Skip navigation

Heart failure risk and alcohol consumption


Clinical bottom line

Drinking moderate amounts of alcohol - up to four drinks a day - reduces the risk of developing heart failure in older people.

Alcohol in moderate amounts is known to have protective effects on heart disease. That is usually meant to mean heart attacks, where part of the heart muscle becomes ischaemic due to clotting. Heart failure, the ability of the heart to pump efficiently, has not usually been included in studies looking at the effects of alcohol. A new study specifically looked for this in older people.


JL Abramson et al. Moderate alcohol consumption and risk of heart failure among older persons. JAMA 2001 285: 1971-1977.


It was conducted in New Haven, as part of an epidemiologic study of the elderly. This is a prospective cohort of persons aged 65 or older in 1982. At baseline 2800 people were asked about their lifestyle and about the presence (actual or suspected) of heart failure. There were 311 people with potential heart failure at baseline, and only 47 who consumed more than 4 alcoholic drinks a day. Information was also collected on demographic variables, other aspects of the medical history, BMI and smoking.

The principal outcome was heart failure (first event where there was more than one occurrence). These were collected through the two area hospitals that looked after most of the people in the study supplemented by other information. International disease classification codes were used for hospital diagnoses. Heart failure (non-fatal) was defined as:

Fatal cases were identified from examining death certificates. The study period was 1982 to 1996, with a maximum follow up of 14 years.


The average age of the population was 74 years, 41% were men and 21% nonwhite. About half drank no alcohol in the month before baseline, 40% consumed up to 1.5 drinks a day, and 10% consumed 1.5 to 4 drinks a day.

During the follow up, 281 of 2235 people developed heart failure. The crude rate of heart failure development was higher in those who drank no alcohol than those who drank a moderate amount of alcohol.

Figure: Crude rates for developing heart failure according to alcohol consumption

However, there were significant associations between alcohol consumption and, for instance, years of education, or history of diabetes, or pulse pressure, or body mass index, and smoking. For this reason the relative risk of developing heart failure was adjusted first for age and sex, and then fully adjusted for all confounding factors. This was also done for total mortality.

Relative risk (95% confidence interval)

Heart failure


Alcohol consumption

Age/sex adjusted

Fully adjusted

Age/sex adjusted

Fully adjusted






Up to 1.5 drinks a day

0.71 (0.56 to 0.92)

0.79 (0.60 to 1.02)

0.76 (0.68 to 0.85)

0.81 (0.72 to 0.91)

1.5 to 4 drinks a day

0.47 (0.29 to 0.76)

0.53 (0.32 to 0.88)

0.74 (0.62 to 0.89)

0.75 (o.62 to 0.90)

Full adjustment was for age, sex, race, education, angina, history of heart attack, heart attack during survey, diabetes, hypertension, pulse pressure, BMI and smoking

The simple answer was that moderate alcohol consumption reduced the risk of developing heart failure, and the risk of dying. A number of secondary analyses made no difference to the outcome, and showed that the effect was substantially the same whether the main type of alcohol consumed was beer, wine, or spirits.


This was a large, well-planned prospective observational study, the type for which no randomised trial could be done. It shows, with others, that moderate alcohol consumption is good for us, and in this instance it shows that it is protective against heart failure. Beware, though, that heavy alcohol intake of more than four drinks a day over long periods may not be beneficial.