wochenschr. 129(5): 176–181.

Kunze, C., Grossmann, U., Stork, W., and Muller-Glaser, K.D. 2002. Application of ubiquitous computing in personal health monitoring systems. Biomed. Tech. (Berl.) 47(Supp. 1, Part 1): 360–362.

Levin, B.E. 2000. Metabolic imprinting on genetically predisposed neural circuits perpetuates obesity. Nutrition 16: 909–915.

Muller, M. and Kersten, S. 2003. Nutrigenomics: Goals and strategies. Nat. Rev. Genet. 4: 315–322.

Mussner, M.J., Parhofer, K.G., Von Bergmann, K., Schwandt, P., Broedl, U., and Otto, C. 2002. Effects of phytosterol ester-enriched margarine on plasma lipoproteins in mild to moderate hypercholesterolemia are related to basal cholesterol and fat intake. Metabolism 51(2): 189–194.

Puska, P. 2002. Nutrition and global prevention on non-communicable diseases. Asia Pac. J. Clin. Nutr. 11(Supp. 9): S755–758.

Reddy, J.K. and Hashimoto, T. 2001. Peroxisomal beta-oxidation and peroxisome proliferator-activated receptor alpha: An adaptive metabolic system. Ann. Rev. Nutr. 21: 193–230.

Shoda, J., Miyamoto, J., Kano, M., Ikegami, T., Matsuzaki, Y., Tanaka, N., Osuga, T., and Miyazaki, H. 1997. Simultaneous determination of plasma mevalonate and 7alpha-hydroxy-4-cholesten-3-one levels in hyperlipoproteinemia: Convenient indices for estimating hepatic defects of cholesterol and bile acid syntheses and biliary cholesterol supersaturation. Hepatology 25(1): 18–26.

Singhal, A., Wells, J., Cole, T.J., Fewtrell, M., and Lucas, A. 2003. Programming of lean body mass: A link between birth weight, obesity, and cardiovascular disease? Am. J. Clin. Nutr. 77: 726–730.

Tai, E.S., Adiconis, X., Ordovas, J.M., Carmena-Ramon, R., Real, J., Corella, D., Ascaso, J., and Carmena, R. 2003. Polymorphisms at the SRBI locus are associated with lipoprotein levels in subjects with heterozygous familial hypercholesterolemia. Clin. Genet. 63(1): 53–58.

Tolstoguzov, V. 2002. Thermodynamic aspects of biopolymer functionality in biological systems, foods, and beverages. Crit. Rev. Biotechnol. 22(2): 89–174.

Ursin, V.M. 2003. Modification of plant lipids for human health: development of functional landbased omega-3 fatty acids. J. Nutr. 133: 4271–4277.

van Ommen, B. and Stierum, R. 2002. Nutrigenomics: Exploiting systems biology in the nutrition and health arena. Curr. Opin. Biotechnol. 13: 517–521.

Ward, R.E., Watzke, H.J., Jimenez-Flores, R., and German, J.B. 2004. Bioguided processing: A paradigm change in food production. Food Technol. 58(5): 44–48.

Watkins, S.M., Hammock, B.D., Newman, J.W., and German, J.B. 2001. Individual metabolism should guide agriculture toward foods for improved health and nutrition. Am. J. Clin. Nutr. 74: 283–286.

Weaver, C.M. and Boushey, C.J. 2003. Milk—Good for bones, good for reducing childhood obesity? J. Am. Dietet. Assn. 103: 1598–1599.

Yoshida, T., Honda, A., Tanaka Net Matsuzaki, Y., He, B., Osuga, T., Kobayashi, N., Ozawa, K., and Miyazaki, H. 1993. Simultaneous determination of mevalonate and 7-hydroxycholesterol in human plasma by gas chromatography-mass spectrometry as indices of cholesterol and bile acid biosynthesis. J. Chromatog. 14: 185-193.

Zhang, J.J., Okutani, F., Inoue, S., and Kaba, H. 2003. Activation of the cyclic AMP response element-binding protein signaling pathway in the olfactory bulb is required for the acquisition of olfactory aversive learning in young rats. Neuroscience 117: 707–713.