Critique 205: A very large population-based study of the association of alcohol consumption with total and disease-specific mortality – 4 September 2017
Xi B, Veeranki SP, Zhao M, Ma C, Yan Y, Mi J. Relationship of Alcohol Consumption to All-Cause, Cardiovascular, and Cancer-Related Mortality in U.S. Adults. J Am Coll Cardiol 2017;70:913–922
Authors’ Abstract
BACKGROUND Previous studies have revealed inconsistent findings regarding the association of light to moderate alcohol consumption with cardiovascular disease (CVD) and cancer mortality.
OBJECTIVES The aim of this study was to examine the association between alcohol consumption and risk of mortality from all causes, cancer, and CVD in U.S. adults.
METHODS Data were obtained by linking 13 waves of the National Health Interview Surveys (1997 to 2009) to the National Death Index records through December 31, 2011. A total of 333,247 participants ≥18 years of age were included. Self-reported alcohol consumption patterns were categorized into 6 groups: lifetime abstainers; lifetime infrequent drinkers; former drinkers; and current light, moderate, or heavy drinkers. Secondary exposure included participants’ binge-drinking status. The main outcome was all-cause, cancer, or CVD mortality.
RESULTS After a median follow-up of 8.2 years (2.7 million person-years), 34,754 participants died of all causes (including 8,947 CVD deaths and 8,427 cancer deaths). Compared with lifetime abstainers, those who were light or moderate alcohol consumers were at a reduced risk of mortality for all causes (light—hazard ratio [HR]: 0.79; 95% confidence interval [CI]: 0.76 to 0.82; moderate—HR: 0.78; 95% CI: 0.74 to 0.82) and CVD (light—HR: 0.74; 95% CI: 0.69 to 0.80; moderate—HR: 0.71; 95% CI: 0.64 to 0.78), respectively. In contrast, there was a significantly increased risk of mortality for all causes (HR: 1.11; 95% CI: 1.04 to 1.19) and cancer (HR: 1.27; 95% CI: 1.13 to 1.42) in adults with heavy alcohol consumption. Binge drinking ≥1 d/week was also associated with an increased risk of mortality for all causes (HR: 1.13; 95% CI: 1.04 to 1.23) and cancer (HR: 1.22; 95% CI: 1.05 to 1.41).
CONCLUSIONS Light and moderate alcohol intake might have a protective effect on all-cause and CVD-specific mortality in U.S. adults. Heavy or binge drinking was associated with increased risk of all-cause and cancer-specific mortality.
NOTE: Included in the same issue of the the journal is an EDITORIAL COMMENT that relates to the Xi et al study: De Gaetano G, Costanzo S. Alcohol and Health. Praise of the J Curves. JACC 2017;70:923-925.
(As Professor de Gaetano is a member of the International Scientific Forum on Alcohol Research, Forum members have decided to not prepare a formal critique of this Editorial Comment to avoid potential conflict of interest. Readers are encouraged to read it for an overview of the J-shaped association between alcohol consumption and mortality.)
Forum Comments (on the publication by Xi et al)
Almost all longitudinal cohort studies have shown that light-to-moderate consumers of alcohol have a lower risk of total mortality and greater longevity of life (e.g., di Castelnuovo et al, Streppel et al, Holahan et al, Ronksley et al, Midlöv et al). However, the present analysis is important as it presents data on the relation of alcohol intake to total mortality as well as to specific mortality from cardiovascular disease (CVD) and cancer for a very large number of subjects in the USA. It is based on 13 waves of National Health Interview Surveys (1997-2009) incorporating data from more than 300,000 subjects, of whom almost 25,000 died during a follow-up period averaging 8.2 years.
The authors conclude that their analysis shows a very clear J-shaped curve for the relation of alcohol to mortality, with lower total, cardiovascular, and even cancer mortality rates for light and moderate drinkers who do not binge drink, with increased total mortality and cancer mortality for those classified as “heavy” drinkers.
Specific comments of Forum members: Most Forum members agreed that this study had a number of strengths: (1) It is based on a very large sample of the USA population, and very few of subjects (8%) had missing data; (2) it had an adequate number of lifetime abstainers (more than 76,000) to use as an appropriate referent group; (3) there were reasonable estimates of exposure (both amount and frequency, and binge-drinking, for alcohol consumption) and for mortality outcomes (using the National Death Index); (4) the analyses were well done, with the ability of the investigators to adjust for a large number of demographic and lifestyle factors as potential confounders; (5) there were adjustments for previous disease and analyses with a 2-year lag-time for outcomes for those reporting chronic diseases (CVD, cancer, diabetes, etc.); these analyses showed no effect of such on outcomes (discrediting worries about “sick quitters”).
Potential weaknesses were also noted: (1) No beverage-specific data were provided, and numerous studies have shown that the beneficial health effects of alcohol are greater with the consumption of wine than of other beverages (e.g., Grønbæk et al, Streppel et al); (2) in their main analysis, the authors included everyone aged 18 years or older, whereas it would have been preferable to see results separately for young and older adults (as effects would be expected only in the latter and the associations might have been more precise); the authors do state in their Results that there was no effect on mortality for adults aged 18-39 years; (3) the investigators classified “heavy” drinking as > 7 drinks/week (women and older men) or >14 drinks/week (younger men), but did not give the distribution of alcohol consumption among subjects in this category; (4) as the authors acknowledge, there remains a real risk for residual confounding, as there were many differences in baseline characteristics between the lifetime abstainers and the subjects in all drinking categories, and even multivariate regression analysis may be inadequate to truly adjust for such factors. Also, while the investigators adjusted for educational level, reviewer McEvoy thought that there may not have been adequate adjustment for socioeconomic status/access to healthcare; these are important modifying factors for the relation of alcohol to health, and could underlie some of the race/ethnicity differences observed.
Reviewer Ellison noted: “It is interesting to me that the lifetime abstainers and the lifetime infrequent drinkers had essentially identical results on mortality; this suggests that valid results can result from using either category, or a combination of both, as a referent group. Further, the ‘light’ and ‘moderate’ drinkers also had essentially the same results. Overall, ‘heavy’ drinking had no significant effects on CVD outcomes but increased cancer risk, but it is not known what the distribution of alcohol intake was in this highest consumption category. Many studies have shown that subjects whose intake is only slightly greater than the guidelines still have a significantly lower risk of death than abstainers (as pointed out by di Castelnuovo et al in their large meta-analysis). It would have been useful to know how many subjects in the highest category in the present study had a defined alcohol use disorder.”
Forum member Skovenborg agreed, writing: “It is unfortunate that the heavy drinking category is defined as women drinking 7+ drinks per week, etc. In Denmark the limit of sensible drinking is 14 drinks/week (12 g alcohol per drink) for women and 21 drinks/week for men. It would have been interesting to include a category of 14-21 drinks/week for men and 7-14 drinks/week for women to study the effects of such rather moderate drinking practices.”
Forum member Mattivi agreed that additional information on what were classified as “heavy” drinkers would have been useful: “In general, I fully subscribe to the several positive comments on this study, which provides data on a very large population. However, I would have preferred to see a more detailed classification of alcohol consumers. In particular, while on the one hand the study classifies (with good reasons) 3 different categories of non-drinkers (lifetime abstainers, lifetime infrequent drinkers, and former drinkers), on the other it compresses consumers into only three categories (light, moderate and heavy). It would have been interesting to have an additional intermediate category between moderate drinkers and heavy drinkers. It is likely that the inclusion of subjects, especially women, still in a range of putative protective consumption (7-14/week), led to an underestimation of the risks for strong drinkers. Could this possibly be the reason why heavy drinking was not significantly associated with risk of all-cause and cancer-specific mortality in women in this study?”
Forum member Finkel commented: “A well-done meta-study, reasserting some truths we’ve come to believe are self-evident but also with additional virtues, notably clean methodology and very large study group. It is even able to address to some extent the adversities of binge drinking. I again would quarrel with the definitions of categories of drinking. These investigators, for example, classify many sober, healthy, and responsible individuals as ‘heavy’ drinkers.”
Skovenborg also noted: “I find it disturbing that light or moderate drinking was not associated with a lower risk of mortality in black subjects. The explanations offered by the authors to explain that finding are not convincing and the racial difference is poorly understood and should be examined in more detail in order to learn what biological pathways might be the cause of this difference. Also, the important question of possible underreporting is not addressed by the authors. Further, there is no plausible biological pathway that would explain why alcohol would reduce the risk of cancer as far as I know, so the finding of an association between light drinking and a reduced risk of cancer mortality might be an indication of residual confounding. I do agree, however, that the possible risk of cancer with even light drinking has been overstated by WHO and other health organizations.” Reviewer Ellison suggested that “Some of the deaths attributed to cancer may have actually been from cardiovascular disease, rather than the cancer itself. If so, a reduced risk of such deaths associated with alcohol might have been more appropriately attributed to CVD rather than to cancer.”
Ellison also pointed out the findings of strong adverse effects of binge drinking: “The study supports the findings of Mukamal et al and many others showing that binge drinking negates the beneficial effects of moderate drinking on CVD. Overall, this excellent paper supports most previous research indicating a J-shaped curve for total mortality and CVD mortality, with little effect of light-to-moderate alcohol intake on cancer mortality but an increase in risk for what were defined as heavy drinkers. The finding of an actual decrease in cancer deaths from light drinking is surprising, but has been reported earlier by others (Grønbæk et al, Breslow et al, Jin et al). The often heard statement that any alcohol consumption increases the risk of cancer may be an exaggeration of the true relation. Instead, the emphasis might better be on focused on preventing excessive drinking.”
Reviewer Estruch commented; “I fully agree with the comments made by the other forum members. I only want to add the importance of the significant reduction in cancer mortality observed in light drinkers compared to lifetime abstainers. The relationship between alcohol intake and cancer continues to be a matter of debate and the results of this large study reinforce the protective effects of light and moderate alcohol consumption on cancer mortality. Unfortunately, dietary habits (other than alcohol consumption) have not been included as covariates in the statistical analysis and consequently the authors could not examine the possible different effects of light and moderate alcohol consumption on cancer mortality depending on the ‘quality’ of the dietary pattern of subjects.”
Forum member Van Velden stressed that we have to take into consideration the genetic background of individuals. “Although a growing body of evidence suggests that light to moderate consumption of antioxidant-rich alcoholic beverages such as red wine may decrease cumulative cardio-metabolic risk by targeting oxidative stress, not everybody may benefit. The benefit depends on the genetic background such as Apoe and HFE single nucleotide polymorphisms (SNP). It is anticipated that insights gathered as a result of research in this field may ultimately help inform public healthcare policy by providing a new approach to promoting safe limits for alcohol consumption in the clinical domain.”
References from Forum critique
Breslow RA, Chen CM, Graubard BI, Mukamal KJ. Prospective Study of Alcohol Consumption Quantity and Frequency and Cancer-Specific Mortality in the US Population. Am J Epidemiol 2011;174:1044-1053. doi: 10.1093/aje/kwr210.
Di Castelnuovo A, Costanzo S. Bagnardi V, Donati MB, Iacoviello L, de Gaetano G. Alcohol Dosing and Total Mortality in Men and Women: An Updated Meta-analysis of 34 Prospective Studies. Arch Intern Med 2006;166:2437-2445. doi:10.1001/archinte.166.22.2437
Grønbæk M, Becker U, Johansen D, Gottschau A, Schnohr P, Hein HO, Jensen G, Sørensen TIA. Type of Alcohol Consumed and Mortality from All Causes, Coronary Heart Disease, and Cancer. Ann Intern Med 2000;133:411-419.
Holahan CJ, Schutte KK, Brennan PL, Holahan CK, Moos BS, Moos RH. Late-Life Alcohol Consumption and 20-Year Mortality. Alcohol Clin Exp Res 2010;34:1961-1771. doi: 10.1111/j.1530-0277.2010.01286.x
Jin M, Cai S, Guo J, Zhu Y, Li M, Yu Y, Zhang S, Chen K. Alcohol drinking and all cancer mortality: a meta-analysis. Ann Oncol 2013;24:807-816. doi: 10.1093/annonc/mds508.
Midlöv P, Calling S, Memon AA, Sundquist J, Sundquist K, Johansson S-E. Women’s health in the Lund area (WHILA) – Alcohol consumption and all-cause mortality among women – a 17 year follow-up study. BMC Public Health 2016;16:22. DOI 10.1186/s12889-016-2700-2
Mukamal KM, Maclure M, Muller JE, Mittleman MA. Binge Drinking and Mortality After Acute Myocardial Infarction. Circulation 2005;112:3839-3845.
Ronksley PE, Brien SE, Turner BJ, Mukamal KJ, Ghali WA. Association of alcohol consumption with selected cardiovascular disease outcomes: a systematic review and meta-analysis. BMJ 2011;342:d671; doi:10.1136/bmj.d671
Streppel MT, Ocké MC, Boshuizen HC, Kok FJ, Kromhout D. Long-term wine consumption is related to cardiovascular mortality and life expectancy independently of moderate alcohol intake: the Zutphen Study. J Epidemiol Community Health. 2009;63:534-540. doi: 10.1136/jech.2008.082198.
Forum Summary
The usual finding in longitudinal cohort studies has been that light-to-moderate consumers of alcohol tend to be at lower risk for total mortality and show greater longevity of life, even when other lifestyle/demographic factors known to affect longevity are adjusted for in the analysis. The present analysis is important as it presents data on the relation of alcohol intake to total mortality as well as to specific mortality from cardiovascular disease (CVD) and cancer for a very large number of subjects in the USA. It is based on data from more than 300,000 subjects, of whom almost 25,000 died during a follow-up period averaging 8.2 years. There was a very large number of lifetime abstainers which could serve as an appropriate reference group for their analyses. Further, by adjusting for a number of chronic diseases, and carrying out sensitivity analyses with a 2-year lag period for mortality, the investigators improved their ability to avoid having their results affected by “sick quitters.”
The authors conclude that their analysis shows that light and moderate drinkers have a lower risk of total mortality, as well as mortality from CVD, heart disease, and cerebrovascular disease. The protective effects of alcohol for such cardiovascular outcomes were not present for subjects who reported binge drinking or for those reporting what was defined as “heavy” drinking (>7 drinks/week for women and older men, 14 drinks/week for younger men). Interestingly, the mortality risk for light and moderate drinking was also significantly reduced for deaths attributed to cancer; Forum members thought that this may have possibly resulted from subjects with cancer who actually died from CVD having their deaths attributed to cancer. Subjects reporting heavy drinking and those with binge drinking showed increased risk of all-cause and cancer mortality, with no significant effect on CVD outcomes. The key results of the study are that there is a very clear J-shaped curve for the relation of alcohol to mortality, with lower total, cardiovascular, and even cancer mortality rates for light and moderate drinkers who do not binge drink. There was increased total mortality and cancer mortality for those classified as “heavy” drinkers.
Forum Members thought it unfortunate that beverage-specific data were not presented (as in many studies wine, and sometimes beer, drinkers have better outcomes than consumers of spirits), that subjects of all ages (≥ 18 years of age) were included in a single analysis (rather than also presenting results specifically for older subjects, when the outcome events studied usually occur), and especially that women and older men who consumed > 7 drinks/week, and younger men consuming >14 drinks/week were all combined into a “heavy” drinking category. It would have been preferable that the investigators also had a category for those consuming only slightly more than the recommended levels, who may well have had different outcomes than heavier drinkers.
Overall, Forum members considered this to be a well-done study providing additional data supporting a J-shaped curve for the association of alcohol consumption with mortality. Thus, data continue to indicate that light-to-moderate intake of alcoholic beverages without binge drinking reduces total mortality as well as death from CVD or cancer. The cumulative scientific data on this topic are well described in the accompanying Editorial Comment by de Gaetano and Constanzo in the same issue of the journal.
* * * * * * * * * * *
The following members of the International Scientific Forum on Alcohol Research have provided comments for this critique:
Andrew L. Waterhouse, PhD, Department of Viticulture and Enology, University of California, Davis, USA
David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa
Dag S. Thelle, MD, PhD, Department of Biostatistics, Institute of Basic Medical Sciences, University of Oslo, Norway; Section for Epidemiology and Social Medicine, Sahlgrenska Academy, University of Gothenburg, Sweden
Pierre-Louis Teissedre, PhD, Faculty of Oenology–ISVV, University Victor Segalen Bordeaux 2, Bordeaux, France
Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway
Creina Stockley, PhD, MSc Clinical Pharmacology, MBA; Health and Regulatory Information Manager, Australian Wine Research Institute, Glen Osmond, South Australia, Australia
Erik Skovenborg, MD, specialized in family medicine, member of the Scandinavian Medical Alcohol Board, Aarhus, Denmark
Linda McEvoy, PhD, Department of Radiology, University of California at San Diego (UCSD), La Jolla, CA, USA
Fulvio Mattivi, MSc, CAFE – Center Agriculture Food Environment, University of Trento, via E. Mach 1, San Michele all’Adige, Italy
Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA
Ramon Estruch, MD, PhD, Hospital Clinic, IDIBAPS, Associate Professor of Medicine, University of Barcelona, Spain
R. Curtis Ellison, MD, Professor of Medicine, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA