Thursday November 19 2009
Several newspapers reported that drinking excessive alcohol is good for the heart. The Independent estimated that “half a dozen beers every day” could cut the risk of heart disease while The Daily Express claimed that a bottle of wine a night can halve the odds.
These reports are based on a large population study that found a link between higher alcohol intake and reduced risk of coronary heart disease. The study has several limitations, including the fact that the participants were not clinically assessed for heart disease before the study began.
The most important limitation, however, is that it did not consider the many other known dangers of excess drinking, including the increased risk of liver disease, obesity, pancreatic diseases and certain cancers.
This research provides no reason for exceeding the recommended alcohol limits of 2-3 units a day for women and 3-4 units a day for men.
Where did the story come from?
This research was carried out by L Arriola and colleagues from a number of health and research institutions in Spain, including the Basque government’s Public Health Department of Gipuzkoa. The study received funding grants from a number of organisations, including the Spanish Ministry of Health, the European Commission and the International Agency for Research on Cancer. The study was published in the peer-reviewed medical journal Heart.
This research has been widely reported in the media, with most news stories highlighting the study’s “positive” finding: that greater alcohol consumption was associated with reduced risk of heart disease. However, while the Daily Mirror and Daily Express predominantly focused on this finding, other news sources have rightly treated the study’s results with caution. Both The Daily Telegraph and The Times warned of the other health risks of excess drinking and featured expert opinions on the flaws of this study.
The Independent noted that Spain is the world's third-largest wine producer and ninth-largest beer producer.
What kind of research was this?
This was a cohort study which investigated the association between alcohol intake and risk of coronary heart disease (CHD). Many previous studies have suggested that moderate alcohol intake reduces the risk of CHD, and this study was designed to examine this theory in more depth.
A cohort study is the best way to investigate the relationship between an exposure and later disease risk. A randomised trial in which people are assigned to drink different amounts of alcohol is clearly not feasible or ethical. Studies examining the link between alcohol and CHD would need to ensure that people did not have CHD at the start of the study, to see whether disease later developed as a consequence.
A drawback of the study is that it is difficult to establish an accurate measure of a person’s alcohol consumption and for researchers to know that this did not change over time. Also, most people underestimate how much they drink when filling in questionnaires for research studies.
What did the research involve?
The study involved participants of the European Prospective Investigation into Cancer (EPIC). Participants (15,630 men and 25,808 women) were recruited from 10 European countries between 1992 and 2000. When they were recruited, participants completed dietary and lifestyle questionnaires and had their weight and height measured.
The researchers asked participants if a doctor had ever told them that they had suffered from a heart attack, angina, blood clot, stroke or bleed in the brain, high blood pressure, high cholesterol or diabetes. Women were also asked if they used hormone replacement therapy. Three years later, the researchers were able to contact 98% of the participants to ask whether they had had any coronary events or diagnoses since recruitment. Their responses were verified against hospital discharge data and mortality registers.
The researchers' analysis estimated participants’ diet and lifestyle habits using data from their earlier questionnaire responses. This can lead to potential limitations:
- Estimates of the size and strength of drinks are likely to vary between participants.
- Information on alcohol consumption was given at a single point in time, but drinking habits are likely to change over time.
- Although the participants were asked to estimate how their intake differed at 20, 30, 40 and 50 years of age, it may be difficult to recall this accurately.
- Cardiovascular diagnoses at recruitment to the study relied on the participants’ self-reports. This is not as accurate as verifying diagnoses using medical records or examinations. It is likely that some participants had unreported or undiagnosed cardiovascular disease at the start of the study.
The analyses of the data can be expected to involve some degree of inaccuracy as relatively few people developed CHD, both overall and in each alcohol category. This decreases the reliability of the risk estimate.
This research also used participants and data from the EPIC cancer cohort, which was not designed to investigate how alcohol intake affects cardiovascular risk. Analysing data from another study and applying it to a different analysis is a potential limitation of the study.
What were the basic results?
At the end of the follow-up period, 481 coronary events had occurred in men and 128 in women (incidence rate of 300.6 per 100,000 person years for men and 47.9 per 100,000 person years for women). The researchers adjusted their results to account for physical activity, waist/hip ratio, energy intake and use of certain medications.
No relationship between alcohol intake and CHD was seen in women.
In men, the researchers found that there was no affect on risk of CHD for former and low-alcohol drinkers. Compared to never-drinkers, moderate alcohol consumption was associated with a 51% decrease in risk of CHD, high intake with a 54% decrease in risk and very high alcohol consumption with a 50% decrease in risk.
Further adjustment for diabetes, blood pressure and cholesterol did not affect the significance of the relationships.
How did the researchers interpret the results?
The researchers concluded that drinking alcohol was associated with a more than 30% lower CHD incidence.
Although this study demonstrated a link between higher alcohol intake and reduced CHD risk, it is has several limitations. Therefore, it cannot be assumed that high levels of alcohol consumption lower the risk of CHD. People are advised not to exceed the recommended daily limits of 2-3 units a day for women and 3-4 units a day for men.
Importantly, the study did not take into account any other adverse effects of alcohol, such as the increased risk of liver disease, obesity, pancreatic diseases, certain cancers, possible addiction, depression, accidental injury or reduced judgement in social situations.
Further limitations to this research include the following:
- Asking someone to recall daily or weekly alcohol intake in the previous 12 months is likely to involve a high degree of estimation, inaccuracy and variability in response.
- Trying to remember alcohol intake from several decades ago is also likely to lead to inaccurate results.
- Most people underestimate how much they drink when filling in questionnaires for research studies. This study used participants’ estimates to assess drinking habits.
- Although this was a large cohort, relatively few people had coronary events during follow-up, and the numbers were even smaller in each of the alcohol categories. This reduces the accuracy of any risk calculations.
- It cannot be accurately established that none of the participants had cardiovascular disease at start of the start of the study as this was not assessed using clinical methods.