It is now clear that coronary heart disease (CHD) risk is lower among light-moderate alcohol drinkers than among abstainers. Consistency in observational studies plus plausible mechanisms for CHD protection by alcohol lead many scientists to accept a causal relationship (1-3). Media dissemination and advice by health professionals has made much of the public aware of benefits of light drinking. There is concern that such knowledge, perhaps especially in persons at CHD risk, might lead some to inappropriately start drinking or to become heavy drinkers (4-5). Except for anecdotal assertions, little data have been presented about this potential problem. A recent study (6) examined this concern.
A 20 Year Follow-Up Study
Subjects were drawn from 63,422 white or black men and women who supplied alcohol drinking information in 1978-85 at voluntary health examinations offered to members of a prepaid health care plan and who remained members in 1999. An examination questionnaire included queries about alcohol intake, beverage choice, smoking, and coffee. Since the objective was to study light/moderate drinkers and abstainers, exdrinkers and heavy drinkers at baseline were excluded. Among lifelong abstainers and three light-moderate drinking groups, 3542 persons were later hospitalized for CHD; 400 of these were randomly selected as survey subjects. Several closely matched controls with no CHD were selected for each CHD subject.
A letter was sent in 2000 which included a copy of the 1978-85 questionnaire.Questionnaire data were returned by 247 CHD subjects and 567 controls (mean age = 70). Responders and nonresponders were similar in demographic traits and baseline habits. The average interval between original and repeat questionnaires was 19.8 years. Telephone interviews were granted by 187 CHD subjects and 382 non-CHD controls with no major demographic or habit differences. At interviews CHD history was confirmed and subjects were queried about the nature and reasons for changes in drinking. Subjects were also asked: “Have you heard of health benefits of light to moderate drinking?” and if “yes”: “What benefits have you heard about?” Responses were coded as “medical - cardiovascular”, “medical non-cardiovascular”, or “non-medical”.
Few Start or Increase Drinking
On questionnaires 55/ 58 baseline non-drinking CHD subjects and 106/117 controls remained nondrinkers in 2000. Of those starting drinking, 1 reported drinking <once/month, 3 reported >once/month but <daily, and none reported daily drinking. Thus, 92% (161/175) of all nondrinkers remained abstainers 20 years later, and only 3 (1.7%) reported drinking > once/month. Among CHD subjects who were drinkers in 1978-85, 18% (34/189) became abstainers by 2000; identical to the proportion of controls that became nondrinkers (82/450). Table 1 shows case/control comparisons of questionnaire data; the only statistically significant difference (p < 0.05) was that a larger percentage of CHD subjects stopped smoking. Proportions of persons decreasing total drinking substantially outnumbered those increasing drinking (33% vs 15%; p < 0.001). Persons reporting decreased liquor or beer drinking outnumbered those reporting increases by > 2/1; for wine this ratio was ~ 1.5/1 (p < 0.001 for each type).
Coronary Disease Results in Decreased Drinking
At telephone interview both cases and controls were more likely than on written questionnaires to report reduced drinking with CHD subjects (52%) more likely than controls (41%) to so report (p = 0.01). This reduction by CHD subjects was substantially attributed to cardiovascular disease, diabetes or medication (Table 2). Nonmedical reasons for reduced drinking were similar in cases and controls (39% vs. 36%, p = 0.6).Reductions because of cardio-vascular disease were reported by persons who previously drank one-two drinks/day (25% of cases vs. 7% of controls; p = 0.003) and those who previously drank <one drink/day (17.9% of cases vs. 2.4% of controls; p < 0.001). Only 6% of telephone interviewees (both CHD subjects and controls) reported increased drinking, fewer than on written questionnaires. Only 3 CHD subjects (and no controls) reported being told by their doctor to increase drinking.
Knowledge of Benefit
Most (81%) responders had heard of benefits of drinking, with men, white persons, college graduates and drinkers (both in 1978-85 and 2000) more likely to have heard. Benefit by wine or red wine was specifically mentioned by 26%; none mentioned liquor or beer. Six CHD subjects and 6 controls stated health benefit as a reason for increased alcohol intake, with one of these 12 reporting >3 drinks/day in 2000. Of all 18 persons (7 CHD cases, 11 controls) reporting increased drinking to > 3 drinks per day in 2000, 13 had reported 1-2 per day and 5 less than daily drinking in 1978-85. Ten of these persons were interviewed; 9 had heard of alcohol’s benefits.
Are CHD Patients Getting the Wrong Message?
Over 20 years these relatively elderly subjects were far more likely to reduce drinking or quit altogether than to increase intake. Only a few reported new heavy drinking; none of these had been abstainers. Even the substantial majority that had heard of medical benefit by alcohol were more likely to decrease than to increase intake over the years. It is noteworthy that persons who had heard of benefit in 2000 were more likely to drink both in 1978-85 and in 2000. It is doubtful that many heard of benefit before 1985; a more likely explanation is that persons are more likely to be receptive to information that is self-serving.
The fact that subjects reported more reduced drinking at telephone interviews than on written questionnaires may reflect a greater tendency to underestimate drinking at interviews. But the latitude of the questionnaire categories allows for changed intake within the categories. E.g., “less than daily” or “one-two drinks per day” each covers an intake range. Clearly, the interview data show that a CHD diagnosis leads to reduced drinking by some persons.
The subjects who reduced intake because of CHD or medications were largely established light drinkers at little risk of alcohol-related problems. In CHD patients light drinking has been associated with reduced mortality and heart failure risk (7) and very light drinkers who modestly increase intake may lower risk of cardiovascular disease (8). It is thus probable that some CHD patients incurred increased risk by reducing or stopping light drinking.
Heavy drinking should always be avoided and alcohol should be avoided in combination with other central nervous system depressants or with drugs having an antabuse-like effect. It is difficult to assess the common injunction to stop light drinking because of potential medication interactions. One review (9) states “most studies assessing alcohol-medication interactions focus on the effects of chronic heavy drinking” and therefore practitioners must “infer potential medication interactions at moderate drinking levels based on observations made with heavy drinkers”. If followed strictly, this inference precludes any drinking by a vast number of older persons taking medications that might potentially interact with alcohol. This under-explored area needs study.
The study subjects were mostly elderly and did not represent a cross-section of the population. Nonresponders possibly included disproportionate numbers both of new heavy drinkers and of very disabled persons likely to be new abstainers (e.g., persons in chronic care facilities). Despite these limitations the data are relevant for practitioners who advise CHD patients or persons at high CHD risk. Problem drinking seems an unlikely result of knowledge about benefit of light/moderate intake.
Arthur L. Klatsky, MD Division of Cardiology, Department of Medicine and Division of Research Kaiser Permanente Medical Care Program, Oakland CA. He is also a valued member of the AIM Social, Scientific and Medical Council.
1. Fagrell B, DeFaire U, Bondy S, Gronbaek M et al. The effects of light to moderate drinking on cardiovascular diseases. J Intern Med 1999; 246:331-340.
2. Rimm E, Klatsky AL, Grobbee D, Stampfer MJ. Review of moderate alcohol consumption and reduced risk of coronary heart disease: Is the effect due to beer, wine, or spirits? British Medical Journal 1996;312:731-736.
3. Klatsky, AL. Drink to your health? Scientific American 2003;288:74-81.
4. Casswell S. Public discourse on benefits of moderation: implications for alcohol policy development. Addiction 1993;88:459-465.
5. Harrison P. Royal College debates whether MDs should promote moderate consumption of alcohol. CMAJ 1998;159:1289-1290.
6. Klatsky AL, Armstrong MA, Landy C, Udaltsova N. The Effect of Coronary Disease on Changes in Drinking in an Older Population. Alcohol Research 2003. In Press.
7. Klatsky AL. Should patients with heart disease drink alcohol? JAMA 2001;285: 2004-2006.
8. Sesso HD, Stampfer MJ, Rosner B, Hennekens CH, Manson JE, Gaziano JM. Seven-year changes in alcohol consumption and subsequent risk of cardiovascular disease in men. Archives of Internal Medicine 2000;25:2605-2612.
9. Weathermon R, Crabb DW. Alcohol and medication interactions. Alcohol Research and Health 1999;23:40-54.