Critique 108: An unusual paper relating alcohol consumption to the risk of stroke — 9 April 2013
Rantakömi SH, Laukkanen JA, Sivenius J, Kauhanen J, Kurl S. Alcohol consumption and the risk of stroke among hypertensive and overweight men. J Neurol 2013;260:534–539. DOI 10.1007/s00415-012-6672-6
Authors’ Abstract
High blood pressure and overweight are risk factors for stroke. The aim of the present study was to examine the association between alcohol consumption and the risk of stroke according to the level of blood pressure and body weight. This study is a population-based sample of men with an average follow-up of 14.9 years from eastern Finland. A total of 2,599 men with no history of stroke at baseline participated. During the follow-up period, 224 strokes occurred, of which 181 were ischemic strokes.
After adjustment for age, year of examination, socioeconomic status, serum LDL cholesterol, body mass index, smoking and energy expenditure of physical activity (kcal/day), there was a significant trend of an increased risk for any and ischemic stroke among hypertensive men. Hypertensive (blood pressure of over 140/90 mm Hg) men, who did not consume alcohol had a 1.72-fold (95 % CI 1.12–2.66; p = 0.014) relative risk (RR) for any stroke and a 1.90-fold (95 % CI 1.15–3.13; p = 0.012) RR for ischemic stroke. Among hypertensive men who consumed alcohol RR was 1.86-fold (95 % CI 1.20–2.89; p = 0.005) for any stroke and 2.02-fold (95 % CI 1.21–3.35; p = 0.007) for ischemic stroke. Men who did not consume alcohol with elevated BMI (C26.4 kg/m2) had a 1.63-fold RR (95 % CI 1.11–2.40; p = 0.013) for any stroke and a 1.33-fold RR (95 % CI 0.87–2.04; p = 0.199) for ischemic stroke after adjusting for risk factors. Overweight men (C26.4 kg/m2) who consumed alcohol had a 1.73-fold RR (95 % CI 1.18–2.54; p = 0.005) for any stroke and a 1.71-fold RR (95 % CI 1.14–2.57; p = 0.010) for ischemic stroke after being adjusted for risk factors. In conclusion, this population-based prospective study shows that hypertensive and overweight men who consumed alcohol had an increased risk for stroke.
Forum Comments
Hypertension is a major risk factor for all types of stroke, including the most common type, ischemic stroke. Obesity is a major risk factor for hypertension, and may also increase the risk of stroke. Most prospective studies have shown that light-to-moderate alcohol consumption is associated with either no increase or only a slight increase in blood pressure, while heavy drinking is a clearly associated with an increase in blood pressure and an increased risk of developing “hypertension.”
In terms of alcohol and stroke, heavy drinking is known to increase the risk of all types of stroke, but almost all prospective studies have shown that light-to-moderate alcohol is associated with a significant decrease in the risk of ischemic stroke. The present paper was undertaken to study the association between alcohol and stroke risk among a population-based sample of middle-aged men in Finland according to whether or not they were obese or hypertensive at baseline.
Specific comments on the present paper: These investigators grouped subjects according to whether or not they drink any alcohol at all, classifying subjects into alcohol consumers or no-alcohol consumers. Given that there is a dose-response ratio between alcohol consumption and stroke, it is baffling why all subjects consuming any alcohol were grouped together. Further, while data were available on some aspects of drinking pattern, the only reference to this is in the last paragraph of the Results, where it is stated that hypertensive binge drinkers had a 66% increased risk of stroke compared with hypertensive drinkers who did not binge, and a similar effect was seen for obese binge drinkers. However, such information is not included in the tables or discussed elsewhere. Specifically, there are no data given that describe the relation of light-to-moderate alcohol consumption with stroke. Also, while data were available on the type of beverage consumed (beer, wine, fortified wine, spirits), no results are presented on potential differential effects according to type of beverage.
The authors conclude that “this population-based prospective study shows that hypertensive and overweight men who consumed alcohol had an increased risk for stroke.” In the opinion of members of the Forum, without reporting the effects according to the amount of alcohol consumed, or the pattern of drinking, this is an unsupported and potentially dangerous statement to present in a scientific journal.
Reviewer Waterhouse commented: “Aside from the lack of drinking categories, it seems that their own data do not show statistical significance for any non-drinker/drinker comparison — in many cases the p values don’t make the appropriate comparisons so it is not possible to question all their points. So, it is hard to see how they could arrive at their conclusion. I think we have to ask what is motivating what appears to be a deliberate misinterpretation of data.” This impression was shared by Forum Reviewer Finkel: “This paper’s results make sense to me only if most of the drinking subjects drank too much, and probably in big gulps. But, shame on the authors and editors for not supporting conclusions with data, assuming they have the data.”
Reviewer Ursini stated: “Usually I refrain from commenting papers on topics that are not in my field of scientific experience. Not this time. It seems clear — even to a non-expert — that this is a badly conducted study. The disappointing feeling is that the reason for publishing this message seems to be different from sound scientific communication.”
Reviewer Skovenborg stated that he was “frustrated and disappointed” to read the paper on alcohol consumption and the risk of stroke. He added: “Reasons for my dissatisfaction are as follows:
(1) In the primary results alcohol is treated as a dichotomous variable, rather than a continuous variable, which is very unusual.
(2) Separate results are reported for binge drinkers; however, no data are presented regarding the proportion of binge drinkers and the alcohol intake and drinking pattern of the group of binge drinkers.
(3) According to the baseline characteristics, the participants are indeed moderate drinkers as a whole with a weekly alcohol intake of 76.7 grams of alcohol per week. (However, the percentage of non-drinkers in this population is not given.) No stratification data of drinkers are presented.
(4) According to the baseline characteristics, the participants are indeed heavy smokers, with an overall average of 173.9 pack-years. So many cigarettes and so few drinks is a highly unusual combination in any population.
(5) Disregarding their own data, the authors make conclusions referring to ‘hypertensive and overweight men who consume alcohol.’ The correct conclusion may rather be that alcohol consumption was not associated with a decreased risk among hypertensive and obese Finnish men, and that binge drinking (6+ drinks per day) increased the risk of stroke among hypertensive men but not among obese men.”
Forum reviewer Keil agreed with other reviewers that this was a poor paper and “should not have been published.” Keil had some interesting further comments: “In my opinion, by far too many manuscripts are being published nowadays and there is an increasing inflation of published papers. In the alcohol field a lot of confusion has been produced in recent years by rather weak or poor papers. Ten years ago scientists and the educated public had a clearer view of the benefits and harms of alcohol than today. Today the public gets confused by being told that the smallest amount of alcohol causes harm.”
Forum reviewer Lanzmann-Petithory adds: ““I agree with all of these comments. The incidence of stroke during the follow up in this study is curiously high, almost 10% cases. The paper lacks a lot of information: there is no definition of what is a drinker or a non-drinker, and only partial allusion to binge drinking. I am surprised that such a paper could be accepted by a respected journal.”
Forum Summary
Numerous well-done prospective epidemiologic studies have shown that while heavy drinking may increase the risk of stroke, light-to-moderate alcohol consumption is usually associated with a significant decrease in the risk of the most common type of stroke, ischemic stroke.
This paper from investigators in Finland states that it was designed to relate alcohol consumption and obesity to the risk of ischemic stroke. Unfortunately, even though it has clearly been shown that the amount and pattern of alcohol consumption strongly modify the effect on the risk of stroke, for some unknown reason the authors have grouped all consumers of alcohol into a single group. Specifically, there are no data given that describe the relation of light-to-moderate alcohol consumption with stroke. Hence, the conclusion of the authors that “alcohol increases the risk of stroke” is not supported by the data presented. In the opinion of members of the Forum, without reporting the effects according to the amount of alcohol consumed or the pattern of drinking, this is an unsupported and potentially dangerous statement to present in a scientific journal.
* * *
Comments on this paper were provided by the following members of the International Scientific Forum on Alcohol Research:
Erik Skovenborg, MD, Scandinavian Medical Alcohol Board, Practitioner, Aarhus, Denmark
R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA
Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway
Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA
Fulvio Ursini, MD, Dept. of Biological Chemistry, University of Padova, Padova, Italy;
Andrew L. Waterhouse, PhD, Marvin Sands Professor, Department of Viticulture and Enology, University of California, Davis; Davis, CA, USA
Ulrich Keil, MD, PhD, Institute of Epidemiology and Social Medicine, University of Münster, Münster, Germany
Dominique Lanzmann-Petithory,MD, PhD, Nutrition/Cardiology, Praticien Hospitalier Hôpital Emile Roux, Paris, France