OUP user menu

Alcohol drinking, consumption patterns and breast cancer among Danish nurses: a cohort study

Lina S. Mørch, Ditte Johansen, Lau C. Thygesen, Anne Tjønneland, Ellen Løkkegaard, Claudia Stahlberg, Morten Grønbæk
DOI: http://dx.doi.org/10.1093/eurpub/ckm036 624-629 First published online: 18 April 2007

Abstract

Background: The aim of this study was to analyse the impact of alcohol intake and drinking pattern on the risk of breast cancer. Methods: A total of 17 647 nurses were followed from 1993 until the end of 2001. At baseline participants completed a questionnaire on alcohol intake and other lifestyle-related factors. Data were analysed using Cox's proportional hazard model. Results: During follow-up 457 women were diagnosed with breast cancer. The relative risk of breast cancer was 2.30 [Confidence interval (CI): 1.56–3.39] for alcohol intake of 22–27 drinks per week, compared to 1–3 drinks per week. Among alcohol consumers, weekly alcohol intake increased the risk of breast cancer with 2% for each additional drink consumed. Weekend consumption increased the risk with 4% for each additional drink consumed friday through sunday. Binge drinking of 4–5 drinks the latest weekday increased risk with 55%, compared with consumption of one drink. A possible threshold in risk estimates was found for consumption above 27 drinks per week. Conclusions: For alcohol consumption above the intake most frequently reported, the risk of breast cancer is increased. The risk is minor for moderate levels but increases for each additional drink consumed during the week. Weekend consumption and binge drinking imply an additional increase in breast cancer risk.

  • alcohol
  • breast cancer
  • drinking pattern

Introduction

The association between alcohol intake and breast cancer has been studied in several cohort and case-control studies.1–5 Most of these suggest a small increase in the relative risk of breast cancer with increasing amount of alcohol intake. Although the increased risk is modest, the association may have a large public health impact because of the large number of women drinking alcohol and the high incidence of breast cancer.

Binge-drinking is known to increase the risk of death from all causes in both men and women,6 and women have been found to be more sensitive to the detrimental effects of non-frequent drinking, compared with men.7 Only few studies have examined the association between frequency of drinking and risk of breast cancer and the results are conflicting.8,,9

It is likely that binge drinking and weekend consumption are associated with peaking levels of the endogenous concentration of estrogens or cause an accumulation of carcinogen bi-products of alcohol. This could lead to an enhancement of the harmful effects of alcohol on breast cancer. The effect of drinking patterns has only been scarcely addressed, probably because questions on drinking patterns are not yet commonly included in epidemiological surveys. No previous study has used alcohol consumption on weekdays as indicator for different drinking patterns, however, the consumption of alcohol is most likely to vary on a day-to-day basis and hence a possible difference in risk of breast cancer may be found. The aim of the present study was therefore to examine risk of breast cancer in this cohort of pre- and post-menopausal women using alcohol intake on different weekdays as an indirect measure for different drinking patterns.

Methods

Study population

The Danish Nurse Cohort Study was established in 1993, when all female members of the Danish Nurse Association above 44 years of age received a mailed questionnaire. In total 19 898 nurses (86%) returned a completed questionnaire, which included detailed information on health-related issues, including alcohol intake among other lifestyle related factors and other possible risk factors for breast cancer. The Danish Nurse Cohort has been described in detail in a previous publication.10

Analysis was stratified according to menopausal status. A woman was considered post-menopausal with cessation of menstrual bleedings, the use of hormone therapy (HT) or age above 55 years. Women were excluded if they had any diagnosis of cancer (except non-melanoma skin cancer) before baseline (N = 1086), missing information on alcohol consumption (N = 798) or other confounders included in the analyses. Some women fulfilled more than one exclusion criterion leaving 17 647 women eligible for analyses of alcohol related breast cancer risk.

Alcohol consumption

Participants were asked about their consumption of beer, wine and spirits the latest weekday (specified as: monday, tuesday, wednesday or thursday) and the previous weekend, which comprises three days (friday through sunday). The consumption of beer was reported as the number of whole bottles and the consumption of wine and spirits was reported as whole glasses. One normal beer in Denmark contains 11.6 g of alcohol, and 12 g is considered to be the amount of alcohol in a standard drink of both wine and spirits.11 The participants were also asked if their reported alcohol intake was typical for their usual alcohol intake. The total amount of alcohol consumed on the latest weekday or weekend was calculated by adding the intake of the specific type of alcohol. To estimate weekly amount of alcohol consumed the number of drinks reported for last weekday were multiplied by four and added to the consumption of the last weekend. The categorization of alcohol consumption is based on sufficient number of cases in each category together with the recommendation by the Danish National Health Board, advising women to drink no more than 14 drinks per week. The subjects were classified into categories according to their weekly total alcohol intake (0, 1–3, 4–7, 8–14, 15–21, 22–27 and >27 drinks per week), according to their alcohol intake on the latest weekday (0, 1, 2, 3, 4–5, 6–7, >7) and according to consumption in the weekend (0, 1–3,4–6, 7–9, 10–15, 16–21, >21). Binge drinking the last weekday was defined as four or more drinks (4–5, 6–7 and >7). Binge drinking on weekends was defined as consumption of 10 drinks or more, as this implies an intake of >4 drinks at least once during the weekend friday through sunday (10–15, 16–21 and 21).

Definition of potential confounders

The following baseline variables were considered as potential confounders among both pre- and post-menopausal women: smoking status (never, former, occasional smokers, daily smokers <15 g/day, daily smokers ≥15 g/day, leisure time physical activity (heavy >4 h/week, moderate <4 h per/week, inactive), Body mass index (BMI: <19 kg/m2, 19–26 kg/m2, 27–30 kg/m2, >30 kg/m2), never or seldom consumption of fruit and vegetables (yes/no), parity (nulliparious, 1–2 children, more than three children), age at birth of first child (nulliparious, <20, 20–29, ≥30 years), age at menarche (≤12, 13–14, ≥15 years), one or more female relatives with breast cancer (yes/no), self reported benign breast disease (yes/no) and night work (yes/no). In the analysis among post-menopausal women the following were additionally considered potential confounders: the use of HT (never, former, current user), age at menopause (continuous variable) and surgical menopause (yes/no).

Follow-up

The personal identification number has been assigned to all Danish citizens since 1968 and allows record linkage between the nationwide Danish registers. The women were followed through the nationwide Danish registers from date of entry (defined as the date the questionnaires were sent out), until date of breast cancer diagnosis, death, emigration or end of follow-up. The vital status was ascertained using the Danish unique identification number of the National Person Register until 30 April 2001. Diagnosis of invasive breast cancers were obtained from thee national registers in order to obtain complete identification of cases: The Danish National Hospital Discharge Register (Diagnose codes ICD10: C50.0–50.9 and ICD8: 174.01–174.09), the Danish Cancer Registry (Diagnose codes ICD7: 1700–1705 when the last digit of morpohology code was 3 or 6) and the Danish Breast Cancer Group (data on operations performed for breast cancer as well as histology based on ICDO: 8500, 8520, 8480, 8510, 8503, 8211, 8200, 8502, 8573). Information on self-reported breast cancer before baseline was only used as exclusion criteria if the event was reported to have occurred before the start of the registers. The diagnosis were manually crosschecked and the earliest date for diagnose was used. The follow-up through the Danish registers were complete.

Statistical analysis

We performed Cox regression by using SAS/STAT software to estimate the association between amount of alcohol intake, drinking pattern and the risk of breast cancer, while taking potential confounders into account. Age was used as the underlying time scale and follow-up was calculated in days. Age at date for mailing the questionnaires was used as entry time as the date of return is not available. (The majority of questionnaires were returned within 28 days after the questionnaires were sent out).10

The following regression models were performed:

  1. The influence of total amount of alcohol consumption in drinks per week (0, 1–3, 4–6, 7–14, 15–21, 22–27, >27), without considering binge drinking.

  2. The influence of total alcohol consumption in drinks the latest weekday (0, 1, 2, 3, 4–5, 6–7, >7) and drinks per weekend (0, 1–3, 4–6, 7–9, 10–15, 16–21, >21) included as separate covariates in the same model. In this way adjustment was made for the amount of alcohol consumed the latest weekday versus weekends.

  3. Linear tests for trends were carried out among the women consuming alcohol, excluding the non-drinkers as other studies have found non-drinkers to have an elevated risk due to other health related lifestyle traits than alcohol consumption. The log likelihood from a model where alcohol consumption was included as a categorical variable was compared to the log likelihood of the model, where the alcohol consumption was included as a continuous variable and compared to a chi-square distribution. If the model where alcohol consumption was included as a continuous variable fitted the data better, the risk estimate for alcohol consumption as a continuous variable was reported.

The same regression models were used for the analysis among pre- and post-menopausal women, respectively, and among participants reporting that their alcohol intake at baseline was typical for their usual alcohol intake. Adjustment was made for the potential confounder effects of age, age at birth of first child, menarche, female relatives with breast cancer and self reported benign breast disease. In the analyses among post-menopausal women adjustment was also made for hormone therapy. The variables were selected according to the likelihood ratio test (5% level) for significant associations with the risk of breast cancer, as none of the potential confounders changed the estimates after adjustment. The value of deciding whether crude and adjusted estimates differed by an important amount was set to 10%.

Results

During 134 796 person-years (py) of follow-up (mean follow-up: 7.6 years) we identified a total of 457 women with breast cancer. Pre-menopausal women accounted for 101 cases (36 348 py), while 348 post-menopausal women were diagnosed with breast cancer (95 610 py).

The majority of the women had a light or moderate intake of alcohol. A total of 3300 (19%) women were non-drinkers and 3829 (22%) had an intake of more than 14 drinks per week, 813 (5%) women had more than 27 drinks per week. Of 14 347 participants who drank alcohol 5660 (39%) did not consume alcohol the latest weekday and 590 (4%) did not consume alcohol the latest weekend. A total of 1536 (10%) were weekday binge drinkers (>4 drinks per day), 1800 (13%) were weekend binge drinkers (>10 drinks per weekend). The majority of women reported that their alcohol intake was typical for their usual alcohol intake (77%) only few women did not respond to this question (2%).

The women were aged 44–93 years (median 56) at baseline, 4,626 were pre-menopausal and 12 644 were post-menopausal, 4% had surgical menopause and the mean age at menopause was 48 years. The mean BMI was 24 kg/m2 (5% of women had BMI above 30 kg/m2), 25% were heavy physically active in their leisure time, 31% were never smokers and 13% were seldom consumers of fruit and vegetables. Only 5% of the women worked usually in night shifts. A detailed description of the baseline characteristics of the cohort is outlined in previous publications.10–12 However, none of these potential confounders were statistically significantly associated with the risk of breast cancer. The potential confounders, which were statistically significantly associated with risk of breast cancer, are shown in table 1. Drinkers of three or less drinks per week were older compared to drinkers of more than three drinks per week. Weekend-drinkers were younger compared with weekday-drinkers (data not shown). More non-drinkers and heavy-drinkers (>27 drinks per week) were nulliparous. Mean age at first birth was similar across categories of weekly alcohol consumption except for non-drinkers, where mean age was higher. Fewer weekend-drinkers were nulliparious and mean age at birth of first child was lower compared with that of weekday-drinkers (data not shown). Mean age at menarche was similar across categories of weekly alcohol intake. No systematically unequal distribution across categories was found for women reporting female relatives with breast cancer or former benign breast disease at baseline. Among post-menopausal woman increased alcohol intake was correlated with never use of HT.

View this table:
Table 1

Baseline characteristics, by different levels of total weekly alcohol consumption

Alcohol intake (No. of drinks per week)No. of subjects (No. of cases)No. of drinks (mean)Mean age (years)Mean age at first birth (years)Mean age at menarche (years)Relatives with breast cancer (%)Benign tumour (%)Pre-menusal (%)Never use of HTa (%)
03300 (92)0.059271422181865
1–32981 (65)2.056261424192657
4–73474 (74)5.055261425203154
8–144063 (95)9.655261425192955
15–212137 (61)16.555261424203151
22–27879 (42)23.455261426203054
>27813 (28)38.255261423222947
  • Weekly consumption of alcohol was defined as total amount in drinks consumed Monday through Sunday

  • a: Only post-menopausal women (N = 12 644)

Weekly amount of alcohol consumed

Women consuming 22–27 drinks per week had an increased relative risk of breast cancer of 2.30 [95% confidence interval (CI): 1.56–3.39], compared to drinkers of 1–3 drinks per week (table 2). Alcohol intake of less than 22 drinks per week compared to 1–3 drinks per week did not increase the risk of breast cancer (table 2). A linear relation was found between increasing alcohol intake and risk of breast cancer, with increased risk of 2% for each additional drink consumed [Relative risk (RR) = 1.02 (CI: 1.01–1.03)]. A possible threshold in the risk estimates was found for consumption above 27 drinks per week (table 2). However, a quadratic term included in the model was not statistically significant. No significant different effect of weekly alcohol intake was found in pre- and post-menopausal women, respectively (data not shown).

View this table:
Table 2

Relative risk of breast cancer, by weekly amount of alcohol intake

Age-adjustedaMultivariate-adjustmentb
N = 17 647(Cases: 457)N = 17 647(Cases: 457)
Alcohol intake (No. of drinks)No. of subjects (No. of cases)RR95% CIRR95% CI
Weekly intakec
03300 (92)1.280.93–1.771.270.92–1.75
1–32981 (65)1.0Reference1.0Reference
4–73474 (74)1.000.71–1.390.990.71–1.38
8–144063 (95)1.090.80–1.501.100.80–1.51
15–212137 (61)1.340.95–1.901.330.94–1.89
22–27879 (42)2.271.54–3.352.301.56–3.39
>27813 (28)1.631.05–2.541.621.04–2.52
Trendd1.021.01–1.03
  • Weekly consumption of alcohol was defined as total amount in drinks consumed Monday through Sunday

  • a: Adjusted for age, as age is underlying timescale

  • b: Adjusted for age, age at birth of first child, menarche, relatives with breast cancer and self reported benign breast disease

  • c: Regression model including total alcohol intake

  • d: Test for linear trend not including non-drinkers

Amount of alcohol consumed in the weekend or on last weekday

The risk of breast cancer increased with 4% for each additional drink consumed on weekends [RR = 1.04 (CI: 1.01–1.07)] (table 3, lower panel). Alcohol intake on last weekday was inconsistently associated with risk of breast cancer.

View this table:
Table 3

Amount of alcohol, binge drinking on latest weekday and weekend and risks of breast cancer

Multivariate-adjustmenta
N = 17 647 (Cases: 457)
Alcohol intake (No. of drinks)No. of subject (No. of cases)RR (95% CI)
On last weekdayc
08960 (210)0.90 (0.68–1.18)
12917 (73)1.00
22794 (70)0.96 (0.69–1.34)
31440 (43)1.07 (0.72–1.57)
4–5b1060 (46)1.55 (1.05–2.29)
6–7b247 (9)1.24 (0.61–2.52)
>7b229 (6)0.92 (0.39–2.14)
Trendd1.02 (0.95–1.10)
On last weekendc
03890 (106)1.34 (1.02–1.77)
1–35299 (111)1.00
4–64523 (119)1.25 (0.96–1.63)
7–92135 (54)1.16 (0.83–1.63)
10–15b1533 (53)1.49 (1.04–2.13)
16–21b216 (13)2.51 (1.37–4.59)
>21b51 (1)0.86 (0.12–6.26)
Trendd1.04 (1.01–1.07)
  • Weekday consumption was defined as alcohol consumption the latest Monday, Tuesday, Wednesday, or Thursday Weekend consumption was defined as alcohol consumption during three days combined; Friday through Sunday

  • a: Adjusted for age, age at birth of first child, menarche, relatives with breast cancer and self reported benign breast disease

  • b: Intake of 4 or more drinks consumed within one day was defined as binge drinking

  • c: Regression model includes weekday and weekend consumption as separate covariates in the same model. This way adjustment was made for the approximate total amount of alcohol consumed the remaining days of the week

  • d: Test for linear trend not including the non-drinkers

Binge drinking in the weekend or on last weekday

Binge drinking in weekends was associated with an increased risk of breast cancer, compared with intake of 1–3 drinks, RR was 1.49 (CI: 1.04–2.13) for 10–15 drinks and RR was 2.51 (CI: 1.37–4.59) for 16–21 drinks. Women who were binge drinking on the last weekday had an increased risk of breast cancer with RR of 1.55 (95% CI; 1.05–2.29) for 4–5 drinks, compared to women consuming one drink the last weekday (table 3, upper panel). Binge drinking was adjusted for the approximate total amount of alcohol consumed (weekend versus weekday, respectively).

Discussion

For alcohol consumers a small increasing risk of breast cancer was found for each additional drink consumed during the week. Alcohol consumption of 22–27 drinks per week was associated with more than two times the risk of breast cancer in women consuming 1–3 drinks per week. A tendency to a threshold in the risk estimates was found for consumption above 27 drinks per week, however based on few cases. Weekend consumption and binge drinking seemed to relate to an additional increased risk of breast cancer.

Prior studies and biological mechanisms

The epidemiological evidence suggests a linear association between alcohol consumption and the risk of breast cancer.1–4,8,,14 In 2002, a pooled analysis by the Collaborative Group on hormonal factors in breast cancer found that the relative risk of breast cancer increases by 7.1% for each additional 10 g per day intake of alcohol.1 A Danish study among post-menopausal women suggests a dose–response relationship between total alcohol intake and breast cancer (rate ratio of 1.10 per 10 g per day (95% CI: 1.04–1.16)).8 Our data further indicate a possible threshold in risk for intake above 27 drinks per week. However, the threshold may be an artifact caused either by chance, resulting from the very small sample size in the highest consumption category, or a real physiologic phenomenon, as others have reported the same.5,15,,16

The biological mechanisms of the effect of alcohol in the causation of breast cancer have been widely discussed. Alcohol consumption seems to increase the endogenous levels of estrogens. As estrogen related factors are known risk factors for breast cancer, it can be assumed that alcohol intake will increase the risk of breast cancer.17 Furthermore, metabolism of alcohol leads to increased production of reactive oxygen radicals, lipid peroxides, acetaldehyde and other toxins, which could influence various stages of breast cancer carcinogenesis.18 It should be suspected that higher alcohol intake would be associated with higher levels of estrogen and therefore increasing the risk of breast cancer. Our finding of a possible threshold is not in accordance with this assumption, however, other factors may influence this relationship. It is possible that the endogenous concentration of estrogens reaches a threshold as response to excessive alcohol drinking which may explain the upper threshold in alcohol associated breast cancer risk.

Our finding that alcohol intake in weekends and binge drinking confer an increased risk of breast cancer supports the hypothesis that consumption of larger amounts of alcohol within a short period of time increases risk of detrimental health outcomes.7 While the multivariate models of weekday- and weekend-drinking are adjusted for differences in the approximate amount of alcohol consumed on the other days of the week, this analytical approach does not take into consideration the amount of alcohol consumed in each drinking episode. It is possible that weekend drinkers consume more alcohol per drinking occasion, which could explain the differential effect of alcohol consumption in weekday versus weekends. The detrimental effect of binge drinking suggests a biological different effect of alcohol or its metabolites, when high serum concentrations of alcohol might be related to peaking levels of estrogens. This could lead to an enhancement of the harmful effects of alcohol on breast cancer. Some genetic mutations in alcohol-metabolizing enzymes cause an accumulation of bi-products of alcohol, which is very similar to what is speculated to occur after binge drinking. Aldehyde dehydrogenase-2 (ALDH2) is the key enzyme for elimination of acetaldehyde, an established animal carcinogen produced after drinking. In persons with inactive ALDH2, the body fails to metabolize acetaldehyde rapidly, leading to excessive accumulation of acetaldehyde. Inactive heterozygous ALDH2, compared with active ALDH2, has been found to enhance the risk of esophageal squamous cell carcinoma in male and female drinkers.19,,20 The biological mechanisms of binge drinking on risk of breast cancer are currently unknown and studies examining associations between different alcohol metabolizing enzymes and breast cancer risk may clarify the biochemical mechanisms underlying the association. The impact of alcohol consumption in weekdays or weekends has not been addressed in previous studies of risk of breast cancer.

Limitations

We have potential measurement error as alcohol intake is measured differently on weekdays and weekends (one-day versus three-days consumption), which could explain the different associations with risk of breast cancer, indicating that weekend assessment may be a more sufficient measurement of the true alcohol intake. We examined if the associations were altered, when women whose alcohol intake was reported to be non-typical for their usual alcohol consumption were excluded. The associations between weekday- and weekend-alcohol consumption, respectively, and the risk of breast cancer were not altered, indicating that our results are not biased by non-typical alcohol consumption.

Binge drinking the latest weekday was adjusted for the weekend consumption, as an approximation for total amounts of alcohol consumed during the week. Thus, conclusions about binge drinking on weekdays are based on the assumption, that adjustment for weekend consumption is a sufficient approximation for the total amounts of alcohol consumed the remaining days of the week. We were not able to adjust weekend binge drinking for the total amounts of alcohol consumed within the weekend and firm conclusions about weekend binge drinking versus total amount of alcohol consumed in weekends can not be made.7,,21

We find the method of using a one-day assessment on the latest weekday an acceptable approximation to the amount of alcohol consumed on other weekdays, because a decrease in reported alcohol use has been found with longer recall period, which indicates problems in correctly reporting alcohol intake for a full week.22 We were not able specifically to examine if the method of multiplying the latest weekday by four and adding to the weekend consumption did result in misclassification of the total weekly amount consumed. We examined if the associations were altered, when women whose reported alcohol intake was non-typical for their usual consumption were excluded. The associations did not change, indicating that our results are not biased by non-typical alcohol consumption.

Methodological issues in alcohol research

As a result of the prospective design, recall bias is unlikely, i.e. all data on alcohol intake were collected before the endpoint occurred. In the present study we excluded women with diagnosis of cancer before baseline (except non-melanoma skin cancer). Therefore, it is not expected that participants had changed their alcohol consumption before baseline due to pre-existing malignant disease. It is an ongoing discussion, whether the increased risk among the non-drinkers, as compared with moderate drinkers is caused by misclassification of prior- or present-heavy drinkers having been categorized in the group of non-drinkers.23,,24 Based on this uncertainty about the non-drinkers, we did not draw any conclusions based on comparisons with the group of non-drinkers. The present study has, unlike many epidemiological cohort and case-control studies, examined the breast cancer risk according to an alcohol intake above 27 drinks per week. The information on consumption of alcohol was reported only once and this was used as an indicator for the general level of exposure, thus providing risk of misclassification of women who changed their alcohol consumption during the follow-up period. This type of misclassification is, however, likely to be non-differential and risk estimates are likely to be biased towards the null value thereby leading to underestimation of risk estimates. On the other hand, our risk estimates would be overestimated if the participants during the follow-up had higher alcohol intake than they reported at baseline. However, as the alcohol consumption among Danish women in general was approximately steady during follow-up period, we do not expect our results to be overestimated.25 The follow-up in this study is shorter than that seen in most other epidemiological studies, and this could imply that the information on alcohol consumption at baseline describes the more factual lifestyle during follow-up, perhaps lowering the risk of misclassification. However, if alcohol intake influences at an early stage of carcinogenesis, the relative rates should increase for longer time since exposure.26 Assuming the average induction time of breast cancer is longer than the follow-up in our study the risk estimates are underestimated. We did not allow for any latency post-poning analysis at baseline, which is a possibility when examining cause and effect. However, we think that most lifestyle-related factors including alcohol consumption most likely do not change on a short-term basis, implying that women consuming alcohol have implemented this habit in their general lifestyle before baseline.

Methodological issues in present study

The present study has a high response rate at baseline (86%) and complete follow-up during the study period of 7.6 years. In our analysis, we were able to take a large number of potential confounders as reproductive and lifestyle factors into account. Still we did not have information on exposure to confounding factors on specific weekdays. Consequently, we were not able to make adjustments for weekday-variations in other risk factors of breast cancer. It is likely that the differential effect of alcohol consumption in weekday versus weekends is explained by other risk factors that are also associated with alcohol consumption on different weekdays. However, none of the other lifestyle factors were statistically significantly associated with risk of breast cancer. Self-reported information at baseline about history of benign breast disease was considered a potential confounder, but analyses without adjustment for benign breast disease did not change the risk estimates significantly.

This cohort of female nurses had a higher weekly alcohol intake compared to the general female population in Denmark.10 The proportion of heavy drinkers was higher and more nurses consumed alcohol the latest weekday and weekend, respectively. Also the distribution of other lifestyle traits was different.10 Consequently, the association between alcohol intake and risk of breast cancer found in present study may not be similar to that in the general female population in Denmark.

In conclusion, these data suggest, that weekly alcohol consumption increases the risk of breast cancer in women reporting alcohol consumption above the average for women in general. The risk is minor for moderate levels of weekly alcohol intake, but increases for each extra drink consumed. Weekend consumption and binge drinking seem to be related to an additional increased risk of breast cancer.

Acknowledgement

We would like to thank Erik Bernhard Obel, MD, Dr Med.Sci and Yrsa Andersen Hundrup, MN, PhD for acquisition of data. The data for the present study were collected and provided by The Danish Nurse Cohort Study, National Institute of Public Health, Copenhagen Denmark. www.niph.dk/The Danish Nurse Cohort Study. This work was supported by Centre for Alcohol Research, National Institute of Public Health.

Conflicts of interest: None declared.

Key points

  • The aim of this study was to analyse the impact of alcohol intake and drinking pattern on the risk of breast cancer.

  • The main finding was that weekend consumption and binge drinking imply an additional increase in breast cancer risk.

  • In addition, it was found that the risk is minor for moderate levels of weekly alcohol intake, but increases for each extra drink consumed.

  • To minimize the risk of breast cancer our data imply that women should not do any binge drinking or drink larger amounts of alcohol.

References

View Abstract