Comments of the Center for Science in the Public Interest Regarding Health Claims on Alcoholic-Beverage Containers and in Advertising
February 18, 2000
James P. Ficaretta
Attention: Notice 884
The Center for Science in the Public Interest (CSPI)1 submits these comments in
response to the Bureau of Alcohol, Tobacco and Firearms' (ATF) proposal to amend its
regulations concerning health claims for alcoholic beverages.2 In sum, CSPI urges ATF to prohibit any and
all health claims and health-related statements in the labeling and advertising of
CSPI urges ATF to ban all health claims for alcoholic beverages for the following reasons:
CSPI also urges ATF to ban all health-related statements, including all directional "health-effects" statements, for alcoholic beverages. Such statements are essentially implied health claims that should be banned for the same reasons outlined above. Because health claims and health-related statements are inherently misleading, the First Amendment does not prevent ATF from banning such claims.
While CSPI acknowledges evidence that demonstrates a reduction in coronary heart disease risk among moderate drinkers compared with non-drinkers3, we contend that the number, consistency, or validity of those studies is not at issue in this rulemaking. Health claims for alcoholic beverages on labels or in advertising should simply not be allowed no matter how well-supported those claims may be. Even if thousands of studies unequivocally proved (which they do not) that alcohol lowered the risk of cardiovascular disease -- or any other disease, for that matter -- health claims should not be allowed.
If ATF nonetheless proceeds against our recommendation and decides to allow health claims and health-related statements for alcoholic beverages, we urge ATF to require the following:
Health claims for alcoholic beverages are inherently misleading for four main reasons: (1) there are serious health risks associated with alcohol consumption, even moderate consumption, (2) the health benefits of moderate alcohol consumption do not apply universally, but only to a discrete segment of the population, (3) there are many groups of people who should abstain from, or minimize their consumption of, alcohol, and (4) allowing health claims for alcohol would undermine the government warning label.
Even if health claims were accompanied by a detailed disclaimer that explained that only certain people may benefit from moderation alcohol consumption and listed the potential health risks, such claims would still mislead consumers. Therefore, because health claims for alcohol are inherently misleading and incurable by a disclaimer, all health claims should be prohibited pursuant to the Federal Alcohol Administration (FAA) Act, 27 U.S.C. §§ 205(e) and (f) (1999).
Health claims that promote the benefits of alcohol consumption are misleading because they fail to acknowledge the serious health risks associated with alcohol consumption. Alcohol is the third leading cause of death in the United States. Men and women who drink alcoholic beverages regularly have, in comparison with abstainers, higher death rates from cirrhosis, cancers of the mouth, larynx, pharynx, esophagus, and liver; from colorectal cancer, breast cancer, hemorrhagic stroke; and from injuries, violence, poisoning, and suicide. Alcohol causes birth defects and can cause inflammation of the pancreas and damage to the brain. In fact, the National Toxicology Program's Board of Science Counselors recently recommended adding alcohol to the list of "known human carcinogens."
Even the moderate use of alcohol has risks for some drinkers. Alcohol's anti-clotting ability, while potentially protective against heart attacks, may increase the risk of hemorrhagic stroke, or bleeding within the brain.4 Alcohol may interact harmfully with more than one thousand medications and worsen advanced heart failure. Studies show that moderate alcohol consumption can increase the risk of congenital problems, migraine headaches, poor sleep, seizures, rectal cancer5, and esophageal cancer.6
New evidence has also emerged linking moderate alcohol consumption and a woman's risk of contracting breast cancer. At least five studies published in the past three years, including one large, pooled analysis of cohort studies involving 322,647 women, identify alcohol consumption as a positive risk factor for breast cancer.7 The relationship between alcohol consumption and breast cancer incidence appears to be linear and manifests at consumption levels as low as one-half glass per day. Although the increased risk is modest (9% for each 10 g/day increase in consumption), it is stronger than the increased risk associated with several reproductive factors and a positive family history of breast cancer.8
Touting the possible health benefits of alcohol in view of those numerous problems associated with alcoholic consumption would be extremely misleading. Consumers who would be encouraged to consume alcohol to reduce their risk of heart disease could ironically increase their risk of numerous alcohol-related diseases and injuries. Therefore, in light of the negative health consequences of alcohol consumption, any health claim for alcohol would be misleading and should therefore be prohibited.
Allowing health claims for alcohol would also be misleading since the health benefits of moderate alcohol consumption apply only to a discrete segment of the population. Most studies show health benefits for men over 45 and post-menopausal women, but there are no net benefits to younger people who are not at risk for coronary heart disease.9
Younger persons apparently have little to gain from beginning to drink or increasing their consumption of alcohol. In fact, young people would likely increase their risk of a wide range of alcohol-related problems. In addition, research demonstrates that the age at which one begins to drink is an important factor in the development of problems with alcohol later in life. If one begins to drink before the age of fourteen, one has a 43% chance of becoming alcoholic later in life; if one waits until 18, the risk drops to 20%, and if one waits until 21 the risk drops to 10%.10
The possible health benefits for a discrete segment of the population fail to justify allowing health claims, which would be disseminated equally to all consumers and would deceive many who stand to gain little or nothing.
Health claims encouraging moderate alcohol consumption would be very confusing to people who should abstain from alcohol entirely: children and adolescents; people who cannot control their drinking; women trying to conceive or who are pregnant; drivers and persons who plan to perform tasks that require attention or skill; and individuals using prescription drugs or over-the-counter medication.11 Those groups of people might erroneously assume that the health benefits of alcohol consumption cited in a health claim outweigh any possible negative consequences that might arise from consuming alcohol.
Health claims promoting moderate alcohol consumption would have no ameliorative effect on heavy drinkers, but would more likely help legitimate their unhealthy behavior. In fact, consumer research on the effect of health claims on wine labels shows that heavy drinkers tend to use their beliefs about the health benefits of wine to justify heavy drinking behavior and resist arguments for reducing their intake.12
Moderate drinking can also lead to heavier drinking for some people, especially alcoholics and individuals with a history of alcoholism in the family. About 50 percent of drinkers have problems with alcohol at some time, and roughly 10 percent of drinkers are alcoholics.13
Health claims encouraging moderate alcohol consumption would also be very misleading to the elderly. Although the benefits of consuming a small amount of alcohol as a way of reducing the risk of coronary artery disease may be more important for older people who are at greatest risk for the disease, there are several risks associated with alcohol consumption that are heightened for the elderly.
With diminishing lean body mass and increasing adipose tissue that accompany aging, the volume of total body water decreases. As a result, alcohol is more quickly absorbed into the bloodstream of elderly persons and a given amount of consumption will produce a higher blood alcohol concentration than it will in younger consumers. Even relatively low blood alcohol concentration levels put older people at risk for crippling falls or car accidents.14
In addition, as the country's largest consumers of legal drugs, older persons are especially vulnerable to adverse alcohol/drug interactions.15 Long-term alcohol consumption activates enzymes that break down toxic substances, including alcohol. Upon activation, these enzymes may also break down some common prescription medications. The average person older than 65 takes two to seven prescription medications each day. Alcohol-medication interactions not only increase the risk of negative health effects, but also potentially influence the effectiveness of the medications. In one study, alcohol was involved in one-third of all drug interactions in the elderly.16
Moreover, increasing attention is being paid to the phenomenon of "late onset alcoholism." It is now recognized that a group of people exist who have not shown problem drinking behavior in earlier life, but who become alcoholics later on.17 Surveys conducted in health care settings have found increasing prevalence of alcoholism among the older population.18 Surveys indicate that 6 to 11 percent of elderly patients admitted to hospitals exhibit symptoms of alcoholism, as do 20 percent of elderly patients in psychiatric wards and 14 percent of elderly patients in emergency rooms.19 The prevalence of problem drinking in nursing homes is as high as 49 percent in some studies.20
Problems affecting the elderly will become even more significant as the population itself ages. According to projections by the U.S. Census Bureau, the elderly population will more than double between now and the year 2050, to 80 million, when as many as one in five Americans is expected to be over the age of 65.21
Allowing health claims for alcohol would seriously undermine the government warning label that informs consumers of some of the risks of consuming alcohol and would mislead consumers by providing conflicting health messages concerning alcohol consumption. When Congress passed the Alcoholic Beverage Labeling Act in 1988, it stated:
As a result of this concern, Congress required that all alcoholic beverage containers bear the following statement on the label:
Health claims on alcoholic beverages that encourage moderate consumption for health reasons contradict the government warning that alcohol consumption "may cause health problems." Health claims would also undercut important guidance about drinking during pregnancy and drinking and driving. Such conflicting health messages would undoubtedly confuse and mislead consumers and would therefore be contrary to Congress' intent to "provide a clear, nonconfusing reminder" about the health hazards that may result from alcohol consumption.
It may be argued that health claims for alcohol are not misleading if they are appropriately qualified, as set forth in ATF's current policy and proposed regulation.24 Accordingly, health claims could be required to include statements that cite the health risks of alcohol consumption, explain that the health benefits apply only to a discrete group of people, and list the groups of people who should abstain from, or minimize their consumption of, alcohol. However, even the most detailed disclaimers cannot cure the inherently misleading nature of health claims for alcoholic beverages.
First, disclaimers that warn certain groups of people not to consume alcohol are inadequate because many of those people may not even know that they are a member of such a group. For example, approximately 8% of all nondrinkers who initiate drinking will become alcoholics. But there is no way for those people to determine who they are until they actually start drinking. Similarly, women who are pregnant may not even know it until several weeks, or even months, into their pregnancy. Therefore, any warning directed to those yet-to-be identified groups of people is useless.
Second, the recommendation to drink moderately depends on a multitude of factors that vary from individual to individual. Relevant factors that must be considered include one's age, sex, family medical history, body mass, activity, and propensity to drink. It would therefore be impossible to create a health claim specific enough to be tailored to a person's unique circumstance. For example, how would a health claim be worded to enable a petite elderly woman with a family history of breast cancer and heart disease to decide whether or not drinking moderately would improve her health? What about people who are not familiar with their family medical history? Because health claims simply cannot provide individually-focused medical advice for the multitude of different scenarios that exist, any such claim will ultimately mislead consumers.
Third, many consumers would probably be confused by such disclaimers, or would simply ignore them. Even the most educated consumer would likely be bewildered by the necessarily lengthy and contradictory statements or would dismiss them altogether and focus only on the general message of the health claim -- that moderate drinking is good for them, which may or may not be the case.
A recent study by the Federal Trade Commission (FTC) supports the notion that explanatory phrases may not be sufficient to clarify a misleading claim.25 The FTC study tested consumers' ability to evaluate health claims for foods that contained high levels of a beneficial nutrient (such as fiber or calcium), but that also contained high levels of a nutrient (such as sodium or saturated fat) that in sufficient quantities can increase the risk of a diet-related disease. The test used ads featuring a health claim for a minestrone soup that was high in fiber, but also high in sodium, and a health claim for a Swiss cheese product that was high in calcium, but also high in saturated fat.
The FTC study revealed that the health claims appeared to interfere with consumers' understanding of the products' saturated fat or sodium content even when the nutrient content was clearly disclosed. Although a sizeable minority of respondents correctly interpreted the absolute disclosure to indicate that the products were high or somewhat high in the problem nutrient, almost as many respondents completely misinterpreted these disclosures to indicate that the products were low or somewhat low in sodium or saturated fat. An additional disclosure, which added an advisory warning of the health consequences of high dietary intake of sodium or saturated fat, exacerbated respondents' confusion. Almost half apparently misconstrued the dietary warning as a favorable commentary on the quantity of sodium or saturated fat in the advertised products.26
This study suggests that a health claim about a single beneficial nutrient interferes with consumers' ability to evaluate the total healthfulness of a product, even when the presence of the risk-increasing nutrient is clearly disclosed. By analogy, it is likely that a health claim for alcohol would interfere with consumers' ability to evaluate the health risks and benefits of alcohol consumption, even if those risks were clearly disclosed.
The inherently misleading nature of health claims for alcoholic beverages is confounded by the wine industry's efforts to provide the American people with "one size fits all" generic public health pronouncements disguised as information about scientific findings. The Wine Institute in particular has launched a massive publicity campaign to sell more wine by sending misleading messages to members, industry groups, policy makers, physicians, researchers, and journalists through its publications (Newsflashes and Research News Bulletins) and Internet site. Those publications have overstated the health benefits of wine, used out-of-context statements, disregarded cautions related to drinking, ignored the contradictory findings of many other studies, relied on single (even unpublished) studies to make broad pronouncements about wine and health, and all but completely failed to acknowledge the health risks of alcohol consumption.27 For example, in May 1997, the same month that an editorial in the scientific journal Epidemiology counseled women to avoid alcohol to reduce their breast cancer risk, the Wine Institute issued a Special Media Advisory suggesting that moderate wine consumption is not associated with increased risk of breast cancer. The FTC investigated those claims and found that the weight of scientific evidence was indeed contrary to the Wine Institute's claims.28
The views of the spirits industry are much the same, although liquor interests have not been nearly so aggressive in promoting the health benefits of alcohol as the wine industry. Tony Greene, chairman of Guinness PLC, answered a rhetorical question about the future of the alcohol industry by suggesting that liquor marketers "chang[e] the sentiment toward the spirits industry through educating consumers to the benefits of alcohol."29
Aided by the industry's overstated pronouncements, the media has contributed to this problem by providing overwhelmingly misleading and inaccurate coverage of the health effects of alcohol consumption. Beginning in 1991 with the widely-publicized report on CBS' 60 Minutes entitled The French Paradox, which touted the health benefits of wine consumption, particularly red wine,30 the media has hyped the results of epidemiological studies, frequently promoting the general notion that all but a few consumers should drink for their health. Information about the downsides of alcohol, about addiction and about the numerous individual factors that should inform a decision to drink have been downplayed or ignored.31 The fact that health researchers almost universally hesitate to provide generic recommendations that consumers begin drinking or increase their consumption of alcohol is most often disregarded or submerged in news coverage of health benefits stories. In the 60 Minutes follow-up of its original 1991 report, promoters of health benefits even resorted to scare tactics to get people to drink when Curtis Ellison warned viewers that "abstinence is a major risk factor for coronary heart disease."32
This "drink for your health" strategy appears to be working, driving wine sales higher. From 1991 to 1998, wine sales increased 9.3%, in contrast to beer sales which increased only 2% and spirit sales which decreased 4.4%.33 "[Health] announcements are increasing consumption more than anything else," stated Stephanie Grubbs, marketing manager for Robert Mondavi Coastal, a brand whose sales increased 86 percent from 1995-1996.34 Howard Jacobson, senior vice president of marketing for Canandaigua Wine Company said that "positive health reports are still having an influence on the market" and acknowledged that "[m]arketers are trying to figure out ways to keep that message in the forefront."35 Other industry observers have noted:
In fact, a group of researchers for the U.S. Centers
for Disease Control and Prevention found that, after decreasing during the late 1980s,
alcohol consumption among pregnant women in the United States began to increase after 1991
(following the 60 Minutes story).37 The lead author hypothesized that the increased drinking by pregnant women
might be due to the recent reports on the health benefits of moderate drinking and to the
heightened media attention given to those reports.38
B. Allowing health claims for alcohol would contradict Congressional policy concerning health claims and conflict with FDA and USDA regulatory schemes.
Allowing health claims for alcohol would directly contradict the policy on health claims for food that was instituted by Congress when it passed the Nutrition Labeling and Education Act (NLEA) of 1990.39 Congress intended that health claims would not merely provide information on particular substance-disease relationships, but would help individuals to maintain healthy dietary practices.40 Thus, section 403(r)(3)(A)(ii) of the Act states that health claims may only be made:
In accordance with this mandate, the Food and Drug Administration (FDA) identified four nutrients -- total fat, saturated fat, cholesterol, and sodium -- whose consumption has been associated with an increased risk of certain diseases or health-related conditions, particularly cancer, cardiovascular disease, and hypertension. For each of those nutrients, FDA regulations establish levels above which foods containing the nutrient are disqualified from bearing health claims.42 For example, even though whole milk is an excellent source of calcium, milk producers are barred from making a health claim concerning osteoporosis because whole milk contains too much saturated fat.
The United States Department of Agriculture (USDA) has voluntarily established identical regulations for health claims on meat and poultry. ATF should also adopt comparable regulations so that the federal government applies the same standard to all health claims. As ATF has previously stated:
If ATF chooses to comport with Congressional policy and
FDA's rationale of precluding certain foods from bearing health claims, it must prohibit
alcohol producers from making any health claim because the consumption of alcohol
increases the risk of other diseases. To allow health claims for alcohol, America's most
devastating drug, while health claims for foods such as whole milk are prohibited, would
be indefensible and would make a mockery of the federal government's health-claim regime.
C. An alcoholic beverage that makes a health claim may be classified as a drug subject to FDA regulatory control.
If an alcoholic-beverage label or advertisement claims that alcohol may reduce the risk of disease, the beverage may be regulated as a drug pursuant to 21 U.S.C. § 321(g)(1) (1999) (an article may be deemed a drug if it is "intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease"44). The actual physical effect is irrelevant -- a substance may be deemed a "drug" if its intended use brings it within the drug definition.45 The intended use may be determined from its label, promotional material, advertising, and any other relevant source.46 Even the most commonly ingested foods and liquids may be classified as drugs depending on their intended use.47 Thus, if an alcoholic beverage makes a claim on its label or in advertising that alcohol can help mitigate or prevent a disease, the FDA may assert its jurisdiction and regulate alcohol as a drug.
Indeed, aside from its regulatory classification,
alcohol is a drug. Depending on a variety of factors such as dose and schedule of use,
individual metabolism, personality factors, and situation, alcohol is variously a
stimulant and depressant, euphorigan and soporific, irritant and anxiety reducer. Alcohol,
like other intoxicants, can produce such dependency phenomena as persistent search
behavior, withdrawal, relapse, and loss of control.48
D. The negative public health consequences associated with increasing alcohol consumption outweigh any potential benefits.
Although moderate drinking might be quite acceptable and harmless for many individuals, the spread of moderate drinking -- which would mean higher levels of alcohol consumption throughout society -- would have significant negative public health consequences. A substantial body of evidence has shown a positive relationship between the aggregate consumption of alcohol in society and population rates of alcohol-related diseases, accidents, criminal violence, and suicide.49 According to the National Institute on Alcohol Abuse and Alcoholism (NIAAA), alcohol abuse and alcoholism cost society more than $166 billion annually and each year over 110,640 deaths have alcohol-related causes.50
A study of per capita alcohol consumption and alcohol-related mortality (liver cirrhosis, alcoholic psychosis, pancreatitis, and cancers of the upper digestive tract and of the pancreas) shows that the temporal variation in this form of mortality was significantly related to changes in per capita alcohol consumption.51 Several studies also show a significant relationship between aggregate consumption of alcohol and alcohol-related traffic crashes.52 The National Highway Traffic Safety Administration (NHTSA) revealed that traffic fatalities in alcohol-related crashes in 1998 totaled 15,935. NHTSA estimates that alcohol was involved in 39 percent of all fatal crashes and projects that about three in every ten Americans will be involved in an alcohol-related crash at some time in their lives.53
Studies have also found positive relationships between overall consumption and suicide rates.54 Scientists have postulated that heavy drinking may lead to a deterioration of social ties, that the acute state of intoxication may reduce the individual's self-control and trigger a suicidal inclination, and that heavy alcohol use is associated with depression, a primary precursor of suicide.55 A link between alcohol and violence is suggested by experimental data, as well as by the fact that a large proportion of violent offenders are reported to have been intoxicated at the time of the crime. Several studies also support the contention that changes in per capita consumption affect violent criminality.56
Thus, the enormous negative public health consequences
of increased alcohol consumption throughout society far outweigh the benefit of possibly
reducing the risk of heart disease for a limited portion of the population.
E. There are safer means of reducing one's risk of disease.
The discrete category of people who may benefit from
moderate drinking could also lower their risk of heart disease by other less risky
alternatives, such as quitting smoking, reducing fat in the diet, getting regular
exercise, taking a daily, low-dose aspirin, or reducing stress. All of those methods are
much less likely to cause accidents or other health problems than consuming alcohol, even
III. ATF SHOULD BAN ALL HEALTH-RELATED STATEMENTS FOR ALCOHOL BEVERAGES ON LABELS AND IN ADVERTISING.
In February 1999, against the advice of many public health authorities, ATF announced its approval of the following directional "health-effects" statements for wine labels:
Those statements are problematic for several reasons. First, a reference to the "health effects of wine consumption" offers no useful information, but simply reinforces existing inaccurate knowledge about the health benefits of alcohol consumption, as spread through the media and the wine industry's misleading publicity campaign, and implies that those benefits are substantial and universal. Thus, those directional "health-effects" statements are implied health claims,57 which should be prohibited for the reasons discussed in Section II of this submission.
Second, referring consumers to a government publication which offers balanced information is only credible if there is a reasonable likelihood that such referral will result. Yet there has been no research to confirm that such directional referrals result in substantial numbers of consumers requesting and reading the government publications, or consulting with a physician. In fact, according to consumer research, few people would actually look at or write for the Dietary Guidelines on the basis of the label language.58
Third, even if consumers actually consulted their family doctor or obtained and read a copy of the Dietary Guidelines, those sources are inadequate to fully inform consumers about the health effects of alcohol consumption. The Dietary Guidelines does not -- nor was it designed to -- provide a complete statement on the risks and benefits of alcohol consumption. The current Dietary Guidelines does not even mention some of the most troubling risks of moderate alcohol consumption, such as the increased risk of breast cancer. Similarly, although some physicians may be able to educate their patients about the risks and benefits of alcohol consumption, many are unlikely to do so. Many physicians have not been trained in this area and do not keep up with the latest research. Physicians are also unlikely to have the time necessary to adequately explain whether moderate alcohol consumption is appropriate for their patients.
Thus, we urge ATF to rescind its approval of the
above-mentioned directional "health- effects" statements and prohibit all future
IV. THE FIRST AMENDMENT DOES NOT PREVENT ATF FROM PROHIBITING THE USE OF INHERENTLY MISLEADING HEALTH CLAIMS.
Even if a claim is entirely truthful, the First Amendment does not prevent the government from prohibiting the claim if it is inherently misleading. Central Hudson Gas & Elec. Corp. v. Public Serv. Comm'n of New York, 447 U.S. 557, 563-564 (1980), In re R.M.J. 455 U.S. 191, 203 (1982), Metromedia, Inc. v. San Diego, 453 U.S. 490, 507 (1981) (plurality opinion). As the Supreme Court has long recognized, "[t]he First Amendment...does not prohibit the State from insuring that the stream of commercial information flows cleanly as well as freely." Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, 425 U.S. 748, 771-72 (1976). The courts have repeatedly upheld government prohibitions on deceptive advertising and labeling of foods and drugs against First Amendment challenges.59 Thus, because health claims for alcohol are inherently misleading, they are not protected by the First Amendment and may be banned entirely.
The issue of prohibiting inherently misleading health claims is easily distinguished from two Supreme Court cases in which the Court struck down prohibitions on commercial speech related to alcoholic beverages. In 1995, the Supreme Court held that a statute prohibiting the display of alcohol content on beer labels was unconstitutional. Rubin v. Coors Brewing Co., 514 U.S. 476 (1995). Because the case concerned regulation of commercial speech that was not deemed misleading, the Supreme Court analyzed the regulation under Central Hudson60 and determined that not only did the labeling ban fail to directly advance the government interest in curbing "strength wars," but that it was more extensive than necessary. Id. at 489.
The following year, the Supreme Court struck down a state restriction on the truthful advertising of alcoholic-beverage prices finding that the ban did not directly advance the state's asserted interest in the promotion of temperance and was more extensive than necessary to serve that interest. 44 Liquormart v. Rhode Island, 517 U.S. 484, 507 (1996). The significant feature of both Coors and Liquormart is that they involved truthful and nonmisleading commercial speech. In contrast, the issue of prohibiting health claims for alcoholic beverages involves inherently misleading speech which has never been held by the Supreme Court to be entitled to First Amendment protection.
This issue is also unaffected by the recent D.C. Circuit court decision concerning health claims for dietary supplements. In Pearson v. Shalala, the court held that health claims for dietary supplements based on preliminary evidence were not inherently misleading and that the FDA was required, under the commercial speech doctrine, to consider whether the inclusion of appropriate disclaimers would negate the potentially misleading nature of health claims for dietary supplements. 164 F.3d 650 (D.C. Cir. 1999).
The issue of banning health claims for alcoholic beverages can be distinguished from the Pearson case in two essential ways. First, the holding of Pearson is limited to health claims for dietary supplements,61 which concededly did not threaten consumers' health or safety. In recognition of this fact, the court noted: "[d]rugs, on the other hand, appear to be in an entirely different category -- the potential harm presumably is much greater." 164 F.3d at 656, note 6.
Second, Pearson dealt with health claims that were not deemed inherently misleading, but only potentially misleading. The Pearson court even acknowledged that disclaimers may not always be sufficient to cure potentially misleading claims: "Nor do we rule out the possibility that where evidence in support of a claim is outweighed by evidence against the claim, the FDA could deem it incurable by a disclaimer and ban it outright." Id. at 659 (citing FTC v. Brown & Williamson Tobacco Corp., 778 F.2d 35, 42-43 (D.C. Cir. 1985) (holding in a false advertising case under the Lanham Act that a proposed disclaimer would not suffice to cure the misleadingness of an advertising claim)).
Thus, because all health claims for alcoholic beverages
are inherently misleading as shown in Section II of this submission, the First Amendment
does not prevent ATF from prohibiting all health claims for alcoholic beverages.
V. IF ATF NONETHELESS DECIDES TO ALLOW HEALTH CLAIMS FOR ALCOHOLIC BEVERAGES, SUCH CLAIMS MUST BE CAREFULLY REGULATED.
CSPI strongly urges ATF to prohibit all health claims and health-related statements for alcoholic beverages as discussed in Sections II and III. If, however, ATF disagrees with our position and decides to allow such claims, we recommend that the claims be strictly qualified to minimize the degree to which consumers will ultimately be misled.
ATF has proposed that "such claim is considered misleading unless it is properly qualified, balanced, sufficiently detailed and specific, and outlines the categories of individuals for whom any positive health effects would be outweighed by numerous, negative health effects."62 However, we recommend that if health claims and/or health-related statements are allowed, ATF should develop more detailed requirements, as outlined below, to minimize the amount of consumer deception and confusion that will inevitably ensue.
A. Health claims and health-related statements should be pre-approved by the FDA.
All health claims should be pre-approved by the FDA to make it easier to prohibit poorly substantiated or improperly worded claims before they actually appear on labels. Prohibiting claims after consumers have already read them does not remedy any lingering misperceptions caused by the claim. Since FDA officials possess the necessary scientific and public health expertise, as well as relevant experience in evaluating health claims for foods, they would be better suited to evaluate health claims for alcoholic beverages than the BATF.
B. Health claims and health-related statements should be supported by "significant scientific agreement."
If ATF allows health claims, the agency should adopt the requirement that health claims be supported by "significant scientific agreement" among qualified experts and that the claim be supported by the "totality of publicly available evidence," as the FDA requires for health claims for foods.63 This evidence should include data from well-designed studies conducted in accordance with recognized scientific procedures and principles. ATF should never allow health claims based on preliminary studies, even if they are accompanied by explanatory disclaimers.64 It is especially important that if health claims for alcoholic beverages are allowed that they be supported by a very high standard of evidence because of the numerous and serious health risks that are associated with alcohol consumption.
C. Health claims and health-related statements should recommend no more than one drink per day rather than encourage people to drink "moderately."
Health claims and health-related statements should not recommend that people drink "moderately" since most people do not know what "moderate" drinking is. When the Center for Substance Abuse Prevention of the Department of Health and Human Services conducted a study of a potential wine label referring to "moderate consumption," it concluded that the word "moderate" has virtually no meaning. The study found that people's conception of moderation ranged from "1 or 2 drinks" to "a bottle of wine in an evening." The study also found that the more a person drinks, the more drinking one thinks is moderate. Thus, heavy drinkers' concept of moderate drinking was much higher than the concept held by light or medium drinkers. The average number of drinks per occasion that heavy drinkers thought was moderate was almost six, which is even higher than the generally accepted definition of binge drinking.65 As Dr. Mary C. Dufour stated in her review of the literature on the risks and benefits of alcohol use, "This definitional vagueness can lead to considerable confusion... For example, people who intend to drink for cardioprotection need to know at what level of alcohol consumption such benefits accrue (i.e., how many and what size drinks constitute 'moderate' drinking)..."66
Most of the studies that have been conducted on the effects of moderate alcohol consumption show the largest protective effect at the level of one drink per day for men, and half a drink per day for women. J. Michael Gaziano, the Director of Cardiovascular Epidemiology at Boston University Hospital, summarized the literature at the 1997 American Heart Association annual meeting: "If you drink half a drink per day, you're not likely to derive greater benefit from drinking two drinks per day."67 Likewise, the study by Camargo et al. of the relationship between moderate alcohol consumption and the risk for angina pectoris or myocardial infarction found the most marked effect at about one drink per day.68 Rimm et al.'s study on the relationship between alcohol consumption and the risk for coronary heart disease in men found the protective effect continued to consumption levels of 50g/day (about four drinks per day). However, two-thirds of that protective effect had been reached by the level of 10-16 g/day (less than one to 1 1/4 drinks per day),69 and the higher levels of alcohol consumption significantly increase risks for other health problems.
When one views the impact of alcohol consumption on total mortality and morbidity rather than narrowly views the impact of alcohol consumption on coronary heart disease, the studies more clearly show that the maximum benefits from consuming alcohol can be reached at levels of about one drink (or less) per day for men, and approximately half that for women. DuFour's review of the literature on all-cause mortality provides a balanced exposition of the variety of risk factors of alcohol consumption based on a person's age, gender, and personal history. Her review suggests that the greatest decline in all-cause mortality occurs at one drink per day, and begins to rise after that.70
Similarly, Thun et al.'s study of alcohol consumption and mortality among middle-aged and elderly U.S. adults found that the lowest level of all-cause mortality for both men and women occurred at the level of one drink per day, and rose thereafter.71 Likewise, Klatsky et al. showed all-cause mortality troughing at levels between less than one drink per month and 1 to 2 drinks per day, and increasing thereafter.72 In a comprehensive review of 29 studies, Poikolainen found that the consumption associated with minimum mortality varied from less than a drink per day to five, with one being the most frequently reported.73
Therefore, in the interest of defining a safe limit of
alcohol consumption that maximizes health benefits and minimizes risks, we recommend that
health claims and health-related statements include a quantitative drinking recommendation
of no more than one drink per day, rather than an ambiguous recommendation that one drink
"moderately." In addition, "drink" should be defined as twelve ounces
of regular beer, five ounces of wine, and 1.5 ounces of 80-proof distilled spirits.
D. Health claims and health-related statements should alert consumers to all potential health risks and accidents that may result from alcohol consumption.
In order to allow consumers to make a more educated decision about whether or not to drink for health reasons, they must be fully informed about all of the potential health consequences. Thus, any health claim promoting the benefits of alcohol consumption should also be accompanied by an equally prominent statement warning consumers of the potential negative health consequences of alcohol consumption, including, but not limited to, the following:
E. Health claims and health-related statements should be addressed only to those groups to whom the particular claims apply.
If a health claim or health-related statement does
not apply universally to all segments of the population, it must be carefully worded so
that it is directed only to those for whom the claim applies. For example, a health claim
concerning lowered coronary heart disease risk should only be directed to men over 45 and
post-menopausal women. The groups of people who should abstain from alcohol should also be
clearly identified: children and adolescents, people who cannot control their drinking,
women trying to conceive or who are pregnant, drivers and persons who perform tasks that
require attention or skill, and individuals using prescription drugs or over-the-counter
medication. We also recommend that all health claims and health-related statements contain
a special cautionary note to elderly consumers warning them of the heightened risks that
they face when consuming alcohol. However, as explained previously, disclaimers that warn
certain groups of people not to consume alcohol are inadequate because many of those
people, such as pregnant women or potential alcoholics, may not even know that they are a
member of such a group.
F. Health claims and health-related statements should not overshadow, contradict, or undermine the government warning label.
Health claims and health-related statements should
be worded and displayed in a manner that does not overshadow, contradict, or undermine the
government warning label. For example, the claim should appear in the same type size and
style as the government warning label, and should not contain any claim that contradicts
any of the statements in the warning label.
G. Health claims and health-related statements in advertisements should be held to the same high regulatory standard as they are on labels.
Health claims and health-related statements should be held to the same high regulatory standard for both advertisements and labels to present consumers with consistent health messages and minimize potential confusion.
In light of the issues discussed above, we strongly
urge ATF to prohibit any and all health claims and health-related statements in the
labeling and advertising of alcoholic beverages. Even the most detailed disclaimers cannot
cure the inherently misleading nature of health claims for alcohol. If ATF, however,
decides to allow such claims, we recommend that the claims be strictly qualified, as
proposed in Section V.
Leila Farzan Leoncavallo
For more information: