Uncertain Risks and the Risks of Certainty

Environmental Health Perspectives Volume 103, Supplement 6, September 1995

Uncertain Risks and the Risks of Certainty

Janet Raloff

Abstract

There are few things as important to parents as their children, and in the search for tips on how to protect the health of their youngsters, parents often turn to the most user-friendly form of continuing education available--the news media. It can be a smart move, but there are also risks attached. This overview points to the media's strengths and weaknesses, with the goal of suggesting how educators might help to improve the natural symbiosis between science and journalism.

While offering a window into how print journalism works, my focus will be on a topical example with special relevance to reproduction and child development-- hormone-mimicking pollutants. -- Environ Health Perspect 103(Suppl 6):00-00 (1995)

Key words: certainty, environmental hormones, hormone mimics, journalism.

This commentary was presented at the Symposium on Preventing Child Exposures to Environmental Hazards: Research and Policy Issues held 18-19 March 1994 in Washington, DC.

Address correspondence to Janet Raloff, Science News, 1719 N St., N.W., Washington, DC 20036. Telephone (202)785-2255. Fax (202)659-0365.

Thirty years ago farmers liberally sprayed DDT on fields throughout America. At that time scientists had not yet recognized the devastating effect this toxic chemical could exert on animals, much less people. Then came troubling reports of the pesticide's feminizing effect on wildlife. Some heavily exposed male birds exhibited at least partially developed female sex organs (1). The most visible symptom of the pesticide's reproductive havoc was a lethal thinning of eggshells in bald eagles and other birds (2). In gulls, which are less susceptible to shell thinning, DDT-exposed females began cohabiting with other females-the so-called "lesbian gulls" (3).

DDT appears to exert these changes by mimicking the female hormone estrogen (1). Although the United States banned the pesticide in 1972, DDT remains a very contemporary threat. Trace levels of the compound continue to contaminate produce in U.S. groceries (4) and water in the Great Lakes. Women throughout the industrial world pass on low levels of this compound--or its even more toxic metabolite DDE--in the breastmilk they are feeding the next generation (4).

The pesticide also continues to rain down throughout North America--even as far north as the Arctic (5)--as a result of DDT's use throughout many developing nations (6).

Twenty years ago when U.S. farmers were being asked to phase out their use of this chemical, most scientists suspected DDT was all but unique in its ability to exert a feminizing influence by mimicking sex hormones. But now, a wealth of new studies indicate that DDT may be only one of hundreds of environmental contaminants, some still widely used in the United States, that possess a hormonal alter ego. These include dozens of pesticides, dozens of PCBs, dioxin, and certain combustion by-products (AM Soto, personal communication; 7-9).

Data collected in 1992 showed that one such pollutant is effectively "burning out" the reproductive tracts of male and female alligators living along Florida's fourth largest body of freshwater (10). DDT and PCBs are largely suspected of fostering the life-threatening congenital bill deformities identified last year in bald eagles around the Great Lakes (10).

One of the most recently identified environmental estrogens is an ingredient in many plastics, dishwashing soaps, condoms and contraceptive jellies [AM Soto, personal communication; (11)]. This pollutant, a nonionic surfactant, has found its way into waterways throughout the world (12). Moreover, new work indicates that it can stunt the growth of a trout's phallus (13). What it can do to the developing males of other species, including our own, remains unknown.

While exposures to most environmental hormones today are small, perhaps at just a fraction of the potential active dose in humans, such contaminants are ubiquitous. We breathe them in urban air, drink them in tap water, consume them as trace contaminants in our food, and even pass them on to newborns via breastmilk. Though skeptics may argue that human exposures to any one of these hormone-mimicking compounds is likely to be insignificant, new research indicates their effects are additive (12).

Many of these pollutants also persist in the environment for decades or more. Some scientists now suspect exposure to such agents might explain several recently observed and very troubling trends (12,14): a) an unexplained rise in testicular cancer (15); b) falling sperm counts in men throughout the industrial world (12); c) increasing rates of undescended testicles and congenital penile abnormalities in boys (16); and d) increasing rates of nonsmoking-related cancers since World War II (17), including an apparent epidemic in breast cancers (18). New studies indicate that in humans, especially in males, the period of greatest sensitivity to these agents may be in utero (12).

Although some of these findings have been published in peer-reviewed journals, others have emerged only at scientific meetings (where attendance has sometimes been restricted to "by invitation only"). As a result, many of these findings, and certainly a synthesis of where the trends appear to point, have been available to the public--and policymakers--only through the media.

How reliable are the media in reporting such stories and in getting the message across? Most would assume that the answer depends on the journalist, but in fact even the audience plays a role in determining how reliable--and truthful--a journalist's message becomes.

It is a fact of human nature that people crave certainty: it drives our curiosity. Indeed, the certainty that comes from understanding our world and how it works provides comfort. By offering the promise of laying some issue to rest, that certainty seductively holds out the prospect that we can safely turn our attention to resolving other unknowns.

Most of us who know nothing about an issue will sensibly express no certainty; once we are experts, we usually become very certain. However, on most topics we will all fall somewhere between these polar extremes. And this presents a potential danger because a little knowledge can be dangerous--if we allow it to lull us into thinking we know enough about a topic to feel certainty about it.

Ironically, most of us not only allow certainty to precede true understanding, but also are willing to act on that certainty.

Consider someone who reads a single magazine article about the types of agents responsible for emasculating wildlife. He or she then tracks down the list of known agents responsible. Upon realizing that none of these agents are in the house or on the shopping list, the reader breathes a sigh of relief. But is this person really safe? Can this person even understand the range of issues and caveats that define this field--and why its practitioners disagree with each other--after reading a single journalistic summary of research spanning a decade or more?

Or consider a woman who recognizes that the spermicide with estrogenic attributes that she read about is the same one in the contraceptive foam she uses, and in the condom lubricant her partner uses. She and her partner promptly switch to alternative products. But have they reduced their exposure to the active ingredient measurably--especially if they don't find out how much of the same agent they ingest monthly from drinking water, or determine how its potency compares to the estrogenicity of pesticides whose residues taint the fruits and vegetables that they buy at the grocery?

A woman might decide not to breastfeed her newborn after learning that measurable levels of DDT lace the breastmilk of U.S. women. Do the potential neonatal risks from these compounds outweigh the nutritional and immunity benefits her breastmilk would have offered?

No one can tell any of these individuals whether the decision they made was wise. Sufficient data simply do not yet exist to make such assessments. But many of us will try to make those assessments anyway on the basis of what limited or poorly understood information we possess. What scientists, educators, the public, and journalists must keep in mind is that, while we relentlessly hunt certainty--often characterized as truth--expecting to find it is unrealistic. There are simply too many factors that conspire to keep truth beyond our grasp.

Several of those factors are endemic to the news business. News reporters seldom get to say as much as they want due to print space. After collecting 150 facts and 20 great quotes, a reporter may have room or time to include just 20% of the information.

To do a credible job, smart reporters will seek to limit the range of what they cover. In other words, instead of reporting on the whole pie, they tackle just a slice. How thin we cut that slice is quite arbitrary; and while cutting the slice paper thin may permit us to be quite complete in our coverage of that selected facet, it also may leave our audience with the impression that we have actually described a much fatter wedge of the pie, or a much more universal picture of the truth.

And what of the stories that do not lend themselves to slicing into manageable bite-size wedges? If space constraints remain, a writer will be forced to focus on highlights.

Our audience will never know what's been left out--and therefore what caveats, exceptions or fudge factors should be applied. In other words, members of the audience cannot know how uncertain they should be about extrapolating beyond the highlights.

Then there are our sources, which vary dramatically in reliability. For instance, researchers or environmental advocates may corner a reporter during a banquet dinner and share information that they perceive is both important and largely ignored. On a 10-point scale, the reliability of this information probably should be ranked between 1 and 2. Presentations at a scientific meeting or press conference may rate a ranking of 3 to 6. And the most trustworthy is perhaps a press release together with a faxed copy of a research paper scheduled to debut in next week's issue of a peer-reviewed journal such as Nature or the New England Journal of Medicine. These may warrant a ranking of 7 or 8.

Even the best sources rate only an 8 on the 10-point reliability scale. Why? None are infallible. More important, few offer all the facts needed to illustrate the full range of what is known and, at least as important, what is not known--for example, a scientific paper and interview of its principal author will usually yield enough information for a news story, but calling a few independent experts on the topic may glean certain critical extras; such as the statistical significance of the findings are marginal and therefore not yet very reliable, or that the mechanism responsible for the observation seen in this experimental model (the rat) doesn't operate in humans, the venue that most people really care about.

Assuming that all the information a reporter picks up from each source is accurate--and that is a big and dangerous assumption--then the more sources we use, the more facts we will acquire. That means that the farther we move away from certainty, the closer we inch toward truth.

Now consider those infamous deadlines. Newsmagazine writers can prepare a feature from research to completed manuscript in 4 to 6 days. Their news stories, by contrast, may be turned around in 24 to 36 hours, and a really brief item may be fully prepared, start to finish, in 45 minutes. Newspaper reporters, by contrast, may have half a day or less to put together a routine news story and just a day or two for an in-depth feature.

The shorter the deadline, the fewer facts that can be acquired--and presented. The resulting stories, prepared under deadline constraints, may convey far more certainty than the data justify.

A reporter has far less control over the interest of the audience. Accuracy and thoroughness are moot if the subscribers of a magazine fail to read the story--or just skim it. Here, the readers' feeling of certainty factors in. If they think they know all about some subject, they may skip over yet another story about it. Whether they have convinced themselves that it is a nonproblem or one of the major banes of industrialized society, they may not risk wasting their time to reaffirm what they know. Reporters recognize this, and that is why they try so hard to make each story fresh, different, or dramatic.

Perhaps we have the least control over the carefulness with which our audience reads (or in the case of broadcast media, listens). Our audience frequently mentally edits out ideas or limitations that we have carefully inserted. We are dealing with qualifiers--phrases which might better be thought of as hedging terms or weasel words.

Editors pray for the story that can make a simple and compelling declaration, like "legal levels of estrogenic pollutants have emasculated tens of thousands of American men." Unfortunately, the best we usually can offer is something like: "legal levels of estrogenic pollutants may be capable of fostering changes that impair the fertility of men." Still, readers tend to distill such qualified statements back down to "hormone-mimicking pollutants emasculate males," and they remember the information that way.

This gets them, and us, in trouble--particularly when later research indicates that one or more of these agents do not work that way in people. Indeed, readers may feel cheated if they are later told that they have worried needlessly about a major problem--that is not really a problem.

Finally, journalists can't define the limits of certainty--and uncertainty--when the jury is still out. Research advances incrementally, often at a snail's pace, and is driven by politics, funding, or curiosity. We are forced to parcel our news in the same quanta that science delivers, and at the same, or slower pace.

How can the research community help?

* Work with reporters and encourage colleagues to do the same. A scientist can warn most journalists away from weak stories or dependence on unreliable sources. Scientists also can emphasize the need for qualifiers--those necessary weasel words--in descriptions of the significance of research findings or the limits on how far new data can be extrapolated safely.

* Point journalists to other experts who not only can round out their understanding of a field, but who are also articulate enough to communicate difficult subjects well.

* Offer to be available by fax or phone on deadline so that a reporter can check the nuances of language when paraphrasing difficult or sensitive information.

* Take the time to be both patient and courteous with reporters. A journalist may have only minutes to pick up the gist of what a specialist has gleaned over years or decades. Moreover, when the media regurgitate it back, it must be in plain English and not in jargon. Moreover, if scientists are abrupt or nasty when a journalist calls, that reporter may not be motivated to ask the followup questions that facilitate understanding--such as what the limits of this work are, how reliable it is, how certain we should feel about its implications.

* Try to resist any innate desire to keep the public in the dark about ongoing work. If a scientist's peers can hear about it, then it should be available for sharing with those taxpayers who fund most university and government research and development.

* Do not wimp out when reporters (and Congress) ask for implications of your findings or those of others. Who knows better than the experts in a given field how far to extrapolate results? Often, change can come swiftly and sensibly if policymakers and the public have been given compelling reasons to act.

* Finally, be careful about what you accept as fact--much less truth. All of us do ourselves and our families a disservice by groping for certainty.

My motto: Seductive as certainty is, learn to live without it. You won't be disappointed.

REFERENCES

1. Fry DM, Toone CK. DDT-induced feminization of gull embryos. Science 213:922-24 (1981).

2. Rattner BA, Eroschenko VP, Fox GA, Fry DM, Gorsline J. Avian endocrine responses to environmental pollutants. J Exp Zool 232:683-689 (1984).

3. Fry DM, Toone CK, Speich SM, Peard RJ. Sex ratio skew and breeding patterns of gulls: demographic and toxicological considerations. Stud Avian Biol 10:26-43 (1987).

4. National Research Council. Pesticides in the diets of infants and children. Washington: National Academy Press, 1993;240-241.

5. Wilson R, Allen-Gil S, Griffin D, Landers D. Organochlorine contaminants in fish from an Arctic lake in Alaska, USA. Sci Total Environ 160,161:511-519 (1995).

6. Iwata H, Tanabe S, Sakai N, Tatsukawa R. Distribution of persistent organochlorines in the oceanic air and surface seawater and the role of ocean on their global transport and fate. Environ Sci Technol 27:1080-1098 (1993).

7. McLachlan JA, Newbold RR, Teng CT, Korach KS. Environmental estrogens: orphan receptors and genetic imprinting. In: Chemically Induced Alterations in Sexual and Functional Development: The Wildlife/human Connection. (Colborn T, Clement C, eds). Princeton:Princeton Scientific Publishing 1992;107-112.

8. Colborn T, vom Saal FS, Soto AM. Developmental effects of endocrine-disrupting chemicals in wildlife and humans. Environ Health Perspect 101:378-384 (1993).

9. Mably TA, Moore RW, Goy RW, Bjerke DL, Gendron-Fitzpatrick A, Peterson RE. In Utero and lactational exposure of male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol 114:97-126 (1992).

10. Raloff J. The gender benders: are environmental "hormones" emasculating wildlife? Science News 145:24-27 (1994).

11. Soto AM, Justicia H, Wray JW, Sonnenschein C. p-nonyl-phenol: an estrogenic xenobiotic released from "modified" polystyrene. Environ Health Perspect 92:167-173 (1991).

12. Raloff J. That feminine touch: are men suffering from prenatal or childhood exposures to "hormonal" toxicants? Science News 145:56-58 (1994).

13. Raloff J. Sperm changes linked to drinking water. Science News. 145:142 (1994).

14. Davis DL, Bradlow HL, Wolff M, Woodruff T, Hoel DG, Anton-Culver H. Medical hypothesis: xenoestrogens as preventable causes of breast cancer. Environ Health Perspect 101:372-377 (1993).

15. Raloff J. Manhood's cancer. Science News 145:138-140 (1994).

16. Giwercman A, Carlsen E, Keiding N, Skakkebaek NE. Evidence for increasing incidence of abnormalities of the human testis: a review. Environ Health Perspect 101 (Suppl 2):65-71 (1993).

17. Raloff J. Non-smoking-related cancers on rise. Science News 145:102 (1994).

18. Raloff J. EcoCancers: do environmental factors underlie a breast cancer epidemic? 144:10-13 (1993).

[Table of Contents]

Last Update: September 15, 1998