Abstract

Background: Alcohol consumption is high in France. Aim: Estimation of alcohol-attributable mortality in France by sex, age and dose, for year 2009. Method: We combined survey and sales data to estimate the prevalence of alcohol consumption by age, sex and dose category. For each cause of death, the relative risk of death as a function of dose was obtained from a meta-analysis and combined with prevalence data to obtain the attributable fraction; this fraction multiplied by the number of deaths gave the alcohol-attributable mortality. Results: A total of 36 500 deaths in men are attributable to alcohol in France in 2009 (13% of total mortality) versus 12 500 in women (5% of total mortality). Overall, this includes 15 000 deaths from cancer, 12 000 from circulatory disease, 8000 from digestive system disease, 8000 from external causes and 3000 from mental and behavioural disorder. The alcohol-attributable fractions are 22% and 18% in the population aged 15 to 34 and 35 to 64, respectively, versus 7% among individuals aged 65 or more. Alcohol is detrimental even at a low dose of 13 g per day, causing 1100 deaths. Conclusion: With 49 000 deaths in France for the year 2009, the alcohol toll is high, and the effect of alcohol is detrimental even at low dose. Alcohol consumption is responsible for a large proportion of premature deaths. These results stress the importance of public health policies aimed at reducing alcohol consumption in France.

Introduction

In many countries, the majority of adults consume alcohol. In France, the amount of alcohol available for consumption was traditionally high and reached a maximum of 65 g per adult per day at the end of the thirties. Since 1960, it has decreased regularly by 1ċ7% per year and was equal to 27 g per adult and per day in 2009.1

The estimation of alcohol-attributable mortality provides a useful indicator for developing national public health strategies. These estimations are available for many countries.2–10 In France, estimations were published by Pignon and Hill for the years 1985 and 1995.11,12 In 2007, the International Agency for Research on Cancer (IARC) published an estimation of alcohol-attributable cancer morbidity and mortality for the year 2000.13 Recently, Rey et al. published estimations of alcohol-attributable deaths for the year 2006.14 The aim of this study was to estimate the number of alcohol-attributable deaths for the year 2009, overall and by cause of death, according to gender, age and alcohol dose.

Methods

To estimate the number of deaths attributable to alcohol, we need to know the distribution of alcohol consumption by gender and age group in the French population, we need to identify all causes of deaths with a risk modified by alcohol consumption and we need to know the dose–effect relationship between exposure to alcohol and the risk for each of these alcohol-related causes of death. This study is based on the French mortality data for the year 2009.

Alcohol consumption by gender and age

A representative sample of 20 178 French individuals aged 15+ were interviewed on their alcohol drinking habits in 2002–03 by the National Institute of Statistics and Economic Studies. The data on alcohol consumption have not been published, but they are available from Réseau Quetelet on request.15 There was a marked discrepancy between the reported average alcohol consumption in this survey (11 g/day) and the average estimated based on the French alcohol sales statistics reported to the tax authorities (27 g/day in 2009). This corresponds to a rate of underreporting equal to 2ċ4 (27/11); therefore, we multiplied the daily alcohol consumption each individual declared in the national survey by 2ċ4 to adjust the declared daily alcohol consumption on 2009 alcohol sales. Supplementary table 1 shows the adjusted results of the survey, by gender and age category, for five dose categories. The alcohol-attributable number of deaths in 2009 was estimated based on these adjusted consumption data.

Causes of deaths associated with alcohol

Conditions wholly attributable to alcohol

A number of conditions, listed in supplementary table 2, are, by definition, wholly attributable to alcohol consumption.

Conditions partly attributable to alcohol

Supplementary table 3 lists all conditions partly attributable to alcohol. Cancer sites with a risk increased by alcohol consumption are the oral cavity, pharynx, oesophagus, colon and rectum, liver, larynx and breast.16–21 Alcohol increases the risk of cardiac arrhythmias,22 haemorrhagic stroke23 and of hypertensive diseases among men and decreases the risk of hypertensive diseases among women for a daily consumption below 12 g of ethanol.24 Alcohol reduces the risk of ischemic heart disease19 and ischemic stroke23 at daily doses below 78 and 35 g of ethanol, respectively. The risk of cholelithiasis is reduced by alcohol consumption.22 Alcohol increases the risk of pancreatitis.25 The risks of accidents, falls, suicides and homicides, that is external causes of death, are increased by alcohol consumption. Low to moderate alcohol consumption is known to have a protective effect on type 2 diabetes.26,27 The risk of death from epilepsy is increased by alcohol.28

Risk functions

For each alcohol-attributable condition except for external causes, the dose–response relationship between alcohol consumption and the risk of death was provided by the most recent meta-analysis of epidemiological studies comparing groups of individuals with at least three categories of alcohol consumption.16–25,27–29 This relationship has the general form of a fractional polynomial of degree two that fits a variety of dose–response relations with only three parameters.

Supplementary table 3 shows, for each cause of death, the estimated fractional polynomial provided by the corresponding meta-analysis and the relative risks for the doses of alcohol of supplementary table 1. The dose–effect relationships are shown in web figure.

Figure 1

Number of deaths attributable and not attributable to alcohol by dose category and gender, for causes other than external

Figure 1

Number of deaths attributable and not attributable to alcohol by dose category and gender, for causes other than external

Supplementary table 3 also shows the dose–effect relationships between alcohol consumption and the risk of liver cirrhosis29 used to distribute the alcohol-attributable deaths for conditions wholly attributable to alcohol between dose categories, see below.

Alcohol-attributable fraction

Partially alcohol-attributable conditions

The alcohol-attributable fraction (AAF) for a specific cause of death is the proportion of deaths from this specific cause due to alcohol consumption. The AAF was estimated for each condition, gender and age group using Levin’s formula30:  

formula
where pi is the proportion of the population exposed to the ith dose category (13, 28, 46 and 117 g of pure alcohol per day for men and 13, 28, 46 and 85 g of pure alcohol per day for women) and RRi is the relative risk associated with the ith dose category compared with non-regular drinkers (see supplementary table 3).

From this formula, one can deduct the AAF associated with the ith dose category (AAFi):  

formula

External causes

For external causes, particularly for accidents which represent a main external cause, the increase in risk depends on acute rather than on average exposure. The risk is also increased for any individual, drinker or non-drinker, involved in an accident caused by a drinker. For these reasons, we have used the attributable fractions proposed by Ezzati et al.31 for the main external causes in Western Europe. We have not included external causes of death in the estimation of mortality according to alcohol dose.

Wholly alcohol-attributable conditions

By definition, all deaths associated with wholly alcohol-attributable conditions are due to exposure to alcohol. To distribute this total number of deaths between the four dose categories, we need a dose–response relationship. Liver cirrhosis is the unique wholly alcohol-attributable condition for which we have an estimation of the dose–response relationship in the literature.29 The AAF associated with each dose category i: AAFi was estimated using the relative risks of liver cirrhosis presented in supplementary table 3 with ‘13 g per day’ as the reference dose category:  

formula
This estimation was used for each wholly alcohol-attributable condition.

Sensitivity analysis

Sensitivity analyses estimated the number of alcohol-attributable deaths in 2009 under the assumptions that 10 or 20% of the alcohol available in 2009 was not consumed, but wasted or undrinkable. The daily alcohol consumption estimated from sales was then equal to 24ċ3 or 21ċ6 g, respectively, of pure alcohol per adult per day. We multiplied the daily alcohol consumption each individual declared in the national survey by 2ċ2 (24ċ3/11) in the first situation and by 2ċ0 (21ċ6/11) in the second one, to adjust the 2003 declared daily alcohol consumption on 2009 alcohol actually consumed.

In another sensitivity analysis, the consumption of alcohol was taken as declared and, therefore, not corrected to account for the amount of available alcohol estimated from sales.

Mortality data

The number of deaths observed in 2009 was obtained from the National Institute of Health and Medical Research for each condition related to alcohol, by gender and 5-year age group (http://www.cepidc.vesinet.inserm.fr) (table 1). The number of deaths attributable to alcohol was calculated for each condition, by gender and age group, for each dose category by multiplying the total number of deaths by the corresponding AAF. For each cause of death and each gender, the age-specific numbers of deaths attributable to alcohol were summed, and the result was divided by the total number of deaths to generate the AAF.

Table 1

Total number of deaths, alcohol-attributable fraction and number of deaths due to alcohol by gender and cause of death

Cause of death Men
 
Women
 
Total
 
 Deaths in 2009 Attributable fraction (%) Deaths due to alcohol Deaths in 2009 Attributable fraction (%) Deaths due to alcohol Deaths in 2009 Attributable fraction (%) Deaths due to alcohol 
Cancera 89 824 12 11 197 62 941 4 003 152 765 10 15 200 
Oral cavity 1 270 75 951 491 36 178 1 761 64 1 129 
Pharynx 2 139 86 1 842 352 49 172 2 492 81 2 014 
Oesophagus 3 402 84 2 860 882 53 467 4 284 78 3 327 
Colon-rectum 9 792 30 2 949 8 523 750 18 314 20 3 699 
Liver 6 203 30 1 850 2 215 201 8 418 24 2 051 
Larynx 1 240 60 745 150 24 35 1 390 56 781 
Breast 220 12 547 18 2 199 12 767 17 2 199 
Circulatory system diseases 66 833 14 9 523 77 510 2 710 144 343 12 233 
Hypertensive disease 3 099 46 1 420 5 834 15 866 8 933 2 286 
Ischemic heart disease 20 774 938 15 765 −3 −529 36 539 409 
Alcoholic cardiomyopathy 74 100 74 10 100 10 84 100 84 
Cardiac arrhythmia 4 236 53 2 250 6 277 22 1 366 10 513 34 3 615 
Haemorrhagic stroke 7 734 48 3 704 10 139 827 17 874 25 4 531 
Ischemic stroke 5 532 19 1 050 8 581 150 14 112 1 200 
Oesophageal varices 88 100 88 20 100 20 108 100 108 
Digestive system diseases 12 543 45 5 627 10 888 20 2 145 23 431 33 7 772 
Alcoholic gastritis 100 100 100 
Alcoholic liver disease 3 543 100 3 543 1 317 100 1 317 4 860 100 4 860 
Chronic hepatitis, not elsewhere classified 15 100 15 10 100 10 25 100 25 
Fibrosis and cirrhosis of the liver 1 684 100 1 684 767 100 767 2 451 100 2 451 
Cholelithiasis 109 −47 −51 201 −10 −19 310 −23 −70 
Pancreatitis 596 74 438 428 17 71 1 024 50 509 
Other diseases 95 382 13 3 441 104 926 634 200 308 4 075 
Type 2 diabetes 3 048 21 638 3 311 −4 −143 6 359 494 
Wernicke’s encephalopathy 100 100 10 100 10 
Mental and behavioural disorders due to alcohol 2 251 100 2 251 608 100 608 2 859 100 2 859 
Degeneration of nervous system due to alcohol 75 100 75 26 100 26 101 100 101 
Epilepsy and status epilepticus 859 53 459 752 18 137 1 611 37 596 
Alcoholic polyneuritis 12 100 12 100 16 100 16 
External causes 22 620 25 5 545 14 748 18 2 609 37 368 22 8 154 
Unknownb 7 671 16 1 250 6 848 363 14 519 11 1 613 
Total 272 253 13 36 584 263 113 12 465 535 366 49 048 
Cause of death Men
 
Women
 
Total
 
 Deaths in 2009 Attributable fraction (%) Deaths due to alcohol Deaths in 2009 Attributable fraction (%) Deaths due to alcohol Deaths in 2009 Attributable fraction (%) Deaths due to alcohol 
Cancera 89 824 12 11 197 62 941 4 003 152 765 10 15 200 
Oral cavity 1 270 75 951 491 36 178 1 761 64 1 129 
Pharynx 2 139 86 1 842 352 49 172 2 492 81 2 014 
Oesophagus 3 402 84 2 860 882 53 467 4 284 78 3 327 
Colon-rectum 9 792 30 2 949 8 523 750 18 314 20 3 699 
Liver 6 203 30 1 850 2 215 201 8 418 24 2 051 
Larynx 1 240 60 745 150 24 35 1 390 56 781 
Breast 220 12 547 18 2 199 12 767 17 2 199 
Circulatory system diseases 66 833 14 9 523 77 510 2 710 144 343 12 233 
Hypertensive disease 3 099 46 1 420 5 834 15 866 8 933 2 286 
Ischemic heart disease 20 774 938 15 765 −3 −529 36 539 409 
Alcoholic cardiomyopathy 74 100 74 10 100 10 84 100 84 
Cardiac arrhythmia 4 236 53 2 250 6 277 22 1 366 10 513 34 3 615 
Haemorrhagic stroke 7 734 48 3 704 10 139 827 17 874 25 4 531 
Ischemic stroke 5 532 19 1 050 8 581 150 14 112 1 200 
Oesophageal varices 88 100 88 20 100 20 108 100 108 
Digestive system diseases 12 543 45 5 627 10 888 20 2 145 23 431 33 7 772 
Alcoholic gastritis 100 100 100 
Alcoholic liver disease 3 543 100 3 543 1 317 100 1 317 4 860 100 4 860 
Chronic hepatitis, not elsewhere classified 15 100 15 10 100 10 25 100 25 
Fibrosis and cirrhosis of the liver 1 684 100 1 684 767 100 767 2 451 100 2 451 
Cholelithiasis 109 −47 −51 201 −10 −19 310 −23 −70 
Pancreatitis 596 74 438 428 17 71 1 024 50 509 
Other diseases 95 382 13 3 441 104 926 634 200 308 4 075 
Type 2 diabetes 3 048 21 638 3 311 −4 −143 6 359 494 
Wernicke’s encephalopathy 100 100 10 100 10 
Mental and behavioural disorders due to alcohol 2 251 100 2 251 608 100 608 2 859 100 2 859 
Degeneration of nervous system due to alcohol 75 100 75 26 100 26 101 100 101 
Epilepsy and status epilepticus 859 53 459 752 18 137 1 611 37 596 
Alcoholic polyneuritis 12 100 12 100 16 100 16 
External causes 22 620 25 5 545 14 748 18 2 609 37 368 22 8 154 
Unknownb 7 671 16 1 250 6 848 363 14 519 11 1 613 
Total 272 253 13 36 584 263 113 12 465 535 366 49 048 

aWe added to each site of cancer, a proportion of deaths from cancer of an ill-defined, secondary or unspecified site (ICD codes: C76–C80) and independent multiple sites (ICD code: C97) equal to the gender- and age-specific fraction of cancers of specified sites attributed to alcohol.

bA fraction of ill-defined and unknown causes of death (ICD codes: R96–R99) equal to the fraction of specified causes of deaths attributable to alcohol was attributed to alcohol for each gender and age group.

Results

Out of 535 000 deaths in 2009, 49 000 were attributable to alcohol: 36 500 among men and 12 500 among women, which represent 13 and 5%, respectively, of total deaths. Table 1 shows by gender for each cause of death, the number of deaths observed in 2009, the estimated AAF and the number of deaths attributable to alcohol or prevented by alcohol consumption. The total of about 36 500 deaths attributable to alcohol among men is obtained after subtracting 50 deaths from cholelithiasis prevented by alcohol. Similarly among women, the 12 500 deaths due to alcohol are obtained by subtracting about 700 deaths, mostly from ischemic heart disease and type 2 diabetes, prevented by alcohol. Alcohol consumption is a major cause of premature death. The AAFs are highest in the populations aged 15 to 34 (22%) and 35 to 64 (18%) versus 7% in the population aged 65 or over (table 2). This is particularly true among men with one death in four attributable to alcohol between ages 15 and 34 and one death in five between ages 35 and 64.

Table 2

AAF according to age, by gender and cause of death

Sex Cause of death Age (years)
 
  15–34
 
35–64
 
65+
 
Total 15+
 
  Deaths due to alcohol AAF (%) Deaths due to alcohol AAF (%) Deaths due to alcohol AAF (%) Deaths due to alcohol AAF (%) 
Men Cancer 14 4 336 16 6 847 11 11 197 12 
Circulatory system 33 10 1 316 13 8 174 15 9 523 14 
Digestive system 33 40 3 371 73 2 223 28 5 627 45 
External causes 1 215 35 2 202 25 2 128 21 5 545 25 
Other diseases 90 11 2 010 17 1 341 3 441 
Unknown 137 26 706 22 407 11 1 250 16 
Total 1 522 26 13 941 21 21 120 11 36 584 13 
Women Cancer 14 1 361 2 628 4 003 
Circulatory system 120 2 588 2 710 
Digestive system 15 1 172 64 966 11 2 145 20 
External causes 153 19 551 18 1 905 18 2 609 18 
Other diseases 468 159 634 
Unknown 17 148 12 197 363 
Total 201 3 820 12 8 443 12 465 
Both Total 1 724 22 17 761 18 29 563 49 048 
Sex Cause of death Age (years)
 
  15–34
 
35–64
 
65+
 
Total 15+
 
  Deaths due to alcohol AAF (%) Deaths due to alcohol AAF (%) Deaths due to alcohol AAF (%) Deaths due to alcohol AAF (%) 
Men Cancer 14 4 336 16 6 847 11 11 197 12 
Circulatory system 33 10 1 316 13 8 174 15 9 523 14 
Digestive system 33 40 3 371 73 2 223 28 5 627 45 
External causes 1 215 35 2 202 25 2 128 21 5 545 25 
Other diseases 90 11 2 010 17 1 341 3 441 
Unknown 137 26 706 22 407 11 1 250 16 
Total 1 522 26 13 941 21 21 120 11 36 584 13 
Women Cancer 14 1 361 2 628 4 003 
Circulatory system 120 2 588 2 710 
Digestive system 15 1 172 64 966 11 2 145 20 
External causes 153 19 551 18 1 905 18 2 609 18 
Other diseases 468 159 634 
Unknown 17 148 12 197 363 
Total 201 3 820 12 8 443 12 465 
Both Total 1 724 22 17 761 18 29 563 49 048 

Table 3 presents mortality in France for 2009, excluding external causes, according to the dose of alcohol. The AAF increases with the dose of alcohol from 0ċ4% at a dose of 13 g per day to 32% at the dose of 117 g per day among men and from 2 to 29% at the dose of 85 g per day among women. Figure 1 shows the total number of deaths and the number of deaths attributable to alcohol among men and women, for each dose category. The number of deaths not attributable to alcohol reflects the distribution of alcohol consumption in the population.

Table 3

France 2009 mortality (excluding external causes): number of deaths due to alcohol and not due to alcohol, according to dose of alcohol, by gender and cause of death

Cause of death Men
 
Women
 
Total 
  Average dose of alcohol in grams per day
 
Total Average dose of alcohol in grams per day
 
Total 
 13 28 46 117  13 28 46 85 
Cancer — 238 553 389 10 017 11 197 — 349 692 261 2 702 4 003 15 200 
Circulatory system — −342 −211 37 10 039 9 523 — −128 262 117 2 459 2 710 12 233 
Digestive system — 265 406 692 4 264 5 627 — 685 397 196 867 2 145 7 772 
Other diseases — −19 −10 56 3 413 3 441 — 22 20 51 542 634 4 075 
Unknownb — 30 57 73 1 090 1 250 — 44 56 28 236 363 1 613 
Deaths attributable to alcohol 172 796 1 247 28 824 31 038 972 1 426 652 6 806 9 856 40 894 
Deaths not attributable to alcohola 69 116 41 492 30 537 17 005 60 446 218 595 153 994 41 909 19 149 7 094 16 363 238 509 457 104 
Total number of deaths excl. ext causes 69 116 41 664 31 332 18 251 89 270 249 633 153 994 42 881 20 575 7 745 23 169 248 365 497 998 
Attributable fraction (%) 0.4 32 12 29 
Cause of death Men
 
Women
 
Total 
  Average dose of alcohol in grams per day
 
Total Average dose of alcohol in grams per day
 
Total 
 13 28 46 117  13 28 46 85 
Cancer — 238 553 389 10 017 11 197 — 349 692 261 2 702 4 003 15 200 
Circulatory system — −342 −211 37 10 039 9 523 — −128 262 117 2 459 2 710 12 233 
Digestive system — 265 406 692 4 264 5 627 — 685 397 196 867 2 145 7 772 
Other diseases — −19 −10 56 3 413 3 441 — 22 20 51 542 634 4 075 
Unknownb — 30 57 73 1 090 1 250 — 44 56 28 236 363 1 613 
Deaths attributable to alcohol 172 796 1 247 28 824 31 038 972 1 426 652 6 806 9 856 40 894 
Deaths not attributable to alcohola 69 116 41 492 30 537 17 005 60 446 218 595 153 994 41 909 19 149 7 094 16 363 238 509 457 104 
Total number of deaths excl. ext causes 69 116 41 664 31 332 18 251 89 270 249 633 153 994 42 881 20 575 7 745 23 169 248 365 497 998 
Attributable fraction (%) 0.4 32 12 29 

aThe total number of deaths not attributable to alcohol is distributed between dose categories according to proportion of drinkers in this dose category. For instance 69 116 = (249 633 − 31 038) × 31.6%.

The sensitivity analyses gave estimations of 45 000 and 42 000 alcohol-attributable deaths in 2009 when 10 and 20%, respectively, of 2009 alcohol sales were supposed not to have been consumed. Using the declared alcohol consumption gave an estimation of 28 000 deaths.

Discussion

We estimate a total of 49 000 alcohol-attributable deaths in France for 2009, 36 500 among men and 12 500 among women. Alcohol consumption is responsible of a large proportion of premature deaths. Even at the lowest dose of 13 g per day, where alcohol is associated with a reduced risk of cholelithiasis, ischemic heart disease, ischemic stroke and type 2 diabetes, the overall effect is detrimental.

These estimates rely on many assumptions and are, therefore, subject to uncertainty. Major sources of uncertainty are the distribution of alcohol consumption, the lag time between alcohol consumption and mortality, the pattern of drinking, the selection of causes of deaths with risk modified by alcohol, the quality of the death certification process and the estimation of the risk functions.

Alcohol consumption estimation

The literature discussing the various sources of bias in alcohol consumption measurement is considerable,32 and there is no consensus on the estimators that should be used to assess alcohol-attributable mortality. Estimations of alcohol consumption from survey data and alcohol available from national statistics are markedly different, resulting in different estimates in number of alcohol-attributable deaths: 28 000 versus 49 000 in our study. The dose–effect relationships are estimated from epidemiological studies based on self-reported alcohol consumption, and therefore, survey data may be preferable. However, underreporting may be less in epidemiological studies than in surveys of a national sample of the population because individuals interviewed in the context of an epidemiological study, knowing its public health objective, may be more truthful about their consumption of alcohol. The aim of this study being a realistic estimation of the number of deaths attributable to alcohol for the year 2009, we chose to use the distribution of drinkers available in a large recent representative survey of the French population15 adjusted on the amount of alcohol available for the year 2009. The underreporting of 2ċ4 is assumed to be independent of the amount of alcohol, but it is quite possible that it depends on the sex, the age and the true consumption of the person. Therefore, the hypothesis of a constant underreporting factor is a crude approximation; however, we are not aware of any data on which to base an improved estimation. The underreporting rate of 2ċ4 may be too high if part of the alcohol available is not drunk: the sensitivity analyses indicate that a reduction of 10 or 20% in the average alcohol consumption leads to a reduction of 4000 or 7000 deaths. Our estimates of 10 or 20% of alcohol wasted or spilled might be overestimates: a recent report33 quoting industry estimates states that less than 10% of alcohol is wasted or spilled.

Lag time

Methodological approaches for analysing the relationship between changes in alcohol consumption and changes in rates of alcohol-related harm have been suggested; however, the problem of time lags is not well addressed.34 The choice of the lag time is more crucial in France than in other countries, because the consumption of alcohol has decreased by 50% over the last 50 years. It was 33 g per adult per day in 1994, 30 g in 2002–03 and 27 g in 2009. For cancer, it would probably have been more appropriate to use alcohol consumption 15 or 20 years before 2009. However, most published estimates combine consumption and mortality data for the same year.2–7,9,10

Characteristics of drinking habits

We studied the effect of alcohol drinking, summarizing consumption by the average daily alcohol intake. This ignores possible different effects of binge drinking versus regular daily exposure. There is only fragmentary information on this factor. A recent meta-analysis showed that the cardioprotective effect of moderate alcohol consumption on ischemic heart disease disappeared when light to moderate drinking was mixed with irregular heavy drinking episodes.35 Mortality from external causes is increased by binge drinking.36

Causes of death associated with alcohol

We did not include conditions for which a link with alcohol is probable but not proven. Recent meta-analyses show no evidence of an association between alcohol consumption and prostate cancer,37 oesophageal and gastric cardia adenocarcinoma,38 and bladder cancer.39 There is some uncertainty in the sorting out of causes of death between the underlying cause and associated causes. We have not taken secondary causes into account.

Previous estimations of alcohol-attributable mortality in France

Several estimations of alcohol-attributable mortality have been published for France. Early studies, estimating alcohol exposure from a 1974 survey by the Institut Français d'Opinion Publique uncorrected for the amount of available alcohol, and ignoring the protective effects of alcohol, estimated 52 000 deaths in 1985 and 45 000 deaths in 1995.12,13 An estimation of alcohol-attributable cancer deaths in France was published by the IARC for the year 2000, using the total amount of alcohol available in 1985 (www.who.int); AAFs for all cancers were estimated at 9% for men and 3% for women in 2000 versus 12 and 6% in our study, and the total number of cancers attributable to alcohol was equal to 9900 versus 15 200 in our study; the difference comes essentially from different dose–effect relationships for oesophageal, colorectal and breast cancer. A recent estimation used data from a 2002 survey: the number of deaths attributable to alcohol for the year 2006 was equal to 33 000 after adjustment for the available amount of alcohol in 20021,4; this estimation is lower than ours due to the use of different surveys and risk equations. None of the previous studies presented the results by dose category.

Comparison with other countries

The proportion of deaths attributable to alcohol in France among men is 13%, much higher than in other countries; for instance, it is 6ċ8% in Scotland,4 5% in Switzerland,6 3% in East Germany3 and Italy,5 1% in West Germany and Denmark3 and −2ċ8 or −0ċ8%, that is a protective effect in the United Kingdom.3,10 Among women, the proportion of deaths attributable to alcohol is 4ċ7%, also higher than the 3ċ3% estimated for Scotland, the 2% observed in Italy, 1ċ4% in Switzerland, 1% in Denmark, 0ċ3% in West Germany and effects varying between 0ċ1 and −0ċ9% in the United Kingdom3,10 and −0ċ3% in East Germany. Liang et al.8 found that 4ċ4% of all cancer deaths were attributable to alcohol consumption in China in 2005, which is to be compared with 10% in our study.

The overall protective effect of alcohol in England and Wales9,10 can be directly compared with the highly detrimental effect of alcohol in France inasmuch as the methods used are quite similar. White et al. use the consumption as declared in the general household survey, whereas we adjust the declared consumption to account for the alcohol available for consumption. When we use the declared consumption, in our sensitivity analysis, we attribute 28 000 deaths to alcohol in France, whereas alcohol has an overall protective effect in England and Wales with 1800 deaths avoided. However, the distributions of causes of deaths in the populations are different. Our results in terms of age and dose effects are nevertheless similar to those observed by White et al.: an increased risk of mortality exists even among people drinking a low dose of alcohol, especially among young people.

Alcohol is a major public health issue in France

This study demonstrates the overall detrimental effect of moderate alcohol consumption, even at the low dose of 13 g of pure alcohol per day. More generally, it shows the major harmful effect of alcohol on health in France, leading to 49 000 deaths in the year 2009, 40% among individuals younger than 65 years. Our study addresses the effect of alcohol on mortality; it ignores the morbidity of drinkers, the morbidity of individuals involved in alcohol-related accidents and the morbidity of newborns exposed in utero. In 2005, France had the fourth highest consumption of alcohol among 48 European Member States after the Czech Republic, Estonia and Ireland.40 Efforts should be made to reduce further alcohol consumption in France and thus avoid many premature deaths.

Supplementary data

Supplementary data are available at EURPUB online.

Acknowledgements

The authors thank Lorna Saint Ange for editing.

Funding

This work was funded by Institut Gustave Roussy and a grant from the French “Institut National du Cancer” 2010

Conflicts of interest: None declared.

Key points

  • Our work presents an estimation of the number of deaths attributable to alcohol in France, in 2009, for each cause of death related to alcohol consumption.

  • The harmful effect of alcohol is larger in the younger population.

  • We estimated the number of deaths and the risks for different dose categories and showed that the overall effect of a moderate alcohol consumption is detrimental in France.

References

1
Hill
C
Laplanche
A
[The French drink too much alcohol]
Presse Med
 , 
2010
, vol. 
39
 (pg. 
158
-
64
)
2
Jones
L
Bellis
MA
Dedman
D
, et al.  . 
Liverpool: Center for public Health, Faculty of Health and Applied Social Sciences
Alcohol-Attributable Fractions for England
 , 
2008
Liverpool John
Moores University
(pg. 
1
-
53
)
3
Britton
A
Nolte
E
White
IR
, et al.  . 
A comparison of the alcohol-attributable mortality in four European countries
Eur J Epidemiol
 , 
2003
, vol. 
18
 (pg. 
643
-
51
)
4
Grant
I
Springbett
A
Graham
L
Alcohol Attributable Mortality and Morbidity: Alcohol Population Fractions for Scotland
 , 
2009
Edinburgh
National Services Scotland
(pg. 
1
-
50
)
5
Corrao
G
Rubbiati
L
Zambon
A
Arico
S
Alcohol-attributable and alcohol preventable mortality in Italy. A balance in 1983 and 1996
Eur J Public Health
 , 
2002
, vol. 
12
 (pg. 
214
-
23
)
6
Rehm
J
Taylor
B
Roerecke
M
Patra
J
Alcohol consumption and alcohol attributable burden of disease in Switzerland, 2002
Int J Public Health
 , 
2007
, vol. 
52
 (pg. 
383
-
92
)
7
Rehm
J
Patra
J
Taylor
B
Harm, benefits, and net effects on mortality of moderate drinking of alcohol among adults in Canada in 2002
Ann Epidemiol
 , 
2007
, vol. 
17
 (pg. 
S81
-
S86
)
8
Liang
H
Wang
J
Xiao
H
, et al.  . 
Estimation of cancer incidence and mortality attributable to alcohol drinking in China
BMC Public Health
 , 
2010
, vol. 
10
 (pg. 
730
-
5
)
9
White
IR
Altmann
DR
Nanchatal
K
Alcohol consumption and mortality: modelling risks for men and women at different ages
BMJ
 , 
2002
, vol. 
235
 (pg. 
191
-
7
)
10
White
IR
Altmann
DR
Nanchatal
K
Mortality in England and Wales attributable to any drinking, drinking above sensible limits and drinking above lowest-risk level
Addiction
 , 
2004
, vol. 
99
 (pg. 
749
-
56
)
11
Pignon
JP
Hill
C
[Estimation of alcohol-related deaths in France in 1985]
Gastroenterol Clin Biol
 , 
1991
, vol. 
15
 (pg. 
51
-
6
)
12
Hill
C
Alcool et risque de cancer
Actualité et Dossier en Santé Publique
 , 
2000
, vol. 
30
 (pg. 
14
-
7
)
13
Boffetta
M
Tubiana
C
Hill
C
, et al.  . 
The causes of cancer in France
Ann Oncol
 , 
2009
, vol. 
20
 (pg. 
550
-
5
)
14
Rey
G
Boniol
M
Jougla
E
Estimating the number of alcohol-attributable deaths: methodological issues and illustration with French data for 2006
Addiction
 , 
2010
, vol. 
105
 (pg. 
1018
-
29
)
15
 
Réseau Quételet, French Data Archives for Social Sciences, Centre Maurice Halbwachs, France. 2012
16
Turati
F
Garavello
W
Tramacere
I
, et al.  . 
A meta-analysis of alcohol drinking and oral and pharyngeal cancers. Part 2: results by subsites
Oral Oncol
 , 
2010
, vol. 
46
 (pg. 
720
-
6
)
17
Rota
M
Bellocco
R
Scotti
L
, et al.  . 
Random effects meta-regression models for studying nonlinear dose-response relationship, with an application to alcohol and esophageal squamous cell carcinoma
Stat Med
 , 
2010
, vol. 
29
 (pg. 
2679
-
87
)
18
Fedirko
I
Tramacere
V
Bagnardi
M
, et al.  . 
Alcohol drinking and colorectal cancer risk: an overall and dose-response meta-analysis of published studies
Ann Oncol
 , 
2011
, vol. 
22
 (pg. 
1958
-
72
)
19
Corrao
G
Bagnardi
V
Zambon
A
La Vecchia
C
A meta-analysis of alcohol consumption and the risk of 15 diseases
Prev Med
 , 
2004
, vol. 
38
 (pg. 
613
-
9
)
20
Islami
F
Tramacere
I
Rota
M
, et al.  . 
Alcohol drinking and laryngeal cancer: overall and dose-risk relation—a systematic review and meta-analysis
Oral Oncol
 , 
2010
, vol. 
46
 (pg. 
802
-
10
)
21
Key
J
Hodgson
S
Omar
RZ
, et al.  . 
Meta-analysis of studies of alcohol and breast cancer with consideration of the methodological issues
Cancer Causes Control
 , 
2006
, vol. 
17
 (pg. 
759
-
70
)
22
Gutjahr
E
Gmel
G
Rehm
J
Relation between average alcohol consumption and disease: an overview
Eur Addict Res
 , 
2001
, vol. 
7
 (pg. 
117
-
27
)
23
Patra
J
Taylor
B
Irving
H
, et al.  . 
Alcohol consumption and the risk of morbidity and mortality for different stroke types-a systematic review and meta-analysis
BMC Public Health
 , 
2010
, vol. 
10
 (pg. 
258
-
69
)
24
Taylor
B
Irving
HM
Baliunas
D
, et al.  . 
Alcohol and hypertension: gender differences in dose-response relationships determined through systematic review and meta-analysis
Addiction
 , 
2009
, vol. 
104
 (pg. 
1981
-
90
)
25
Irving
HM
Samokhvalov
AV
Rehm
J
Alcohol as a risk factor for pancreatitis. A systematic review and Meta-analysis
JOP
 , 
2009
, vol. 
10
 (pg. 
387
-
92
)
26
Baliunas
DO
Taylor
BJ
Irving
H
, et al.  . 
Alcohol as a risk factor for type 2 diabetes: a systematic review and meta-analysis
Diabetes Care
 , 
2009
, vol. 
32
 (pg. 
2123
-
32
)
27
Koppes
LL
Dekker
JM
Hendriks
HF
, et al.  . 
Moderate alcohol consumption lowers the risk of type 2 diabetes: a meta-analysis of prospective observational studies
Diabetes Care
 , 
2005
, vol. 
28
 (pg. 
719
-
25
)
28
Samokhvalov
AV
Irving
H
Mohapatra
S
Rehm
J
Alcohol consumption, unprovoked seizures, and epilepsy: a systematic review and meta-analysis
Epilepsia
 , 
2010
, vol. 
51
 (pg. 
1177
-
84
)
29
Rehm
J
Taylor
B
Mohapatra
S
, et al.  . 
Alcohol as a risk factor for liver cirrhosis: a systematic review and meta-analysis
Drug Alcohol Rev
 , 
2010
, vol. 
29
 (pg. 
437
-
45
)
30
Hanley
JA
A heuristic approach to the formulas for population attributable fraction
J Epidemiol Community Health
 , 
2001
, vol. 
55
 (pg. 
508
-
14
)
31
Rehm
J
Room
R
Monteiro
M
, et al.  . 
Ezzati
M
Lopez
AD
Rodgers
A
Murray
CJL
Alcohol use
Comparative Quantification of Health Risks: Global and Regional Burden of Diseases Attributable to Selected Major Risk Factors
 , 
2004
Geneva
World Health Organization
(pg. 
959
-
1108
)
32
Greenfield
TK
Kerr
WC
Alcohol measurement methodology in epidemiology: recent advances and opportunities
Addiction
 , 
2008
, vol. 
103
 (pg. 
1082
-
99
)
33
Robinson
M
Thorpe
R
Beeston
C
McCartney
G A
review of the validity and reliability of alcohol retail sales data for monitoring population levels of alcohol consumption: a Scottish perspective
Alcohol Alcohol
 , 
2013
, vol. 
48
 (pg. 
231
-
40
)
34
Holmes
J
Meier
PS
Booth
A
, et al.  . 
The temporal relationship between per capita alcohol consumption and harm: a systematic review of time lag specifications in aggregate time series analyses
Drug Alcohol Depend
 , 
2012
, vol. 
123
 (pg. 
7
-
14
)
35
Roerecke
M
Rehm
J
Irregular heavy drinking occasions and risk of ischemic heart disease: a systematic review and meta-analysis
Am J Epidemiol
 , 
2010
, vol. 
171
 (pg. 
33
-
644
)
36
Taylor
BJ
Shield
KD
Rehm
JT
Combining best evidence: a novel method to calculate the alcohol-attributable fraction and its variance for injury mortality
BMC Public Health
 , 
2011
, vol. 
11
 pg. 
265
 
37
Rota
M
Scotti
L
Turati
F
, et al.  . 
Alcohol consumption and prostate cancer risk: a meta-analysis of the dose-risk relation
Eur J Cancer Prev
 , 
2012
, vol. 
21
 (pg. 
350
-
9
)
38
Tramacere
I
Pelucchi
C
Bagnardi
V
, et al.  . 
A meta-analysis on alcohol drinking and esophageal and gastric cardia adenocarcinoma risk
Ann Oncol
 , 
2012
, vol. 
23
 (pg. 
287
-
97
)
39
Pelucchi
C
Galeone
C
Tramacere
I
, et al.  . 
Alcohol drinking and bladder cancer risk: a meta-analysis
Ann Oncol
 , 
2012
, vol. 
23
 (pg. 
1586
-
93
)
40
WHO Regional Office for Europe
 
European status report on alcohol and health, Geneva, 2010. Available at: http://apps.who.int/gho/data/?showonly=GISAH&region=euro# (18 February 2013, date last accessed)

Supplementary data

Comments

0 Comments