Critique 163: The association of alcohol consumption with the risk of developing heart failure — 5 May 2015

This critique comments on two recent papers on alcohol and heart failure.  The authors’ abstracts of the two papers are given below:

(1) Larsson SC, Orsini N, Wolk A.  Alcohol consumption and risk of heart failure: a dose–response meta-analysis of prospective studies.  European Journal of Heart Failure 2015;17:367–373.  doi:10.1002/ejhf.228

Authors’ Abstract

Aims  The aim of this study was to conduct a meta-analysis of prospective studies assessing the relationship between alcohol consumption and risk of heart failure (HF).

Methods and results  We searched the PubMed database from inception to September 2014 and reviewed the reference list of relevant articles to identify prospective studies assessing the association between alcohol consumption and risk of HF. Study-specific relative risk (RR) estimates were combined using a random-effects meta-analysis. The meta-analysis included eight prospective studies, with a total of 202 378 participants and 6211 cases of HF. The pooled adjusted RRs of HF were 0.85 [95% confidence interval (CI) 0.78–0.93] for light to moderate alcohol consumption (<14 drinks/week) and 0.90 (95% CI 0.72–1.13) for high alcohol consumption (≥14 drinks/week) compared with non-drinkers. In a dose–response meta-analysis, we observed a non-linear relationship between alcohol consumption and risk of HF (P for non-linearity=0.001). Compared with non-drinkers, the RRs (95% CI) across levels of alcohol  consumption were 0.90 (0.84–0.96) for 3 drinks/week, 0.83 (0.73–0.95) for 7 drinks/week, 0.84 (0.72–0.98) for 10 drinks/week, 0.90 (0.73–1.10) for 14 drinks/week, and 1.07 (0.77–1.48) for 21 drinks/week.

Conclusion Alcohol consumption in moderation is associated with a reduced risk of HF.

 

(2).  Dorans KS, Mostofsky E, Levitan EB, Håkansson N, Wolk A, Mittleman MA.  Alcohol and Incident Heart Failure Among Middle-Aged and Elderly Men: The Cohort of Swedish Men.  Circulation Heart Failure 2015; pre-publication.  DOI:10.1161/CIRCHEARTFAILURE.114.001787.

Authors’ Abstract

Background—Compared with no alcohol consumption, heavy alcohol intake is associated with a higher rate of heart failure (HF) whereas light-to-moderate intake may be associated with a lower rate. However, several prior studies did not exclude former drinkers, who may have changed alcohol consumption in response to diagnosis. This study aimed to investigate the association between alcohol intake and incident HF.

Methods and Results—We conducted a prospective cohort study of 33,760 men 45–79 years old with no HF, diabetes mellitus or myocardial infarction at baseline participating in the Cohort of Swedish Men Study. We excluded former drinkers. At baseline, participants completed a food-frequency questionnaire and reported other characteristics. HF was defined as hospitalization for or death from HF, ascertained by Swedish inpatient and cause-of-death records from January 1, 1998 through December 31, 2011. We constructed Cox proportional hazards models to estimate multivariable-adjusted rate ratios (IRRs). During follow-up, 2916 men were hospitalized for (n=2139) or died (n=777) of incident HF. There was a U-shaped relationship between total alcohol intake and incident HF (p=0.0004). There was a nadir at light-to-moderate alcohol intake: consuming 7 to less than 14 standard drinks per week was associated with a 19% lower multivariable-adjusted rate of HF compared with never drinking (IRR: 0.81, 95% CI: 0.69, 0.96).

Conclusions—In this cohort of Swedish men, there was a U-shaped relationship between alcohol consumption and HF incidence, with a adir at light-to-moderate intake. Heavy intake did not appear protective.

Forum Comments

The association between alcohol intake and the risk of developing heart failure (HF) is unclear, but most cohort studies have shown a lower risk of HF among moderate drinkers but a possible higher risk among heavy drinkers.  The recent meta-analysis included in this critique (Larsson et al) was based on eight prospective studies, with more than 200,000 subjects and 6,211 cases of HF.  Those authors report that “In a dose–response meta-analysis, we observed a non-linear relationship between alcohol consumption and risk of HF (P for non-linearity=0.001).  Compared with non-drinkers, the RRs (95% CI) across levels of alcohol consumption were 0.90 (0.84–0.96) for 3 drinks/week, 0.83 (0.73–0.95) for 7 drinks/week, 0.84 (0.72–0.98) for 10 drinks/week, 0.90 (0.73–1.10) for 14 drinks/week, and 1.07 (0.77–1.48) for 21 drinks/week.”

Meta-analyses provide important information because they are usually based on very large numbers of subjects and cases, and can detect smaller increases in risk than seen in individual studies.  However, single large prospective cohort studies tend to provide more detailed information on, and allow better adjustment for, potential confounders.  The present paper from Sweden also reviewed in this critique (Dorans et al) is based on 33,760 men who were followed for 14 years, with almost 3,000 cases of fatal or non-fatal HF.

There are a number of strengths of both papers, the meta-analysis because of its large size and breath of coverage, and the cohort report which had consistent data on a number of risk factors that may have confounded the results.  However, the overall results are very similar in showing a reduction in HF from moderate alcohol consumption.

Specific comments on Larsson et al:  The first paper, the meta-analysis, used appropriate techniques to summarize data from eight prospective studies that included more than 200,000 subjects with 6,211 incident cases of HF.  The degree of control for potential confounding varied among the studies included in the meta-analysis, but the authors chose the most fully adjusted risk ratios provided by the original authors.

This paper included a two-stage random-effects dose-response analysis, using a restricted cubic spline analysis to judge for non-linear effects.  The investigators also pooled the study-specific RRs from all of the studies.  Their key findings were reduced risks of HF for consumers of 3 and 10 drinks/week (a 10% and 17% decreased risk, respectively) and a RR of 1.07 (95% CI 0.77 – 1.48) for consumption of 21 drinks/week.  The authors note that their analyses cannot judge the effect on the risk of HF among alcoholics or other heavy drinkers.

Specific comments on Dorans et al:  The second study is based on a population-based cohort from central Sweden with excellent follow up for the development of HF; for example, more than 99% of inpatients in Sweden are included in the dataset used for the ascertainment of HF.  The investigators excluded ex-drinkers from the study, so that they would not potentially influence the association by including subjects who may have stopped drinking because of certain diseases or advice from their physicians.

This cohort study had an adequate number of never drinkers (n=1,396) to use as a comparison group.  The analyses were well-done, with appropriate sensitivity analyses.  Further, these investigators had data permitting adjustments for potential confounders, including age, BMI, physical activity, education, smoking, marital status, family history of premature myocardial infarction, history of hypertension and hypercholesterolemia, and diet quality, with the latter assessed as degree of compliance with a DASH diet (Levitan et al).  The authors carried out chart reviews of reported cases, with confirmation of 95% of suspected cases.  They also had data on specific types of alcohol consumed by subjects.

A weakness, as acknowledged by the authors, was the lack of information on the pattern of drinking (regular versus binge).  In beverage-specific analyses, there were similar reductions for consumers of beer, wine, and spirits, although there was a tendency for a slightly greater reduction in risk for consumers of beer, the most common beverage in this cohort.  However, for beverage-specific analyses, the authors included non-drinkers in their comparisons with consumers of each type of beverage, so that subjects not consuming a specific beverage included drinkers of other beverages plus non-drinkers, making it somewhat difficult to judge the results.

The results show that drinkers, in comparison with never-drinkers, tended to have a higher family history of heart disease, more hypertension and high cholesterol, much higher rates of smoking, less physical activity, and a poorer diet (all of which should have increased their risk of HF).  However, the risk of HF for most categories of drinkers tended to be lower than that of non-drinkers, with the lowest risk among consumers of 7 – 14 drinks/week (HR=0.81, 95% CI 0.69 – 0.96).  [This estimate is very similar to that shown in the meta-analysis by Larsson et al: 0.83 (0.73–0.95) for 7 drinks/week and 0.84 (0.72–0.98) for 10 drinks/week.  In both the present study and the meta-analysis, drinkers in the highest category of drinking had an estimated HR greater than 1.00, but the differences from non-drinkers was not statistically significant.

Forum member Finkel commented: “This study seems to me well done and of adequate power.  It confirms the previously determined protection against heart failure by light-moderate alcohol consumption in a U-shaped relationship, and counters the age-old objection that reported benefits of moderate consumption were based on ‘sick quitters.’”

Overall comments on alcohol and HF:  Forum member Thelle raised some important points on the subject: “My main concern with these papers is that heart failure covers a plethora of mechanisms and underlying diseases, among which coronary heart disease is the most important.  Other causes are high blood pressure, valve disorders, cardiomyopathy, myocarditis and arrhythmias.  The U-shaped curve could be due to the association with coronary heart disease, that the authors comment upon.  But as the cohort paper excluded patients with previous known myocardial infarction, this would seem less likely.  What then caused heart failure in these subjects: a non-coronary factor affecting the myocardium, and likely to be prevented by alcohol?  Or, are the HF cases un-diagnosed or newly diagnosed coronary patients where HF is a manifestation?”

Arthur Klatsky, a member of this Forum, published an editorial accompanying the paper by Larsson et al.  He pointed out that key issues on studies of alcohol and HF include the following: (1) a realization that alcohol-health relationships are non linear; (2) results of studies may be affected by whether or not ex-drinkers are included in the reference group; (3) there have been no randomized trials of alcohol and major health outcomes, such as HF; (4) it is difficult to judge results from different types of beverage because of confounding by other lifestyle factors among consumers of a particular beverage; (5) in many studies of HF, only patients who are hospitalized or patients with HF who die are included, so milder cases are not evaluated; (6) it is difficult to separate effects of alcohol on the development of coronary heart disease (the underlying factor for much HF) and effects directly on the myocardium; (7) data support increased risk of many diseases from heavy drinking, but moderate drinking may decrease risk for middle-aged and older adults; thus, advice should be individualized.”  Klatsky also commented that “Since there is some presumed misclassification of intake by under-reporting, the inverse alcohol-HF association may be stronger than the data show.”

Cardiologist and Forum member Goldfinger agreed with the comments of Klatsky.  “As always, I appreciate Arthur’s editorializing and stated positions on the role of moderate alcohol consumption.  His comment ‘HF is pluricausal, with multiple risk factors and not a unitary cause’ is most important when reviewing these papers.  HF is a syndrome and not a disease.  It may result from coronary artery disease and multi-infarct/ischemic cardiomyopathy, viral or toxic cardiomyopathy, congenital heart disease, valvular heart disease, pericardial disease, volume overload syndromes, diastolic dysfunction of aging, etc.

“Certainly moderate alcohol consumption has a robust role in prevention of atherosclerotic cardiovascular disease (ASHD) and thus its consequences, such as HF.  And, ASHD is the number 1 cause of heart failure, thus a beneficial effect, as has been shown in both papers, is of little surprise and to be expected.  Moderate alcohol drinking, particularly wine drinking, has been associated with an overall healthier lifestyle, and thus may be associated with a reduced risk of infection, exposure, and healthier aging.  This has impact on the likelihood of developing acquired conditions that may lead ultimately to the syndrome of HF.

“Another critical consideration is how the HF patients in these papers were identified: almost exclusively by hospitalizations.  The precipitating cause of decompensation needs to be considered, e.g., infection (particularly respiratory and urinary tract infections in the elderly), dietary indiscretion, excessive salt intake, non-compliance with chronic medication, etc.  This certainly complicates the analysis, as a great many patients with chronic heart failure are not hospitalized.  Several studies have linked modest alcohol with a decreased risk of infections (e.g., Kamholz), so, an important effect may be on the triggers of decompensation in a more expansive stable population, rather than an effect on the instigation of the disease that caused it in the first place.  In general, the results/consensus are consistent with the fund of knowledge we currently have, that is, safe and moderate alcohol consumption is associated with a healthier life and one that is more likely free of disease and disability.”

References from Forum review

Dorans KS, Mostofsky E, Levitan EB, Håkansson N, Wolk A, Mittleman MA.  Alcohol and Incident Heart Failure Among Middle-Aged and Elderly Men: The Cohort of Swedish Men.  Circulation Heart Failure 2015; pre-publication.  DOI:10.1161/CIRCHEARTFAILURE.114.001787

Klatsky AL.  Editorial: Alcohol drinking and heart failure: where do we stand? European Journal of Heart Failure 2015;17:348-350.

Kamholz SL.  Wine, spirits and the lung: Good, bad or indifferent?  Trans Am Clinical and Climatological Assoc 2006;117:129-145.

Larsson SC, Orsini N, Wolk A.  Alcohol consumption and risk of heart failure:a dose–response meta-analysis of prospective studies.  European Journal of Heart Failure 2015;17:367–373. doi:10.1002/ejhf.228.

Levitan EB, Wolk A, Mittleman MA.  Relation of consistency with the dietary approaches to stop hypertension diet and incidence of heart failure in men aged 45 to 79 years.  Am J Cardiol 2009;104:1416-1420. 

*                     *                      *                      *                      *

Forum Summary

Two new papers have provided data on the association of alcohol consumption with the risk of developing heart failure (HF).  The first is a meta-analysis by Larsson et al that is based on eight prospective studies, with more than 200,000 subjects and 6,211 cases of HF.  Meta-analyses provide important information because they are usually based on very large numbers of subjects and cases, and can detect smaller increases in risk than seen in individual studies.  However, single large prospective cohort studies tend to provide more detailed information on, and allow better adjustment for, potential confounders.  The second paper discussed in this critique by Dorans et al, is based on 33,760 men who were followed for 14 years, with almost 3,000 cases of fatal or non-fatal HF.

There was a high degree of consistency between these two studies: both concluded that moderate alcohol consumption, in comparison with non-drinking, is associated with a lower risk of developing HF.  The meta-analysis (Larsson et al) concluded that there was a dose-response association between alcohol and HF; for example, a reduced risks of HF for consumers of 3 and 10 drinks/week (a 10% and 17% decreased risk, respectively), but not for those reporting the consumption of 21 drinks/week (a RR of 1.07; 95% CI 0.77 – 1.48).  In the report from the single large Swedish prospective study (Dorans et al), the risk of HF for most categories of drinkers tended to be lower than that of never drinkers, with the lowest risk among consumers of 7 – 14 drinks/week (HR=0.81, 95% CI 0.69 – 0.96).

Forum members considered both of these studies to have been very well done, and the similar results not unexpected.  HF has many causes, but in developed countries the most common is coronary artery disease (CAD), and moderate drinking has long been known to be associated with a lower risk of developing CAD.  In their critique, Forum members discuss other conditions that relate to the development of HF, including conditions (such as respiratory infection, excessive salt intake, non-compliance with prescribed medication, etc.) that may trigger an attack of symptomatic HF.

To summarize, based on previous research and these two excellent papers, current data suggest that the moderate intake of an alcoholic beverage lowers the risk of the development of symptomatic HF.  This undoubtedly relates, as least to some extent, to the demonstrated protection of alcoholic beverages against CAD.  As for giving advice regarding the consumption of alcohol, Forum members agree with the conclusions stated by Klatsky in his editorial accompanying one of the papers: “All persons should avoid heavy drinking and many persons should avoid all alcohol.  However, it is the author’s opinion that many middle-aged and older persons at risk of CAD or HF should be told that he or she is better off as a light to moderate drinker.  Dispassionate objectivity plus common sense should dictate advice about drinking.”

*                     *                      *                      *                      *

Comments on these publications were provided by the following members of the International Scientific Forum on Alcohol Research:

Andrew L. Waterhouse, PhD, Marvin Sands Professor, Department of Viticulture and Enology, University of California, Davis; Davis, CA, USA

David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa

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

Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway

Ian Puddey, MD, Professor, Faculty of Medicine, Dentistry & Health Sciences, The University of Western Australia, Nedlands, Australia

Tedd Goldfinger, DO, FACC, Desert Cardiology of Tucson Heart Center, Dept. of Cardiology, University of Arizona School of Medicine, Tucson, Arizona, USA

Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA

R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA