Critique 167: What determines a person’s belief that “alcohol is heart-healthy?” — 22 July 2015
Whitman IR, Pletcher MJ, Vittinghoff E, Imburgia KE, Maguire C, Bettencourt L, Sinha T, Parsnick T, Tison GH, Mulvanny CG, Olgin JE, Marcus GM. Perceptions, Information Sources, and Behavior Regarding Alcohol and Heart Health. Am J Cardiol 2015; pre-publication. http:// dx.doi.org./10-1016/j.amjcard/2015.05-029.
Authors´ Abstract
Despite the equipoise regarding alcohol’s cardiovascular effects and absence of relevant rigorous controlled trials, the lay press frequently portrays alcohol as “heart healthy.” The public perception of alcohol’s heart effects, the sources of those perceptions, and how they may affect behavior are unknown. We performed a cross-sectional analysis of data obtained from March 2013 to September 2014 from consecutive participants enrolled in the Health eHeart Study.
Of 5,582 participants, 1,707 (30%) viewed alcohol as heart healthy, 2,157 (39%) viewed it as unhealthy, and 1,718 (31%) were unsure. Of those reporting alcohol as heart healthy, 80% cited lay press as a source of their knowledge. After adjustment, older age (odds ratio 1.11), higher education (odds ratio 1.37), higher income (odds ratio 1.07), US residence (odds ratio 1.63), and coronary artery disease (odds ratio 1.51) were associated with perception of alcohol as heart healthy (all p <0.003). Ever smokers (odds ratio 0.76, p 0.004) and those with heart failure (odds ratio 0.5, p 0.01) were less likely to cite alcohol as heart healthy. Those perceiving alcohol as heart healthy consumed on average 47% more alcohol on a regular basis (95% confidence interval 27% to 66%, p <0.001).
In conclusion, of >5,000 consecutive Health eHeart participants, approximately 1/3 believed alcohol to be heart healthy, and the majority cited the lay press as the origin of that perception. Those with a perception of alcohol as heart healthy drink substantially more alcohol.
Forum Comments
Reviewer Barrett-Connor considered that “This paper was an interesting idea about where the public gets its information about health and risks of common behaviors and prescriptions. It reminds me of the first 30 years of ‘estrogen is good for you’ papers and books intended for a lay audience. In this case, the authors have interesting questions and interesting answers and a reasonable study design.
“I think that my biggest problem with the paper is there is no attempt to understand the different kinds of associations or to do multiply adjusted analyses. The actual number of people is still small; it would have been smart for them to wait another couple of years in order to get more people so that they could have a much better study.”Forum member Ellison had a number of concerns about the paper. “First, who are the people whose opinions are given in this survey? They are certainly not a random sample of the population. The ways that people were recruited are not described very well, and differences in response by recruitment techniques (which varied markedly among the participating centers) are not given. (It would be assumed that advertising in a wine magazine would give very different responses than if recruitment for the survey was in a conservative or evangelical paper.) Ellison also thought that “The leading question: ‘Do you believe alcohol is good for your heart?’ would be answered very differently from what Forum members considered to be a more appropriate question: ‘Do you believe that the moderate consumption of alcoholic beverages is good for your heart?’
“While the authors report that people who think alcohol is good for your heart tend to drink slightly more, I fail to find any details on the results of an addtional and important question: ‘How does your understanding of the potential health effects of alcohol influence the amount of alcohol you drink?’ For this question, the potential choices included statements that there is no influence, I drink more because of health reasons, and I drink less because of health reasons. It seems unusual that comments are not given in the paper for the responses to this question; if perceived health effects are only a minor reason for drinking, this would tend to undermine the arguments against giving what the authors consider to be ‘premature or even erroneous’ advice to the public.”
Reviewer Skovenborg commented that this article is interesting as not many studies of this type have been published. However, he pointed out that “The statement of the authors that the participants perceiving alcohol as heart healthy consumed 47% more alcohol may sound alarming; however, this real difference is from 0-3 to 1-5 glasses of wine/week and from 1-7 to 2-8 drinks/week. These are not ‘alarming’ differences, and all fall within usual guidelines for ‘sensible drinking.’”
Skovenborg continued: “Further, the authors have misrepresented the scientific literature when they suggest that ‘Even low to moderate levels of alcohol consumption have been shown to impose harm including a number of conditions.’ For example, they state that there was a ‘higher risk of mortality from <2 drinks/day,’ whereas the reference they quote (Gronbaek et al) actually states: ‘Compared with stable light drinkers, stable nondrinkers and heavy drinkers (>13 drinks/week) had similarly increased risks of dying (RR 1.29 and 1.32 respectively). There was little difference in mortality between light and moderate drinkers.’ And Gronbaek et al conclude: ‘Persons with stable patterns of light and moderate alcohol intake had the lowest all-cause mortality.’ Almost all other prospective studies have shown the same result.
“For their statement that atrial fibrillation increases at 1-2 drinks/day, the reference they quote (Conen et al) states: ‘Among healthy middle-aged women, consumption of up to 2 alcoholic beverages per day (15 – 30 g per day) was not associated with an increased risk of incident atrial fibrillation.’ And, for ‘increased risk of hypertension at 3/drinks/day,’ the reference they quote (Paulin et al) states: ‘The mean systolic and diastolic pressures of male heavy alcohol users (300 g or more per week = >43 g per day) were, respectively, 9.8 and 8.9 mmHg higher than those of male nondrinkers. No relationship between alcohol intake and blood pressure was found in the women.’” Skovenborg concludes: “I do not know if these quotes are examples of intentional or unintentional bias; however, I find it strange that so many authors have underwritten these statements that do not reflect current scientific knowledge and yet are presenting information to help doctors give advice to patients regarding alcohol and health.” Added Forum member Barrett-Connor: “I agree that the bias of study authors can lead to bias of the responders; this was never analyzed in this paper. Nor did the authors address the huge difference between heavy drinking and social drinking.”
Reviewer Ursini stated: “It is clear we are dealing with a hidden prohibitionism when authors conclude their abstract with a sentence (obviously not scientific) saying that if you believe that alcohol is good for your heart you drink more (and presumably should expect more health problems). What can be the relevance of this for the general population and the community of experts? I’m more convinced every day that there must be an ‘ethical commitment’ rather than just ‘scientific data’ behind many publications.”
Reviewer Barrett-Connor commented: “It is clear that from the way the text is written that the authors do not believe that alcohol is good for you. It would have been nice to have known how many of the participating authors think of alcohol as a drug that should be avoided, possibly altogether. Clearly there is a bias in the authors which I think would lead to respondents’ bias. This is the most problematic aspect for the understanding of this study; in other words, those who do not believe that alcohol can have benefits when used in moderate doses would presumably have been more likely to have responded to the questionnaire.
”However, I think the article was interesting in that it demonstrated where people get their medical information from—the lay press.” Added Forum member Mattivi: “It is not too surprising that people get information from the lay press. The population has little access to scientific/technical data that are published for ordinary folk, and are very unlikely or unable to read the full text from scientific journals. Even skilled researchers sometimes tend to be poorly informed about complex scientific issues, and some citations in this paper seem a good example of that. In the real scientific world (not that of the media), the truth is not black or white, or simple or comfortable.”
As pointed out by a number of reviewers, the Discussion in this paper suggests anti-alcohol sentiments of the authors which results in a gross misrepresentation of scientific evidence. The authors suggest that their survey results suggesting an ‘optimal level of ≤ 14 drinks/week’ (a level that was reported by 98% of their subjects) is too high. However, this level of drinking has been found in most studies to have the greatest health benefit, both in reducing heart diseases and overall mortality. In addition, this consumption rate is in accordance with the recommendations of most national agencies in North America and Europe.
Reviewer Ellison added: “Statements such as those suggesting that people who think it is heart healthy to consume alcohol are misinformed reinforces the clear impression that it is the authors who are apparently misinformed. Statements suggesting that perceptions that alcohol decreases heart disease are ‘premature or even eroneous’ and that such ‘misperceptions’ may affect behavior in an unhealthy manner further indicate that the authors are not basing thcir conclusions on sound scientific data.”
Forum member Goldfinger points out that “Different populations with different demographics may (obviously) present different perceptions on alcohol and health. Depending on the subjects recruited for such a survey, results could potentially present a dangerous misrepresentation of scientific data.” All Forum members agreed that data on this topic from a random sample of the population could be very informative.
References from Forum critique
Conen D, Tedrow UB, Cook NR, Moorthy MV, Buring JE, Albert CM. Alcohol consumption and risk of incident atrial fibrillation in women. JAMA 2008;300:2489e2496.
Gronbaek M, Johansen D, Becker U, Hein HO, Schnohr P, Jensen G, Vestbo J, Sørensen TI. Changes in alcohol intake and mortality: a longitudinal population-based study. Epidemiology 2004;15:222e228.
Paulin JM, Simpson FO, Waal-Manning HJ. Alcohol consumption and blood pressure in a New Zealand community study. N Z Med J 1985;98:425e428.
Forum Summary
This paper, based on data from more than 5,000 adults participating in an internet-based survey, sought to determine what were the perceptions of subjects on the relation of alcohol to heart disease, and how these perceptions resulted in particular behaviors related to alcohol consumption. Partricipants for the survey were recruited by a variety of methods, including announcements “in the lay press, promotional events, word of mouth, social media, e-mail, and clinic visits.” While the goal of the study is commendable, the results of any such survey obviously depend on many social and cultural factors of the people who respond to the survey; it was not a random sample of the population used in this analysis. This is critical because an interpretation of the data requires an understanding of the population surveyed.
Forum members had other concerns about this paper, including the specific survey questions regarding the perceptions of alcohol effects. For example, the key question asked in the survey was “Do you believe alcohol is good for your heart?” Forum members pointed out that it is obvious to nearly everyone that this cannot be answered without quantifying the amount of alcohol or the pattern of drinking. They consisdered that a more appropriate quesiton might be something such as “Do you believe that the moderate consumption of alcoholic beverages can benefit your heart?”
The authors make statements such as “Participants perceiving alcohol as heart healthy consumed 47% more alcohol.” This may sound alarming until it is revealed that the observed difference was from a baseline of 0-3 to 1-5 glasses of wine/week and from 1-7 to 2-8 total drinks/week. These higher amounts are not ‘alarming’ at all, and in fact fall within usual guidelines for “sensible drinking.”
Forum members were particularly upset that the authors misquote a number of scientific papers to support their claim that even low levels of drinking increase the risk of certain diseases. The authors also use such misrepresentations of data to justify their concerns that current guidelines for alcohol consumption may not be appropriate. Forum reviewers found it especially disturbing that the peer review process by the journal failed to uncover such misrepresentations of data.
The authors repeatedly give the impression that they are against any amount of alcohol consumption. They do not comment on the consistent findings in almost all well-done prospective studies indicating that moderate drinkers live longer than abstainers. Further, statements by the authors that suggest that people who think that alcohol is heart healthy are misinformed does not stand up to scientific data, which indicate that moderate and responsible drinking is associated wtith better health and longevity.
It is interesting that the report indicates that older, higher-income, and more highly educated subjects are more likely to consider alcohol to be heart healthy than younger and less-educated subjects. While a question was asked by the investigators as to whether or not the responder’s perception of the health effects of alcohol influences their decision to drink or to not drink, the specific responses to this question are not presented in the paper. It remains unclear to what extent beliefs about alcohol’s potential health effects relate to alcohol use or, more importantly, to alcohol abuse.
Overall, Forum members thought that the idea behind this study was of interest, but emphasized that it is important to collect data based on a random and well described sample of the population if one is to provide significant and understandable results. Further, the report is compromised by what appears to be a deliberate misrepresentation of the prior scientific literature to support the authors’ contention that alcohol is not healthy; there are a number of obvious errors that should have been detected during the review process by the journal. In the opinion of the Forum, the most reliable result from these analyses relates to where people obtain their information regarding alcohol and health: it does not come from physicians or scientific publications, but primarily from the lay press.
* * *
Comments on this paper have been provided by the following members of the International Scientific Forum on Alcohol Research:
Elizabeth Barrett-Connor, MD, Chief of the Division of Epidemiology, Distinguished Professor in the Departments of Family and Preventive Medicine & Medicine, University of California, San Diego, La Jolla, CA, USA
Giovanni de Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy
R. Curtis Ellison, MD. Section of Preventive Medicine & Epidemiology, Department of Medicine, Boston University School of Medicine, Boston, MA, USA
Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA
Tedd Goldfinger, DO, FACC, Desert Cardiology of Tucson Heart Center, University of Arizona School of Medicine, Tucson, AZ, USA
Maritha J. Kotze, PhD, Human Genetics, Dept of Pathology, University of Stellenbosch, Tygerberg, South Africa.
Dominique Lanzmann-Petithory,MD, PhD, Nutrition/Cardiology, Praticien Hospitalier Hôpital Emile Roux, Paris, France
Fulvio Mattivi, PhD, Head of the Department of Food Quality and Nutrition, Research and Innovation Centre, Fondazione Edmund Mach, in San Michele all’Adige, Italy
Erik Skovenborg, MD, specialized in family medicine, member of the Scandinavian Medical Alcohol Board, Aarhus, Denmark
Creina Stockley, PhD, MSc Clinical Pharmacology, MBA; Health and Regulatory Information Manager, Australian Wine Research Institute, Glen Osmond, South Australia, Australia.
Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway
Dag S. Thelle, MD, PhD, Senior Professor of Cardiovascular Epidemiology and Prevention, University of Gothenburg, Sweden; Senior Professor of Quantitative Medicine at the University of Oslo, Norway
Gordon Troup, MSc, DSc, School of Physics, Monash University, Victoria, Australia
Fulvio Ursini, MD, Dept. of Biological Chemistry, Universityof Padova, Padova, Italy
David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa
Andrew L. Waterhouse, PhD, Department of Viticulture and Enology, University of California, Davis, USA
Yuqing Zhang, MD, DSc, Clinical Epidemiology, Boston University School of Medicine, Boston, MA, USA