Critique 186: Changes in alcohol and effects on risk of breast cancer and heart disease — 19 May 2016

Dam MK, Hvidtfeldt UA, Tjønneland A, Overvad K, Grønbæk M, Tolstrup JS.  Five year change in alcohol intake and risk of breast cancer and coronary heart disease among postmenopausal women: prospective cohort study. BMJ 2016;353:i2314.

Authors’ Abstract

Objective: To test the hypothesis that postmenopausal women who increase their alcohol intake over a five year period have a higher risk of breast cancer and a lower risk of coronary heart disease compared with stable alcohol intake.

Design Prospective: cohort study.

Setting: Denmark, 1993-2012.

Participants: 21 523 postmenopausal women who participated in the Diet, Cancer, and Health Study in two consecutive examinations in 1993-98 and 1999-2003. Information on alcohol intake was obtained from questionnaires completed by participants.

Main outcome measures:  Incidence of breast cancer, coronary heart disease, and all cause mortality during 11 years of follow-up.  Information was obtained from the Danish Cancer Register, Danish Hospital Discharge Register, Danish Register of Causes of Death, and National Central Person Register.  We estimated hazard ratios according to five year change in alcohol intake using Cox proportional hazards models.

Results: During the study, 1054, 1750, and 2080 cases of breast cancer, coronary heart disease, and mortality occurred, respectively. Analyses modelling five year change in alcohol intake with cubic splines showed that women who increased their alcohol intake over the five year period had a higher risk of breast cancer and a lower risk of coronary heart disease than women with a stable alcohol intake. For instance, women who increased their alcohol intake by seven or 14 drinks per week (corresponding to one or two drinks more per day) had hazard ratios of breast cancer of 1.13 (95% confidence interval 1.03 to 1.23) and 1.29 (1.07 to 1.55), respectively, compared to women with stable intake, and adjusted for age, education, body mass index, smoking, Mediterranean diet score, parity, number of births, and hormone replacement therapy. For coronary heart disease, corresponding hazard ratios were 0.89 (0.81 to 0.97) and 0.78 (0.64 to 0.95), respectively, adjusted for age, education, body mass index, Mediterranean diet score, smoking, physical activity, hypertension, elevated cholesterol, and diabetes. Results among women who reduced their alcohol intake over the five year period were not significantly associated with risk of breast cancer or coronary heart disease. Analyses of all cause mortality showed that women who increased their alcohol intake from a high intake (≥14 drinks per week) to an even higher intake had a higher mortality risk that women with a stable high intake.

Conclusion: In this study of postmenopausal women over a five year period, results support the hypotheses that alcohol intake is associated with increased risk of breast cancer and decreased risk of coronary heart disease.

Forum Comments

Innumerable epidemiologic studies have evaluated the association between reported alcohol intake and the risk of breast cancer, coronary heart disease (CHD), and total mortality. The typical findings from such studies are that alcohol intake is associated with a slight increase in the risk of breast cancer (Willett et al, Longnecker et al, Neilsen et al, Morch et al), although this association has not been demonstrated in all cohort studies (Zhang et al).  In one large study in which perceived underreporting of alcohol was taken into consideration, no effect on breast cancer risk was seen for light or moderate drinkers who had no evidence elsewhere in their medical records suggesting heavier alcohol intake (Klatsky et al, 2014).  It has been found in many studies that women who binge drink, have a low folate intake, and/or have taken hormonal replacement therapy have greater increases in cancer risk associated with alcohol (Larsson et al, Colditz et al, Nielsen et al, Morch et al).  Further, other study characteristics have been shown to modify the association between alcohol consumption and breast cancer risk (Ellison et al).

For the association of alcohol consumption with CHD, the data are extremely consistent: in essentially every prospective study since the 1970s, light or moderate drinkers have been found to have a significantly lower risk of disease (Klatsky et al, 1974; Coate; Doll), and most studies show a clear “J-shaped” curve (Corrao et al). Further, most recent studies also show a significant decrease in the risk of all-cause mortality among moderate drinkers when compared with abstainers or only occasional drinkers (McCaul et al, Midlöv et al). Di Castelnuovo et al have reported also that moderate alcohol consumption decreases the risk of total mortality, stating: “Low levels of alcohol intake (1-2 drinks per day for women and 2-4 drinks per day for men) are inversely associated with total mortality in both men and women.  Our findings, while confirming the hazards of excess drinking, indicate potential windows of alcohol intake that may confer a net beneficial effect of moderate drinking, at least in terms of survival.”

The present large study from Denmark sought to determine among postmenopausal women whether a change in their level of alcohol intake during a follow-up period (averaging about 11 years) affected their subsequent risk of breast cancer, CHF, or total mortality.  With more than 1,000 incident cases of breast cancer, more than 1,700 of CHD, and more than 2,000 deaths during the follow-up period, the study had adequate power to examine not only the effects of baseline alcohol intake on these outcomes but to judge the effects of increases or decreases in reported intake over a 5-year period when compared with drinkers who had stable patterns of drinking.

Overall, using baseline data on alcohol consumption, women reporting 7 or more drinks/week had a tendency to have a higher risk of subsequent breast cancer, while the lowest category (< 1 drink/week) and those reporting 14 or more drinks/week had a lower risk of CHD. These findings are consistent with most other epidemiologic studies.

As for changes in intake between the two assessments (an average of 5.4 years apart), most drinkers remained stable in their alcohol intake; 7% of subjects reported that they decreased their intake by at least 7 drinks/week while 9% reported that they had increased their intake by 7 or more drinks/week.  For subsequent breast cancer risk, with alcohol consumption stratified into categories of < 7, 7-13, and ≥14 drinks/week, those initially in the lowest category who moved into a higher category of intake showed an increase in risk subsequent to the change; however, those in the highest category at baseline who decreased their alcohol intake also tended to increase their risk of breast cancer.

In terms of CHD risk subsequent to a change in alcohol intake, the risk tended to be lower with increasing alcohol intake and higher with decreasing intake. For total mortality, risk tended to be similar in direction to the changes in risk seen for CHD, but with wide confidence limits.  The authors conclude that “In this study of postmenopausal women over over a five-year period, results support the hypotheses that alcohol intake is associated with increased risk of breast cancer and decreased risk of coronary heart disease.”

Specific comments by Forum members: Forum members considered this to be a very well-done attempt to evaluate the effects of changes in alcohol intake.  Reviewer Ellison noted that both diabetes and self-assessed health status were evaluated as potential confounders.  For CHD, both of these factors have been shown to relate to alcohol consumption (a decrease in the risk of diabetes and higher self-perceived health status for moderate drinkers), so they could potentially be mechanisms of effect.  However, the authors did sensitivity analyses when such factors were excluded and state that they ‘resulted in only minor changes’.”

When usual alcohol intake in 1999-2003 was taken as the baseline intake, there was a significant (but non-linear) higher risk of breast cancer with greater amounts of alcohol. Subjects reporting <1 and those reporting 1-6 drinks/week had essentially the same relative risk (0.99 and 1.00, respectively), indicating no increase in breast cancer risk for light drinkers.  There were very small percentages of women (2% in 1993-1998 and 5% in 1999-2003) who were abstainers from alcohol, making it difficult to use them as a referent group.

When evaluating the effects of changes in drinking, a question was raised by some reviewers as to whether or not it was appropriate to adjust for earlier alcohol consumption.  The two values are closely linked, and there have been problems described when using a prevalence estimate of any exposure that may affect the disease profile of subjects at “baseline” (who may have different disease profiles from subjects not previously exposed to that agent).  Regardless of what approach is taken, either including or not including baseline alcohol consumption data (both approaches were reported in this study), it can still cause difficulties in interpreting the results.

Reviewer de Gaetano considered it appropriate to adjust for baseline alcohol consumption in the present paper: “I consider that it is correct and important to adjust the changes in alcohol intake taking the baseline values into account.  For example, when evaluating the effect on blood pressure change of an anti-hypertensive drug, it is necessary to evaluate changes in relation to the basal levels of blood pressure.  In the case of alcohol, it is most likely different if you change from 1 drink to 2 drinks or from 3 to 4, not only in terms of amounts of change, but also in terms of motivation.  The observation that cancer risk increases whether you either increase or decrease the amount of alcohol could be partly due to alcohol reduction because of health/disease problems (which would lead to biased estimates).  It is not clear to me whether changes included people who did not drink at baseline but started to drink later on.  Overall, this paper confirms that alcohol protects against cardiovascular events, even at doses higher than the moderate amounts suggested, as described by Costanzo et al.”

Reviewer Thelle stated: “This is probably one of the best studies undertaken on the effect of changing drinking habits, but to assess the effects of changes is always difficult as people have different starting levels. The authors had access to two sets of data which made it possible to adjust for this, a remarkable asset.  Still, it is an observational study, and we will never know why people changed their drinking habits. The authors very thoughtfully performed sensitivity analyses by excluding the first three years of follow-up, thereby reducing the likelihood of reversed causality.  They also excluded abstainers and others that may have confounded the associations.  But what we don’t know, we just cannot adjust for.”

Thelle concluded: “The authors confirmed what we already believe we know, but we remain ignorant as to whether the increased risk of breast cancer is due to the total alcohol consumption (life-time exposure) or because of increased postmenopausal consumption.”  Reviewer Lanzmann-Petithory regretted that beverage-specific data were not addressed, stating “In our studies in Nancy, France, we found a greater decrease in total mortality for wine drinkers than for spirits drinkers.  However, this very interesting study could maybe lead to better define recommendations for drinking among postmenopausal women, based on personal risk for breast cancer and for cardiovascular disease.”

Forum member Skovenborg also agreed with other reviewers that this was a well-done study. He stated: “Regarding the validity of the data on alcohol intake, it is my impression that the data from the Danish Diet, Cancer and Health Study are as good as such data get.  The study results give rise to a few comments:

(1)  A primary breast tumor starts from one single cell that multiplies exponentially with time. At the time of diagnosis the tumor consists of 109 cells after about 30 cell generations and has reached a volume of about 1 cm3.  The growth rates for primary breast cancers show great inter-individual variability, however, the average tumor volume doubling time is 280 days, which means that more than 18 years are required from the first tumor cell to produce a tumor above the lowest detection level.  The results of the present study thus associate alcohol with tumor promotion and not tumor initiation (Friberg & Mattson).

(2) The paper states:  ‘Women who drank seven to 13 drinks per week in the 1993-98 examination and reduced their intake by 3.6 drinks per week or more had a significantly higher risk of mortality risk than women with a stable alcohol intake’. This is an interesting result considering the fact that the Danish Health Authority 6 years ago reduced the sensible drinking limit for women from 14 units (1 unit = 12 g alcohol) per week to 7 units per week.”

Forum member Finkel had some questions about the reported changes over the 5.4 years between the two alcohol assessments: “I am a bit uneasy, particularly about the validity of the subgroup that increased its alcohol consumption. Intuitively (I am just a clinician), I’d suspect that the number of such subjects making real increases in their intake at this stage of life would be very small, perhaps too small to give the results significance.  Further, some other, likely confounding, event might have stimulated the change in consumption.  Another possibility is that the consumption had not changed much at all, but that the subjects’ perception or willingness to admit the true quantity they consumed had led to the supposed change.”

Several Forum members noted the relatively wide ranges for each category of dinking used when assessing for the effects of changes. For example, women consuming 7 drinks per week were in the same category of those consuming 13 drinks/week.  [Reviewers Ursini and Finkel also wondered if subjects who regularly consumed small amounts of wine as part of religious exercises (e.g., communion, kiddush) would consider themselves as “drinkers” or not.]

Stated reviewer Zhang, “In the stratified analysis the range of alcohol consumption before the baseline is wide, so within the strata, it is difficult to consider that the subjects are an homogenous group. I think that it is very unusual that postmenopausal women will start to drink more than 14 drinks/week if they were non-drinkers before.  Those who increased their intake to move into a higher category (say to 7 – 13 drinks/week) are likely to be those who were drinkers of 5 or 6 drinks/week before.”  Reviewer Zhang also stated: “It is unclear what the reason was that some women apparently increased or decreased their number of drinks/week so much.  Even though the authors did sensitivity analyses omitting the first 3 years of follow up for disease outcomes, they were unable to note why the women changed their intake.  The reasons for change in alcohol consumption may be the consequence of the disease under study (except for death) or confounders of the disease under study, even when the disease had not yet been diagnosed.”

References from Forum review

Coate D: Moderate drinking and coronary heart disease mortality: evidence from NHANES I and the NHANES I Follow-up. Am J Public Health 1993;83:888.

Colditz GA, Stampfer MJ, Willett WC, et al: Prospective study of estrogen replacement therapy and risk of breast cancer in postmenopausal women, JAMA 1990;264:2648.

Corrao G, Rubbiati L, Bagnardi V, et al: Alcohol and coronary heart disease: a metaanalysis, Addiction 2000;95:1505.

Costanzo S, Di Castelnuovo A, Donati MB, Iacoviello L, de Gaetano G. Alcohol Consumption and Mortality in Patients With Cardiovascular Disease: A Meta-Analysis.  J Am Coll Cardiol 2010;55:1339-1347.  doi:10.1016/j.jacc.2010.01.006.

Di Castelnuovo A, Costanzo S, Bagnardi V, Donati MB, Iacoviello L, de Gaetano.  Alcohol dosing and total mortality in men and women: an updated meta-analysis of 34 prospective studies. Arch Intern Med. 2006;166:2437-2445.

Doll R: One for the heart. BMJ 1997;315:1664.

Ellison RC, Zhang Y, McLennan CE, Rothman KJ: Exploring the relation of alcohol consumption to risk of breast cancer, Am J Epidemiol 2001;154:740.

Friberg S, Mattson S.  On the growth rates of human malignant tumors: Implications for medical decision making. J Surg Oncol 1997;65:284-297. 

Klatsky AL, Friedman GD, Siegelaub AB: Alcohol consumption before myocardial infarction. Results from the Kaiser-Permanente epidemiologic study of myocardial infarction, Ann Intern Med 1974;81:294.

Klatsky AL, Udaltsova N, Li Y, Baer D, Tran HN, Friedman GD. Moderate alcohol intake and cancer: the role of underreporting.  Cancer Causes Control 2014;25:693-699. doi: 10.1007/s10552-014-0372-8.

Larsson SC, Giovannucci E, Wolk A: Folate and risk of breast cancer: a meta-analysis.  J Natl Cancer Inst 2007;99:64.

Longnecker MP: Alcoholic beverage consumption in relation to risk of breast cancer: meta-analysis and review, Cancer Causes Control 1994;5:73.

McCaul KA, Almeida OP, Hankey GJ, Jamrozik K, Byles JE, Flicker L. Alcohol use and mortality in older men and women. Addiction. 2010 Aug;105(8):1391-400. doi:10.1111/j.1360-0443.2010.02972.x.

Midlöv P, Calling S, Memon AA, Sundquist J, Sundquist K, Johansson SE. 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;12;16:22. doi: 10.1186/s12889-016-2700-2

Morch LS, Johansen D, Thygesen LC, et al: Alcohol drinking, consumption patterns and breast cancer among Danish nurses: a cohort study, Eur J Public Health 2007;17:624

Nielsen NR, Gronbaek M: Interactions between intakes of alcohol and postmenopausal hormones on risk of breast cancer, Int J Cancer 2008;122:1109.

Willett WC, Stampfer MJ, Colditz GA, et al: Moderate alcohol consumption and the risk of breast cancer, N Engl J Med 1987;316:1174.

Zhang Y, Kreger BE, Dorgan JF, Splansky GL, Cupples LA, Ellison RC. Alcohol consumption and risk of breast cancer: the Framingham Study revisited. Am J Epidemiol 1999;149:93.

Forum Summary

This large study from Denmark was designed to test the hypothesis that women who increase their alcohol intake over a five year period have a higher risk of breast cancer and a lower risk of coronary heart disease (CHD) compared with women who exhibit a stable alcohol intake. It consisted of more than 20.000 postmenopausal women who had two assessments of alcohol intake, about 5 years apart; changes during that period were related to their subsequent risk of developing breast cancer or CHD, or dying, during the subsequent follow-up period that averaged 8 years.

For the risk of breast cancer, the baseline alcohol consumption reported by the women showed that higher alcohol intake was associated with a greater risk of developing breast cancer during follow up, but a lower risk of developing CHD and for total mortality. For relating changes in alcohol intake in the 5 years between the two alcohol assessments, those who increased their reported alcohol intake showed an increased risk of subsequent breast cancer, while those who decreased their intake also had a tendency for greater breast cancer.  In terms of CHD and total mortality, the data were consistent with an increase in consumption lowering risk, while there was a tendency for a decrease in consumption to increase risk.  Hence, the directionality for risk from changes in intake reflected the risks associated with baseline risk, and the authors conclude that their results “ . . .support the hypotheses that alcohol intake is associated with increased risk of breast cancer and decreased risk of coronary heart disease.”

Forum reviewers considered this to be a very well-done study. They, and the authors, appreciated the innate difficulty in separating effects on health of usual alcohol intake over time and changes over a limited period of time.  Reviewers were also concerned by these problems, suggesting that some of the reported changes in alcohol consumption in this study may have just reflected individuals replying differently to questions regarding their drinking at different points of time.  Further, the reason why some women may have actually increased or decreased their intake are not known.

Nevertheless, Forum members applaud the attempt of the authors to judge if changes in alcohol intake relate to cancer, CHD, and total mortality. This is a difficult task given that baseline and life-time alcohol consumption clearly relate to these outcomes.  Further, data on the pattern of drinking (regular versus binge), any estimate of underreporting of alcohol, the type of beverage consumed and, as stated by the authors, certain known risk factors of breast cancer, were not available to be included in their analyses.  Thus while the results of this study add important information on the relation of alcohol to disease, there remain questions about the specific relevance of changes in intake for such outcomes.

*                      *                      *                      *                      *                      *                      *         

Comments on this critique by the International Scientific Forum on Alcohol Research were provided by the following members:

Yuqing Zhang, MD, DSc, Clinical Epidemiology, Boston University School of Medicine, Boston, MA, USA

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

Fulvio Ursini, MD, Dept. of Biological Chemistry, University of Padova, Padova, Italy

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

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

Dominique Lanzmann-Petithory, MD, PhD, Nutrition Geriatrics, Hôpital Emile Roux, APHP Paris, Limeil-Brévannes, France

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

R. Curtis Ellison, MD, Professor of Medicine & Public Health, Boston University School of Medicine, Boston, MA, USA

Giovanni de Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy