Chapter 2

Lighting Up the World

The fictional Douglas Spaulding, a mighty 12 years old, stood in early morning darkness at the window of a cupola at his grand-parents' house. He smiled, pointed a finger, and performed his magic. As he willed it, the town began to awaken. He ordered the old people to wake up, the lights to turn on, his grandma and great-grandma to fry hotcakes, birds to sing, and the sun to rise. With a final snap of his fingers, he willed the world awake, and it followed his bidding.

"Yes, sir, he thought"—as author Ray Bradbury composed in young Douglas' voice—"everyone jumps, everyone runs when I yell. It'll be a fine season."

The same adolescent narcissism with which Bradbury's character felt all-powerful and invulnerable in the world offers psychological protection to many adolescents who meander and sometimes screech toward adulthood. It is a feature of many adolescents to believe that they can run and not be weary, drive fast and not crash, stay up all night and be normal the next day, and do dangerous things without consequences. If there are consequences, adolescents typically believe that the consequences are so distant that they don't matter now. Adolescents have greater tasks than worrying about a faraway future: They are becoming someone and becoming separate, a process somehow made simpler by becoming indistinguishable from those around them. They are struggling with relationships, acceptance, hormones, identity, school, religion, death, success, and, all too often, tobacco.

It is a jarring reality that the feeling of invulnerability that can protect them against the slings and arrows of an outrageous world also makes them vulnerable to behaviors that can haunt them throughout their lives. A few minutes of unplanned and unprotected intimacy can become an infant nine months later. A few beers and a brief car ride can become an incarceration for drunk driving. A single cigarette can lead to emphysema or a heart attack in thirty or forty years. Adolescence isn't a time for existing that far ahead. It is a today time. At best a tomorrow or next week time. And because of this, adolescence has become the prime time for tobacco initiation.

Tobacco's initial toxicity seems to provide adolescents with little protection against its continued and perpetual use. No known interventions seem to work consistently either in preventing tobacco use or in stopping it among adolescents. Tobacco prevention specialists find themselves outmaneuvered by the clever wizards behind the tobacco industry's advertising and promotions. Tobacco control countermeasures, however expensive and well planned, all too often backfire.

An example of this is a recent Arizona state initiative to combat teen smoking with T-shirts and other paraphernalia emblazoned with this slogan: "Tobacco. Tumor causing, teeth staining, smelly puking habit." Predictably, many young persons wore the slogan openly as they proudly and defiantly smoked. (As lyricist Tom Jones [The Fantasticks] wrote, "Dog's got to bark, a mule's got to bray/. . . Children, I guess, must get their own way/The minute that you say 'no.'")

As the strategies falter, the figures climb. Tobacco use among young people has steadily increased throughout the decade of the '90s. Smoking has increased among teenagers and young adults ages 18-25; smokeless tobacco use, following a boom in the '80s, has not shown the sharp decline many experts predicted; cigar use has become wildly popular and also shows no sign of diminishing.

It is difficult to dismiss these youthful indulgences in view of the statistic that some 90 percent of all smokers begin the practice during adolescence. Every day, about 6,000 U.S. young people try smoking and about 3,000 become regular smokers. Of these, scientists predict that one-third to one-half eventually will die of smoking-related causes. The path from experimentation to becoming a regular smoker usually takes between two and three years. Young persons who take up smoking are likely to keep smoking for up to 20 years.

The statistics on youth use of tobacco are sobering. Even in states with relatively low overall smoking rates among adults, we commonly see groups of young people walking along, laughing, talking, and smoking. Somewhat invisible to us are the sizable numbers of young users of smokeless tobacco, primarily moist oral snuff. Some smokeless users also smoke, but many do not.

Tobacco use is growing faster among adolescents than among any other age group. Even though tobacco use levels off among other age groups, it continues to rise among the young. A 1998 report indicated that more than half of white male U.S. high school students used tobacco in the previous month. Among high school females, the rate was more than 40 percent. Cigarette smoking doubled among African-American high school males from 14 percent in 1991 to 28 percent in 1997. These findings from the U.S. Centers for Disease Control and Prevention (commonly called the CDC), reported results of the 1997 Youth Risk Behavior Survey. About 40 percent of all high school students surveyed in 1997 said they used cigarettes. About 21 percent of high school white males reported using smokeless tobacco, leading all other groups studied. More than 30 percent of all male high school students used cigars.

Recently published data show that urban-rural differences in youth smoking rates have shifted. As researchers Christine E. Cronk and Paul Sarvela of Southern Illinois University explained, "These findings contrast with the popularly held notion that rural youth are more protected against the use and abuse of drugs by their distance from the factors supporting drug use in urban environments." They noted that substance availability has changed in rural areas, that prevention is less common and less effective, and that social factors no longer protect youth from drug use. The risk factors for adolescent substance use are not always obvious or well known. For example, tobacco use is higher among adolescents with learning disabilities than among their nondisabled peers. Researcher John W. Maag at the University of Nebraska-Lincoln and his colleagues speculated that the increased risk for delinquency among those who are learning disabled could contribute to their higher rates of tobacco and marijuana use.

Among homeless and runaway youth, tobacco use is common. About 81 percent of homeless youth living on the streets used tobacco, as reported in a 1997 study from Jody Greene and colleagues of the Research Triangle Institute. Among those living in shelters, the smoking prevalence was 71 percent. Use of other substances, including alcohol and marijuana, was also high.

Nor is adolescent smoking limited to the United States. The usual pattern in a country is for men to take up smoking first, followed by boys, women, and girls, according to Ann Charlton of the International Union Against Cancer (UICC). In many developing countries, cigarette smoking is at the men/boys stage, thus creating what she calls "two different scenarios with regard to young people and smoking." Young persons in Western countries are aware of the health risks of smoking but pay little attention to them. Young persons in developing countries are less aware of potential health risks from tobacco use, and sometimes are receptive to educational programs informing them about the dangers of tobacco.

If smoking rates don't change among today's children, some 30 million Europeans and 50 million Chinese who are now children could die of tobacco-related disease later in life, according to Charlton, who cited World Health Organization estimates.

Too Young to Vote

How do young persons get tobacco, since they can't buy it legally in most places? According to a 1996 Youth Risk Behavior Surveillance survey published by the CDC, half of high school seniors under age 18 reported simply buying cigarettes at a store or gas station. Some 39 percent reported that they usually bought cigarettes in a store. One-third routinely borrowed cigarettes from someone else, and 16 percent gave someone else money to buy cigarettes for them. Across all high school years, at least three-fourths of the youthful smokers reported not being asked to show proof of their age when they bought cigarettes. Availability varied widely among states: 18 percent of Idaho high school students under age 18 bought cigarettes at a store or gas station, yet 49 percent of New Hampshire students reported such purchases. More than 90 percent of students from some cities, including Washington, D.C., and Detroit, reported not having to show proof of age to purchase cigarettes.

Camels and Caveats

Research into tobacco use, or into any other aspect of human behavior, generally is conducted in one of a several ways. The most typical approach is a cross-sectional study, in which a statistically random sample of subjects is chosen from throughout the "population" of the group under study. This would involve, for example, randomly selecting groups of young persons from throughout a particular state or region and surveying their tobacco use by administering questionnaires. The information cited in the preceding paragraphs was collected through use of an 84-item questionnaire filled out by high school students during class time in selected schools throughout the country. The questionnaire could be filled out anonymously, and the researchers got parental consent before administering the questionnaires to the young people.

By its nature, this procedure involves some data collection challenges. Those conducting the research have to ask these questions: Did those filling out the questionnaire tell the truth? Did they understand the questionnaire? Were the respondents sufficiently typical so that their responses could generalize to the population at large? The researchers collecting the data employ sophisticated techniques from the fields of epidemiology and statistics to work around those potential hazards. (Epidemiologists, by the way, study not only contagious epidemics such as cholera or measles, but also other conditions within a particular population, such as numbers of deaths by traffic accident, cases of head lice infestation, or occurrence of multiple births.)

Go, Joe, Go

Cross-sectional studies can provide useful information about the relationships between factors under study, but they provide little help in determining causality or predicting outcomes. For more powerful predictive analyses that can suggest causation, researchers need to follow people over a period of time. This longitudinal research can be expensive to conduct, because of the need for tracking the research subjects and maintaining a research facility. It can be simulated, to some degree, by cross-sectional research that looks back retrospectively to possible precursors to tobacco use, such as school performance or tobacco use within the family where the person grew up. Nonetheless, some aspects of longitudinal research cannot be duplicated through any other approach.

Two words are key in this type of research: prediction and association. A study that involves prediction (which is actually a precise statistical term, unrelated to prophecy or to a "psychic" phone line) can explain to what extent one factor, such as depression, predicts, but does not necessarily cause, another. In other words, the process provides a numerical indication of how much the presence or absence of a factor can be used statistically to predict the likelihood that another factor will be present. Weather patterns (temperature, humidity, clouds, wind, etc.) can be used to predict whether or not it will snow in Minneapolis or blow in Idaho. As we all know, such prediction isn't a sure thing in weather forecasts, and neither is it in research into human behavior.

As a hypothetical and scientifically unexamined case, consider this example: Imagine that a preference for long baths as a child predicts adult obesity. The baths themselves would not be a cause of obesity, but they may be related to other factors (disinterest in exercise, for example) that are directly related to obesity. A layperson unfamiliar with statistics might see data connecting the two and wonder whether soaking in a tub for a long time in childhood might actually be causing eventual obesity. Once the news spreads, mothers everywhere would be whisking their youngsters through a light lather and a sprinkling of water, hoping to avoid whatever might cause obesity years down the road.

The other key concept, association, also can be misinterpreted. Association is generally demonstrated by some variation or expansion of a statistical procedure called correlation. Similar to its usage in everyday language, the word correlation indicates an association between two elements. Even if the statistics "work" by meeting preset levels of mathematical significance, the relationship is meaningful only if it makes explainable sense.

It's also important to consider another caveat to understanding statistical prediction and association: That is, we can never know that we're accounting for all the factors that come into play in a person's life. Even if we find, for example, that depression in childhood strongly predicts use of tobacco years later, it is only one of many variables in a person's life. In human behavioral research, we can never identify all the meaningful variables. Scientists can, at best, identify some that matter.

Another aspect of research that applies particularly to work with adolescents is the fact that usually studies are run and data are collected by adults. Unfortunately, what adults expect to affect a young person can be as unrelated to the adolescent experience as the Arizona T-shirt slogans. Another dud with adolescents was an ad in which a for-real "talking" camel announced its displeasure with being used as a symbol for smoking. Adults (including this author) found it clever. Focus groups of young persons, however, weren't amused.

Is the Outcome Inevitable?

Remember candy cigarettes? Remember how we used to "puff" on them as if they were real, all the time performing a grand, sweeping gesture with the hand that held the cigarette? Remember how adults used to tell us not to buy them, lest we become inured to the ways of tobacco and become smokers? Remember those olden days?

Actually, those days are right now. Candy cigarettes and others of their ilk still hold their own in the confectionary business, which also offers children cigars and pipes made from bubble gum, chocolate, and other candies. These sales are predictably controversial, enough so that at least one candy maker removed the "tip" from its candy cigarette and renamed it Candy Stix. Several states have tried, without success, to outlaw candy tobacco-look-alikes; some convenience stores refuse to sell them even if they are legal.

Just what draws a young person toward tobacco? If we had better answers to that question, perhaps we would have better ways to prevent pediatric tobacco use in the first place. Sometimes, it can help to reverse the question and determine what draws people away from tobacco. For example, the state of Utah, which is heavily populated by persons with a religious prohibition against smoking (Mormons), has the nation's lowest adult smoking rate at 13 percent. Even though Utah ranks low in youth cigarette use, some 44 percent of female and 52 percent of male Utah high school students have tried smoking. Those figures are lower than similar data from other states, but they still represent a substantial number of young persons. Fewer Utah youths reported current smoking when questioned for a 1995 survey (17 percent for both females and males), and less than half of those were considered frequent users. The teen smoking rates in neighboring Wyoming were more than double those of Utah, and in Nevada they were nearly double.

Does the anti-tobacco religious influence cause these regional differences? Is the youth smoking rate lower among children who are less exposed to adult use of tobacco? In contrast to the other western United States statistics, a greater percentage of young people in California reported having tried tobacco (63 percent of females and 65 percent of males), but only 7 percent of females and 8 percent of males were frequent cigarette users. Smokeless tobacco rates among young persons were also lower in California than in Utah. While California does have a sizable number of Mormons and other nonsmoking groups, such as Seventh Day Adventists, their representation among the overall state population is not sufficient to swing the numbers that dramatically. A more likely explanation is the aggressive anti-tobacco campaigns carried on in California in recent years, coupled with the basic California ardor for health.

Social Promotion

A 1998 longitudinal report concluded that advertisements and promotions lead one-third of teenagers to try tobacco. John P. Pierce and his colleagues interviewed a sample of nonsmoking California adolescents in 1993 and re-interviewed them again in 1996. Although the teenagers stated in 1993 that they had no intention of smoking, those who had a favorite cigarette advertisement in 1993 were twice as likely to later start smoking or be willing to start as were those with no favorite ad. Those owning or willing to use a tobacco promotional item in 1993 were nearly three times as likely to be smoking by 1996 as those who were unwilling to use a promotional item. About half of the nearly 1,600 teenagers sampled moved closer to becoming smokers between 1993 and 1996. Nearly 30 percent had experimented with smoking during the three-year interval. Of those who expressed a preference for a favorite ad in the first interview, 83 percent favored Camel or Marlboro.

Also, what pulls children and adolescents toward or away from smoking differs from group to group. Reports from Gilbert Botvin and colleagues at Cornell and Columbia universities, who studied predictors of smoking among inner-city Latino and African-American youth, indicated that the most important social influences promoting smoking were friends and peers. Additionally, feelings of hopelessness, lack of efficacy in basic life skills, and low self-esteem appeared to contribute to the likelihood of smoking. In this research, measures of socioeconomic status were unrelated to the extent of smoking. These reports are particularly pertinent because smoking rates among African-American youth are lower than the rates for whites until the trend flip-flops as both groups reach adulthood, when whites' smoking rates fall below those of African-Americans.

The concept of "group self-identification" has also been used to predict adolescent cigarette smoking. A longitudinal study by Steve Sussman and colleagues at the University of Southern California and the University of Illinois at Chicago noted that the peer groups with which seventh graders identified themselves predicted (statistically) smoking in eighth grade. These were the group categories as derived from previous self-descriptions by youth: (1) high-risk youth, including stoners, heavy metalers, and bad kids; (2) skaters, including skaters and surfers; (3) hotshots, including brains and socials; (4) jocks, composed of jocks and cheerleaders; (5) regulars, including new wavers and actors; and (6) others. The highest rate of smoking was among the high-risk youth. Although group self-identification and seventh-grade smoking did significantly predict smoking a year later, the authors were careful to clarify that self-identification was "a fair predictor" with its own merits, but it did not describe the total picture.

The predictive relationship between smoking and the negative emotions of depression or anxiety has been the object of considerable research in adults, but fewer studies have examined it in children and youth. Paul Rohde and a research team at the Oregon Research Institute determined that adolescents who had experienced an episode of depression experienced "psychosocial scars" that included cigarette smoking, in addition to increased health problems and excessive emotional reliance on others. "One implication is that although the rate of smoking in the general population may be decreasing, it may be increasing in adolescents who have experienced an episode of depression," they wrote. "Another possibility...is that...depressed adolescents progress from experimentation to more serious levels of tobacco use."

Children's competence and their parents' behaviors were linked to early tobacco use in research reported by Christine Jackson at the University of North Carolina at Chapel Hill and her colleagues. They found that children who rated themselves as less competent and whose teachers also rated them as less competent were more likely to use tobacco at an early age. Children were also more likely to use tobacco at an early age if their parents were nonaccepting and if their parents were less skilled at setting rules and supervising behavior. Additionally, children of at least one smoking parent were twice as likely to smoke as were their peers whose parents did not smoke.

Children's personality styles at ages six and ten contributed to early use of cigarettes, according to research by Louise Mâsse at the University of Texas-Houston and Richard Tremblay at the University of Montreal. They found that "novelty-seeking" and "low harm avoidance" predicted the early use of tobacco and other substances. They were investigating dimensions of personality postulated by Robert Cloninger in the 1980s. Novelty-seeking is believed to be an inherited tendency toward "exploratory activity and exhilaration" prompted by novelty, or by things that appeal to the appetites. Harm avoidance refers to an inherited tendency to "react intensively to aversive stimuli," which controls the learning mechanisms that enable inhibition.

Scientists attempting to understand the onset of tobacco use also watch for trigger events and environmental risk factors, in addition to vulnerable personality styles. Young smokers themselves state that they smoke because they enjoy it. A survey of more than 10,000 British adolescents, questioned by J.R. Eisner and colleagues in the 1980s, found that they attributed their cigarette use to the experience of smoking itself, rather than to peer pressure. They said that they found smoking to be an enjoyable, calming act that helped them deal with stress. They were more inclined to reject notions that they were smoking because it was a grown-up thing to do, or because of how it made them look among their peers.

University of Reading, England, researchers David Warburton and colleagues in 1991 summarized such findings to date: "It may not be simple exposure to nicotine that results in adolescent smoking, but that smoking results from the situations in which the young people find themselves at this most stressful time of life." The scientists identified smoking as "a coping strategy" for both younger and older smokers.

Evidently, many young white women use smoking as a way to control body weight. Findings from Memphis State University's Diane Camp and her fellow investigators indicated that nearly 40 percent of the surveyed adolescent female smokers stated that they used smoking to depress appetite and control weight. About one-fourth of young male smokers reported the same. Amazingly, not even one African-American teenager who was questioned, either male or female, reported using smoking to control weight or appetite. The researchers concluded that the racial differences might be explained in view of young white women's greater vulnerability to pressure that they be slender. Young African-American women, on the other hand, expressed less discrepancy between their actual weight and their ideal weight than do white adolescent girls. The person most likely to use smoking as a means of weight control was a white female who chronically diets, the researchers determined. They raised the possibility of an intertwining of depression, restrained eating, and smoking in young women. (See chapter 6.)

An Unnatural History

A natural history museum often endeavors to explain the course of events in the life of the Earth, such as the evolution of a dinosaur or a horse, tracing the species' change across time as it moves toward maturation or extinction. Tobacco use also has its own history. This history usually starts in childhood or adolescence and often progresses across the life span of a tobacco user.

A longitudinal study that followed more than 4,000 adolescents into adulthood found that smoking in adolescence was a powerful predictor of smoking in adulthood. Once the adolescents had become adults, many who were smokers attempted cessation and experienced relapse. Few individuals studied actually started smoking as adults. Researchers Laurie Chassin and colleagues from Arizona State University and Indiana University found that young persons with less education and with parents who smoked were less likely to quit smoking themselves. Those who adopted the adult social roles associated with marriage, parenting, and employment tended to quit smoking more than those who did not take on those roles.

The natural history of tobacco use is part of a larger picture of legal and illegal drug use. Columbia University investigators Kevin Chen and Denise B. Kandel followed 1,160 teenagers for 19 years, contacting them first in 1971 and then at three subsequent follow-ups. They reported that the major risk period for commencing the use of alcohol, cigarettes, and marijuana was before age 20, with cigarette initiation peaking at age 16. Smoking peaked in the early 20s and stabilized by age 22. The proportion of participants using tobacco had decreased only slightly by the time the participants were in their mid-30s. While heavy drinking decreased substantially in the study group as they aged from their 20s to their 30s, smoking showed virtually no decline.

Quoting Chen and Kandel: "Two behavioral features of drug histories continue to be strongly associated with persistence of use throughout adulthood: recency and frequency of use at an earlier period." Frequent use of tobacco at an earlier age predicted persistent use well into adulthood. Of all the substances measured, including so-called hard drugs as well as marijuana and alcohol, persistent use of cigarettes was one of the most serious drug-related health problems the study group faced.

The biggest risk of adolescent smoking is that it won't stop once adolescence ends. John P. Pierce and Elizabeth Gilpin of the University of California at San Diego posed the following question in the title to a report on the life span of smoking: "How Long Will Today's New Adolescent Smoker Be Addicted to Cigarettes?" Their answer, summarized, was this: "[T]hese data predict that smoking will be a long-term addiction for many adolescents who start now." Their findings can be coupled with those of Naomi Breslau and Edward L. Peterson of the Henry Ford Health Sciences Center, who determined that the likelihood of eventual cessation was higher in those smokers who started smoking after the age of 13. They suggested that efforts aimed toward delaying the onset of smoking could increase the potential for quitting, and thus could help reduce the death and disease associated with tobacco use.

Pierce and Gilpin found the escalating social undesirability of smoking to be encouraging, even though smoking has become increasingly acceptable among the young. Prevention programs could help teens who are still in the "initiation process" of becoming acquainted with tobacco and who have not yet become dependent smokers, the authors noted. They also placed a high priority on efforts to prevent initial experimentation.

Tobacco use in childhood and adolescence may predispose people to use other substances. Teenagers who drink alcohol and smoke cigarettes even infrequently are 30 times more likely to use marijuana than are those who neither drink nor smoke. Also, among teenagers with no other problematic behaviors, using tobacco, alcohol, and marijuana even to a limited extent increases the risk of other drug use (heroin, cocaine, LSD, etc.) by 17 times.

Cigarette smoking among U.S. adolescents was associated with risky sexual behaviors, marijuana use, binge drinking, and engaging in physical fights, according to a 1997 CDC analysis. Among young persons in Australia, tobacco and alcohol were both deemed to be "gateway" substances leading to other drug use, as reported by Debra Blaze-Temple and Sing Kai Lo of the National Centre for Research into the Prevention of Drug Abuse at Curtin University of Technology in Western Australia.

All of this may seem distant and irrelevant to someone who considers anyone over 21 to be ancient, and who suspects that turning 40 is a near-death experience. This perceived irrelevance is a major reason that smoking prevention programs often are ineffective with young persons and that cessation programs are difficult to design. Donald J. Reid and colleagues of the Association for Public Health in London summarized teenage smoking in Western countries: "Interventions aimed primarily at youth are likely to have a delaying effect only, and sophisticated school programmes, though potentially valuable, have proved difficult to implement effectively on a large scale." They proposed, instead, community interventions that covered all age groups, changes in fiscal policy, smoking restrictions, advertising bans, and media campaigns.

An article in the American Psychological Association's monthly Monitor about the ineffectiveness of substance-use prevention programs indicated that U.S. schools lack access to effective programs and don't use "the best prevention science." Those programs with good track records, including a Life Skills Training program developed at Cornell University Medical College and the Midwestern Prevention Project designed at the University of Southern California, are not widely used. They haven't been "touted to the public," as the author Bridget Murray lamented, laying responsibility on school administrators, researchers, and funding agencies. Shekeh Kaftarian of the U.S. Center for Substance Abuse Prevention's National Center for Advancement of Prevention stated, "You can't just pull any program off the shelf and hope it works."

Both people and tobacco, it seems, are more complicated than that.

The Spitting Image

While the prevalence of smoking—the most common nicotine-delivery system—has stabilized or decreased in many demographic groups in recent years, the use of smokeless tobacco has increased dramatically, particularly among teenagers and young adults of both sexes. The expanding popularity of smokeless tobacco products dates to about 1980, mostly reflecting increased consumption of moist oral snuff.

While the use of smokeless tobacco products has risen, public knowledge about smokeless tobacco has been marked by incorrect notions that smokeless tobacco use is an outdated practice, or a practice with negligible health risk. Compared to what they know about smoking, scientists have learned far less about smokeless tobacco. We know little about how its effects differ from those of smoked tobacco, for instance. We know it has the potential for dependence, as does smoked tobacco, but researchers have only begun studying the ways that dependence on smokeless tobacco differs from dependence on cigarettes. We do not yet know what factors come into play in fostering the continued use of smokeless tobacco over many years.

We probably know more about smokeless tobacco's history than we know about its effects. Tobacco has been used in many forms, including oral and nasal forms, in many cultures, from Sweden to southeast Asia, for centuries. Smokeless tobacco, which was both chewed and used as a nasal snuff, was common among upper-class Europeans. As the practice of spitting out saliva laden with tobacco juice came into disfavor because of its potential for spreading disease, smokeless tobacco use declined. Cigarette use, on the other hand, increased. Although the European Union in 1992 prohibited the sale of moist oral snuff in all countries except Sweden (where, according to Swedish scientist Gunilla Bolinder, it was regarded as a tradition), the prohibition was not yet implemented at this writing. The seesaw trade-off in tobacco products has continued; as cigarette usage has declined for the last several decades, smokeless tobacco use has increased.

If smoking is baffling to non-smokers, the use of smokeless tobacco is an even greater puzzle. Non-users of smokeless tobacco often have no idea how snuff is used, or why. One researcher of smokeless tobacco reported that such questions as these were common, even among smokers:

• Do people stuff snuff up their noses? (Answer: They can, but that isn't a common practice in the United States.)

• How do people use it? (Answer: Users park a "dip" or "chew" or "plug" in their mouths, usually in their cheek or alongside their gums, and the nicotine is absorbed through the lining of the mouth.)

• Does snuff consist of bags of loose tobacco leaves? (Answer: Usually it is sold in little cans or pouches. Those cans of Kodiak, Copenhagen, and Skoal are all smokeless tobacco, or "moist oral snuff," which is the most popular form of smokeless tobacco used in the United States. Chewing tobacco, on the other hand, is sold in bags.)

• Isn't using snuff just something old people do, sitting on their porches somewhere in the Ozarks? (Answer: Not any longer. Some areas report use among teenage boys to be as high as 25 percent, and among girls about one-fifth of that or less. In some locations, such as some Native American reservations, as many as 70 percent of both men and women use smokeless tobacco.)

• Is it purely tobacco? (Answer: No, moist oral snuff consists of cut or ground tobacco leaves, water, and flavoring.)

• Do users swallow it? (Answer: Not generally; they spit out the tobacco-laden saliva.)

• Where do they spit? (Answer: Usually in a cup or other receptacle, but look around before you go barefoot . . .)

The majority of smokeless tobacco research has examined initiation of smokeless tobacco use in teenagers younger than 18 years old. Little research has focused on continued adult use. Consequently, we do not know whether the factors that led to initiation of the practice are involved in maintaining the behavior. We also do not know the role of nicotine dependence in maintaining smokeless tobacco use. Additionally, it is unclear to what extent research findings about psychological factors leading to smoking are applicable to smokeless tobacco use.

Smokeless tobacco involves a distinct pattern of use, often being consumed gradually over a period of hours. Nicotine's effects differ according to the speed and intensity of its absorption. This is believed to account for the differences in psychological impact between cigarettes and smokeless tobacco. Such factors as depression, anxiety, risk-taking, peer and family influences, and perceived health risk have been associated with adolescents' use of smokeless tobacco. University of Minnesota scientist Dorothy Hatsukami, an authority on this and other forms of tobacco, and her associates noted that precursors to smokeless tobacco use included depression and relaxation, and that the self-reported feelings of depression were directly related to feeling bored. Other researchers have found that among smokeless tobacco users of middle-school age, anger is commonly reported.

Work by Elbert Glover and his colleagues indicated that college students using smokeless tobacco were likely to have family members who also used snuff. Users said that smokeless tobacco helped them relax. They considered it less harmful than cigarettes. The tendency to take risks and to get into trouble have been associated with smokeless tobacco use among adolescents in several reports, as has a lack of concern about negative social consequences of smokeless tobacco use.

None of the psychological factors that are believed to lead to the initial use of smokeless tobacco or to contribute to its continuing use are diagnosable psychological conditions. They are all merely normal human traits that vary in degree from person to person. In continued use, alleviation of feelings of depression may contribute to persistent use of snuff, but users of smokeless tobacco are not necessarily pathologically depressed. As is the case in many conditions that occur in human existence, the combination of otherwise benign vulnerabilities and situations may be what leads a person to years of using a potentially hazardous substance.

Some researchers and clinicians encourage smokers to shift from cigarettes to smokeless tobacco as a means of reducing health risks. Although the risks associated with smokeless tobacco use appear to be less than those associated with smoking cigarettes, they remain substantial. As Gunilla Bolinder concluded: "It seems beyond doubt that the use of chewing tobacco together with different mixtures of flavoring, alkaline substances, and a variety of natural products, when used in populations with poor mouth hygiene, and inadequate nutritional status, is causally related to the development of oral cancer."

Chew 'bacca

And lest those not in developing countries think they are immune, consider data indicating that the relative risk of oral cancer is significantly higher in users of moist oral snuff. As for whether snuff use increases the risk of other cancers, the evidence is somewhat inconclusive. Nonetheless, Bolinder pointed out, even if smokeless tobacco is less risky than smoked tobacco, it presents considerably greater potential risk to health than such hazards as pollution, UV radiation, or tainted food. And, as he soberly added: "One must still keep in mind that the use of smokeless tobacco involves the exposure to one of the most addictive substances known."

Epilogue

And what of Douglas Spaulding?

At the end of Bradbury's 1957 book Dandelion Wine, Douglas is still invincible and all-powerful. As the deep of night descends, he commands the town to get ready for bed, to brush their teeth, and to turn out their lights. With his sleep, the summer ends. He is, after all, the mighty age of 12.

If only he had thought, during his moments of power, to suggest that all the smokers lay down their cigarettes and stop smoking.