The myth of the smokefree health miracle

Recent reports of a ‘dramatic’ fall in the number of heart attacks in England after July 2007 represented the latest in a long line of attempts to find immediate health benefits from smoking bans. But a serious examination of this body of evidence suggests that the effect of smoking bans is either tiny or non-existent.

The worldwide search began in 2004, when the British Medical Journal reported a 40 per cent decline in ‘acute myocardial infarction’ (AMI), the medical term for heart attack, in the small town of Helena, Montana. Subsequent ‘heart miracles’ claimed drops in AMI of 47 per cent (Bowling Green, Ohio), 27 per cent (Pueblo, Colorado) and 17 per cent (Scotland).

As previously reported on spiked, the widely touted Scottish figure of 17 per cent was at odds with hospital admissions data showing an eight per cent drop in the first year of the ban followed by an eight per cent rise in the second year. When this inconclusive evidence is combined with hospital admissions data from Wales, Denmark, New Zealand and Australia showing smoking bans having no effect on the heart attack rate (1), the most striking aspect of this field of research is the tendency to find dramatic results in small communities and practically nonexistent effects over large populations.

The counterintuitive conclusion was that secondhand smoke was ferociously lethal in one-horse towns in the mid-West, but strangely benign in whole nations. The alternative, if more cynical, explanation was that obscure destinations like Helena and Bowling Green were brought to the world’s attention because anti-smoking campaigners had dredged the data for unusual blips that roughly coincided with provincial smoking bans.

That question seemed set to be resolved when The Sunday Times announced in September 2009 that the smoking ban in England (population 49million) ’caused a fall in heart attack rates of about 10 per cent’. The source of this claim was never disclosed and the anti-smoking campaign Action on Smoking and Health (ASH) quickly downplayed it, insisting that the 10 per cent figure was ‘not based on any research conducted to date’ (2). Nevertheless, the research was underway and it finally bore fruit a fortnight ago in the form of another British Medical Journal study.

Led by Dr Anna Gilmore, a member of ASH and the director of the Tobacco Control Research Group, the study found a post-ban drop in AMI of not 10 per cent, let alone the 40 per cent found in Helena, but of just 4.3 per cent. A welcome decline, to be sure, but since the final years of ‘smoky’ England saw similar declines of 3.2 per cent and 5.2 per cent, the evidence for a heart miracle in the most populous nation yet studied was less than compelling (11).

Faced with data that unequivocally showed heart attacks falling at the same pedestrian rate as before the ban, Gilmore and her team turned to computer modelling. After making adjustments to the data, they concluded that, despite appearances, the smoking ban had a profound effect on the nation’s hearts. Of the 4.3 per cent drop in AMI admissions, Gilmore attributed more than half (2.4 per cent) to the smoking ban. The study concluded that ‘the implementation of smoke-free public places is associated with significant reductions in hospital admissions for myocardial infarction’. A press release was then issued, headlined ‘Smokefree legislation linked to drop in admissions for heart attacks’.

To make life simpler for busy journalists, the press release chose not to mention that this was a computer-generated estimate, instead flatly stating: ‘A 2.4 per cent drop in the number of emergency admissions to hospital for a heart attack has been observed following the implementation of smokefree legislation in England.’ As was helpfully pointed out, this 2.4 per cent drop equated to 1,200 heart attacks being ‘prevented’ by the 2007 legislation. There was no mention of the downward trend in AMI that long predated the smoking ban.

Since the 2.4 per cent figure exists only on a laptop at Bath University, the calculations that led to it can be neither verified nor debunked. The possibility that the smoking ban contributed to part of the drop in AMI admissions after July 2007 cannot be ruled out, particularly if it led to a significant drop in the number of smokers (the jury is still out on whether this happened). But since the number of heart attacks fell at a similar rate after July 2005 and July 2006, the burden of proof rests on Gilmore & Co. Without it, it is as if they were doing a rain dance in the middle of a thunderstorm and demanding credit for the rain. The onus is on them to convince us that the skies would have cleared if they hadn’t showed up, not the other way round.

Gilmore’s case rests on making adjustments for three relevant but hardly decisive confounding factors that might disguise the effect of the smoking ban: surface air temperature, population size and Christmas holidays. This is all good practice, but more significant risk factors such as smoking status, diet, statin use, exercise and stress go unaddressed. It could not be otherwise. Hospital admissions data reveal no personal information about any of the patients beyond their age and gender. This only highlights the immense difficulty of making specific assumptions from a mass of nameless aggregate data.

The only thing that can be said with any confidence is that there were 2,300 fewer heart attacks in 2007/08 than in the year before. With heart attacks and heart disease having hundreds of risk factors interacting with each other in complex and unpredictable ways, using raw data to single out any one of them is like listening out for a kazoo in a stadium full of vuvuzuelas. Any estimate made against this noisy statistical background can only be speculative to the point of wishful thinking.

Perhaps this underestimates the power of the team’s computer model, but if they have truly devised a formula that can predict the number of heart attacks by taking the temperature and seeing what day Christmas falls on, it is not one they are prepared to share with us. Effectively, the reader is told: ‘We know it doesn’t look like the smoking ban had any effect on AMI admissions but we’ve run it through a computer model and it has. Take it or leave it.’ In the context of the dubious and frequently bizarre history of ‘heart miracle’ studies, the reader could be forgiven for leaving it.

If it does nothing else, the English study confirms that the wilder claims of heart miracles in Helena and elsewhere were way off base. In the course of six years, the ‘smoking ban effect’ on heart attacks has fallen from over 40 per cent to less than five per cent. And since the heart attack rate was known to fall by more than five per cent in some years before smokefree legislation was introduced, attributing any part of the secular decline to the smoking ban becomes a matter of interpretation and conjecture.

Not that the hypothetical nature of Gilmore’s study ever impinged on the news coverage devoted to it. No one reading the newspapers two weeks ago could have gone away thinking anything other than that there were 1,200 fewer heart attacks after the smoking ban and that this decline in numbers was an unusual and remarkable event.

As was the case in Scotland two years ago, the statisticians who painstakingly collected admissions data from English hospitals might as well not have bothered. The true figures vanished, replaced by unseen adjustments and unspoken assumptions from the gatekeepers of knowledge at the UK Centre for Tobacco Control Studies. Once again we had findings erroneously leaked to the media months in advance, a press release which failed to get the most basic facts straight, and a study with no verifiable evidence to support its central conclusion. And all published just in time for the government’s review of the smoking ban. If this doesn’t warrant a little scepticism, what does?

Christopher Snowdon is author of Velvet Glove, Iron Fist: A History of Anti-smoking, published by Little Dice. (Buy this book from Amazon(UK).) and, most recently, The Spirit Level Delusion: Fact-Checking the Left’s New Theory of Everything, published by Democracy Institute/Little Dice (Buy this book from Amazon(UK).)

(1) Related articles: Denmark; Australia; New Zealand; Wales

(2) ASH Daily News for 15 September 2009