Living with Dementia October 2008

In August's magazine in a letter entitled Fad or fact? a member asked for clarification on possible links between alcohol and dementia. Here, Head of Research Dr Susanne Sorensen, responds.

What does research tell us about drinking alcohol and the risk for developing dementia?
Most of the studies that look at what people eat and drink, what they do in their spare time and how this relates to the risk of developing dementia are retrospective.

This means that researchers may ask people who are developing dementia, and members of their family, what their lifestyle has been like. This information is then compared to data from groups of people who have not developed dementia at an equivalent age.

Other types of retrospective studies may involve investigating a group of people who have been carefully followed for another reason, and finding out how many people in this group develop dementia . This leaves us with a number of research projects that have collected different types of lifestyle data at different ages in people's lives.

No ready conclusions
It is very difficult to draw strong conclusions on alcohol consumption and risk of developing dementia from individual studies. The studies may have collected some data of interest, but omitted other information that would now have been useful in a complete analysis.

Other lifestyle factors, such as whether the participants were smokers, how much fish, fruit and vegetables people consume, and the type of alcohol they choose to drink, are all confounding factors when analysing large sets of complex data pertaining to the risk of developing dementia. For example, people who prefer wine may also eat more fish and not be smokers, whereas people who prefer spirits may smoke and not like fish.

A review published by a research group at Imperial College London this summer2 compared information on alcohol consumption from 23 studies which looked at the links between lifestyle and the risk of dementia. The group concluded that limited alcohol intake may protect you from developing dementia. Limited alcohol intake means different things in different studies, but the research group concluded that one or two units of alcohol per day may be beneficial.

Choice of drink, and how much?
Only some of the studies included in the summer review contained data on the type of alcoholic drink chosen by the people taking part. There is a trend suggesting that drinking wine may be more beneficial than other alcoholic drinks, but this is not conclusive2,3. Some studies indicate that one unit of any type of alcohol a day has a risk-reducing effect on the development of specifically vascular dementia, whereas more than one may be detrimental4.

It looks as though drinking up to three units of alcohol per day - preferably wine - may have a risk-reducing effect on the development of Alzheimer's disease. Drinking more than three or five units per day (depending on which type of data was collected in the individual studies) seems likely to increase the risk of developing dementia4.

When discussing alcohol consumption, we must always remember the risk of developing an addiction and that alcohol has other immediate effects on how our brain functions. All studies show that heavy drinking or regular binge-drinking increases the risk of cardiovascular disease. Cardiovascular disease includes angina, heart attack and stroke.

People with high blood pressure, heart problems, high cholesterol, a history of stroke and diabetes have an increased risk of developing vascular dementia, as these conditions increase damage to the vascular system. If the vascular system, which supplies blood to the brain through a network of blood vessels, is damaged, blood cells in the brain will eventually die. This can lead to vascular dementia.

References:
1.Ngandu T et al (2007) Dement Geriatr Cogn Disord 23(3): 140-9
2.Peters R et al (2008) Age Ageing 37(5): 502-512
3.Luchsinger JA, et al. Diet and Alzheimer's disease. Curr Neurol Neurosci Rep. 2007 7(5): 366-72.
4.O'Keefe JH et al (2007) J Am Coll Cardiol. 50(11):1009-14