3

Safety Signals and Surveillance

In addition to estimating exposure, another important first step to assessing the public health safety of the growing number of new caffeine-containing food and beverage products entering the marketplace is conducting surveillance and identifying incidents that warrant further investigation. The panelists of the Day 1, Session 2, panel, moderated by Steven E. Lipshultz, M.D., Department of Pediatrics, University of Miami, Florida, considered the types of surveillance in place and the safety signals being detected. This chapter summarizes their panel presentations and the discussion that followed. Box 3-1 describes the key points made by each speaker.

CAFFEINE AND ENERGY DRINK EXPOSURE CALL SURVEILLANCE

Presented by Alvin C. Bronstein, M.D., Rocky Mountain Poison Center

The United States, Puerto Rico, and the U.S. Virgin Islands, plus three Pacific jurisdictions, are served by 56 U.S. poison centers, with all the information from callers contemporaneously entered into a computer database and uploaded approximately every 19 minutes to the National Poison Data System (NPDS). Alvin C. Bronstein described the type of data collected during those calls and presented results from a descriptive analysis of caffeine exposure calls from January 2000 through July 2013 and energy drink exposure calls from June 2010 to July 2013.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 31
3 Safety Signals and Surveillance In addition to estimating exposure, another important first step to assessing the public health safety of the growing number of new caffeine- containing food and beverage products entering the marketplace is con- ducting surveillance and identifying incidents that warrant further investi- gation. The panelists of the Day 1, Session 2, panel, moderated by Steven E. Lipshultz, M.D., Department of Pediatrics, University of Miami, Flori- da, considered the types of surveillance in place and the safety signals be- ing detected. This chapter summarizes their panel presentations and the discussion that followed. Box 3-1 describes the key points made by each speaker. CAFFEINE AND ENERGY DRINK EXPOSURE CALL SURVEILLANCE Presented by Alvin C. Bronstein, M.D., Rocky Mountain Poison Center The United States, Puerto Rico, and the U.S. Virgin Islands, plus three Pacific jurisdictions, are served by 56 U.S. poison centers, with all the information from callers contemporaneously entered into a computer database and uploaded approximately every 19 minutes to the National Poison Data System (NPDS). Alvin C. Bronstein described the type of data collected during those calls and presented results from a descriptive analysis of caffeine exposure calls from January 2000 through July 2013 and energy drink exposure calls from June 2010 to July 2013. 31

OCR for page 31
32 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS BOX 3-1 Key Points Made by Individual Speakers Alvin Bronstein described the National Poison Data System (NPDS), a near real- time national public health database, which collects call infor- mation from all 56 U.S. poison centers. Although centralized, national medi- cal databases are not perfect, but they can nonetheless serve as a valuable resource for early safety signals. NPDS can provide critical insight into early recognition of trends in chemical, pharmaceutical, and commercial products including caffeine and energy drink exposures. Bronstein presented the methods and results of a descriptive analysis of poison center calls showing an initial increase but recent stabilization in caffeinated energy product calls between 2010 and 2013. Bronstein opined that it is too early to predict future trends in energy product calls. Eighteen percent of the energy product calls resulted in no effect, with agitation, irri- tability, and tachycardia being the most frequently reported clinical effects. In addition to serving as a resource for early safety signals, Bronstein emphasized, the U.S. National Poison Control Center database could be used for more systematic and prospective data collection and analysis re- lated to caffeine exposure. Ashley Roberts described how industry has used the GRAS process to determine the safety of caffeine as a food or beverage ingredient, including as an ingredient in energy drinks. In addition to NHANES data, industry has relied on data from a wealth of animal toxicity and human safety studies. He disparaged case reports publicized by the media that are not supported by the weight of scientific evidence. Poison Center Calls and Data Bronstein described the NPDS as a passive system, not an active sys- tem. Calls come from both consumers and health care professionals, with most centers using registered nurses or poison specialists to field the calls and others using PharmDs. All are backed by medical and clinical toxicologists or physicians. Poison centers field two types of calls: in- formation calls, which are questions about substances, events, or medical conditions, and exposure calls. Data collected during the calls include information on 131 clinical effects, or “signs and symptoms,” with all data collected since 2000 available online. According to Bronstein, in addition to the contemporaneous record- ing of the calls by poison specialists, the poison centers also maintain a robust products database containing brand name and composition infor- mation for more than 390,000 products. The information typically comes directly from the companies that produce the products. Included are

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 33 4,401 caffeine products, 262 energy drinks, and 36 energy products (i.e., typically energy bars). The database can be searched by ingredient (e.g., taurine), as well as by product name. Since 2000, poison centers have received about 2.5 million exposure calls per year (see Figure 3-1), with peak times consistently being June, July, and August (Bronstein et al., 2012). The change over time for vari- ous types of agents has been very consistent, according to Bronstein, with steady increases during the past 12 years in exposure calls related to analgesics, sedatives/hypnotics/antipsychotics, cardiovascular drugs, and antihistamines. An Analysis of Caffeine-Related and Energy-Product Exposure Calls, 2000–2013 Bronstein described the methods and results from an analysis of sin- gle substance, closed, human exposure calls related to either caffeine (from January 1, 2000, through July 22, 2013) or energy products (from June 18, 2010, through July 22, 2013). For the purposes of the analysis, energy drinks and energy products were combined. July 22, 2013, was chosen as the end date because that was the day Bronstein and his col- leagues started analyzing the data. The analysis of energy product calls began on June 18, 2010, because energy product generic codes were not added to the generic code vocabulary until mid-2010. (Products are cate- gorized using a controlled hierarchical vocabulary, with seven energy product generic codes.) Bronstein and his colleagues calculated descrip- tive statistics; examined changes over time via either linear regression or spline fit; and correlated clinical effect frequencies. They used SAS JMP 9.0 for their analysis. The researchers identified 48,177 caffeine calls between January 1, 2000, and July 22, 2013, and 6,724 energy product calls between June 18, 2010, and July 22, 2013. In these calls, the caller said that either the caller or his or her child or other person had ingested what was identified as either a product containing caffeine or an energy product. As shown in Figure 3-2, caffeine exposure calls per day, by year, were not typical of the change over time that other products have shown over the past 12 years (i.e., as shown in Figure 3-1). Also shown in Fig- ure 3-2, energy product exposure calls per day, by year, increased initial- ly and then appeared to level off, although Bronstein stressed that it is too early to predict whether the leveling off will persist.

OCR for page 31
34 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS   — Human Exposures = –106,142 + 56.3* Year –21.0* (Year – 2006)^2 10,000 — Information Calls = –379,566 + 191* Year – 40.5* (Year – 2006)^2 — Animal Exposures = 8,840 – 4.22* Year – 4.23* (Year – 2006)^2 Encounters per Day 8,000 6,000 4,000 2,000 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Year FIGURE 3-1 U.S. poison center calls per day, 2000–2012. SOURCE: Bronstein et al., 2012. Data on the total number of poison center calls received related to caf- feine exposure were collected from January 1, 2000, through July 22, 2013, and calls related to energy product consumption from June 18, 2010, through July 22, 2013. Data were analyzed as change over time with linear regression, and correlation with clinical effect frequency. As shown in Ta- ble 3-1, of the 6,724 energy product calls analyzed, about 40 percent were for energy drinks containing caffeine only (i.e., without any other sub- stance contributing to caffeine production, so no guarana, kola nut, tea, yerba mate, cocoa, and so forth). Initially, there were a small number of calls for energy drinks containing caffeine and ethanol (alcohol), an obser- vation that led Bronstein to note that he and his colleagues have learned over the course of time that poison center calls can be used to show the effects of product withdrawals (i.e., energy drinks containing both alcohol and caffeine were withdrawn from the U.S. market after the FDA sent warning letters in 2010 to the manufacturers of those products). With respect to reasons for energy product calls, most calls analyzed (55 percent) were unintentional. Unintentional calls are usually expo- sures in young children who do not demonstrate intent. The next two most common reasons for energy product calls were for intentional mis- use (12 percent) and intentional abuse (8 percent). With respect to age categories, the highest mean exposures per month were for children aged 2 years (26.50 mean exposures/month), children aged 6 to 12 years (17.50 mean exposures/month), and adults

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 35 Caffeine Exposure 13 Year Single 2000 4523 12 2001 3997 2002 3449 11 2003 3219 Exposures/day 2004 3479 10 2005 3617 2006 4455 9 2007 4292 2008 3766 8 2009 3482 Spline fit, lambda = 0.0631 2010 3426 7 2011 2696 6 2012 2488 2001 2004 2007 2010 2013 2013 1288 Year Total 48177 Energy Product Consumption 9 8 7 Exposures/day 6 5 4 Year Exposures 3 2010 496 Spline fit, lambda = 1 2011 2309 2 2012 2471 1 2013 1448 Total 6724 0 2011 2012 2013 2014 Year FIGURE 3-2 Caffeine and energy product exposure calls per day over time. SOURCE: Alvin Bronstein. Presented to the Planning Committee for a Work- shop on Potential Health Hazards Associated with Consumption of Caffeine in Food and Dietary Supplements on August 4, 2013. aged 20 to 59 years (42.00 mean exposures/month). Among the 2-year- old children, which Bronstein noted is a prime age for unintentional in- gestion of substances, a linear regression revealed that the number of

OCR for page 31
36 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS ingestions increased over time between June 18, 2010, and July 22, 2013, but with an r2 value of only 0.41 (p < 0.0001). Medical outcomes associated with energy product calls received are listed in Table 3-2. Bronstein highlighted that 18 percent of the 6,724 energy product calls between June 18, 2010, and July 22, 2013, showed no effect. About 10 percent had a moderate effect, which includes symp- toms such as tachycardia. About 20 percent had minor effects—for ex- ample, hyperactivity or other effects that generally resolve. There was one death associated with consumption (2011). Among 2-year-old chil- dren, 34 percent of exposures had no effect, 12 percent had a minor ef- fect, and only three children experienced what would be considered a moderate effect. When Bronstein and colleagues compared the rank order of energy drink central nervous system clinical effects versus the rank order of cen- tral nervous system effects for all caffeine-related exposures, calls that did not involve energy products, they noticed similar patterns. For exam- ple, the most common central nervous system clinical effect, “agitat- ed/irritable,” for both types of calls ranked third among all clinical ef- fects for energy drinks and fifth for other caffeine-related calls. The next TABLE 3-1 Energy Product Calls by Generic Code, June 18, 2010, Through July 22, 2013 N % Energy Product Generic Code 2,664 39.6 Energy drinks: caffeine only (without guarana, kola nut, tea, yerba mate, cocoa, and so forth) (0200606) 1,870 27.8 Energy drinks: caffeine containing (from any source including guarana, kola nut, tea, yerba mate, cocoa, etc.) (0200605) 1,092 16.2 Energy drinks: unknown (0200608) 711 10.6 Energy products: other (0200609) 318 4.73 Energy drinks: ethanol and caffeine containing (from any source including guarana, kola nut, tea, yerba mate, cocoa, etc.) (0200610) 62 0.92 Energy drinks: no caffeine (from any source) (0200607) 7 0.10 Energy drinks: ethanol and caffeine only (without guarana, kola nut, tea, yerba mate, cocoa, etc.) (0200611) SOURCE: Alvin Bronstein. Presented to the Planning Committee on Potential Health Hazards Associated with Consumption of Caffeine in Food and Dietary Supplements on August 4, 2013.

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 37 TABLE 3-2 Medical Outcomes for 6,724 Energy Product Calls Received Between June 18, 2010, and July 22, 2013 N % Medical Outcome Not followed; minimal clinical effects possible (no more than mi- 2,548 37.9 nor effects possible) 1,301 19.3 Minor effecta 1,205 17.9 No effect 668 9.93 Moderate effectb 482 7.17 Unable to follow; judged as potentially toxic exposure 254 3.78 Not followed; judged as nontoxic exposure (clinical effects not expected) 216 3.21 Unrelated effect; exposure was probably not responsible for the effect(s) 29 0.43 Major effectc 20 0.30 Confirmed nonexposure 1 0.01 Death a NOTES: minor effect = symptoms that generally resolve, for example, hyperac- tivity; bmoderate effect = possible harm with symptoms such as tachycardia; c major effect = serious harm or death. SOURCE: Alvin Bronstein. Presented to the Planning Committee for a Work- shop on Potential Health Hazards Associated with Consumption of Caffeine in Food and Dietary Supplements on August 4, 2013. most common central nervous system clinical effect for both types of calls, “dizziness/vertigo,” ranked seventh among clinical effects for energy drink calls and eighth for caffeine calls. Next for both types of calls was “tremor,” followed by “headache.” In a similar comparison of the rank order of cardi- ovascular effects, again there appeared to be a good correlation in rank or- der between the two types of calls. The most common cardiovascular effect, tachycardia, ranked fourth for both energy drink and caffeine calls. The overall correlation of clinical effect frequencies (i.e., all clinical effects rec- orded) between energy product and caffeine exposure calls was 0.942 (p < 0.0001) (Bronstein et al., 2012). Bronstein remarked that he and his col- leagues would like to investigate this trend more thoroughly in an effort to identify which components of energy products are responsible for the clini- cal effects so similar to those being seen with caffeine exposure calls.

OCR for page 31
38 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS Conclusions About Energy Drink Exposure Calls In summary, energy product exposure calls to U.S. poison centers in- itially increased but appear to have stabilized, although without a full year (2013) of data, it is difficult to know whether the trend has in fact stabilized. Most energy product exposure calls are unintentional, fol- lowed by misuse and abuse. The most frequently reported clinical effects were agitation, irritability, and tachycardia. But 18 percent of energy product calls were recorded as having no effect. Bronstein and his col- leagues believe that the energy product clinical effect pattern is very sim- ilar to that of caffeine ingestions, but more work in that area needs to be done. Similarly, although poison center data allow for analysis of calls by specific active agents, more work in that area needs to be done as well. Limitations of poison center data include the fact that they represent self-reported exposures, whether from health care professionals or the public, so underreporting is likely. Also, exposures are confirmed with clinical laboratory data but only when clinically indicated. Generally, calls are not confirmed. On the basis of his and his colleagues’ findings, Bronstein made sev- eral recommendations for moving forward. Poison centers can assist with data gathering on energy product and caffeine exposure calls by (1) de- veloping a prospective-focused system for data collection in collabora- tion with the FDA and the IOM; (2) focusing further analyses and sur- veillance on products with specific active ingredients (e.g., taurine- containing energy products), on vulnerable populations, and on clinical effect profiles of substances (e.g., via cluster analysis); and (3) providing real-time alerts of exposure calls of public health significance. SAFETY ASSESSMENT OF CAFFEINE IN FOODS AND BEVERAGES Presented by Ashley Roberts, Ph.D., Intertek Cantox The industry approach to supporting the safety of food and beverage ingredients, such as caffeine, which includes energy drinks, relies on the GRAS determination process. Ashley Roberts provided an overview of

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 39 the GRAS process as it pertains to caffeine and briefly considered some current international regulatory positions on the safety of energy drinks. The GRAS Determination Process The procedure of establishing the use of a substance as GRAS is FDA regulated. The GRAS status of a food ingredient is based on the consensus of experts qualified by scientific training and experience to evaluate its safety. GRAS status determination requires the same quality and quantity of scientific evidence needed to obtain food additive ap- proval from the FDA, with the major difference being that the safety data used to determine GRAS status must be generally available to the scien- tific community. Equally important, Roberts said, the mere showing of safety is not sufficient. Rather, the use of the ingredient must be general- ly recognized as safe, which is normally achieved through publication of research in peer-reviewed journals. Estimating Human Exposure A key component in the safety evaluation process is estimating hu- man exposure. According to Roberts, estimates of caffeine consumption can be calculated using the NHANES database.1 On the basis of the data from the 2007–2008 and 2009–2010 NHANES surveys, 348 mg is the estimated daily intake at the 90th percentile for caffeine from back- ground diet (i.e., caffeine in coffee) plus caffeine from energy drinks (i.e., at an incorporation level of up to 120 mg/8 ounces) and dietary sup- plements for the entire population. The highest estimated daily intake for caffeine from background diet plus energy drinks and dietary supple- ments at the 90th percentile is for male adults: 444 mg caffeine/day. Roberts observed that, compared to estimates for background exposures, adding energy drinks and dietary supplements to the diet results in a min- imal increase in caffeine intake. Roberts noted that the 90th percentile is an FDA benchmark for high consumers and that, other than for male adults, estimates of caffeine intake at the 90th percentile among all other 1 See Chapter 2 for a discussion of how the NHANES database has been used to estimate caffeine exposure among different age groups and changes in caffeine exposure over time.

OCR for page 31
40 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS groups are below the 400 mg safety threshold level indicated by the FDA to be a safe level. Roberts identified several limitations of the NHANES database. First, the intake methodology generally overestimates exposure by as- suming that all energy drink products in the marketplace have caffeine incorporated at the maximum use level. Exposures are further overesti- mated by assuming that consumption on a given day is always associated with consumption of caffeine from other sources, including dietary sup- plements, which may not be the case in real-life situations. Also, the NHANES food survey is a short-term survey over 2 days, which over- estimates consumption over the longer term. Despite these limitations, exposure results from NHANES data are similar to those obtained from postmarketing surveillance data, including data from surveys conducted by the FDA and on behalf of the International Life Science Institute (RSEQ, 2011; FDA, 2012; ISQ, 2012; Mitchell et al., 2013). Moreover, two of those other surveys (RSEQ, 2011; ISQ, 2012), both conducted in Canada, collected data on up to 70,000 teenagers, showing that energy drink consumption 4 times or more weekly was found in about 2 percent of subjects, consistent with what the NHANES survey found. Safety Data The GRAS determining process for the safe use of caffeine as an added ingredient of food and beverages takes into account a range of data and information from metabolic, pharmacokinetic, animal, and human studies, including data on carcinogenicity, cardiovascular effects, sei- zures, and other specific effects and data on children and adolescents. Roberts highlighted studies that he thought most relevant. With respect to the metabolism and pharmacokinetics of caffeine, Kempf et al. (2010) reported that high daily intakes of caffeine from cof- fee, up to 1 gm/day, do not result in bioaccumulation. Arnaud (2011) reported that caffeine kinetics appears to follow a first-order process, and Hodgson et al. (2013) demonstrated that the kinetics of naturally occur- ring caffeine in coffee is similar to those of anhydrous caffeine. Roberts remarked that the latter finding was especially interesting because anhy- drous caffeine is what is added to products such as energy drinks. Roberts noted that because the major metabolites produced in hu- mans are almost identical to those found in rodents, this finding supports the use of rodents as animal models for assessing human safety. A com-

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 41 plete range of toxicity studies has been conducted with animal models, with a no-observed-adverse-effect level (NOAEL) of approximately 100 mg per kg per day for carcinogenicity and reproductive effects (Brent et al., 2011). According to Roberts, that dose is considerably higher than what humans are exposed to on a daily basis and equates to somewhere in the region of 40 to 80 cups of coffee per day. Human studies suggest that toxic doses of caffeine range from 3 g in young children to 10 g in adults (Iserson, 1990; Nawrot et al., 2003), which is equivalent to the consumption of between 28 and 93 8-ounce servings of energy drinks per day. With respect to specific effects, ac- cording to Roberts, caffeine has not been consistently linked with any adverse reproductive consequences. Nevertheless, both the European Commission’s Scientific Committee on Food and Health Canada rec- ommend that the consumption of caffeine during pregnancy be no more than 300 mg per day. To support these recommendations, energy drink companies have labeled their products as not being recommended for pregnant and nursing mothers. With respect to carcinogenic potential, the International Agency for Research on Cancer has concluded that there is inadequate evidence to confirm that caffeine, as present in coffee, is carcinogenic, a viewpoint that has been confirmed in several large prospective trials (Nawrot et al., 2003; Nkondjock, 2009; Freedman et al., 2012). Roberts noted that, in- terestingly, the largest prospective study to date (Freedman et al., 2012), with roughly 620,000 subjects, included about 10,000 men and 5,000 women who consumed at least 6 cups of coffee per day, which Roberts opined could result in consumption levels far greater than 1 g caffeine per day. Human studies on cardiovascular effects have shown that, while caf- feine produces a slight but transient increase in blood pressure, other electrocardiogram parameters are not affected. Several reports, including recent meta-analyses of prospective studies, have concluded that the ef- fects of caffeine on cardiovascular function are not clinically significant (IOM, 2001; Pelchovitz and Goldberger, 2011; Bohn et al., 2012; Freedman et al., 2012; Mostofsky et al., 2012). Similarly, for cardiac ar- rhythmias, results have shown that normal daily caffeine consumption, even when administered as a bolus dose in persons with arrhythmias, was well tolerated (Myers and Reeves, 1991; Pelchovitz and Goldberger, 2011; Menzes et al., 2013). Nevertheless, conflicting results have been reported among excessive consumers of coffee (i.e., more than nine cups per day) (Frost and Vestergaard, 2005; Gronroos and Alonso, 2010).

OCR for page 31
42 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS Although a number of case reports regarding seizures may be linked to energy drink consumption (e.g., Iyadurai and Chung, 2007; Trabulo et al., 2011), Roberts opined that these case reports have significant limita- tions that preclude establishing causality. A large and robust prospective study using data from the Nurses’ Health Study cohort did not identify caffeine as a risk factor for developing seizure (Dworetsky et al., 2010). Human studies have shown that acute caffeine administration nega- tively impacts calcium balance and glucose tolerance. Studies also show, however, that the effects on calcium balance can be mitigated through consumption of calcium in the diet (van Dam et al., 2004; Roberts and Rogerson, 2008; Bhupathiraju et al., 2013). Similarly, evidence from prospective studies suggests that caffeine from coffee protects against the risk of type 2 diabetes, thereby mitigating the short-term effects on glu- cose tolerance (Huxley et al., 2009; Wedick et al., 2011; Freedman et al., 2012; Bhupathiraju et al., 2013). Available evidence from studies in children (12 years and younger) and adolescents suggests that caffeine results in similar pharmacokinetic and pharmacodynamic effects in younger individuals as those observed in adults (ANZFA, 2000; Meltzer et al., 2008). According to Roberts, the effects of caffeine are therefore considered to be more a function of body weight rather than age. Health Canada has suggested that caffeine con- sumption by children be limited to no more than 2.5 mg per kg body weight per day (Nawrot et al., 2003; Rotstein et al., 2013). Roberts re- marked that caffeine has a long history of safe use in children for the treatment of apnea or prematurity and attention deficit disorder. In the case of caffeine use in energy drinks, GRAS determinations have also taken into consideration potential interactions with alcohol. Although some researchers have hypothesized that caffeine may antago- nize the effects of alcohol, resulting in greater alcohol consumption, the European Scientific Committee on Food and the UK Committee on Tox- icity of Chemicals in Food Consumer Products and the Environment have reported that the totality of evidence is currently insufficient to con- clude that co-consumption of energy drinks with alcohol is unsafe (SCF, 2003; COT, 2012). In Roberts’s opinion, the increasing concerns being expressed re- garding the use of caffeine in energy drinks come from a number of case reports. All published case reports, however, including incidences of

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 43 adverse events reported in the FDA’s Adverse Event Reporting System2 database, have been reviewed as part of the GRAS determination process. The expert panels reviewing this information agreed with the FDA’s posi- tion that an adverse effect report itself reflects only information as reported and does not represent any conclusion regarding a causal relationship with the product or ingredient. Roberts pointed out that, in addition, the Drug Abuse Warning Network (DAWN) report called into question the causal link between energy drink consumption and emergency room visits be- cause of a number of limitations in the data (SAMHSA, 2013). In sum, according to Roberts, on the basis of an overall assessment of the safety information, experts have determined that there is no differ- ence in the systematic handling of naturally occurring caffeine and caf- feine added to foods and beverages, such as energy drinks. Nor is the premise that adolescents react differently than adults following caffeine consumption supported by the scientific evidence. Roberts noted that the caffeine content in energy drinks is similar to or below that of a number of marketed coffee products and the incremental use of caffeine in ener- gy drinks combined with other sources of caffeine results in total human caffeine exposures below 400 mg/day in all age groups besides adult males. International regulatory positions, including those of the FDA and Health Canada, indicate that caffeine intake up to 400 mg daily is not associated with adverse effects. In Roberts’s opinion, the concern about caffeinated energy drinks is based on case reports that have been publicized in the media and are not supported by the weight of published scientific evidence on caffeine. The scientific experts who reviewed this evidence thought that the case re- ports were of limited relevance to a GRAS determination process and that the weight of scientific evidence, in combination with exposure data, supports the safety of caffeine in food and beverage products such as energy drinks. PANELIST DISCUSSION WITH THE AUDIENCE Following Roberts’s presentation, the moderator invited the work- shop audience to ask questions of either panelist. Most of the discussion 2 The FDA’s Adverse Event Reporting System is a computerized information data- base designed to provide support for postmarketing safety surveillance for all approved drug and therapeutic biologic products.

OCR for page 31
44 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS revolved around the interpretation of poison center data and the potential for future analyses of those data; the idea of a national registry to track health outcomes related to caffeine-containing energy drink exposure; the lack of data on interactions between caffeine and other ingredients in energy drinks (which is the subject of Chapter 6); and limitations of the available data. Questions About Poison Center Data There was some confusion about the reasons for poison center calls being categorized as intentional (misuse or abuse) versus unintentional. For example, how is intentional misuse different from unintentional use? Bronstein explained that an example of intentional misuse would be somebody consuming two or three doses of a product when the direc- tions indicated that only one dose should be consumed. Intentional abuse would be the use of a drug or nondrug product for pharmacological effect or intoxication. There were questions about reporting and whether the analysis that Bronstein presented included reports from both consumers and health care professionals. Bronstein said that the analysis did include both. Most health care professional reports are from emergency department physi- cians. He observed that, in general, calls from emergency departments are increasing for all substances. Although he and his colleagues did not examine emergency department data specifically, he said, it could be done. Finally, there was some discussion about the 18 percent of no-effect cases that Bronstein described. An audience member described that spe- cific conclusion as “misleading” because it implied that essentially 82 percent of cases had an effect. However, according to the audience member, on closer examination of children aged 12 years and younger, when cases that were not followed are included, in fact about 82 percent of cases in the pediatric population were asymptomatic and only 18 per- cent showed symptoms, the majority of which were minor. Bronstein explained that the 18 percent figure he reported was just one piece of information, that is, that 18 percent of all cases, including both pediatric and adult cases, were recorded as no effect. The audience member sug- gested that pulling and displaying data for children only, that is, for the pediatric population, would be helpful. When asked how helpful it will be in the future to incorporate into

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 45 the NPDS information on habitual versus nonhabitual use of caffeine, as well as information on actual caffeine levels, Bronstein replied that that type of information could certainly be collected. Currently, exposures are assigned chronicity, acute, acute-on-chronic, chronic, or unknown. Call history is taken in a manner similar to how physicians talk to their pa- tients, with everyone doing it a little differently. In his opinion, the da- taset could be standardized further to collect additional data. Underreporting and the Need for a National Registry Bronstein’s mention (during his presentation) of underreporting prompted an audience member to observe that, when looking for infor- mation on myocardial infarctions, it became evident that not all emer- gency departments were capturing those data. Thus, a registry was de- veloped and is now very well established within the United States such that anyone who experiences a myocardial infarction is well documented. Bronstein was asked whether a similar nationwide registry to track expo- sures and health outcomes might be beneficial in terms of “getting closer to the actual real data.” Bronstein replied, “Absolutely.” He remarked that the beginnings of such a registry already exist. Ninety-five percent of callers to poison centers provide phone numbers. If a nationwide reg- istry could be structured, it would not be difficult to call those callers back and provide them with the option of being entered into a national registry. A national registry would be especially helpful for obtaining information on the effects of chronic exposures. The Interaction Between Caffeine and Other Ingredients in Caffeine-Containing Energy Drinks Roberts was asked to describe studies on interactions between caf- feine and taurine, which the questioner noted seems to be a popular in- gredient in energy drinks, or with any of the other chemical and herbal ingredients in energy drinks. Roberts replied that there are no data on the combination toxicology of the various ingredients in energy drinks. Hav- ing said that, all of the other ingredients have undergone the GRAS de- termination processes. With respect to taurine in particular, he noted that taurine occurs naturally in the diet and is a component of meat, poultry, seafood, and dairy products. It is also present in human breast milk and is

OCR for page 31
46 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS added to some infant formulas. The level of taurine in energy drinks has been evaluated in the context of that background exposure scenario and is considered safe. The Importance of Anecdotal Experience An audience member disagreed with Roberts’s remark during his presentation that case studies are irrelevant and argued that anecdotal experience helps to generate hypotheses. Use of Animal Models to Study the Effects of Caffeine Session 4 (Day 1) speaker Sergi Ferré commented on Roberts’s claim that caffeine metabolites in rats versus humans are the same and that rats therefore serve as a good model for studying the safety of caf- feine exposure in humans. He noted that paroxetine, a caffeine metabo- lite with very strong psychostimulant effects, is a much more active me- tabolite in humans than it is in rats. Limitations of the Data An audience member who described himself as a practicing board- certified emergency physician remarked that he has not been seeing seri- ous side effects from caffeine from any source other than with intentional overdose from over-the-counter pills. He suggested that the approximate- ly 6 billion energy drinks being sold every year in the United States, with an average of only about 2,000 calls per year to poison centers, almost none of which report serious side effects, is pretty strong evidence that energy drinks are in fact very safe products. Bronstein replied, “Let the data speak for themselves.” Whether the level of analysis he and his colleagues conducted for the purposes of this workshop, with the data they had, presents a “total picture” is unclear at this point. Furthermore, he noted that more studies are needed to deter- mine what, if any, clinical effects are related to caffeine-containing ener- gy drink exposures. Audience members also offered some additional comments about the limitations of all datasets and how different datasets can lead to slightly different conclusions. For example, Session 3 (Day 2) panelist Steven

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 47 Lipshultz pointed out that the DAWN data suggest that the rate of ad- verse effects, as indicated by emergency room visits following energy drink exposure, is slightly higher than what the poison center data sug- gest. In response, another audience member commented that, even with the DAWN data, a numerator of approximately 10,000 (i.e., half of the approximately 20,000 emergency visits were for energy drink exposures alone, that is, not in combination with alcohol or another toxic ingestion) over an estimated 130 million emergency room visits per year suggests that the evidence for serious side effects is lacking. Lipshultz opined that, to the contrary, the DAWN data, as well as NHANES, ILSI, and poison center data, all point to the need for a more formal assessment. Each da- taset has its limitations, but each also has it strengths. They all deserve further discussion. REFERENCES ANZFA (Australia New Zealand Food Authority). 2000. Report from the Expert Working Group on the Safety Aspects of Dietary Caffeine. Canberra, Australia/Wellington, New Zealand. http://foodsafety.govt.nz/elibrary/ industry/ caffein-info-sheet.pdf (accessed November 26, 2013). Arnaud, M. J. 2011. Pharmacokinetics and metabolism of natural methylxan- thines in animal and man. Handbook of Experimental Pharmacology (200):33–91. Bhupathiraju, S. N., A. Pan, V. S. Malik, J. E. Manson, W. C. Willett, R. M. van Dam, and F. B. Hu. 2013. Caffeinated and caffeine-free beverages and risk of type 2 diabetes. American Journal of Clinical Nutrition 97:155–166. Bohn, S. K., N. C. Ward, J. M. Hodgson, and K. D. Croft. 2012. Effects of tea and coffee on cardiovascular disease risk. Food and Function 3(6):575–591. Brent, R. L., M. S. Christian, and R. M. Diener. 2011. Evaluation of the repro- ductive and developmental risks of caffeine. Birth Defects Research. Part B, Developmental and Reproductive Toxicology 92:152–187. Bronstein, A. C., D. A. Spyker, L. R. Cantilena, B. H. Rumack, and R. C. Dart. 2012. 2011 annual report of the American Association of Poison Control Centers’ National Poison Data Systems (NPDS): 29th annual report. Clini- cal Toxicology 50:911–1164. COT (Committee on Toxicity of Chemicals in Food, Consumer Products, and the Environment). 2012. COT statement on the interaction of caffeine and alcohol and their combined effects on health and behavior. http://cot. food.gov.uk/pdfs/cotstatementcaffalco201204.pdf (accessed December 25, 2013).

OCR for page 31
48 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS Dworetzky, B. A., E. B. Bromfield, M. K. Townsend, and J. H. Kang. 2010. A prospective study of smoking, caffeine, and alcohol as risk factors for sei- zures or epilepsy in young adult women: Data from the Nurses’ Health Study II. Epilepsia 51:198–205. FDA (U.S. Food and Drug Administration). 2012. FDA comments on “Caffeine Intake by the U.S. Population,” prepared by Laszlo P. Somogyi, 2010. Sil- ver Spring, MD: U.S. Food and Drug Administration, Center for Food Safe- ty and Applied Nutrition, Oakridge National Laboratory. http://www.fda. gov/downloads/AboutFDA/Centers/Offices/OfficeofFoods/CFSAN/CFSAN FOIAElectronicReadingRoom/UCM333191.pdf (accessed November 26, 2013). Freedman, N. D., Y. Park, C. C. Abnet, A. R. Hollenbeck, and R. Sinha. 2012. Association of coffee drinking with total and cause-specific mortality. New England Journal of Medicine 366:1891–1904. Frost, L., and P. Vestergaard. 2005. Caffeine and risk of atrial fibrillation or flutter: The Danish diet, cancer, and health study. American Journal of Clinical Nutrition 81(3):578–582. Gronroos, N. N., and A. Alonso. 2010. Diet and risk of atrial fibrillation— epidemiological and clinical evidence. Circulation Journal 74:2029–2038. Hodgson, A. B., R. K. Randell, and A. E. Jeukendrup. 2013. The metabolic and performance effects of caffeine compared to coffee during endurance exer- cise. PloS ONE 8(4):e59561. Huxley, R., C. M. Lee, F. Barzi, L. Timmermeister, S. Czernichow, V. Perkovic, D. E. Grobbee, D. Batty, and M. Woodward. 2009. Coffee, decaffeinated coffee, and tea consumption in relation to incident type 2 diabetes mellitus: A systematic review with meta-analysis. Archives of Internal Medicine 169:2053–2063. IOM (Institute of Medicine). 2001. Caffeine for the sustainment of mental task performance: Formulations for military operations. Washington, DC: Na- tional Academy Press. Iserson, K. V. 1990. Caffeine and nicotine. In Clinical management of poisoning and drug overdose, 2nd ed., edited by L. M. Haddad and J. F. Winchester. Philadelphia, PA: W.B. Saunders Co. Pp. 869–876. ISQ (Institute de la Statistique du Quebec). 2012. Tableau A3.2: Frequence de consummation habituelle de certaines boissons sucrees, eleves du sec- ondaire, Quebec, 2010–2011. Dans: L’Enquete quebecoise sur la santé des jeunes du secondaire 2010–2011: TOME 1: Le visage des jeunes d’aujourd’jui: leur santé physique et leurs habitudes ve vie. Chemin Sainte- Foy, Quebec: Institut de la Statistique du Quebec. http://www.stat.gouv. qc.ca/publications/sante/eqsijs.htm (accessed November 26, 2013). Iyadurai, S. J., and S. S. Chung. 2007. New-onset seizures in adults: Possible association with consumption of popular energy drinks. Epilepsy and Be- havior 10:504–508.

OCR for page 31
SAFETY SIGNALS AND SURVEILLANCE 49 Kempf, K., C. Herder, I. Erlund, H. Kolb, S. Martin, M. Carstensen, W. Koenig, J. Sundvall, S. Bidel, S. Kuha, and J. Tuomilehto. 2010. Effects of coffee consumption on subclinical inflammation and other risk factors for type 2 diabetes: A clinical trial. American Journal of Clinical Nutrition 91:950– 957. Meltzer, H. M., T. O. Fotland, J. Alexander, E. Elind, H. Hallstrom, H. R. Lam, K.-H. Liukkonen, M. A. Petersen, and E. J. Solbergsdottir. 2008. Risk as- sessment of caffeine among children and adolescents in the Nordic coun- tries. Copenhagen, Denmark: Nordic Council of Ministers, TemaNord. http://www.nrden.org/en/publications/publikationer/2008-551/at_download/ publicationfile (accessed November 26, 2013). Menzes, A. R., C. L. Lavie, J. J. DiNicolantonio, J. O’Keefe, D. P. Morin, S. Khatib, and R. V. Milani. 2013. Atrial fibrillation in the 21st century: A current understanding of risk factors and primary prevention strategies. Mayo Clinic Proceedings 88(4):394–409. Mitchell, D. C., C. A. Knight, J. Hockenberry, R. Teplansky, and T. J. Hartman. 2013. Caffeine intakes from beverages in the U.S. Abstract. http://www. fasebj.org/cgi/content/meeting_abstract/27/1_MeetingAbstracts/848.18?sid +eb47194e-54d1-40cb-b977-a6dfbd25280a (accessed November 26, 2013). Mostofsky, E., M. S. Rice, E. B. Levitan, and M. A. Mittleman. 2012. Habitual coffee consumption and risk of heart failure: A dose-response meta- analysis. Circulatory Heart Failure 5(4):401–405. Myers, M., and R. Reeves. 1991. The effect of caffeine on daytime ambulatory blood pressure. American Journal of Hypertension 4:427–431. Nawrot, P., S. Jordan, J. Eastwood, J. Rotstein, A. Hugenholtz, and M. Feeley. 2003. Effects of caffeine on human health. Food Additive and Contami- nants 20:1–30. Nkondjock, A. 2009. Coffee consumption and the risk of cancer: An overview. Cancer Letters 277:121–125. Pelchovitz, D. J., and J. J. Goldberger. 2011. Caffeine and cardiac arrhythmias: A review of the evidence. American Journal of Medicine 124(4):284–289. Roberts, A., and R. Rogerson. 2008. Chinese approach on regulating food addi- tives, novel foods, functional foods and dietary supplements. In Nutraceuti- cal and functional food regulations in the United States and around the world, edited by D. Bagchi. New York: Elsevier. Pp. 291–303. Rotstein, J., J. Barber, C. Stowbridge, H. Hayward, R. Huang, and S. Benrejeb Godefroy. 2013. Energy drinks: An assessment of the potential health risks in the Canadian context. International Food Risk Analysis Journal 3(4):1–29. http://cdn.intechopen.com/pdfs/45380/InTech-Energy_drinks_as_assessment_ of_the_potential_health_risks_in_the_canadian_context.pdf (accessed No- vember 26, 2013).

OCR for page 31
50 CAFFEINE IN FOOD AND DIETARY SUPPLEMENTS RSEQ (Reseau du sport etudiant du Quebec). 2011. Enquete Quebecoise sur le marketing de la malbouffe: 10,000 jenunes se prononcent. http://ll.rseq.ca/ download/attachments/15958040/Rapport+d'enquete-FRA-1-page.pdf?version =1&modificationDate=1328122709903 (accessed November 26, 2013). SAMHSA (Substance Abuse and Mental Health Services Administration). 2013. The DAWN report. http://www.samhsa.gov/data/Short_Reports.aspx (ac- cessed November 26, 2013). SCF (Scientific Committee on Food). 2003. Opinion of the Scientific Committee on Food (SCF) on additional information on “energy” drinks. ACF/CS/PLEN?ENDRINKS/16 Final. Brussels, Belgium: European Com- mission, Health & Consumer Protection Directorate-General, Scientific Committee on Food. http://ec.europa.eu/food/fs/sc/scf/out169_en.pdf (ac- cessed November 26, 2013). Somogyi, L. P. 2010. Caffeine intake in the U.S. population. http://www. fda.gov/downloads/AboutFDA/CentersOffices/OfficeofFoods/CFSAN/ CFSANFOIAElectronicReadingRoom/UCM333191.pdf (accessed January 8, 2014). Trabulo, D., S. Marques, and E. Pedroso. 2011. Caffeinated energy drink intoxi- cation. BMJ Case Reports 2011 (February 2). doi:10.1136/bcr.09.2010.3322. van Dam, R. M., W. J. Pasman, and P. Verheof. 2004. Effects of coffee con- sumption on fasting blood glucose and insuling concentrations: Random- ized controlled trials in healthy volunteers. Diabetes Care 27(12):2990– 2992. Wedick, N. M., A. M. Brennan, Q. Sun, F. B. Hu, C. S. Mantzoros, and R. M. van Dam. 2011. Effects of caffeinated and decaffeinated coffee on biologi- cal risk factors for type 2 diabetes: A randomized controlled trial. Nutrition Journal 10:93.