'Do not let them fool you' (1999-2003)

The 1990s saw the most draconian anti-tobacco policies since the days of the Anti-Cigarette League and yet the US smoking rate stubbornly refused to recede. At the start of the decade, 25.5% of the US population smoked and David Krogh, writing in 1991, predicted that this would fall to 22% by the end of the millennium. In fact, despite all the smoking bans, tax hikes and anti-smoking commercials, it rose to 28%. Far from reviewing their policies of denormalisation, segregation and stigmatisation, the anti-smoking lobby responded by calling for more of the same.

It had been hoped that the millions spent warning adults and educating children about the dangers of cigarettes would be enough to virtually eliminate teenage smoking and ultimately bring about a smoke-free world. The educated classes had been the first to give up smoking and there was an assumption that if the rest of society had access to accurate information about the risks of smoking, mass abstinence would ensue. Millions of smokers around the world had indeed quit. Per capita cigarette consumption in the US and much of Europe halved between the 1950s and the 1990s (1) but while education had a powerful effect on smoking behaviour it had not eliminated the habit, and it did not seem likely that it ever would.

Ignorance was no excuse. A 1974 study found that 99.2% of American 7-8 year olds believed smoking caused cancer, rising to 100% for the 10-11 age group and three quarters of them believed cigarettes to be addictive (2). By 1990, these youngsters were in their mid-twenties and many had become parents themselves but smoking prevalence was at 25.5% and rising. Between 1991 and 1997 the smoking rate amongst high school students rose from 27.5% to 36.4% (3).

One flaw in the campaign to eliminate teen smoking was the presupposition - by now widely held - that smoking was simply not pleasurable. Similar efforts against LSD, cocaine and marijuana did not pretend that users would not enjoy the physical sensations, only that the consequences for the mind and body were not conducive to a happy life. But anti-smoking education for high school youths endorsed the prevalent view that tobacco was worthless, pointless and disgusting and that people only smoked because of peer pressure, 'image' and addiction. This approach backfired when they tried cigarettes, discovered that they provided a pharmacological 'kick' and became sceptical about other information they had been given by authority figures, including that about the health hazards of cigarettes. Worse still, they were inclined to view addiction as more of a challenge than a threat.

By the 1990s, it was no longer tenable to argue that people smoked because they did not understand the risks. An assessment by W.Kip Viscusi found that smokers and nonsmokers alike had responded to thirty years of anti-smoking messages by hugely overestimating the risks of tobacco use. The scientific consensus was that around 10% of lifelong smokers would die of lung cancer but the American public believed, on average, that 38% of smokers would suffer that fate. While smokers tended to give a lower estimate (31%) than nonsmokers (42%), both groups had an exaggerated perception of the risks (4). This was not so surprising. State-funded public health campaigns often led people to believe that they are at greater risk than they really are. The UK's AIDS awareness campaign, for example, created a panic about HIV in heterosexuals that was grossly disproportionate to the number of cases involved (171 in 15 years)(5). Similarly, heavy news coverage of sudden death in dramatic incidents such as aeroplane accidents, terrorist attacks and child murders produce an exaggerated fear of what are, in reality, miniscule risks to the individual. Lung cancer and heart disease are significantly less rare, of course, but with an understanding of how people adapt their behaviour by balancing perceived hazards against perceived benefits, Viscusi estimated that if people had a true understanding of the hazards of smoking, the number of smokers would rise by 7.5% (6).

For public health campaigners, smoking had no benefits and so there was no trade-off to be made and, in any case, they did not consider the indulgence of a few unhealthy pleasures to be worth the possible sacrifice of a few years of existence in old age. Nor did they any longer judge the success of anti-smoking programmes by how well-equipped the public were to make risky decisions but by how many gave up smoking. That over a billion smokers worldwide continued to take what tobacco's enemies emphatically saw as the wrong option seemed incomprehensible, even insane, but health education was only one tool and if it was insufficient to build a smoke-free society, there were other options available. Having tried the carrot, they would have to use the stick.

This shift from persuasion to coercion was resonant of America's 19th century crusade against alcohol. Long before Carry Nation and Lucy Page Gaston were born, there existed a Temperance movement which successfully campaigned for sobriety through education and moral suasion. It emerged in the 1820s, when Americans were drinking an average of 7 gallons of pure alcohol each year but through a combination of education and evangelism they helped to reduce per capita consumption to 3 gallons in the space of just ten years. By 1850, this had dropped to just 2 gallons - around the same as US consumption today. It was a spectacular example of a popular reform movement moderating the habits of a nation and it was achieved without the passage of a single law. But success bred a new militancy. Having begun life as a campaign against hard liquor, the Temperance crusade began demanding abstinence from all forms of alcohol and petitioned lawmakers for total prohibition. Maine was the first state do go 'dry' in 1851 and in the next year another 13 states followed suit. These included Minnesota, New York, Massachusetts, Connecticut and Ohio and several other East Coast states (7).

That all of these states would be first off the mark to ban smoking in public places in the next century is not the only parallel between the two crusades. Just as alcohol consumption dropped dramatically between 1830 and 1850, smoking prevalence in the US halved between 1950 and 1970, at a time when there were no smoking bans, no 'denormalisation' programmes, indeed no anti-smoking laws of any kind other than a rather tame warning label on cigarette packs. Militant anti-smoking groups emerged only after this unprecedented fall in tobacco consumption had taken place but once they had arrived the political lobbying began, the educational programmes took a back seat and an absolutist mentality developed which dictated that the battle be fought until the last smoker threw away his cigarettes.

Traditional methods of tobacco control were shed in favour of more aggressive tactics. Simon Chapman of ASH (Australia) argued that stop-smoking clinics should be closed down (8) and Stanton Glantz even recommended that efforts to stop young people buying cigarettes be abandoned (9) along with plans to prosecute minors caught in possession of tobacco, albeit on the somewhat predictable premise that "The real problem is the cigarette companies"(10). In their place came further smoking bans and the whole-hearted adoption of the denormalisation campaign. Anti-smoking advertising completed its transition into anti-smoker advertising that portrayed smokers as smelly, unattractive inadequates with slogans like the pithy 'If you smoke, you stink'. This, they asserted, was the most effective way of reducing teen smoking.

Adults did not always agree. When hundreds of advertisements were put up in the state of Ohio showing a smoker next to the slogan 'Welcome to LOSERVILLE. Population: YOU', objections came less from smokers than from the local tourist industry who felt that visitors might be put off since the most prominent billboard was positioned outside the local airport. In New York, an advert showed a smouldering cigarette alongside the words 'If you can smell it, it may be killing you'. In a Californian billboard, a woman responds to the familiar request "Mind if I smoke?" with the words "Mind if I die?" Such campaigns served mainly to cement the passive smoking theory in the minds of the public and, in contrast to earlier quit-smoking adverts, were not really aimed at smokers at all. Demonising the tobacco industry and accusing smokers of manslaughter did little to persuade smokers to give up and Stanton Glantz readily admitted that their function was to "enlist" nonsmokers in the cause. The state-funded American Legacy Foundation, like the Californian anti-smoking groups were forbidden from using money collected from cigarette taxes for political lobbying but they were free to launch a $150m advertising campaign that vilified the tobacco industry to such an extent that the state government eventually called a halt to it. No one could accuse these adverts of subtlety. One commercial showed body bags being dragged out of Philip Morris's headquarters and another depicted a tobacco businessman in Hell picking up an award for the 'most deaths in a year' in front of an applauding crowd that included Adolf Hitler. So entrenched was anti-industry sentiment in the movement that when the commercials were pulled, David Kessler asked in exasperation: "How can you run an anti-smoking campaign and not vilify the industry?"(11)

The World Health Organisation passive smoking study

With all eyes on California, far-reaching anti-smoking legislation was being pencilled in by governments around the world. The Californian bar and restaurant ban may have been swung more by emotion than science but two of the largest studies ever carried out into secondhand smoke were expected to produce the killer proof that would end debate on the passive smoking issue for good. The World Health Organisation commissioned France's International Agency for Research on Cancer (IARC) to carry out a huge Europe-wide study of the effects of passive smoking. It was eagerly awaited, said the International Epidemiology Institute, "because of the size of the study, the special attempts to minimise misclassification of cigarette smoking status and the ability to control for various potential confounding factors." (12) With 1,008 female lung cancer cases to examine, the sample group was twice as large as any study to date and no one was in any doubt about the importance of the report, including the tobacco industry who made plans to challenge it in the event that it condemned secondhand smoke as a hazard. They considered this to be the likely outcome after finding out that the IARC's director was a "'fervent antismoker' who believes that 'passive smoking is more dangerous than active smoking.'"(13)

By March 1998 the study had been completed and written up but remained unreleased, creating speculation that the results did not support the WHO's view of secondhand smoke as a genuine health hazard. This suspicion was heightened when the Sunday Telegraph found a summary of the results buried in an internal WHO report. On March 8 1998 the newspaper published an article revealed that the researchers had found no statistically significant elevation in risk for those exposed to secondhand smoke as adults and found a statistically significant reduction in risk for those exposed in childhood. Entitled "Passive smoking doesn't cause cancer - official", the article not only reported the lack of association but, referring to the data on childhood exposure, reported that passive smoking "might even have a protective effect." This sent anti-smoking groups into a frenzy but what was good for the goose was good for the gander. For years, they had used weak associations to 'prove' the passive smoking theory and now they were left in the position of having to dismiss a statistically significant risk reduction of 22% for children while promoting a smaller and statistically insignificant risk elevation of 11% for the wives of smokers (14).

The usual personal insults were dished out to the Sunday Telegraph journalists who had reported the story. The British Medical Association labelled them "mesmerised hacks" while an apoplectic ASH (UK) called them "dupes of the tobacco industry" (15) and filed a complaint with the Press Complaints Commission about both the article and the accompanying editorial which had been entitled 'Setback for Nanny' (it was not upheld). The WHO intervened themselves, issuing a press release headlined "Don't let them fool you, passive smoking is a killer" even though their study had said no such thing. Stanton Glantz searched for tobacco industry involvement but, finding none, rather desperately lamented the fact that people focused on the lower confidence interval when they could focus on the higher one and implied that the whole concept of statistical significance was some sort of tobacco industry trick.

The medical journals were more circumspect but insisted that the report changed nothing. The Lancet accepted that the study provided evidence of a link between secondhand smoke and lung cancer risk which was "tenuous at best" (16) and fell back on older studies and the pronouncements of the Environmental Protection Agency and Surgeon General. Suddenly one of the most thorough studies ever carried out into Environmental Tobacco Smoke (ETS) was "too small to show statistically significant differences" and "underpowered."(17)

If it did nothing else the IARC study ended all talk of ETS posing any more than a very modest risk to nonsmokers. It seemed a long time since Hirayama and others had speculated about secondhand smoke doubling nonsmokers' chance of getting lung cancer. For the time being the anti-smokers were put on the back foot and instead of predicting evidence of higher and higher risks from ETS, worked to protect the credibility of lower, insignificant and/or meta-analysed figures ranging from 1.10 to 1.30. In this they referred back to smaller studies which had not made "special attempts" to minimise smoker misclassification nor had the "ability to control" confounding factors.

Britain's own meta-analysis, conducted by the Scientific Committee On Tobacco or Health (SCOTH), came out a week later and diverted attention away from the World Health Organisation's embarrassment. Timed to coincide with National No Smoking Day, the report was never likely to contain many surprises. A similar endeavour had recently been carried out in Australia where secondhand smoke was officially denounced as a major cause of lung cancer on the basis that it killed eleven nonsmokers a year (in a country of 19 million) (18). A dress rehearsal for the British version had been performed in the pages of the British Medical Journal in 1997. Under the stewardship of Professor Nicholas Wald, the journal collated the evidence to date and concluded that secondhand smoke exposure resulted in a 1.26 (26%) increased risk of lung cancer. After reading that report, the Swedish toxicologist Dr Robert Nilsson suggested that the BMJ must be "innocent of epidemiology" to have published it but with the fiercely anti-smoking Wald also in charge of the SCOTH report there was little doubt that the committee would reach similar conclusions. They announced that ETS exposure increased risk by 1.24 for lung cancer and 1.25 for heart disease. The SCOTH report was designed to guide government policy for the years ahead and so gave anti-smokers something to cheer about after the shock of the IARC study. It even inspired some newspapers to tell their readers that 300 Britons were dying each year as a result of passive smoking (19). But there was more trouble in store.

The American Cancer Society passive smoking study

The WHO's IARC study had been the largest retrospective survey of nonsmokers with lung cancer ever conducted. The largest prospective study - monitoring people over years rather than asking patients about past exposure to ETS - had been running for decades. Commissioned by the American Cancer Society in 1959, the Cancer Prevention Study was originally designed to study the effect of tobacco on smokers, but it also held data about secondhand smoke exposure and included a huge sample group of never-smokers married to smokers. Amongst them were 35,561 nonsmoking women from California and it was they who provided the raw data for Dr. James Enstrom of UCLA who analysed it with his colleague Dr Geoffrey Kabat. The project was also funded by the Tobacco-Related Disease Research Program (TRDRP), another Californian anti-smoking organisation paid for with Prop 99 money.

By 1997 virtually all the data was in but, again, the indications were that this would be a less than ringing endorsement of the passive smoking theory. Faced with the prospect of funding a study which exonerated passive smoke, the TRDRP suddenly withdrew its funding, the ACS withdrew its support and Enstrom and Kabat were forced to seek alternative funding. With the word out on what their research was likely to reveal, no one within the medical community wanted to be associated with it and, in desperation, James Enstrom accepted a grant from the tobacco industry funded Center for Indoor Air Research to finance the final years' work. The report, published in the British Medical Journal in 2003, found no increase in lung cancer or coronary heart disease risk with statistically insignificant negative associations of 0.75 (CI: 0.42 to 1.35) and 0.94 (CI: 0.85 to 1.05) respectively. The American Cancer Society and anti-smoking groups around the world again reacted with horror when it became clear that the ACS's own study had every chance of bursting the passive smoking bubble. The report was due to be published in Britain on March 17 and, as was their standard practice, the BMJ issued an embargoed press release which prevented comments about the story being published before one minute past midnight on March 16. The ACS set about preparing a press release to undermine the report to be circulated as soon as the embargo ended. This statement seems to have been passed to Stanton Glantz who forwarded it to his supporters by e-mail and by the evening of March 15 he had arranged a press conference to, as put it in his e-mail, "debunk" the paper. James Repace was enlisted to give his support and the event took place in Miami at 11am the following morning.

Inevitably, it was the funding from the tobacco industry that gave them their best chance of rubbishing the study in the eyes of the public. Never mind that the funding had covered a fraction of the study's 39 years or that to all intents and purposes the research had been completed before they were involved. Never mind either that the scientists had only been forced into asking around for money to complete their work because organisations with an agenda of extending smoking bans had tried to pull the plug.

Both researchers were respected epidemiologists whose integrity had never before been questioned. Both were nonsmokers and Geoffrey Kabat had assisted in the production of the EPA's 1992 report on secondhand smoke, hardly a pro-tobacco publication. For his part, James Enstrom had produced several papers that had convincingly condemned smoking as a major health hazard as far back as 1975 and had helped write the 1983 Surgeon General's report. No one from the tobacco control movement was able to show how Enstrom and Kabat were supposed to have twisted the data. In addition to the question 'how?' was the question 'why?' Why - if they were mercenaries - did they not tailor their conclusions to meet the agenda of the ACS or the TRDRP when they were paying the bills? And pay the bills they did. The ACS provided 90% of the funding while the TRDRP and the tobacco industry provided just 5% each.

Still, Stanton Glantz described the paper as "tobacco industry funded"(20), called Enstrom "a damn fool" and said that "the science that the UCLA study did was crap." He flatly denied that the American Cancer Society had funded the study and instead placed the blame at the door of Philip Morris(21). Both scientists were dubbed "tobacco industry consultants" by the ACS (who at one time had provided 100% of Enstrom's funding) while ASH accused them of "deliberately downplaying the findings to suit their tobacco paymasters." Michael Thun, the ACS's head of epidemiology, called the paper "fatally flawed" and "not reliable or informative."

Details to back up these allegations were thin on the ground. In lieu of any serious flaws in the paper, Thun employed a spoiler argument recently used by Glantz to undermine the IARC study(22). There was, he claimed, no such thing as a reliable control group since everybody was subject to "considerable amount of environmental smoke before the late 1990s when Californian public places became smoke-free" (23). As a result, since no one was free from exposure, it was impossible to differentiate between one group and another. There was effectively no control group because those not married to smokers were breathing in so much smoke from other sources. It was an extraordinary piece of reasoning because it could equally be applied to Hirayama's 1981 study, and indeed to every study of its kind ever undertaken. In his eagerness to discredit Enstrom and Kabat, Thun risked throwing the baby out with the bath water. By such logic, no study had ever been legitimately carried out since the supposed omnipresence of ETS made it impossible to categorise anyone as unexposed. But even if this dubious idea was taken at face value, one would expect those exposed both at home and outside the home to have a greater risk of disease than those who were only exposed in occasional social settings. This was doubly true of the many housewives monitored in the study who could not be exposed at work (24).

The whole controversy inspired a lively debate in the letters pages of the BMJ but the study's supposed 'fatal flaws' remained elusive. For some, the veracity of the study was not the issue. One correspondent urged the BMJ to print a 'retraction' simply because "unless it is retracted by the BMJ the tobacco industry will use it to promote their vigorous opposition to antismoking legislation" (25) Things were more vicious on the BMJ website's message board, this entry being not untypical:

"Thanks for turning back the clock on public health decades or more. We don't need this kind of negligence from what used to be a professional medical publication. I seriously wonder who got paid off at the BMJ to publish this utter garbage. Dale Jackman, Seriously annoyed I won't dignify this rag with my credentials" (26)

Under fire for having had the temerity to publish the study, BMJ editor Richard Smith went to the unprecedented lengths of publishing its pre-publication history and peer review procedure. In the following week's issue he wrote that although the journal was "passionately anti-tobacco" it was not "anti-science" and that he found it "disturbing that so many people and organisations referred to the flaws in the study without specifying what they were. Indeed, this debate was much more remarkable for its passion than its precision."(27)

Enstrom and Kabat vigorously defended themselves through the pages of the BMJ during this period. They called Michael Thun's attack on them "character assassination of the worst kind" and reasserted their unimpeachable track record which had led to the ACS funding them time and again to carry out smoking research. Three years later, still under attack from the anti-smokers, Dr Enstrom formed the Scientific Integrity Institute to put junk science under the spotlight and published a lengthy rebuttal to his critics that read, in part:

"It is very disturbing that a major health organisation like the ACS made false and misleading statements in a press release about our study before even reading our full paper and then cooperated with Glantz in distributing these defamatory statements on a wide scale basis in violation of the strict BMJ press embargo policy. It is very disturbing that our study continues to be condemned,even though we have presented extensive evidence to refute the unsubstantiated claim that our paper is "fatally flawed." In addition, it is reprehensible that the ACS and Glantz have continued their campaigns to discredit us and "silence" honest research when this research is entirely valid. These actions must be kept in mind when evaluating the honesty and integrity of the ACS and Glantz...Hopefully, epidemiology can continue as a field in which all legitimate research findings can be published and objectively evaluated, including those findings considered to be controversial. However, this will happen only if advocacy organizations like the ACS and activists like Glantz refrain from unethically smearing honest scientists and putting out false and misleading statements about their research."(28)

As a parting shot, the American Cancer Society changed their funding policy in the wake of the study, refusing to grant money to any researcher who had ever received a cent from the tobacco industry or any of its associated organisations. This left Enstrom and Kabat out in the cold along with many other impartial scientists had accepted money from the industry in some form or other in the previous fifty years and went some way to shutting out any dissenting voices. Anti-smoking campaigners turned researchers had always avoided tobacco industry grants and thanks to the anti-smoking industry's own millions often did not need it anyway, and they were free to continue to win funding. This lock-out mirrored the policy of a growing number of medical journals who refused to publish any study that was felt to be tainted by tobacco money (though not the BMJ, who viewed it as a form of censorship). The net effect was that the number of people entitled to participate in the passive smoking debate shrank further still and a clear warning was sent to those who might be tempted to ask whether the emperor was wearing any clothes in the future.

The hastily arranged Miami press conference was an indication of how worried the anti-smoking groups were that the Enstrom & Kabat study would cast doubt in the public's mind about the passive smoking peril. As Enstrom has recalled (above), the counter-attack began before the paper was even published and the volley of press releases and interviews from the tobacco control lobby ensured that news reports were weighted towards assurances that the passive smoking peril was very real. The story received some coverage for a day or two but very few newspapers referred to it in more than one article or revisited the topic later. Some, like the New York Times. did not report it at all and where it was covered, the study's findings were tempered with shovel-loads of anti-smoking reaction. The Sacramento Bee, for example, wrote: "A new study downplaying the effects of second hand smoke on the health of smokers' spouses is being condemned even before it has appeared in print."(29) The rebuttal, rather than the report, became the story.

But the Enstrom & Kabat study did not 'downplay' the risks of secondhand smoke. It explicitly showed them to be a convenient illusion and in so doing was in line with the IARC report and three of the largest recent studies of spousal exposure. Nyberg (1998), Zhong et al (1999) and Kreuzer et al (2001) found relative risks of 0.94, 1.1 and 0.96 respectively(30), as close to a risk of zero as one could reasonably expect. In Europe, a further three papers reported the findings of long running studies of airline cabin staff. Begun in 1960, they covered the period when smoking was allowed on European aircraft and the health of tens of thousands of air hostesses and stewards was monitored. All of them confirmed that cabin crew had a lower rate of cancer than the general population and that lung cancer, in particular, was significantly less common (31). Meanwhile, America's Occupational Safety & Health Administration, the government agency who - unlike the EPA - had the authority to regulate indoor air quality, found themselves unable to condemn secondhand smoke using normal scientific criteria. They already had limits on dangerous toxins and air contaminants, including all the major suspects in tobacco smoke, and after years of investigation concluded that "in normal situations, exposures would not exceed these permissible exposure limits."(32) For those who had eyes to see, the passive smoking theory was unravelling with every year that passed.

The Helena Miracle

While anti-smokers made unsupported and unwarranted attacks on epidemiology that went against their preconceptions, their own science was plumbing new depths. In 2003, three researchers declared that secondhand smoke in bars and restaurants might be responsible for not just some but most heart attacks. At a tobacco control conference in Chicago, they announced that their study of hospital admissions in the little town of Helena, Montana found that incidence of myocardial infarction (heart attacks) fell by 'nearly 60%' in the six months following a ban on smoking in public places. The unavoidable conclusion, as publicised in a worldwide press release by the Americans for Nonsmokers' Rights, was that reducing exposure to secondhand smoke immediately saved lives. Released on April 1, one might have suspected an April Fool's prank. If true, these findings would make secondhand smoke in hospitality venues the principle cause of heart attacks in America and governments who banned smoking in bars and restaurants would have gone more than halfway towards eliminating heart attacks entirely.

The glaring problem was that active smoking was not responsible for anything close to 60% of heart attacks - it is said to be to responsible for 17% of heart disease cases. It was therefore risible to suggest that 'nearly 60%' of heart attacks were triggered by secondhand smoke exposure in bars and restaurants. Not so much questionable as downright impossible, this notion could be dismissed by any layman with a second's thought. Nonetheless, it was picked up by the international media and august news-gathering organisations such as the BBC, The Guardian and the New York Times singularly failed to ask even the most basic questions of the report. New Scientist, to its shame, covered it with the headline 'Public smoking ban slashes heart attacks'(33), a phrase so close to the title of the ANR press release that it is most charitable to assume the journal had not seen the (still unpublished) study and instead took the ANR's claim on trust.

It was another twelve months before it was published (online) by the British Medical Journal. By this time the supposed decline in cardiac admissions had dropped from 'nearly 60%' to 40%, a discrepancy that was never explained. The study was so flawed as to be comical. No questions had been asked of any of the patients regarding their exposure to secondhand smoke, their lifestyles, diet or age and two-thirds of them turned out to be either current or former smokers. It transpired that the authors of the paper had merely counted heart attack cases and ascribed their own explanation to it. The numbers involved were extremely small; there had been an average of seven cases a month prior to the ban and four cases after it.

One vaguely plausible explanation for the drop in heart attack admissions was that the ban, as intended, had caused smokers to give up, cut down or smoke out of town but the authors did not entertain any of these ideas (and, again, no data was given for smoking prevalence or cigarette consumption). Nor did they consider by far the most obvious interpretation: that the decline was a unrepresentative blip in a small, isolated community that had been seized on by anti-smoking activists desperate to strengthen the case for smoking bans. The size of the town (population 66,000), and the small number of cases involved, inevitably made year-to-year changes appear more pronounced when put into percentage terms.

Two of the researchers were doctors at one of the town's hospitals and, according to the ANR, had been "at the forefront of the smoke-free policy movement in Montana for four years"(34). They had only decided to conduct the study at the tail-end of the period in question and so had presumably already noticed the blip. The other member of the team was none other than Stanton Glantz who had publicly declared that his criteria for working on scientific papers was whether they would further his cause (see chapter 7). Americans for Nonsmokers' Rights called the study "a landmark", encouraged tobacco control groups around the world to quote the report's findings and awarded the two Helena activists-cum-epidemiologists the inaugural 'Smokefree Advocate of the Year Award'.

It seemed odd that Glantz was spending his time studying an obscure town in Montana when he lived in San Francisco, the city that had practically invented the smoking ban and one that had a far larger population. So far he has not used the data from his hometown to carry out a similar study, although two researchers (funded by neither the tobacco industry nor the anti-smoking lobby), found that following the ban on smoking in bars on January 1 1998, Californian hospitals dealt with 6% more cases of myocardial infarction than they had in 1997. By 2001, with the lowest smoking rate in the US and a state-wide smoking ban in place, California still ranked a lowly 33rd in a list of the US states with the best record on heart disease(35).

Discrimination in the workplace

"We have to treat them like human beings, I suppose," said the general manager of Cincinnati's Westin Hotel, having reluctantly allowed his staff to smoke outside the building in their breaktime (36). Just as they had in the early years of the century, employers played a part in 'discouraging' smoking among the working class. Paternalism was back in fashion and was no longer anything to be ashamed of. "Developed societies are paternalistic," insisted tobacco control advocate Nigel Gray (37), in an article which argued that companies should refuse to employ smokers altogether. By 1991, 17% of American companies stated that they preferred to hire smokers and 2% simply refused to employ them at all(38). History was repeating but employment law had changed since Henry Ford's day. No longer was it acceptable to discriminate on the grounds of race, sex or disability but discrimination against smokers remained legal. Said one company president: "I'm a firm believer in protecting people from themselves. Employees were given the option of staying with us or leaving because of the new policy. No one left."(39)

Although couched in terms that emphasised the health of nonsmokers, the medical community was well aware that smoking bans on semi-public and private property would help force, or in their words 'encourage', some smokers to cut down or quit and that they would do so by making their habit so inconvenient that smoking was no longer a practical option. In 1989, the Surgeon General remarked that smoking bans "may have the side effect of discouraging tobacco use by reducing opportunities to smoke and changing public attitudes about the social acceptability of smoking" (40). He was surely being disingenuous, treading lightly so as not to be accused of shameless paternalism. For many in the tobacco control movement coercing, inconveniencing and humiliating smokers to the point where they had little choice to give up was not so much a "side effect" as the primary goal with the passive smoking theory acting as the spoonful of sugar to overcome liberal objections. At the 'Revolt Against Tobacco' conference in 1992, Americans for Nonsmokers Rights' spokesman Glenn Barr explained that the movement's aim was to "force [smokers] to do the right thing for themselves" (41) and the medical community was inclined to agree. Lung surgeon William Cahan remarked that: "People who are making decisions for themselves don't always come up with the right answer."(42) It was up to others to make these decisions for them.

In the 1980s, for the first time in decades, companies were firing employees for smoking off the job. Fortunoff's department store in New York State was one of the first firms to refuse to hire smokers at all. Challenging this in the courts, a failed applicant - Amy Lipson - lost her case and company spokesman explained that "We're a health oriented company and we're committed to preserving the health of our employees." He did not elaborate on whether the company would also be refusing to hire people who were overweight, used sunbeds, went skiing, drove motorcycles or drank alcohol but an important precedent had been set. A few years later, Janice Borne of Indiana was subjected to a drugs test at the company she worked for and was judged to have failed it when nicotine showed up in her urine sample. She had smoked six cigarettes at the weekend and was sacked on the spot (43). In 1998 John Dixon worked at a packaging plant in Leeds, England, until he was caught lighting a cigarette in his car after work and was fired (44).

Discrimination took other forms. Smoking employees at a South Wales interior design firm were paid £1 an hour less than nonsmokers, Thurrock Council proposed making smokers work an extra two and a half hours a week and Stockport City Council were only stopped from marking smokers' ID cards with a red dot when a union official complained of discrimination (45). In 1999, the smokers' rights group FOREST counted over 300 job ads that suggested that smokers would be discriminated against and the words "smokers need not apply" were appearing in situations vacant advertisements all over the UK, perfectly legally. Sophie Blinham was fired within 15 minutes of starting her job at Dataflow Communications when her bosses found out she smoked. As she was not smoking on the job she believed she had been unfairly dismissed and assumed she would have the same rights as someone discriminated against for engaging in any other legal habit in her own time. Blinham felt her case was particularly strong since the firm had advertised for 'healthy' staff and had therefore already breached Britain's tough anti-disability laws. She was wrong. Blinham took the company to court in 2005 but lost her case after the judge ruled that smokers were not protected by existing legislation. The court ruled that the disability laws would have only applied had Blinham been 'unhealthy' and that the Disability Discrimination Act specifically excluded nicotine addiction as a disability. A spokesman for the firm told the press that they didn't want people coming into the building "smelling of smoke" and that smokers did not present "a good image" (46).

Bar humbug

Anti-smoking activity was entering an era of absurdity and America was in a league of its own, pushing the boundaries of how far the crusade could go. The US Postal Service commemorated the legendary bluesman Robert Johnson by issuing a stamp in his honour. Finding that the only extant photograph of Johnson showed him smoking a cigarette, they did what the Nazis had done with Stalin's pipe and airbrushed it from his mouth. Paul McCartney's cigarette was airbrushed from the Beatles' iconic Abbey Road album cover when it was turned into a poster and the UK tourist board followed suit in 1998, airbrushing a cigar from a photograph of Victorian engineering genius Isambard Kingdom Brunel (47). Tobacco Control, the British Medical Association's international journal dedicated to anti-smoking advocacy, argued that Franklin D Roosevelt's distinctive cigarette holder should be expunged from commemorative images and even claimed that FDR would have approved of such historical revision. Roosevelt was a paraplegic who died of a brain haemorrage but Tobacco Control insisted that it was cigarettes that killed him (48).

In 1998, Congress took revenge of a sort on the First World War soldiers who had done so much to popularise cigarettes by voting to withhold millions of dollars set aside to pay for their smoking-related diseases on the basis that smoking on duty constituted "willful misconduct"(49).

The story of the death row inmate who was denied his last wish of a cigarette on health grounds sounded so good a joke that some assumed it to be apopcryphal. It was not. Gary Lee Davis, executed in Colorado in 1997, was, like others on death row, legally protected from the passive smoke that could shorten his life. Those who dispensed justice were soon subject to the same rules as those who were incarcerated. In 2006, lawyers representing Phillip Elmore, a man sentenced to death for a murder in Ohio, appealed the verdict on the basis that the jury had rushed to judgment because the judge had refused their requests to be allowed to smoke whilst deliberating (50).

New York City built on a number of anti-smoking efforts in the 1990s by banning smoking in bars, restaurants and even under parasols in March 2003, a law that owed less to the weight of scientific evidence on passive smoke than it did the crusade of a few wealthy individuals. Joe Cherner earned a fortune as a Wall Street trader before devoting his energy to the anti-smoking campaign as a senior member of Smoke-Free Educational Services and the Coalition for a Smoke-Free City. As a gay man with two adopted children, Cherner was a great advocate of tolerance unless tobacco was involved. He spent much of his time in the South of France, reportedly only allowed his family to speak French at home(51) and the New York Post described him as "the world's most annoying human" and an "anti-smoking nazi."(52) He worked tirelessly to make New York smoke-free and had sufficient money and influence to bend the ear of the city's new mayor, Michael Bloomberg.

Bloomberg was a billionaire businessman who had turned his back on a 60-a-day cigarette habit and become an ardent anti-smoker, so he needed little persuasion from Cherner and his friends. A life-long Democrat, he had avoided the competitive selection process of that party by running on a Republican ticket and, having spent $73 million on his electoral campaign, secured a narrow victory in 2002. Upon taking office Bloomberg passed a number of zero tolerance restrictions in the name of combatting crime and protecting public health, of which the most infamous was his far-reaching smoking ban. Not only was it an offence to smoke in any indoor setting but Bloomberg made it a crime to display an ashtray in public and those who went outside to smoke were liable to be booked for 'loitering outside a business.'

The New York ban was opposed by the tobacco industry, a number of journalists, a small band of libertarians and, above all, the hospitality and bar industry. Health groups insisted that banning smoking would be good for business; their argument was that there was an inherent market failure which only comprehensive legislation could correct. Those in the trade were not convinced and a year after the New York ban took effect, numerous studies showed that the ban had a negative impact on bars and restaurants. Anti-smoking groups dismissed them as tobacco industry propaganda since some of them had been part-funded by the industry. For good measure, they cast doubt on figures coming from the hospitality industry, assuming their opposition to the ban signified pro-smoking sentiment. If one used the same presumption of bias to dismiss reports paid for by anti-tobacco organisations there was little evidence left to go on.

One report, with which the anti-smokers could find no tobacco ties, came from Ridgewood Economic Associates which showed the law had cost New York City bars $37,000,000 in revenue and that associated industries (eg. suppliers to bars) lost a further $34,500,000. In total, they reported, 2,650 jobs had been lost as a direct result of the ban in its first year (53), a dramatic downturn for an industry that had previously been booming. On the other hand, Mayor Bloomberg reported that bar and restaurant tax receipts had leapt by 8.7% and that 2,800 jobs (after seasonal adjustment) had been created. The Campaign for Tobacco-Free Kids claimed the ban resulted in growth of 12% in the sector.

Finding the truth amongst all the figures cited in this controversy was no easy task and was muddied further by the World Trade Centre attacks. Throughout 2002, with a significant part of downtown Manhattan in rubble and tourists staying away, New York's economy was in an unusually troubled period. Bloomberg's predecessor, Rudolph Giuliani, estimated that business was damaged to the tune of $50 billion that year. One would expect the city's trade to improve in 2003 regardless of anti-smoking legislation, but such an improvement was not far from obvious.

The discrepancies between the reports mentioned above may be explained by Ridgewood only examining bars while Bloomberg included restaurants, many of whom, like McDonalds were already smoke-free before the ban took effect. They could be explained by changes to the tax system and the effects of inflation (positive views of the ban always focused on tax receipts rather than sales). Then again, they may be explained by bias. The anti-smokers' favoured method was to carry out surveys asking if people would go out more or less if there was a smoking ban, a notoriously unreliable method that amounted to little more asking people if they liked smoking. On the other hand, the tobacco industry tended to ask bar owners if they expected trade to go down and was equally unsound. There are too many partisan interests involved, and too many contradictory figures, to be sure of how much damage - if any - was done to the New York economy by Bloomberg's law.

The picture was no less murky in California. Bar takings rose in the years following the ban but this coincided with a general upturn in the economy and bar revenue in the state rose at a slower rate than in the rest of the country. Stanton Glantz had no doubt that the ban was good for bars and no doubt either that those who questioned it were connected with the tobacco industry. With the help of a grant from the Robert Wood Johnson Foundation he founded TobaccoScam, an organisation which presented the idea that smoke-free laws damaged business as a lie propagated by the tobacco industry. TobaccoScam launched 22 different adverts using photographs of pretty barmaids and waitresses to persuade the public that, as their tagline went, "Big Tobacco is lying again." The TobaccoScam slogan represented the by-now distinctive anti-smoking argument of 'The tobacco industry says it, so it must be a lie', an effective approach that acted as a convenient substitute for evidence.

One final tale is worth telling for the light it shines on the influence of vested interests. In 2006, two Minnesotan anti-smoking groups, the Association of Nonsmokers and ClearWay Minnesota, invited the press to relay the results of their study of air quality in the newly smoke-free bars of the city of St. Paul. Air samples had been taken before and after the ban and in the interests of fairness they monitored the same bars, on the same day of the week and at the same time, counting the number of patrons therein. With discernible glee, they reported a 93% fall in particulate levels since the ban came into effect. Having duly reported this story under the headline "Smoking Ban Clears Air in Bars, Study Says", St. Paul's Pioneer Press requested the raw data for review. At this point the anti-smokers suddenly became cagey and it took a request under the Minnesota Data Practices Act and an intervention from the Minnesota Department of Administration for them to hand over the full results. The reason for their reticence then became clear. The figures showed 38% fewer people in the bars after the ban than had been present before. James L. Repace inadvertently showed a similar effect in the same year. A man of finite talents, he had been once again wandering around pubs measuring how much smoke was in the air, this time being paid to do so with another RWJF grant. So excited did he become with his findings - which proved the less than earth-shattering fact that there was less smoke around after a smoking ban than before - that he failed to comment on the fact that the pub had attracted an average of over 100 customers before the ban but just 71 a year later (54).

In March 2004, the Republic of Ireland became the first country in the modern age to ban smoking in all its bars and restaurants and the downturn in pub revenue was so sharp and unmistakable that it soon became very difficult for anti-smoking groups to pretend otherwise. The ban had been sold to the Irish pub industry on the basis that it would fill their establishments with nonsmokers who had previously avoided them. Nonsmokers consistently told pollsters that they would indeed go out more often if a ban was brought in but, alas, it was easier for them to agree with someone holding a clipboard than it was to spend enough money in Ireland's pubs to make up for the departing smokers. Within three months, the Vintners' Association of Ireland reported that business was down by 15% in urban areas and 25% in rural areas. A year after the ban was enforced, the Centre for Economics and Business Research found that overall trade in bars, clubs and pubs had fallen 10.7%; this at a time when the country was experiencing sustained economic growth and where GDP was rising at 7% a year. Landlords experimented with patio heaters, awnings and even double decker buses parked outside their pubs but not everyone was able to find ways of getting around the law. Those on the border were worst hit since their customers had the option of travelling over the border to smoker-friendly Northern Ireland and within a year of the ban taking effect many had gone bust or sold up.

Evidence of a severe drop-off in trade was so strong in Ireland that anti-smoking groups retreated to a position of declaring the damage was not too bad. Tobacco Control magazine claimed that trade had fallen by 5.8% but maintained that the ban was a "runaway success"(55). Naomi King of ASH gave an unsupported figure of a 3.5% fall in pub revenue and insisted that "we have figures showing that it's been a huge success and everybody is happy with it."(56)

One person who wasn't happy with it was Gareth Kendellen, the landlord of a Cork pub. After the ban was imposed he saw takings fall by 30% and took drastic action, cutting a hole fifty square feet wide in his roof, thereby making it no longer an "enclosed space". Kendellen saw his customers return and profits recovered to pre-ban levels although, as he explained to the press, "we have to keep a stock of free umbrellas behind the bar."(57)

More diseases linked to passive smoking

It was a peculiar feature of the passive smoking story that when, in the first years of the new millennium, the evidence against passive smoking was at a low ebb, public belief in it was at its highest. In Britain, ASH and the British Medical Association (BMA) continued to preach the gospel of passive smoking in a bid to ratchet up the battle against tobacco. Hopes that a study of thousands would provide a clear indictment of secondhand smoke had been dashed but since public opinion was broadly sympathetic to the passive smoking theory there was nothing to be gained from opening up a debate. In the absence of satisfactory evidence the movement adopted a policy of repetition: Passive smoking was a killer and the case was closed. With neither the time nor inclination to delve into the scientific data, this was enough to persuade much of the press and large swathes of the public that the case against passive smoke was strong and proven. Lazy or sympathetic newspaper editors published full page ASH press releases as news stories and by the middle of the decade British newspapers regularly stated that 1,000 people were dying in the UK each year as a result of secondhand smoke.

The movement did not retreat an inch in the weight of recent evidence and added cot death and asthma to the list of diseases supposedly caused by secondhand smoke, ignoring the fact there was no geographical or historical correlation between smoking prevalence and incidence of either illness. Indeed the recent upsurge in asthma was occurring at a time when there was less smoke in the air than ever before and cot death was rare in countries like Russia and Hong Kong where smoking was most prevalent (58). Such common sense observations were not enough to deter the anti-smoking industry from touting mathematical and theoretical death counts which showed passive smoking to be wreaking ever-greater carnage.

In the United States, ASH claimed that 3,000 children a year were dying from Sudden Infant Death Syndrome (SIDS) as a direct result of passive smoking. This was a low blow. Cot death charities had spent years trying to help parents not feel guilty about a syndrome whose cause remained a mystery and those involved in SIDS support groups could see with their own eyes that their members were not predominantly smokers. This was at odds with ASH's message that parents were responsible for killing their offspring and the Sudden Infant Death Syndrome Alliance sent Banzhaf a strongly worded letter criticising his "use of misleading data and terminology when linking Sudden Infant Death Syndrome to your cause", adding:

"The sensational heading for one of your recent Internet reports [07/30] "Smoking Parents Are Killing Their Infants" has gone too far. The fact is, researchers still do not know what causes SIDS...Insensitive generalisations about SIDS broadcast through print or the electronic media serve only to perpetuate the public's misconceptions...Your literature states that smoking 'kills more than 2,000 infants each year from SIDS.' Any published figures are sheer speculation, or guesses, not grounded in actual experimentation...we respectfully request that you adjust your message as far as SIDS is concerned. While we support your cause, we can not do so at the expense of the tens of thousands of families we represent. Thank you for your consideration of our concerns. A copy of our latest information brochure is enclosed. We welcome your reply."(59)

Banzhaf did not reply and ASH continued to quote the statistic.

The real holy grail for the anti-smoking lobby was finding a link between secondhand smoke and breast cancer. Despite the dramatic rise of lung cancer in the previous century, and despite it only being able to affect half the population, breast cancer continued to be by far the most common form of cancer in the US. For those who liked to extrapolate tiny relative risks over vast populations, finding any association with secondhand smoke would be an invaluable public relations tool. It was somewhat infuriating, then, that despite numerous attempts, no consistent or convincing link had ever been made with active, let alone passive, smoking. A 1994 paper published in the British Journal of Cancer found no link between breast cancer and first or secondhand smoke (60), nor did a massive assessment of 53 studies that encompassed 55,515 breast cancer patients in the British Journal of Cancer(61). The Centers for Disease Control, the American Cancer Society, the IARC, the Australian Medical Journal, British Medical Journal and the US Surgeon General all agreed that there was no link. Geographical and historical spread of cigarette consumption showed no correlation with breast cancer prevalence and while lung cancer rates in women began rising in the US from the mid-1960s breast cancer rates were unaffected by the post-1940 surge in female smoking.

In any other field of research, this would surely have been enough to lay the matter to rest but, unhappy with the idea that not every death could be prevented by banning smoking, the Californian Environmental Protection Agency (Cal-EPA) conducted a meta-analysis of 15 studies. This was their second attempt. The first, in 1997, failed to find any link. The second, released in 2004, showed a small but statistically significant 1.40 (1.17-1.68) relative risk for nonsmoking women exposed to cigarette smoke. This result only applied to young women; the middle-aged and elderly remained mysteriously unaffected. As with the infamous 1993 EPA report, Cal-EPA cherry-picked the evidence. None of the five cohort studies backed up their hypothesis and the case-control studies that did were based on breast cancer victims filling out questionnaires in the most anti-smoking state in America and were therefore wide-open to recall bias. A large cohort study of 1,150 cases found a relative risk of 0.93 but apparently arrived too late to be included in the meta-analysis (62). Had it been, the overall risk ratio would have landed at 1.01.

Instead, and for the benefit of the media, the Cal-EPA gathered together the five studies that best supported the ETS theory and meta-analysed this mini-group to come up with a figure of 1.90. This made passive smoking more dangerous than smoking itself; this was no surprise since there was no link with active smoking and the Cal-EPA was effectively comparing one placebo with another. Nor was it any surprise to anyone except those deeply entrenched in the anti-smoking movement that null studies frequently produced apparent risks of between 1.0 to 2.0. If it proved anything, the Cal-EPA report proved how easy it was to create a statistical association out of nothing.

Rather than questioning the validity of the report, Banzhaf issued a credulous ASH press release entitled: 'Secondhand tobacco smoke more dangerous than smoking itself - implications for women especially frightening'(63) Even the American Cancer Society, who seldom worried about overstating the dangers of secondhand smoke, declined to back the meta-analysis and their director, Michael Thun, politely remarked that the "lack of an association with active smoking weighs heavily against the possibility." Incensed by this betrayal, Stanton Glantz all but accused them of siding with the tobacco industry and blamed "pro-tobacco forces" for casting doubt on the reliability of Cal-EPA's efforts (64). His TobaccoScam organisation wasted no time in producing a poster of three pretty waitresses telling the public that passive smoking doubled their chance of contracting breast cancer.

Every disease is a smoking related disease

While secondhand smoke was being fingered as the cause of virtually any malady, teams of eager epidemiologists put together statistics to implicate smoking in the few diseases that had not yet been prefixed with the words "smoking-related." Historically, anti-smokers had reeled off long lists of diseases supposedly caused by the demon weed based on little more than a hunch and a burning desire to scare the public off the habit. History was repeating itself again. A distinct lack of plausible biological mechanisms did not stop smoking being 'linked to' diabetes, skin cancer, lower back pain, cervical cancer, prostate cancer, brain damage, obesity, hair loss and stupidity (65). One study announced that smokers had a 60% increased risk of developing colon cancer and a 500% greater chance of rectal cancer (66). Its authors blamed tobacco for 1 in 10 deaths from colorectal cancers despite the fact that this went against decades of previous research. Several studies into secondhand smoke had even used colon cancer patients as a control group for the very fact that "colon-rectum cancers have been shown in epidemiological studies not to be related to cigarette smoking."(67) Such studies, including those of Garfinkel and Fontham, had shown that colon cancer victims were no more likely to have been around tobacco smoke than random population samples but if, as was now suggested, smoking increased colorectal cancer risk by 310%, smoking was much more dangerous to the colon than it was to the heart. Two years later another study came along which served mainly to emphasise the inevitability of chance in the science of statistics, when it found that passive smoking increased rectal cancer risk by just 50% in men, a tenth of the original estimate and, oddly, not at all in women (68). One could believe the association to be unsound or one could believe that a major killer had been discovered; a killer that had not aroused the least bit of suspicion until the end of the millenium, by which time individual epidemiological studies based primarily on questionnaires came to be taken at face value however unlikely their conclusions.

Another 'epidemic' that had gone curiously unnoticed for centuries was given rather more prominence. The opportunity to link smoking with male impotence was too good to pass up and ASH (UK) wasted no time in announcing that 120,000 men between 30 and 49 were suffering from erectile dysfunction as a direct result of smoking. In the late 1990s and early 2000s, several sniggering advertising campaigns on both sides of the Atlantic were launched with the intention of shaming men into quitting and convincing women that nonsmokers made better lovers. Men who smoked, they claimed, were 50% more likely to be impotent than those who did not. Again, this was hardly supported by empirical evidence and would have been news to the young men of the 1940s who defied their prodigious consumption of cigarettes to create a baby boom. When one delved a little deeper, this statistic turned out to be entirely based on a mid-1980s survey of Vietnam veterans which apparently showed that a few more smokers than nonsmokers were impotent. The total number of impotent smokers in the study numbered just 74 and the inherent risks of extrapolating from such small numbers over populations of millions - particularly when they come from just one study - were well known to responsible epidemiologists. The estimate of 120,000 men came not from the original researchers but from ASH (UK), who did some quick sums - in reality, wild extrapolations - that converted the findings of one small study onto the population of the UK for the purposes of a press release. They went on to announce that "though full or partial recovery is possible, this assumption must be regarded as optimistic." This was a perverse statement. Since they were disposed to base health statistics on one twenty year old study they could have pointed out that it had also shown that former smokers were less likely to be impotent than nonsmokers. They could have also reminded the public that self-reported ailments from surveys provided shaky evidence at best and that smokers, as David Krogh found, are less likely to lie to scientists, less concerned about what people think of them and are "more honest than nonsmokers in the view of themselves that they present to others"(69); all characteristics that have particular relevance in this case.

Where there was smoke there was inevitably a passive smoking study to follow it. A fresh bout of data dredging - this time of 50 to 80 year olds - produced a study that showed that 33% of men subjected to secondhand smoke at home and at work were impotent. It also purported to demonstrate that 30% of cigar smokers and 24% of cigarette smokers were similarly affected in this area. The unlikely conclusion, therefore, was that cigar smoking was particularly dangerous. "There seems to be a message here for old and young men," announced The Center for Social Gerontology(70). And what message was that? That in any smoky bar there will be more impotent nonsmokers that impotent smokers? That cigarettes are less harmful than cigars? Or perhaps simply that older people are more likely to smoke cigars and are more likely to be impotent?

Even AIDS was been bracketed as a smoking-related disease. Andrew Furber, a doctor working in Sheffield, South Yorkshire, put together a meta-analysis with the help of Google which showed that smokers were more likely to contract the HIV virus(71). This in itself was hardly a shock. Furber admitted that smokers tended to be risk-takers and - although he did not broach the subject - smoking is twice as common amongst gay men than it is amongst straight men (72). Such subtleties were lost on the media ('Smokers at greater risk of HIV' declared the BBC) and Dr Furber insisted that a biological explanation was likely. As a further lesson in how statistics can be bent to fit any purpose, some AIDS activists interpreted the findings in quite the opposite way. "People with AIDS have increased lung cancer risk and it's not all due to smoking", reported the journal AIDS, which suggested that having AIDS made an individual more susceptible to lung cancer, even as it accepted that 80% of those with HIV smoke cigarettes (73).

All of this was like saying that the sun came up in the morning as a response to birds singing. The complusion to confuse cause with effect led ASH to trumpet the supposed power of cigarettes to cause young people to use marijuana, not floss daily and engage in other "risky teenage behaviour" (74). A report in the journal Addiction even showed that those who have boyfriends or girlfriends early in life were many times more likely to become smokers (75). While such associations may well exist, the question remains 'So what?'. As discussed in Chapter Three, smokers do not always behave like nonsmokers. Is there anything to be added by conducting expensive epidemiological studies to prove that smokers are less likely to join a gym than nonsmokers or more likely to smoke cannabis? All that is left is a pile of associations that are of no use to doctors and are too obvious be to of interest to sociologists. Absurdly, such findings have been used by anti-smokers, including ASH, to 'prove' that tobacco somehow twists smokers' brains and changes their behaviour.

This thinking reached its nadir when the world's press reported that smoking made people kill themselves. Three British researchers published a study that showed smokers to be more than twice as likely to commit suicide than nonsmokers (76). Again, this was no surprise; those suffering from depression often turned to tobacco. The twist here was that the whole point of the paper was to illustrate the challenges facing epidemiologists and to highlight the danger of putting the cart before the horse when making conclusions from statistics. It was designed more as satire than advocacy. But irony can be a foreign language in the arena of public health and it was not long before another team published a similar study and claimed that smoking really did make people kill themselves. They even cited the earlier paper as supporting evidence. The British team responded in the letters page of the American Journal of Epidemiology to remind them that their study had also shown that smokers were twice as likely to be murdered but that "unless health promoters have moved onto a direct action phase, during which they shoot smokers, this association is unlikely to be causal"(77). Their critique of anti-smoking research that ignores other aspects of smokers' lifestyles bears repeating:

"We live in a world in which associations are more common than lack of associations, and the former are only worth drawing attention to if they increase our understanding of why the world is the way it is. The 'independent' association between smoking and suicide is about as interesting as the equally strong 'independent' association in which never wearing a seatbelt apparently increases the risk of dying of all-cause or respiratory disease mortality. These findings merely show for the thousandth time that smokers are different to nonsmokers and that people who wear seatbelts are different from people who don't wear seatbelts."(78)

These words represented an oasis of reason in a field of science which seemed intent on presenting tobacco as the source of all mortality but they fell on deaf ears. The Robert Wood Johnson Foundation and the state-funded Centers for Disease Control immediately paid for more research to be carried out into the smoking-suicide connection.

Copyright: Christopher J. Snowdon 2007-09