Kari Poikolainen

Finnish Foundation for Alcohol Studies, P.O. Box 220, FIN‑00531 Helsinki, Finland


Also published partly in:

Poikolainen K. Quién pierde en esa apuesta: consecuencias del consumo de las drogas, In: La Drogadicción en México - indiferencia o prevención, ING Seguros, 2001, pp. 175-194.


Last updated June 25, 2002







Consumption of alcoholic beverages may influence the risk of death, disease, injuries and social problems in many ways. These effects are briefly reviewed here. Relative risks are presented either as risk ratios (RR) or as percentage increase or decrease relative to the control group. The RR for control group is taken to be one (RR=1). Alcohol intake is expressed as grams (g) of alcohol. One 33 cl bottle of beer, a 12 cl glass of non‑fortified wine and a 4 cl shot of liquor contains approximately 11‑14 g of alcohol. The emphasis is on causal effects, incidence of problems, and long‑term average intake of alcohol. Most results pertain to average intake of alcohol, without consideration of drinking patterns, and an underlying implicit assumption is that the self‑reported alcohol intake estimates reflect the true long‑term average intake. However, drinking patterns, blood alcohol concentrations and beverage type will also be considered. Understanding the art of measuring alcohol intake is paramount to evaluating the results of the studies reviewed.


Measuring alcohol intake


Alcohol consumption is a surprisingly complex behaviour. There are no simple answers for complex problems. The most appropriate way of asking questions depends on many things, such as

‑ the purpose of the study

‑ the resources available

‑ the educational level of the study subjects

‑ the drinking patterns of the subjects

‑ the willingness of the subjects to cooperate.

The best option for a certain study is always a compromise between several demands and possibilities. The best in the ideal world may be the enemy of good in the real world.

Basically, there are two traditions of measuring alcohol intake, the tradition asking about the customary or usual alcohol intake and another tradition mapping out the actual drinking occasions. A comprehensive review has earlier been published by Robin Room (Room, 1991).




In general population, the relation between alcohol intake and total mortality is curvilinear (resembling the letter "J"). Abstainers have slightly higher mortality than moderate drinkers, and heavy drinkers much higher mortality than the former two groups. A meta‑analysis of follow‑up studies showed that mortality was lowest among men consuming on the average 10‑19 g of alcohol per day and among women imbibing 1‑9 g of alcohol per day. Compared with abstainers, men drinking at that level had 16% and women 12% lower mortality. Men with intake of 40 g and women with 20 g of alcohol had death rates similar to abstainers (English et al., 1995). A very large and well‑controlled follow‑up study from the U.S.A. broadly agrees with the above findings (Thun et al., 1997). In Finland, increased mortality has been found in relation to increased frequency of intoxication, increased frequency of hangover and the frequency of morning drinking (Poikolainen, 1983).


In most large studies, the mortality of alcoholics after treatment is from two‑ to six‑fold compared with the general population (Bruun et al., 1975). Such death rates shorten the life span by 6‑18 years. Excess mortality is mostly due to alcohol, tobacco, and unhealthy living habits. Approximately one half of deaths are tobacco‑related (Hurt et al., 1996).


People with alcohol dependence have more concentrated drinking pattern that the other drinkers. In a large study in the U.S.A., the average frequency of intoxication was 13‑fold, the frequency of drinking 5 or more drinks per day was 7‑fold, the prevalence of morning drinking 5‑fold, and the average intake 4‑fold among the alcohol dependent compared with the not dependent (Dawson, 2000b). This study also found that among the alcohol dependent population the mortality increased linearly by alcohol intake in contrast to the curvilinear pattern among the rest. Adjusted for smoking and other risk factors, subjects with alcohol dependence (alcohol intake > 52 g/day) sampled for general population showed 1.65‑fold mortality compared with never‑drinkers (Dawson, 2000a).




In agreement with findings on mortality, several studies have found that the relation between alcohol intake and the overall occurrence of diseases, hospital admissions and leaves of absence is curvilinear (Poikolainen, 1996; Haapanen‑Niemi et al., 1999).

Heavy alcohol intake increases strongly the incidence of liver cirrhosis, aerodigestive cancer, haemorrhagic stroke, hypertension and injuries. Less strong significant increases can be found for chronic pancreatitis and cancers of the liver, colon, and rectum. Alcohol is not related to the incidence of ischemic stroke, peptic or duodenal ulcer. These findings are based on a meta‑analysis of 123 high‑quality studies and adjusted for major confounders (Corrao et al., 1999). Heavy drinking also causes cardiomyopathy, peripheral neuropathy, myopathy and hepatitis (English et al., 1995). Case‑control studies of prevalent cases suggest that alcohol may cause psoriasis and infectious eczema (Poikolainen et al., 1990; Karvonen et al., 1992).


Increased risk for the (female) breast cancer have been found in several studies among drinkers compared with abstainers (Corrao et al., 1999), but the low relative risk and multitude of potential confounding factors make it difficult to draw any firm conclusions on causality. Moreover, there is no clear dose‑response relationship nor any plausible biologic mechanism.


Risk of spontaneous abortion, intrauterine growth retardation, prematurity and fetal alcohol syndrome are increased by alcohol intake (English et al., 1995). There is no increase of fetal malformations at the level of <24 g/day (Polygenis et al., 1998). In the U.S.A., one follow‑up study found out that alcohol intake and especially drinking early in pregnancy and binge drinking were positively related to lower speed of information processing at the age of four years, and lower arithmetic and educational performance at the age of seven to eight years(Streissguth, 1986). In France, another study found out that children exposed during the first trimester to 45 g of alcohol or more per day had lower cognitive function than those with less exposure at the age of four and a half years (Larroque et al., 1995). The risks were adjusted for several potential confounders in both these studies. Children exposed to alcohol in utero have also been found to have higher alcohol intake in adolescence (Baer et al., 1998) and as adults to have higher alcohol, drug, and nicotine dependence symptom counts than those not exposed (Yates et al., 1998). Available data are not sufficient to indicate whether there is a safe limit or not for the cognitive developmental deficits(Larroque and Kaminski, 1997). The only absolutely safe course is to abstain during pregnancy.


Compared with never‑drinkers, the lowest relative risk of coronary heart disease has been found to be 0.78 at the level of 29 g/day according to a meta‑analysis of 51 high‑quality studies (Corrao et al., 2000). Increasing alcohol intake from the very low level of <12 g/week to the moderate level <72g/week has been found to decrease the incidence of cardiovascular diseases by 29% (Sesso et al., 2000). Human angiographic studies have shown that abstainers have more atherosclerotic lesions than moderate drinkers. Likewise, incident carotic atherosclerosis has been found to be less common among moderate drinkers (1‑50 g of alcohol a day) than among irregular drinkers (Kiechl et al., 1996). In contrast, heavy drinkers have been found to have increased atherosclerosis  (Gruchow et al., 1982), increased risk of tachyarrhythmias, and decreased variability of the heart rhythm (Koskinen and Kupari, 1992).


 The present evidence strongly supports the view that moderate alcohol intake decreases the risk of coronary heart disease. This decrease seems to be confined to steady moderate drinking. In an Australian case‑control study decreased coronary heart disease risk was found at the level of 10‑20 g/day on five or six days a week among both men and women. In contrast, increased coronary heart disease risk was found among men who consumed 90 g/day or more a day on 1‑2 days a week. Also women having 50 g/day or more daily had increased coronary heart disease risk compared with lifelong abstainers (McElduff and Dobson, 1997).


Most of the decrease in coronary heart disease risk is due to an increase in HDL‑cholesterol. A meta‑analysis of randomized controlled studies showed that at the level of 30 g/day the mean increase in HDL‑cholesterol is on the average 13% (1.0 mmol/l). Smaller beneficial changes were due to decreased blood clotting tendency (Rimm et al., 1999). This increase is under genetic control. Moderate drinkers who are homozygous for the slow‑oxidizing ADH3 allele of alcohol dehydrogenase have the highest HDL‑cholesterol levels and highest reduction in the risk of myocardial infarction (Hines et al., 2001).  Blood pressure is likely to increase by alcohol intake after the consumption exceeds 25 g/day (Puddey et al., 1985; Corrao et al., 1999). This effect is likely to increase the risk of coronary heart disease and haemorrhagic stroke. If the average intake level and pattern are moderate, the beneficial changes surpass the harmful ones.


Compared with abstainers, the incidence of thromboembolic stroke (due to blocked blood vessels) and gallstones is decreased among moderate drinkers (English et al., 1995). One study suggests that moderate drinking protects from chronic obstructive pulmonary disease (Tabak et al., 2001).


There are some experimental observations suggesting that wine might be especially beneficial for the heart because its flavonoid and other antioxidant content. In epidemiologic studies, some results favour wine, some beer and some liquor while some studies have not been able to find any differences between these beverage types (Rimm et al., 1996). No firm conclusions can be drawn at present. The differences between beer, wine and liquor might be explained by varying drinking patterns related to different beverages.


Cross‑sectional population studies have found lower insulin levels among drinkers than abstainers (Kiechl et al., 1996; Lazarus et al., 1997). One study has found that the incidence of adult‑type diabetes was 39% lower among male moderate drinkers than among abstainers (Rimm et al., 1995). However, heavy drinking may impair glucose balance and lead to impotence, retinopathy and peripheric neuropathy(Swade and Emanuele, 1997).




Alcohol weakens sensorimotor coordination. Thus alcohol, especially intoxicating drinking, increases the risk of accidents, violence and self‑harm. Alcohol has been found to increase accidental falls, bicycle accidents, pedestrian accidents, fire injuries and deaths, drowning, occupational injuries, suicide, assaults and homicide (English et al., 1995).


The probability of become involved in a crash and that of a fatal traffic accident increases by blood alcohol concentration (BAC). Some studies show an exponential increase in risk by BAC, others do not (Chafetz, 1975; Zador, 1991). Some case‑control studies controlled for the site and time of the accident, while some did not. Data for fatal single vehicle accidents on the risk difference between zero and <0.05% BAC is inconclusive. A simulation study suggests that some driving‑related functions are impaired at 0.02% and all at 0.04% (Moskowitz et al., 2000). According to a meta‑analysis of nine Australian case‑control studies the risk of road traffic accident increases exponentially by BAC (English et al., 1995). The latter data were adjusted were made for sex, age, time of the day and day of the week. The control data and that for fatal accidents came from different states in Australia, and no adjustments were made for the accident location and the state of the vehicle (Lloyd, 1992).


Self‑reported health and functional capacity


Self‑reported health is a fairly good indicator of disease and life expectancy (Kennedy et al., 2001). Compared with lifelong non‑drinkers, self‑reported health has been found to be better among moderate drinkers and worse among heavy drinkers in never smokers in Finland (Poikolainen et al., 1996). The group drinking from 12 to 36 g/day had the best self‑reported health. The risk ratios were adjusted for sex, age, education, marital status, lack of friends, disability pension, smoking, being an ex‑drinker, and decrease of alcohol intake during the past 12 months due to health problems.

In the Netherlands, chronic conditions, self‑reported general health, and health complaints were studied in a general population sample. Abstainers were excluded. Drinking 6 glasses a day (alcohol content not reported) on 1‑2 days a week was associated with more reports of chronic diseases, compared to the reference group in which the pattern was 1‑2 glasses per day on 1‑2 days a week. Drinking 3‑5 glasses per day on 3‑5 days a week as well as drinking 1‑2 glasses a day on 6‑7 days a week was associated with better self‑reported health compared with the reference group (San Jose et al., 2000).


In one U.S. study, moderate drinkers were found to have better physical fitness than abstainers (Braun et al., 1995). However, heavy drinking is likely to decrease fitness (Urbano‑Marquez et al., 1995). Even small amounts of alcohol may cause harm in during the exercise, sports or work activities.


Mental health


Many cross‑sectional studies have found that moderate drinkers have better emotional and social adjustment and less psychiatric hospital admissions than abstainers (Chick, 1999). These differences may, however, be due to selection of people with mental problems into the group of abstainers. At present, there is no clear evidence supporting the idea that moderate alcohol intake would improve mental health. Alcohol is likely to be harmful for the mentally ill, because it may aggravate symptoms and interfere with drug treatment. Heavy drinking, on the other hand, increases the risk of depression (Poikolainen, 1994). One possible mechanism is stimulation of the release of stress hormones by heavy drinking. In both case‑control and cohort studies, the risk of dementia has been found to be lower among moderate drinkers than among abstainers (Orgogozo et al., 1997; Eckardt et al., 1998). This association is strong and not refuted by control of confounding.


Family problems


In the family, alcohol intake during pregnancy may impair the health of the child. Drinking by parents may impair interpersonal relations, and increase the risk of heavy drinking and alcohol dependence in the child. Excessive alcohol intake can bring about financial problems, domestic violence, abuse and neglect. A problem drinker in the family may not fulfill his or her expected role and duties, increasing the tasks of the other spouse and the children. Social life may suffer because of shame, anxiety and depression. These problems may increase because of a divorce (Eurocare, 1998). In a British study, the degree of psychological problems among the grown‑up children depended less on problem drinking among the parents than on the level of disharmony in the family (Eurocare, 1998). In a Mexico City sample, the level of mental distress and open conflict in the family correlated closely with the level of tolerant‑inactive coping, i.e. excusing, covering up, defending, denying, pretending that everything is normal and making threats that were not carried out (Orford et al., 2001).


Social problems


 The frequency of alcohol intoxication has been found to relate to belligerence, as well as to several legal, family and social problems (Harford et al., 1991). Increased frequency of alcohol intoxication has been found to relate to more drunken fights and more injuries (Rossow, 1996). A cross‑sectional Nordic study found that the frequency of alcohol intoxication associated more strongly than average alcohol intake with negative consequences, such as reckless behaviour but also with positive consequences, such as more easy‑going social interaction (Hauge and Irgens‑Jensen, 1990).

In a Swiss follow‑up study,  weekly heavy drinking occasions, i.e. the intake of 50 g of alcohol or more, was found to relate to increased problems with the police and other authorities as well as with problems in personal relationships while light occasional or light steady drinking were not. Daily intake of 40 g or more was related to an increased risk for problems with the police but not with personal problems. No pattern of drinking was related to the risk of unemployment (Rehm and Gmel, 1999).


"Safe" and optimal levels


Unfortunately, it is unknown what is the average safe level of intake and what is the optimal moderate intake. The self‑reports of alcohol intake tend to be underestimates of actual consumption. Although we know that there is some underestimation, it is difficult to correct for it because the degree of underestimation varies between studies and between individuals. The average percentage of underestimation varies greatly (from 29% to 83%) between studies (Poikolainen, 1995). However, a conservative estimate, based on the face value of the available evidence, suggests that the optimal level for a healthy adult man may be around one to three drinks a day (10‑40g/day), and for a woman slightly less, perhaps one to two drinks a day (10‑20 g/day). This pertains both to the risk of CHD and all‑cause mortality. Thus, all advice is approximate and open to criticism. Cross‑nationally, recommendations for sensible or "safe" drinking have varied between 13.5‑70 g/day for men and 10.8‑50 g/day for women (ICAP, 1996).


Practical advice


In advising individual persons, the best course of action might be to compare carefully the potential benefits and costs of alcohol intake. Protection from CHD by alcohol seems to be especially beneficial for men, for the old and middle‑aged, for patients with CHD, and for healthy individuals with a family history of CHD. On the other hand, abstaining might be the healthiest choice for many patients with certain diseases, such as depression, hypertension, liver cirrhosis, or peptic ulcer, when these are alcohol‑related. Risk of alcohol dependence should also be considered.

Familial alcoholism increases the risk of alcoholism 3 to 4‑fold among the offspring compared with persons without alcoholism in the family. Persons suffering from major depression, anxiety disorder, antisocial disorder, or conduct disorder are also at increased risk. Future alcoholics tend to be heavy drinkers already in adolescence. Of all patients aged 18‑62 years with the diagnosis of alcohol dependence, 97% had started drinking alcoholic beverages before the age of 18 years (Prescott et al., 1999). There are only a few studies that help us to estimate the potential risk of alcoholism among moderate drinkers. Among Swedish male conscripts aged 18‑21 years, the adjusted risk of hospital inpatient admission for alcoholism over a follow up of 15 years was not significantly different between men reporting abstinence from alcohol at the baseline examination lower and those reporting an intake of 1‑100 g of alcohol a week (Andreasson and Allebeck, 1991). In a cohort of adult Californians, both men and women changing from abstinence to moderate drinking had nonsignificantly lower age‑adjusted all‑cause mortality than adults who continued to abstain (Lazarus et al., 1991). In the assessment of risks and benefits, monitoring of blood pressure and other risk factors seems to be advisable. Drinkers should know what moderation means, watch for any harmful effects of alcohol, and avoid intoxication.




 Andreasson, S. and Allebeck, P. (1991) Alcohol and psychiatric illness: longitudinal study of psychiatric admissions in a cohort of Swedish conscripts. Int J Addict 26, 713‑28.


 Baer, J. S., Barr, H. M., Bookstein, F. L., Sampson, P. D. and Streissguth, A. P. (1998) Prenatal alcohol exposure and family history of alcoholism in the etiology of adolescent alcohol problems. J Stud Alcohol 59, 533‑43.


 Braun, B. L., Wagenaar, A. C. and Flack, J. M. (1995) Alcohol consumption and physical fitness among young adults. Alcohol Clin Exp Res 19, 1048‑54.


 Bruun, K., Edwards, G., Lumio, M., Mäkelä, K., Pan, L., Popham, R. E., Room, R., Schmidt, W., Skog, O.‑J., Sulkunen, P. and Österberg, E. (1975) Alcohol control policies in public health perspective. , Vol. 25. Forssa: Finnish Foundation for Alcohol Studies, Forssa.


 Chafetz, M. E. (1975) Second Special report to the U.S. Congress on alcohol and health. U.S. Department of health, education, and welfare, Washington, D.C.

 Chick, J. (1999) Can light or moderate drinking benefit mental health? Eur Addict Res 5, 74‑81.


 Corrao, G., Bagnardi, V., Zambon, A. and Arico, S. (1999) Exploring the dose‑response relationship between alcohol consumption and the risk of several alcohol‑related conditions: a meta‑analysis. Addiction 94, 1551‑73.


 Corrao, G., Rubbiati, L., Bagnardi, V., Zambon, A. and Poikolainen, K. (2000) Alcohol and coronary heart disease: a meta‑analysis. Addiction 95, 1505‑23.


 Dawson, D. A. (2000a) Alcohol consumption, alcohol dependence, and all‑cause mortality. Alcohol Clin Exp Res 24, 72‑81.


 Dawson, D. A. (2000b) Drinking patterns among individuals with and without DSM‑IV alcohol use disorders. J Stud Alcohol 61, 111‑20.


 Eckardt, M. J., File, S. E., Gessa, G. L., Grant, K. A., Guerri, C., Hoffman, P. L., Kalant, H., Koob, G. F., Li, T. K. and Tabakoff, B. (1998) Effects of moderate alcohol consumption on the central nervous system. Alcohol Clin Exp Res 22, 998‑1040.


 English, D. R., Holman, C. D. J., Milne, E., Winter, M. G., Hulse, G. K., Codde, J. P., Bower, C. I., B., C., de Klerk, N., Knuiman, M. W., Kurinczuk, J. J., Lewin, G. F. and Ryan, G. A. (1995) The quantification of drug caused morbidity and mortality in Australia, 1995 edition. . Commonwealth Department of Human Services and Health, Canberra.


 Eurocare (1998) Alcohol problems in the family ‑ A report to the European Union. . Eurocare, St. Ives, Cambridgeshire.


 Gruchow, H. W., Hoffmann, R. G., Anderson, A. J. and Barboriak, J. J. (1982) Effects of drinking patterns on the relationship between alcohol an coronary occlusion. Atherosclerosis 43, 393‑404.


 Haapanen‑Niemi, N., Miilunpalo, S., Vuori, I., Pasanen, M. and Oja, P. (1999) The impact of smoking, alcohol consumption, and physical activity on use of hospital services. Am J Public Health 89, 691‑8.


 Harford, T. C., Grant, B. F. and Hasin, D. S. (1991) The effect of average daily consumption and frequency of intoxication on the occurrence of dependence symptoms and alcohol‑related problems. In Alcohol in America: drinking practices and problems, Clark, W. B. and Hilton, M. E. eds, pp. 213‑237. State University of New York Press, Albany, New York.


 Hauge, R. and Irgens‑Jensen, O. (1990) The experiencing of positive consequences of drinking in four Scandinavian countries. Br J Addict 85, 645‑53.


 Hines, L. M., Stampfer, M. J., Ma, J., Gaziano, J. M., Ridker, P. M., Hankinson, S. E., Sacks, F., Rimm, E. B. and Hunter, D. J. (2001) Genetic variation in alcohol dehydrogenase and the beneficial effect of moderate alcohol consumption on myocardial infarction. N Engl J Med 344, 549‑55.


 Hurt, R. D., Offord, K. P., Croghan, I. T., Gomez‑Dahl, L., Kottke, T. E., Morse, R. M. and Melton, L. J., 3rd (1996) Mortality following inpatient addictions treatment. Role of tobacco use in a community‑based cohort. Jama 275, 1097‑103.


 ICAP (1996) Safe alcohol consumption: a comparison of Nutrition and Your Health: Dietary Guidelines for Americans and Sensible Drinking, Report 1. International Center for Alcohol Policies, Washington, D.C.


 Karvonen, J., Poikolainen, K., Reunala, T. and Juvakoski, T. (1992) Alcohol and smoking: risk factors for infectious eczematoid dermatitis? Acta Derm Venereol 72, 208‑10.


 Kennedy, B., Kasl, S. and Vaccarino, V. V. (2001) Repeated hospitalizations and self‑rated health among the elderly: a multivariate failure time analysis. Am J Epidemiol 153, 232‑41.


 Kiechl, S., Willeit, J., Poewe, W., Egger, G., Oberhollenzer, F., Muggeo, M. and Bonora, E. (1996) Insulin sensitivity and regular alcohol consumption: large, prospective, cross sectional population study (Bruneck study). BMJ 313, 1040‑4.


 Koskinen, P. and Kupari, M. (1992) Alcohol and cardiac arrhythmias. BMJ 304, 1394‑5.


 Larroque, B. and Kaminski, M. (1997) Larroque and Kaminski respond. Am J Public Health 87, 299‑301.


 Larroque, B., Kaminski, M., Dehaene, P., Subtil, D., Delfosse, M. J. and Querleu, D. (1995) Moderate prenatal alcohol exposure and psychomotor development at preschool age. Am J Public Health 85, 1654‑61.


 Lazarus, N. B., Kaplan, G. A., Cohen, R. D. and Leu, D. J. (1991) Change in alcohol consumption and risk of death from all causes and from ischaemic heart disease. BMJ 303, 553‑6.


 Lazarus, R., Sparrow, D. and Weiss, S. T. (1997) Alcohol intake and insulin levels. The Normative Aging Study. Am J Epidemiol 145, 909‑16.


 Lloyd, C. J. (1992) Alcohol and fatal road accidents: estimates of risk in Australia 1983. Accid Anal Prev 24, 339‑48.


 McElduff, P. and Dobson, A. J. (1997) How much alcohol and how often? Population based case‑control study of alcohol consumption and risk of a major coronary event. BMJ 314, 1159‑64.


 Moskowitz, H., Burns, M., Fiorentino, D., Smiley, A. and Zador, P. (2000) Driver Characteristics and Impairment at Various BACs. U.S. Department of Transportation National Highway Traffic Safety Administration, Washington, D.C.


 Orford, J., Natera, G., Velleman, R., Copello, A., Bowie, N., Bradbury, C., Davies, J., Mora, J., Nava, A., Rigby, K. and Tiburcio, M. (2001) Ways of coping and the health of relatives facing drug and alcohol problems in Mexico and England. Addiction 96, 761‑74.


 Orgogozo, J. M., Dartigues, J. F., Lafont, S., Letenneur, L., Commenges, D., Salamon, R., Renaud, S. and Breteler, M. B. (1997) Wine consumption and dementia in the elderly: a prospective community study in the Bordeaux area. Rev Neurol (Paris) 153, 185‑92.


 Poikolainen, K. (1983) Inebriation and mortality. Int J Epidemiol 12, 151‑5.


 Poikolainen, K. (1994) Depression in alcoholism: a review. Psychiatria Fennica 25, 75‑87.


 Poikolainen, K. (1995) Alcohol and mortality: a review. J Clin Epidemiol 48, 455‑65.


 Poikolainen, K. (1996) Alcohol and overall health outcomes. Ann Med 28, 381‑4.


 Poikolainen, K., Reunala, T., Karvonen, J., Lauharanta, J. and Kärkkäinen, P. (1990) Alcohol intake: a risk factor for psoriasis in young and middle aged men? BMJ 300, 780‑3.


 Poikolainen, K., Vartiainen, E. and Korhonen, H. J. (1996) Alcohol intake and subjective health. Am J Epidemiol 144, 346‑50.


 Polygenis, D., Wharton, S., Malmberg, C., Sherman, N., Kennedy, D., Koren, G. and Einarson, T. R. (1998) Moderate alcohol consumption during pregnancy and the incidence of fetal malformations: a meta‑analysis. Neurotoxicol Teratol 20, 61‑7.


 Prescott, E., Gronbaek, M., Becker, U. and Sorensen, T. I. (1999) Alcohol intake and the risk of lung cancer: influence of type of alcoholic beverage. Am J Epidemiol 149, 463‑70.


 Puddey, I. B., Beilin, L. J., Vandongen, R., Rouse, I. L. and Rogers, P. (1985) Evidence for a direct effect of alcohol consumption on blood pressure in normotensive men. A randomized controlled trial. Hypertension 7, 707‑13.


 Rehm, J. and Gmel, G. (1999) Patterns of alcohol consumption and social consequences. Results from an 8‑year follow‑up study in Switzerland. Addiction 94, 899‑912.


 Rimm, E. B., Chan, J., Stampfer, M. J., Colditz, G. A. and Willett, W. C. (1995) Prospective study of cigarette smoking, alcohol use, and the risk of diabetes in men. BMJ 310, 555‑9.


 Rimm, E. B., Klatsky, A., Grobbee, D. and Stampfer, M. J. (1996) Review of moderate alcohol consumption and reduced risk of coronary heart disease: is the effect due to beer, wine, or spirits. BMJ 312, 731‑6.


 Rimm, E. B., Williams, P., Fosher, K., Criqui, M. and Stampfer, M. J. (1999) Moderate alcohol intake and lower risk of coronary heart disease: meta‑analysis of effects on lipids and haemostatic factors. BMJ 319, 1523‑8.


 Room, R. (1991) Measuring alcohol consumption in the U.S.: methods and rationales. In Alcohol in America: drinking practices and problems, Clark, W. B. and Hilton, M. E. eds, pp. 26‑50. State University of New York Press, Albany, N.Y.


 Rossow, I. (1996) Alcohol‑related violence: the impact of drinking pattern and drinking context. Addiction 91, 1651‑61.


 San Jose, B., van Oers, J. A., van de Mheen, H., Garretsen, H. F. and Mackenbach, J. P. (2000) Drinking patterns and health outcomes: occasional versus regular drinking. Addiction 95, 865‑72.


 Sesso, H. D., Stampfer, M. J., Rosner, B., Hennekens, C. H., Manson, J. E. and Gaziano, J. M. (2000) Seven‑year changes in alcohol consumption and subsequent risk of cardiovascular disease in men. Arch Intern Med 160, 2605‑12.


 Streissguth, A. (1986) dose‑response study of the enduring effects of prenatal exposure to alcohol: birth to 14 years. In Alcohol, pregnancy and the developing child, Spohr, H.‑L. and Steinhausen, H.‑C. eds, pp. 141‑168. Cambridge University Press, Cambridge.


 Swade, T. F. and Emanuele, N. V. (1997) Alcohol & diabetes. Compr Ther 23, 135‑40.

 Tabak, C., Smit, H. A., Rasanen, L., Fidanza, F., Menotti, A., Nissinen, A., Feskens, E. J., Heederik, D. and Kromhout, D. (2001) Alcohol consumption in relation to 20‑year COPD mortality and pulmonary function in middle‑aged men from three European countries. Epidemiology 12, 239‑45.


 Thun, M. J., Peto, R., Lopez, A. D., Monaco, J. H., Henley, S. J., Heath, C. W., Jr. and Doll, R. (1997) Alcohol consumption and mortality among middle‑aged and elderly U.S. adults. N Engl J Med 337, 1705‑14.


 Urbano‑Marquez, A., Estruch, R., Fernandez‑Sola, J., Nicolas, J. M., Pare, J. C. and Rubin, E. (1995) The greater risk of alcoholic cardiomyopathy and myopathy in women compared with men. Jama 274, 149‑54.


 Yates, W. R., Cadoret, R. J., Troughton, E. P., Stewart, M. and Giunta, T. S. (1998) Effect of fetal alcohol exposure on adult symptoms of nicotine, alcohol, and drug dependence. Alcohol Clin Exp Res 22, 914‑20.


 Zador, P. L. (1991) Alcohol‑related relative risk of fatal driver injuries in relation to driver age and sex. J Stud Alcohol 52, 302‑10.