A bleary-eyed attitude to alcohol research02-Feb-2010
A bleary-eyed attitude to alcohol research
It is time we started evaluating studies on the effects of booze on their merits rather than on who happened to fund them.
Basham and Luik
Ever since the House of Commons Health Select Committee grilled alcohol industry chief executives last summer, it has been open season on alcohol, especially alcohol advertising. Leading this campaign has been the British Medical Association (BMA), whose Board of Science recently called for a complete ban on all forms of alcohol advertising and marketing.
Now the campaign against alcohol promotion has been taken up by the BMA-owned British Medical Journal (BMJ), which recently published three articles attacking alcohol promotion and spawned headlines like Alcohol industry is ‘targeting young people’ in the national press. There are many claims within these pieces that are troubling, but two in particular deserve special attention.
First, there is the claim, explicit in the BMJ’s Lobby Watch article, that researchers funded by the UK government are objective in the debate over alcohol advertising policy because they are ‘independent’, whereas researchers funded by the alcohol industry are not. Second, there is the claim that it is ‘established’ that alcohol advertising ‘encourages young people to drink alcohol sooner and in greater quantities’.
The effect of these two claims is to suggest to the public and policymakers that, first, the debate over the effects of alcohol advertising is over and, second, only those people funded by the alcohol industry question this reality.
Is this true? The idea that people whose research is funded by the taxpayer are independent, and therefore objective in their findings and policy prescriptions, is naive at best, disingenuous at worst. Public Choice theory explains that the interests of those who claim to represent the public are themselves conflicted and self-serving, and certainly neither independent nor objective. Whoever supports their work, researchers have biases shaped by their education, experience, ideology, and the politics of their profession.
In the case of alcohol research, the public health research community is largely supported by a state that is Janus-faced over the appropriate place of alcohol in society, decrying and controlling on the one hand whilst happily taking its share of an immensely profitable undertaking. Are we seriously to believe that researchers from the public health community are unaware of the government’s policy positions on alcohol when they undertake their research?
Indeed, they approach their research for the state knowing the policy outcome that will please their patron in the same fashion as an industry-supported researcher does. The recent controversy involving Professor David Nutt, chair of the Advisory Council on the Misuse of Drugs, who was fired by the home secretary Alan Johnson following his challenge to the current drug classification scheme, shows that the government neither wants nor appreciates policy-neutral, independent science.
The BMJ’s Lobby Watch columnist, Claire Harkins, herself displays the absurdity of this position. In her statement of competing interests, she notes that she is ‘negotiating over a report with Scottish Health Action on Alcohol Problems’. How is the fact that her research is funded by Scottish Health Action on Alcohol Problems different than, for example, a report that is funded by the alcohol industry? In both cases, the funders have interests and agendas.
Since all research, except for the rare independently wealthy researcher, is funded by someone, the fact that research is funded cannot by itself be taken as a reason for rejecting it. This suggests that the BMJ’s real objection is not to the funding of research, but the fact that some funding is by ‘good’ funders, that is, funders who guarantee independence and objectivity, and some funding is by ‘bad’ funders, that is, funders who demand subservience and bias. Since the BMJ clearly thinks that the state funds research without strings attached, government-supported health research is good by definition; industry supported research is bad by definition. End of story.
But the history of alcohol policy research disproves this.
For example, the Framingham project, a multi-generational American epidemiological study financed by the US government, found early on that there were significant positive effects of moderate drinking on heart disease. Yet, the government denied the study’s key researcher permission to publish these results because they would be ‘socially undesirable in view of the major health problem of alcoholism that already exists in the country’. (See Seltzer, Journal of Epidemiology, 1997 50: 627-629.)
Therefore, the issue is not the black-and-white one that the BMJ posits. Both kinds of research have the potential to be shaped by the interests of the funder and both begin with biases, conscious or not. To simply reduce research to a question of funding avoids the hard question of the quality of the research itself. Rather than evaluating each piece of research on its individual merits, evidence which doesn’t fit what the public health community wants to hear is simply dismissed or ignored on the grounds that it is done and funded by the wrong people.
As a case in point, our new research on the connection between alcohol advertising and youth consumption and overall consumption is shortly to be published in a peer-reviewed journal. It has not been funded by the alcohol industry, yet it profoundly challenges the views set out in the BMJ by Professor Gerard Hastings and others.
Here is what we found:
- Only a handful of the 30 econometric studies of alcohol advertising and consumption and drinking initiation done over the last 20 years support the claim that alcohol advertising leads to drinking initiation or increases total consumption;
- Virtually all of the studies of alcohol advertising exposure and recall fail to meet the minimal standards of science in that they are unable to warrant the integrity of their measurements. Further, their results, even when statistically significant, demonstrate only weak associations and fail to show causal connections between advertising and consumption, drinking initiation, and alcohol-related harm;
- Of the 17 cross-sectional and longitudinal studies of the effects of advertising restrictions and bans (the draconian measures supported by the BMJ) on drinking initiation and consumption, only three find that such measures have a statistically significant effect on either initiation or consumption. Further, the evidence from jurisdictions that have removed bans shows that consumption has not increased when advertising has resumed.
Since we are not funded by the alcohol industry, we will leave it to the BMJ to tell us where we have gone so horribly wrong.