Tuesday May 5 2009
Wine drinkers lived longer, but more research is needed
“Half a glass of wine a day can add five years to your life” The Daily Telegraph has said, claiming that new research shows that that light, long-term consumption boosted longevity, ‘with the biggest increase caused by wine’.
The study behind this article ran from 1960 to 2000, and enrolled 1373 men born between 1900 and 1920. The results suggest that men consuming an average of 8g of alcohol from wine per day (equivalent to about half a small glass) may live about 5 years longer than those who don’t consume alcohol.
This is not a precise estimate, and the results suggest the real difference could lie anywhere between 1.6 and 7.7 years. This researchers took care to consider other factors that may be linked to both wine consumption and mortality, i.e. factors that mean people consuming wine are systematically different from those who do not. The researchers took into account the obvious factors of smoking habits and social/ economic class but others important factors were not investigated. This justifies the researchers’ cautious conclusion that more studies are needed to verify their results.
Where did the story come from?
The study was undertaken by Drs Streppel and colleagues from the National Institute for Public Health and the Environment in Bilthoven and the Division of Human Nutrition at Wageningen University, both in the Netherlands.
It was published in the peer-reviewed Journal of Epidemiology and Community Health, and partially funded by the former Inspectorate for Health Protection and Veterinary Public Health (at present integrated in the Food and Consumer Product Safety Authority).
What kind of scientific study was this?
This was a cohort study looking at male alcohol consumption and rates of mortality among 1373 Dutch men within a 40 year period.
This Zutphen Study (named after a town in eastern Holland) began in 1960, when a random sample of men who were born between 1900 and 1920 and living in Zutphen for at least 5 years were drawn from the town.
The initial sample was 1088 men, although only 878 (81%) of them participated in the study. They were initially interviewed in 1960, and then again in 1965, 1970, 1985, 1990, 1995 and 2000. By 1985 there were only 554 survivors, so the researchers increased their sample size by recruiting an additional random sample of men (also born between 1900 and 1920).
Of the 1266 men invited in 1985, 939 men participated. In total 1817 there were participants, although researchers excluded those men who did not take part in dietary and physical examinations at each round, leaving 1373 in their analyses.
This second group of men were assessed four times; upon joining the study in 1985 and then again in 1990, 1995 and 2000. During the assessments prior to 1985 all participants were asked about alcohol and food consumption in the 12 months preceding the interview, while alcohol intake in the previous month was assessed for interviews after 1985.
Food consumption tables were used to determine the daily intake of alcohol, energy and other nutrients. For those participants who joined the study in 1985, their missing consumption data were imputed, i.e. modelled to fill in the gaps.
The link between alcohol consumption (beer, wine or spirits) and mortality was determined by categorising men by alcohol consumption levels of 0g, 0-20g and more than 20g per day. Many factors (confounders) may be linked to both alcohol consumption and mortality outcomes. The researchers measured some of these potential confounders including;
- smoking status (never or long-term ex-smokers, recent ex-smokers, current smokers)
- length of any smoking,
- health history including any history of heart attack, stroke, diabetes or cancer,
- a measure of socioeconomic status (manual workers, non-manual workers, small business owners and professionals) which was determined by the participants occupation at baseline.
They also considered dietary variables including vegetable, fruit and fish consumption as well as saturated fats.
Researchers were interested in assessing the link between long-term intake of alcohol and cardiovascular and all-cause mortality, according to each type of alcoholic beverages consumed. The life expectancy of the men at baseline (average age 50 years) was compared across consumption types and categories. Death certificates were used to determine official causes of death.
What were the results of the study?
During the 40 years of follow up, 82% (1130 of 1373) of the men died. In total 628 deaths were due to cardiovascular disease; 348 from coronary heart disease and 139 from cerebrovascular events.
Numbers using alcohol increased from 45% in 1960 to 86% in 2000, and from an average of 8g/day to 14g/day (with a peak of 18g/day in 1985). Long-term use of alcohol was linked to a reduced risk of death; compared to men who did not drink, those drinking between 0-20g/day had a 57% lower cerebrovascular mortality risk, a 30% lower risk of death due to cardiovascular disease (CVD) and a 25% reduced risk of death due to any cause.
When assessing alcohol by type, wine had the strongest effect, with those consuming an average of 0-20g of wine per day having a 39% reduced risk of coronary heart disease (CHD), a 32% reduced risk of CVD and a 27% reduced risk of death from any cause. Conversely, beer or spirits did not reduce mortality risk. This effect was similar across all socioeconomic categories.
Overall, men consuming an average of 6g/day of alcohol (from beer, wine or spirits) could expect to live 2.3 years longer from age 50 than those who did not consume alcohol. When limiting the calculation to wine drinkers, those who drank an average of 8g alcohol/day from wine (less than half a glass per day) lived 4.7 years longer than those who didn’t use alcohol.
What interpretations did the researchers draw from these results?
The researchers conclude that long-term light alcohol intake reduces cerebrovascular, total cardiovascular disease and all-cause mortality.
What does the NHS Knowledge Service make of this study?
The authors have raised several weaknesses of their study:
- Firstly, they were unable to investigate the effects of drinking frequency, which is important because some studies suggest that frequency of alcohol consumption is linked to a reduced risk of coronary heart disease independent of the amount consumed.
- The estimates of missing data for those participants who joined the study in 1985 may have led to some biases. The researchers considered this a good approach as consumption around the time these men joined was higher than in the preceding years. They say that they repeated their analyses in only the men who had been in the study since 1960, and found their results were similar to those from analysing the whole sample.
- Their finding that the highest exposure to alcohol did not lead to increased mortality is in contrast to other studies. The researchers explain this by saying that their highest intake category was ‘relatively low’ (29g/day), and this may explain why there was no link with this category and death.
The researchers have adjusted for obvious confounders – smoking and socioeconomic status – and conclude that the association found between wine consumption and mortality is not explained by these factors.
It is important to note that although those drinking an average of 8g alcohol from wine per day lived about 5 years longer than those who didn’t drink alcohol, this was not a very precise estimate: the researchers were confident the difference was somewhere in the range of 1.6 to 7.7 years.
The conclusions from cohort studies such as this are sensitive to the consideration of other factors that may potentially be linked to both consumption and mortality. Although researchers took into account the major factors of smoking and socioeconomic status, there may be other differences between wine drinkers and teetotallers that are not captured in the analysis. Physical activity, which is linked to mortality, may be one such factor.
As the researchers cautiously conclude, ‘more studies are needed to draw conclusions on the strength of the association between wine consumption and mortality’.