• Advertising and marketing;
  • alcohol;
  • bans;
  • exposure;
  • longitudinal studies;
  • policy analysis

I must apologize for citing the wrong study in my paper. In referring to research on advertising expenditure and restrictions I had meant to cite Nelson [1]. Incidentally, I recommend researchers interested in this area to look at Dr Nelson's summary (Appendix Table A1) of the history of restrictions on alcohol advertising 1975–2000 in 17 OECD countries. This provides a good insight into the difficulties inherent in categorizing countries into binary codes (e.g. bans/no bans), given the complexity and range of restrictions in the measured advertising media. (Of course, increasingly in this research field we also have to take into account the unmeasured media: digital, including social media, place of purchase, branded merchandise and product placement, to name a few.) Dr Nelson refers to this complexity when he points out, on finding some unexpected positive relationships between advertising bans and consumption, that this might reflect ‘substitution toward permitted media and marketing methods, lack of enforcement and ease of evasion…’ ([1], p.814).

In his letter, Dr Nelson [2] relies on several of his own papers to critique longitudinal research with young people. I am grateful for the opportunity to highlight a problem in one of these publications [3] which came to my attention some time ago, but which I failed to address at that time. In this paper Dr Nelson cites my own research under the heading ‘Four Examples of Dissemination Bias’. He summarizes the results of this study [4] as showing that only three of 48 estimates of the effect of alcohol advertising exposure showed significant positive effects and critiques my claim that ‘supporting evidence has been found in a series of longitudinal analyses of data from a cohort of New Zealand teenagers’ ([5], p.473).

His summary of the findings in this analysis (a study of alcohol in the mass media) does not provide a useful overview. Of the 24 analyses relating to recall of commercial advertising, significant positive relationships were found between reports of advertisements by the boys at age 15 years and the quantities of beer they drank (average and maximum) at age 18 years. These findings, as we pointed out in the discussion, seemed to make good sense, as the newly introduced TV commercials were predominantly for beer (the beverage of choice for New Zealand men at the time) and aimed clearly at a male audience. The lack of a relationship with wine and spirits consumption among girls and with frequency of drinking was in line with the context and previous literature.

More serious, however, is Dr Nelson's omission in his paper of any mention of the other publications resulting from this longitudinal study, which were summarized and cited in the same paragraph of my paper as part of the ‘supporting evidence’ for a causal relationship. These analyses, carried out at later ages of the cohort, continued to find evidence of positive relationships, independent of earlier drinking behaviour and, using a measure of liking of alcohol advertising, stronger relationships [6, 7]. The omission of any mention of these studies in a paper on dissemination bias seems somewhat surprising.

Dr Nelson's final point is to critique the policy implications I draw from the case study of alcohol marketing. My suggestion that there is an urgent need to regulate the ubiquitous alcohol marketing to which young people are exposed relies on the evidence of an impact of marketing on recruitment of the young to become drinkers and of heavier drinking among those most exposed. Restricting alcohol marketing will have no adverse effects on citizens, either drinkers or non-drinkers. It is the producers of alcohol whose interests are affected, and they are engaged actively in maintaining an unregulated marketing environment (including funding research activity and consultancy). Nothing in Dr Nelson's letter has led me to believe that advocacy for regulation of alcohol marketing is other than evidence-based.

Declarations of interest

The author has no connection with the tobacco, alcohol, pharmaceutical or gaming industries or any body funded substantially by one of these organizations.

  • Sally Casswell

  • SHORE (Social and Health Outcomes Research and Evaluation), SHORE andWhariki Research Centre, School of Public Health, Massey University, PO Box 6137, Wellesley Street, Auckland 1141, New Zealand. Eߚmail:


  1. Top of page
  2. References