Chapter 1 Chemistry and Health Benefits of Caffeinated Beverages: Symposium Overview D. G. Steffen Scientific Relations, Kraft Foods, Inc., Glenview, IL 60025 This symposium provided a unique opportunity for food scientists, analytical chemists, and biological researchers to engage in an interdisciplinary conversation exploring the potential health benefits inherent in the world's most popular beverages. This continued a long-standing tradition of the Division of Agricultural and Food Chemistry wherein advances in food chemistry knowledge have been described in relation to the food itself as well as to the physiological outcome in an end-user organism. Introduction Caffeine, a 1,3,7-trimethyixanthine, is naturally present in over one hundred plant species. However, over the centuries, humans have selected only a few plant varieties as common sources for consumption: coffee beans, tea leaves, cocoa beans, kola nuts, guarana seeds, and mate. These natural sources of caffeine are generally consumed as beverages; cocoa and chocolate are enjoyed as beverages or solid confections; and fresh kola nuts are sometimes chewed. If the common sources of naturally occurring caffeine listed above are considered as one "commodity," the global economic impact of this one substance is enormous. Indeed, green (raw) coffee beans alone are the second most widely traded commodity in the world (oil being first). Thus, understanding the basic chemistry that contributes to the desirable organoleptic qualities of these source plants is an important scientific and economic endeavor. Although caffeine is a unifying theme of this workshop, it is by no means the total focus of the papers that follow. The occurrence of caffeine and other methylxanthines such as theobromine and theophylline in the tropical and subtropical plants of interest here may represent a common evolutionary strategy to 2 2000 American Chemical Society
3 deter predators. However, caffeine and a host of other chemical compounds also account for the complex flavor and aroma profiles that make the food and beverage products of these plant species so pleasing to humans. Many of these plant-based chemicals, or phytochemicals, are currently under intense study by medical and health researchers for their putative beneficial health effects. For example, part of the well-recognized health benefit of fruits and vegetables appears to be a result of polyphenols compounds acting as antioxidants in the body. Several papers in this book suggest that the food and beverage products made from caffeine-containing plants can deliver substantial levels of antioxidant polyphenols to the human diet. This overview continues with a brief review of the various chapters in a slightiy different grouping than as published. The purpose is to separate papers dealing with a food chemistry topic (e.g., new analytical method, effect of processing step) from those dealing with a physiological topic (e.g., tumor formation, blood lipoprotein oxidation, behavior). Due to the nature of the research database, the majority of the papers discuss caffeine, coffee, or tea; several address cocoa, but only a few cover guarana and kola. Chemistry Knowledge of the mechanism and chemistry of compositional changes during growth and storage of caffeine-containing plants, and their processing to final products, may permit enhancement of compounds with beneficial organoleptic or health properties or reduction of less desirable compounds. Many papers concentrate on the aroma and flavor compounds that are important to consumer acceptance and enjoyment of these products. However, a few papers go beyond the traditional hedonic endpoints to consider these same compounds as potential contributors to human health based on their antioxidant capabilities. Analysis Modern analytical methods increase knowledge of the chemical profile of complex foods and can identify specific compounds useful for quality control purposes. Stadler extends work on C-8 hydroxylated methylxanthines as markers of oxidation in raw coffee, tea and cocoa materials and finished products. Ames demonstrates the use of capillary electrophoresis to examine the color fraction of roasted coffees. Hammerstone presents a method (HPLC-MS with API-ES chamber) to identify a variety of flavonoids, especially procyanidins, in tea, cocoa and chocolate products. Speer notes that the unique coffee diterpenes kawheol and cafestol may be used to distinguish among Arabica blends, while 16-0- methylcafestol may distinguish Robusta coffee in Arabica mixtures and that the distribution of these compounds in roasted coffee tends to remain similar to that found in the respective green bean type.
4 Flavor and aroma play special roles in consumer acceptance of caffeinated beverages. Using HPLC/UV PDA and HPLC/DC Voltammetry, Cohen determines specific green coffee phenols and phenolic acids which may impact the flavor profile. Likewise, Glazier describes statistical methods to correlate traditional laboratory evaluations of cocoa bean quality with sensory panel data to develop a predictive model of cocoa bean flavor based on instrumental measures of cocoa bean volatiles generated during roasting. General Considerations Coffee and Tea Waller describes recent work on caffeine metabolism in coffee and tea plants. Caffeine biosynthesis uses a common pathway from xanthosine through theobromine to caffeine, but theophylline does not appear to be methylated to caffeine. Caffeine accumulation generally results from a slow degradation of caffeine to theophylline; the caffeine degradation sequence then proceeds by conversion of theophylline to xanthine via 3-methylxanthine and entry into the purine catabolic pathway. Caffeine metabolism is related to the stage of plant development, time of year, and species. Segall compares and contrasts coffee and tea from farm to cup. The similarity of preparation of the final beverages (hot water infusion) belies the botanical, agronomic, processing, and compositional differences between coffee and tea. Coffee Several papers examine the chemistry of coffee flavor development. Parliment outlines the major chemical changes occurring during the roasting of coffee beans that lead to desirable flavor compounds and discusses the most important flavor compounds in roasted coffee. Grosch finds that a quantitative mixture of 27 key odorants of roasted Arabica coffee produced an odor profile close to a real sample, but only 8 of these odorants were essential for the coffee flavor. A key thiol flavor component, 2-furfurylthiol, is present in coffee brews at about one-third the amount found in roasted coffee. Blank suggests that most of this loss may result from oxidative processes. The same sulfur-containing compound exhibits a distinct coffee-like aroma characteristic of brewed coffee. Rizzi examines model systems that suggest the key precursors in formation of this thiol may be pentose sugars and coffee bean proteins, the latter being responsible for the unique aroma developed on roasting. In a somewhat different approach, Steinhart attempts to model the effect on volatile aroma compounds of various dairy-based additives in typical coffee beverages. Many processing conditions affect final beverage characteristics. Ko concludes that roasted bean quality improved with high temperature/short time roasting based on aromatic impact, cup strength, taste intensity, and extraction parameters.
Cammenga describes a filter bypass system that allows full caffeine extraction at a lower total solubles level, giving a "lighter" but folly stimulating brew. 5 Tea In his review of tea chemistry, Ho considers changes in the major chemical compounds, catechins and methylxanthines, during fermentation from green to black tea relative to the distinctive flavor and color of various teas. High levels of catechins in green tea decrease with fermentation, such that black tea contains predominantly undefined polyphenols. Catechins contribute to the astringency of the final brew, while bitterness may be a function of caffeine complexation with tea polyphenols. Tea aroma derives from critical changes in several compounds, but the predominant chemicals in green tea aroma and black tea aroma are different. Masuda describes flavor changes in canned green and black teas after retorting. Although significant variations in the volatile fraction contribute to off-flavors, some may be generated from nonvolatile precursors. An elegant report by Sakata suggests that the action of a specific enzyme, β-primeverosidase, during fermentation produces most of the alcoholic tea aroma (floral tea note) from stored disaccharide glycosides. Green tea catechins are important to the taste of green tea. Yoshida reports that lower ph improves extraction of major catechins and that ph decreases with increasing tea concentration; however, at higher tea concentrations, gallate catechins are less well extracted than non-ester catechins. Cocoa, Guarana and Kola Nuts As with coffee, cocoa beans undergo a lengthy preparation process to produce the usable fractions, cocoa butter and cocoa powder. Schieberle follows the important chemistry of cocoa from growing conditions through manufacturing in relationship to the bitter taste, melting behavior and, importantly, characteristic aroma of cocoa products. Hashim argues that the complex flavor of cocoa develops mainly through the Maillard reaction during the roasting process. Chen further elucidates cocoa flavor development as reactions of sugars, amino acids, lipids and other precursors and identifies two compounds imparting strong chocolate-like notes to the overallflavorprofile. Much less is known about the chemical changes in guarana and kola nuts during growth and processing. Walker notes that guarana, mainly grown in Amazonian Brazil, is likely the richest known vegetable source of caffeine. Compositional changes on processing appear minimal, and little is known about the flavor profile. Most guarana is used commercially as a concentrate for beverages. Ringleib speculates that the characteristic flavor notes of the kola nut, a flavor ingredient in cola beverages, are likely due to borneol, geosmin and selected terpenoid compounds; however, data on processing chemistry are scarce.
6 Antioxidant Properties Several papers consider the antioxidant properties of the polyphenols compounds in coffee, tea, and cocoa. Detailed studies provide information not only on the types and concentrations of these components in raw materials and final products, but also on mechanisms of action which are likely important to human health after ingestion. Engelhardt presents data showing that tea, coffee and cocoa drinks can contribute substantial amounts of total polyphenols to the diet, but only tea appears to be a significant source of flavonoids. Similarly, Lee notes that the total antioxidative activity of aroma chemicals in brewed coffee may be comparable to that of recognized antioxidant vitamins. Using a DPPH radical system, Bungert examines the relationship between phenolic compound structure and radical scavenging activity, while Suzuki begins to unravel the catechol type catechin radical scavenging mechanism and stoichiometry. Health As with all organisms, the unique biochemistry and physiology of the caffeinecontaining plants are the result of evolutionary strategies which ensure their survival the chemistry is for the benefit of the plant. Although humans have long enjoyed the products of these plants for their taste and mildly stimulating effect, only during the latter half of this century have there been intense investigations of these plant components in relation to human health. Because of its widespread consumption in beverages, caffeine is one of the most studied chemicals in the world. Today, most of the concerns about the adverse effect of caffeine on major health problems (e.g., cardiovascular disease, osteoporosis, cancer) have been refuted by carefully controlled studies. Now, many investigators are looking at potential beneficial impacts on human health that may derive from caffeine itself or from other compounds, particularly the polyphenols, in these plants. Perhaps the foods and beverages from caffeine-containing plants are just an early example of what in contemporary parlance are called "functional foods"--in this case, a common food whose unique phytochemicals may impart potential health benefits. Behavior Controversy remains among researchers about the long-term effects of caffeine use on behavioral endpoints. Smith reviews a decade of work on the effects of caffeine on mood and cognitive performance. He suggests that caffeine at relatively low levels, comparable to those achieved with average consumption patterns, rapidly improves mood and sustained performance tasks, especially in low alertness situations. In a different interpretation of similar data, Rogers contends that consumers' liking of the taste, aroma, and flavor of caffeinated beverages is
7 reinforced by the stimulating effects of caffeine. Over time, consumers lose the expected beneficial stimulant effect and continue to drink these beverages to avoid unwanted consequences of caffeine deprivation and maintain a normal level of alertness. While the opposing conclusions of Smith and Rogers derive from similar psychological testing paradigms, both agree that there are conditions under which caffeine consumption is beneficial. Another contentious area of behavioral psychology is whether or not caffeine should be considered a drug of dependence. In one approach to this question, Nehlig uses a sensitive animal model of cerebral energy metabolism to understand the behavioral effects of caffeine. Her data indicate that areas of the brain controlling locomotor activity and the sleep-wake cycle are very sensitive to low concentrations of caffeine which are in the general range of human consumption. However, brain structures involved with addiction, reward and motivation respond only to levels of caffeine that induce global brain stimulation and general adverse effects. Importantly, these results are being replicated in human studies. Thus, the data do not support a potential dependence on caffeine at normal levels of intake. Cancer The influence of caffeine and coffee on carcinogenesis has been closely followed for many years. Although many reports suggest that some particular chemical found in coffee may produce a mutagenic or carcinogenic outcome in an isolated test system, Adamson reports that coffee is not mutagenic in vivo and that coffee or caffeine are not carcinogenic in long-term animal bioassays. Likewise, both animal and epidemiological studies suggest that caffeine and coffee may inhibit tumor induction and are not associated with various cancers. Ho continues these observations with detailed results on the reduction by caffeine of chemically-induced skin, stomach, mammary and lung tumorigenesis in animal models. Extending previous work on green coffee beans and fraction thereof, Miller demonstrates in a hamster model that consumption of roasted coffee beans or fractions inhibits the development of tumors. In a succinct summary of recent cancer research, an invited paper by Sivak concludes that coffee consumption does not affect the occurrence of cancer in humans; furthermore, coffee intake clearly reduces the risk of colon cancer, possibly due to the antioxidant-like activity of several compounds in coffee. Similarly, Lin attempts to elucidate mechanism for cancer prevention of tea based on studies with tea polyphenols. Antioxidants A final series of papers may be grouped by a common focus on health-related outcomes based on antioxidant properties of coffee, tea or cocoa products. Elevated plasma cholesterol, particularly the low-density lipoprotein (LDL) fraction, is one of many recognized risk factors for cardiovascular disease; recent data suggest that
8 oxidized LDL may actually be the important fraction in development of arterial wall plaques. Richelle compares the ability of coffee, tea and cocoa beverages, under normal preparation conditions, to inhibit LDL oxidation in an in vitro test system. All beverages prolong the lag time before initiation of oxidation, based on polyphenol content. Other measures indicate that antioxidant polyphenols are likely absorped at levels within an active physiological range. In similar work, Schmitz examines the structures of procyanidins in chocolate and cocoa in relationship to their activity in various in vitro tests related to potential cardiovascular health benefits. Chen concludes that the hypolipidemic activity of green tea catechins fed to hamsters is related to reduced absorption of cholesterol and triglycerides. Osakabe presents results from a variety of in vitro tests and animal feeding experiments with a crude polyphenol extract from cocoa liquor. The data corroborate and extend the results from other investigators. The extract shows antimutagenic and anti-tumorigenic activity, reduction of oxidative stress and LDL oxidative susceptibility, and a slight level of anti-ulcer activity. Since catechins are only partially absorbed after ingestion, Hara offers a unique perspective on green tea catachins as antibacterial agents in the gastrointestinal tract. Studies suggest that isolated catechins, given at amounts present in normal tea consumption ranges, can improve fecal parameters associated with health benefits in a reversible fashion. Conclusions Sophisticated analytical techniques continue to reveal in ever greater detail the complex chemistry of caffeine-containing plants and their products. Flavor and aroma chemistry remains a most important area of research. Effects of processes such as fermentation, drying, roasting and grinding on precursor compounds and mechanisms of action of flavor and aroma development are subjects of intense study. Coupled with sensory panels and statistical models, such research attempts to identify chemicals of importance to consumer enjoyment of tea, coffee and cocoa food and beverage products and to predict which raw materials may produce the best final products. Research on potential health benefits of these products is in its early stages. Current data on polyphenols suggest that the protective antioxidative effects of these compounds observed in source plants may be transferred to humans. This exciting hypothesis must be confirmed by sound science. Human research needs to establish that these substances are absorbed in physiologically significant amounts and active forms. Meaningful in vivo endpoints of antioxidative activity need to be developed so that controlled human trials can corroborate experimental animal work and epidemiological observations. With more definitive human data, the foods and beverages from caffeinated plants may be prized for their health benefits in the future just as they are cherished for their organoleptic qualities today.