Is It Worth the Worry? 
Determining Risk
by Marian Segal

Of the many health concerns challenging the public trust in recent years, 
food safety looms large. Pesticide use is the issue most familiar to many of
us, but recent news stories reflect a multitude of other worries: dioxin
leaching into milk from paperboard milk containers; lead in ceramic products
leaching into food; Salmonella bacteria in eggs causing outbreaks of
illness; color additives banned for possible carcinogenicity (cancer-causing
ability); and possible risks from hormones given to cows to increase milk 
production. These are but a few issues raising the level of public concern
and confusion.

Consumers are not alone in their frustration. "Scientists, managers and 
regulators who study risks for a living are constantly dismayed because the 
public seems to worry about the 'wrong' risks," says Robert Scheuplein, 
Ph.D., of FDA's Center for Food Safety and Applied Nutrition. He notes, for 
example, that "for several decades, food and color additives have topped the
list of perceived consumer risks among the substances FDA regulates, despite
the view of FDA and other professionally qualified groups that objectively
they belong at the bottom." 

What Is "Safe?" 

Just what substances in foods do or do not pose a safety risk, how are the
risks assessed, and how should they be managed? These are the questions with
which government agencies, special interest groups, and consumers continue
to grapple. 

Remarkable advances in scientific knowledge and technology over the past
half century have heightened public expectations for a risk-free environment
and quick solutions to public health problems. But these very same
scientific advances have raised new questions that confound how we are to 
define what is "safe." (See "Weighing Food Safety Risks" in the September 
1989 FDA Consumer.) 

A 1949 FDA monograph stated: "A substance proposed for use in foods should
show no chronic toxicity in animals in an amount equivalent to 100 times the
proposed human use level, i.e., a safety factor of 100 should be present."
Thus, the safety standard for a substance was set to be at least 100 times
lower than the highest dose at which the chemical causes no ill effects in
animals. Later, the Delaney Clause of the 1960 Color Additive Amendments to 
the Food, Drug, and Cosmetic Act prohibited approval of any product shown to
have a cancer-causing effect, no matter how small.

New, highly sensitive chemical methods for detecting minuscule quantities of
cancer-causing agents--in parts per billion or parts per trillion--have 
complicated decision-making about safety and placed new pressures on
regulatory agencies. Also, highly sensitive methods of measuring toxic
changes in animals have further complicated the interpretation of long-term 
study results. Scientists can now detect multiple subtle biochemical and
physiological changes that had previously gone unobserved. However, the 
health significance to humans of these changes is often unclear, creating a 
regulatory dilemma. 

How accurately can results from animal studies--the current standard for
evaluating toxicity--be applied to humans? This question has perplexed
scientists for a long time. 

At the turn of the century, Harvey Wiley, father of the Food and Drugs Act
of 1906, conducted food additive experiments with human volunteers. Then, as
now, scientists were not certain how conclusively results of animal studies 
could be extrapolated to humans. Wiley's volunteers, dubbed the "poison 
squad," consumed graduated doses of suspect chemicals, such as borax, sodium
benzoate, formaldehyde, and salicylic acid. Wiley's methods and conclusions 
were controversial, however--disputed both by the affected industries 
because of possible economic repercussions and among scientists who 
disagreed with his studies. 

Testifying before the House Subcommittee on Interstate and Foreign Commerce 
in 1906, Wiley had the following exchange with Congressman James D. Mann
concerning the use of borax as a preservative:

Mr. Mann: Does your report show that in your opinion the use of borax has a 
deleterious effect upon the organs of the body? 

Dr. Wiley: Of course, you understand, Mr. Mann, the tests that we have made 
are not the same as those made upon animals fed for pharmacological 
experiments because after a given time the animals are killed and their 
organs are examined, and the changes in the cells are studied by the
microscope. We were precluded from doing that.

Mr. Mann: Is that your conclusion?

Dr. Wiley: My conclusion is that the cells must have been injured, but I had
no demonstration of it, because I would not kill the young men and examine
the kidneys.

Today, nearly a century later, scientists are still trying to refine methods
of assessing risk to humans without endangering human life. 

Adding to the confusion is frequent disagreement among experts. In his book 
News and Numbers, medical reporter Victor Cohn quotes Tim Hammonds of the 
Food Marketing Institute: "The public has become used to conflicting opinion

opposite Ph.D." 

On the Other Hand 

In EPA's publication "Explaining Environmental Risk," Peter Sandman, Ph.D., 
director of Rutgers University's Environmental Communications Research
Program, points to another source of frustration--the qualifications, 
conditions and limitations that seem always to accompany experts' 
statements. He explains that everyone outside his or her own field prefers
simplicity, precision and certainty to complexity, approximation and
tentativeness. Sandman tells about Senator Edmund Muskie's complaint about
experts who kept qualifying their testimony with the phrase, "on the other
hand . . .". "Find me an expert with one hand," Muskie said.

"The dichotomization of risk," says Sandman, "distorts the reality that 
nothing is absolutely safe or absolutely dangerous, and polarizes 
more-or-less' disagreements into 'yes-or-no' conflicts."

The complexity of assessing and managing risk can be illustrated by last
year's Alar scare. Alar is a growth-regulating chemical that was used on
apples. The environmental group Natural Resources Defense Council charged 
that children exposed to Alar were at increased risk for cancer. After much 
media attention, the chemical was pulled from the market. Its manufacturer
stopped selling it, and the Environmental Protection Agency proposed to 
phase out all allowable residue levels. 

Yet, according to Bruce Ames, chairman of biochemistry and director of the
Department of Environmental Health Sciences at the University of California 
at Berkeley, the human cancer risk from Alar is about the same as that from 
tap water (which contains the carcinogen chloroform) and about 30 times 
lower than from peanut butter (which can contain aflatoxin, a natural 
carcinogen).

A bill sponsored by Representative Henry Waxman (D-Calif.) and Senator
Edward Kennedy (D-Mass.) would allow the use of additives and permit
pesticide residues that present no more than a risk of one cancer case in a 
million over a lifetime of exposure. This "negligible risk" standard would
replace the much debated Delaney Clause, whose "zero-risk" standard is no 
longer practical. 

When FDA assesses a cancer risk at less than one in a million, it means that
at most, there will be one cancer in a million. The risk, in fact, may be 
closer to zero, causing no additional cancers. "The upper bound nature of 
these risk estimates are generally misperceived by the public, however,"
says Scheuplein. "For example, if I say I have less than $1 million in my 
pocket, people will generally assume that I have closer to about $5 than to 
$999,000. But with risk, it's different--people generally assume the worst."

Also, risk estimates are averages for a population; they are not tailored to
individuals. Rather, an individual's entire exposure history and genetic
background probably affect susceptibility to any single carcinogenic
chemical. 

Natural Versus Man-Made 

Scheuplein stresses that the primary threat of cancer from food is the food 
itself, not pesticides and other contaminants: "The nation's food supply
contains a lot of natural carcinogens that dwarf all the synthetic sources."

Carcinogens are found naturally in many spices, in smoked or salted fish, 
pickled vegetables, corn, peanuts, and broiled or fried protein-rich foods
such as beef, pork, eggs and chicken, for example. On the other hand, many
foods contain substances such as vitamins A, C and E that seem to have a
protective effect against cancer. 

In 1981, British investigators Richard Doll and Richard Peto published a
117-page report in the Journal of the National Cancer Institute on "The 
Causes of Cancer: Quantitative Estimates of Avoidable Risks of Cancer in the
United States Today." In their report, commissioned by the U.S. Congress, 
the researchers attributed 35 percent of cancer deaths at least in part to
diet. 

Dietary guidelines issued by the National Cancer Institute, the American
Cancer Society, the federal government (U.S. Departments of Agriculture and 
Health and Human Services), and the American Heart Association are all
similar in their conclusions about the roles of various foods in promoting
or helping to prevent cancer. Emphasis is on reducing fat intake, increasing
fiber intake, avoiding obesity, and limiting consumption of alcoholic 
beverages and of salt-cured, salt-pickled, and smoked fish. 

Based on the Doll and Peto numbers and an analysis of the quantity of 
cancer-causing agents in the diet, Scheuplein concluded that the risk of
dying of cancer from dietary exposure to both natural and man-made
carcinogens was approximately 7.7 percent. Even this figure reflects a
combined effect of the carcinogenic substances in the food and the diet 
itself--that is, the cancer-causing effect of fats, smoked foods, alcohol,
and other substances. The risk from naturally occurring carcinogens alone 
was at least 7.6 percent, and probably much closer to 7.7 percent.

"Most of the dietary risks are people's personal choices," Scheuplein noted.
"They are not imposed on people by corporations. Apparently that's a hard 
lesson. People want to blame somebody." He added, "I think we should be 
doing more about diet and less about specific chemical residues. The notion 
that you can ban one or two of the carcinogens in ordinary food and improve 
your health doesn't make sense."

This is not to dismiss all concerns about food contaminants, but rather to
put these risks in perspective and to examine how public perception of risk 
differs from scientific assessment of risk. For example, as mentioned 
earlier, the lifetime risk of death from natural carcinogens in the diet is 
at least 7.6 percent. Yet people are overwhelmingly more willing to accept
these dietary risks than the much smaller risk of cancer from Alar. 

Hazard + Outrage = Risk 

Why is that? Rutgers' Sandman attributes this seeming paradox to a disparity
in what scientists define as "hazardous" and what the public perceives as 
risk. The environmental risks that will kill us often don't match up with 
those that most anger and frighten us, he writes in a November 1987 article 
in the EPA Journal. 

"To the experts," Sandman explains, "risk means expected annual mortality.
But to the public (and even the experts when they go home at night), risk 
means much more than that." The "much more" is what Sandman calls "outrage."
He maintains that the public pays too little attention to "hazard" or death 
rate, and the experts pay absolutely no attention to outrage. (See
accompanying article.) Not surprisingly, he says, they rank risks 
differently.

People's concerns are often more a function of outrage than hazard. The 
risks associated with a high-fat, low-fiber diet are more acceptable in the 
public's mind than the risk posed by Alar. For one thing, there is no 
"villain"--no one to blame. Second, food choices are voluntary, not forced, 
so that people can, quite literally, "pick their own poison." 

Third, the issue of chemicals in food has been transformed into a "moral" 
issue. It is no longer a simple matter of harmful versus innocuous, but of
good versus evil. Things "natural" are persistently viewed as "good," 
whereas chemicals added to foods are seen as "bad"--this despite the fact 
that, taken together, natural carcinogens overwhelm synthetic ones in their 
harmful effects.

Sandman emphasizes that outrage factors are not "distortions in the public's
perception of risk [but rather] intrinsic parts of what we mean by risk," 
and must be considered in forming policy about risk management. He contends 
that when a risk manager continues to ignore these factors, and continues to
be surprised by the public's response of outrage, "It is worth asking just
whose behavior is irrational."

Medical reporter Victor Cohn echoes this belief: "The public is not entirely
illogical," he says. "It is easier to cope with the known than the unknown
and mysteriously threatening. We decide for ourselves whether to accept the 
risks of driving, drinking, smoking, or hang gliding. We may feel very
different about a risk someone imposes on us--or a risk that could decimate 
a population if the worst happens." 

Sandman examines the role of the media in contributing to the confusion 
concerning risk. An analysis of news stories on environmental risk submitted
by newspaper editors in New Jersey showed that reporters focused on the 
politics of risk rather than the science of risk, politics being more 
newsworthy. 

"Only a handful of the articles told readers what standard [if any] existed 
for the hazard in question, much less the status of research and technical
debate surrounding the standard," Sandman points out. "Yet the public needs 
to understand abstractions like the uncertainty of risk assessments, the
impossibility of zero risk, the debatable assumptions underlying
dose-response curves and animal tests." Sandman advises journalists not to
assert that the issue is "risky or not," but "how risky" it is. 

Managing Risk 

Health protection agencies have a responsibility to carry out risk
assessments and use them judiciously to protect the public on the basis of
the data available. Therefore, agencies sometimes must act before there is
much more than even a "glimmer of certainty," as Scheuplein puts it, about
the scientific accuracy of their conclusions. It is important to remember 
that the published risk estimates are inherently protective and not 
predictive. 

Former FDA toxicologist W. Gary Flamm, Ph.D., in a chapter in Risk and
Reason: Risk Assessment in Relation to Environmental Mutagens and 
Carcinogens, notes that there are few places outside of the United States 
where quantitative risk assessment is used for determining safe levels of 
carcinogenic substances. Flamm, who is now with Science Regulatory Services,
International, in Washington, D.C., writes, "We could ask ourselves, is that
because the United States is ahead of everyone else . . . or is it that the 
rest of the world knows something that we do not know and that ultimately we
will come to realize that there are better ways of controlling risks to 
carcinogenic substances than the methods we are currently developing and
using?" 

The answer to this question may be a long time coming. In the meantime, 
Sandman proposes that "First, we need to teach people about hazard, to help 
them understand what the serious risks are. That's the long-term solution.
Second, we have to do everything we can to make serious hazards outrageous; 
the furor over second-hand smoke, for example, has probably saved thousands 
of smokers' lives. And third, we have to stop goosing the outrage of
insignificant hazards--environmentalists have to stop doing it on purpose,
and government has to stop doing it by mistake."

Marian Segal is a member of FDA's public affairs staff. 

For More Information

For more information about risk assessment in foods, see these FDA Consumer 
articles: 
* An Unwanted Souvenir: Lead in Ceramic Ware, December 1989-January 1990. 

* Deciding About Dioxins, February 1990 

* Perspectives on Food Biotechnology, March 1990

* EBDCs: Becoming a Household Word, March 1990

* Keeping Up with the Microwave Revolution, March 1990

* Bovine Growth Hormone: Harmless for Humans, April 1990

* Salmonella Enteritidis: From the Chicken to the Egg, April 1990 

* Red No. 3 and Other Colorful Controversies, June 1990 

Outrage Factors 

Risk perception scholars have identified more than 20 "outrage factors" 
that, according to Rutgers University's Peter Sandman, risk managers cannot 
ignore in making policy decisions about managing environmental risks. 
Sandman defined the following nine in the November 1987 EPA Journal.

"* Voluntariness: A voluntary risk is much more acceptable to people than a 
coerced risk, because it generates no outrage. Consider the difference
between getting pushed down a mountain on slippery sticks and deciding to go
skiing. 

* Control: Almost everybody feels safer driving than riding shotgun. When 
prevention and mitigation are in the individual's hands, the risk (though 
not the hazard) is much lower than when they are in the hands of a
government agency.
* Fairness: People who must endure greater risks than their neighbors,
without access to greater benefits, are naturally outraged--especially if 
the rationale for so burdening them looks more like politics than science.
Greater outrage, of course, means greater risk. 
* Process: Does the agency come across as trustworthy or dishonest, 
concerned or arrogant? Does it tell the community what's going on before the
real decisions are made? Does it listen and respond to community concerns?
* Morality: American society has decided over the last two decades that 
pollution isn't just harmful--it's evil. But talking about cost-risk
tradeoffs sounds very callous when the risk is morally relevant. Imagine a
police chief insisting that an occasional child molester is an ?acceptable
risk.'
* Familiarity: Exotic, high-tech facilities provoke more outrage than 
familiar risks (your home, your car, your jar of peanut butter).
* Memorability: A memorable accident--Love Canal, Bhopal, Times Beach--makes
the risk easier to imagine, and thus (as we have defined the term) more 
risky. A potent symbol--the 55-gallon drum--can do the same thing.
* Dread: Some illnesses are more dreaded than others; compare AIDS and
cancer with, say, emphysema. The long latency of most cancers and the 
undetectability of most carcinogens add to the dread. 
* Diffusion in time and space: Hazard A kills 50 anonymous people a year
across the country. Hazard B has one chance in 10 of wiping out its 
neighborhood of 5,000 people sometime in the next decade. Risk assessment 
tells us the two have the same expected annual mortality: 50.  Outrage
assessment' tells us A is probably acceptable and B is certainly not."

Peter M. Sandman
EPA Journal 
November 1987