Tobacco Documents Online

RISK PERSPECTIVES One of the biggest problems in communicating information about risk is the simple fact that the public and members of the media often have mistaken impressions about what risk is and what it means. Far too often, the words "risk, ""risk factors" and "cause" are used interchangeably, as if they were the same. They are not -- and it is important to understand how and why they differ. Defining Risk Understanding science-based risk communications can be extremely difficult, especially for people without specialised training. The most common • misunderstandings relate to terms such as "risk", "risk factor", "statistical association" and "cause". If a valid statistically significant association is observed between a factor and a disease, the factor is said to be a "risk factor" for that disease. To say that something is a risk factor does not necessarily mean that it is a cause of the disease. To take a simple example, having a drivers license is a risk factor for car accidents because possessing a driving license is statistically associated with having an accident while driving a car. But drivers licenses do not cause accidents. Bad driving does. Epidemiologic studies are designed to assist in the identification of risk factors. Many scientists have cautioned against jumping to causal conclusions based on epidemiological studies. • Confusing Risk with Cause The tendency to confuse risk with cause sometimes leads to the incorrect belief that a reduction in exposure to "risk factors" will reduce death and disease. In fact, many of the risk factors that are identified in epidemiologic studies turn out, on further investigation, not to be related.causally to the disease or other health problem being investigated -- so eliminating these factors has no effect. Oversimplification of Risk Information Because of the huge number of risk factors that have been identified for potentially life-threatening diseases, the process of avoiding these "risks" can be challenging indeed. A 1981 article by Hopkins and Williams, for example, listed 246 different risk factors for heart disease alone. Since that time, the number of risk factors that has been identified for heart disease has continued to grow. Risk Perspectives March. 96 Page 1

The heart disease risk factors that have been identified by Skrabanek (1991), include age, male gender, cold weather, speaking English as a mother tongue, short stature, illegitimate birth, slow beard growth, widowhood, snoring, baldness and having an intelligent wife. It seems unlikely, however, that one could avoid heart disease by switching from English to French -- or, indeed, by divorcing an intelligent wife. With maddening frequency, no sooner are people confronted with one type of advice based on an epidemiologic study than they are given contradictory advice based on an even more recent study: • They substitute margarine for butter, only to be told that margarine may be worse for their arteries (Willett & Ascherin, 1994). • They are told to eat oat bran to lower their cholesterol (Kirby et al, 1981; Anderson et al, 1984) but later are told it may be useless (Swain et al, • 1990). • They substitute low-calorie saccharin for high-calorie sugar, only to hear that some researchers have found an association between saccharin and bladder cancer (Howe et al, 1977), while others have not (Hoover & Strasser, 1980). • They exercise because they are told it will protect them from heart disease, only to hear that exercise may increase the immediate risk of sudden death (Willich et al, 1993; Curfman, 1993). • To cite a final example, vitamin E and beta-carotene, long said to be beneficial in preventing cancer, were reported recently to be no better and possibly worse than a placebo (The Alpha-tocopherol, Beta-carotene Cancer Prevention Study Group, 1994). It has been suggested that the problem is not in the research but in the way • epidemiologic findings sometimes are reported by scientists and interpreted by the media (Angell & Kassirer, 1994): "[The media] are serving a public that believes passionately that the more we can learn about what to eat or how to live, the longer we will live. And neither the public [n]or the media are inclined to wait for conf a"rmatory studies. Often, the media reports are exaggerated or oversimplif ied. Even when a report itself is circumspect, headline writers may sensationalize the story. " Risk Perspectives March. 96 Page 2

0 Risk Perceptions and Politically Correct Science In October 1994, the U.S. Journal of the National Cancer published an epidemiological study suggesting that women who have abortions are 50% more likely to develop breast cancer than those who do not. That corresponds to a relative risk figure of 1.5. Dr. Eugenia Calle, director of analytical epidemiology for the American Cancer Society, emphasised at the time that a relative risk ratio of 1.5 "is not strong enough to call induced abortion a risk factor for breast cancer". It is, however, noteworthy that when the United States Environmental Protection Agency (EPA) calculated a relative risk ratio of 1.19 for statistical studies of environmental tobacco smoke and lung cancer, the result was sufficient to convince the Agency to conclude that ETS is a "Group A" carcinogen -- a substance known to cause cancer in humans. This perception seems to have held even though a more recent review that combined the results of the studies considered by EPA plus two new studies reported a reduced -- statistically non-significant -- risk ratio of 1.09 (Layard, 1995). This would suggest that scientific information is not always interpreted in an objective manner. Too often, the analysis and reporting of scientific data reflects a.bias in favour of what is deemed to be "politically correct" -- regardless of the quality or state of the underlying science. Consider the following two points, as they compare to the 1.19 or 1.09 risk ratios reported for ETS: • The U.S. EPA has declined to regulate electromagnetic fields emanating from power sources because, according to the Agency, "the relative risks in the published reports have seldom exceeded 3.0 ". • When the EPA investigated diesel emissions in 1989, it estimated a relative risk of 2.6 for the population exposed to diesel fumes but rated such emissions only a "probable human carcinogen". Problems With Risk Assessment Few governments have yet developed risk assessment guidelines, although some are seriously considering doing so. An analysis prepared recently for the U.S. Department of Energy has noted that, without such guidelines, conservative assumptions having little or no scientific basis tend to be used in performing risk assessments. That almost always results in over-regulation. Done properly, risk assessments should follow a set of rules that allow pertinent scientific information to be judged in a consistent and objective manner. Without objective guidelines, risk assessment is vulnerable to inconsistent and misleading interpretation. Risk Perspectives Ma.rch. 96 Page 3

Conclusions According to one eminent author (Wells, 1994): "Plagued by an obsession with danger and risk, people are now not only afraid of dying, they are also afraid of living!... Fear of Living is the idea that every person's life should be completely risk and discomfort free and that no-one bears responsibility for his own actions. " The public has a right to information concerning issues that can affect health. But they need to receive information that is both accurate and consistent -- free from the biases that result from political correctness. The public clearly is not served by the "scare of the week" tactics. Neither is the public served when risk factors, statistical associations and demonstrated causes of ill health are discussed as if they were the same. 0 ### 0 Risk Perspectives March. 96 Page 4