Experts agree that long-term alcohol abuse is detrimental to memory function and can cause neuro-degenerative disease. However, according to a study published in Age and Ageing by Oxford University Press today, there is evidence that light-to-moderate alcohol consumption may decrease the risk of cognitive decline or dementia.

Estimates from various studies have suggested the prevalence of alcohol-related dementia to be about 10% of all cases of dementia. Now researchers have found after analyzing 23 longitudinal studies of subjects aged 65 years and older that the impact of small amounts of alcohol was associated with lower incidence rates of overall dementia and Alzheimer dementia, but not of vascular dementia and cognitive decline. It is still an open question whether different alcoholic beverages, such as beer, wine, and spirits, all have a similar effect. Some studies have shown a positive effect of wine only, which may be due either to the level of ethanol, the complex mixture that comprises wine, or to the healthier life-style ascribed to wine drinkers.

A total of 3,327 patients were interviewed in their homes by trained investigators (physicians, psychologists, gerontologists) and reassessed one and a half years and three years later. Information on the cognitive status of those who had died in the interim was collected from family members, caregivers or primary care physicians.

Among the 3,327 patients interviewed at baseline, 84.8% (n=2,820) could be personally interviewed one and a half years later and 73.9% (n=2,460) three years later. For the vast majority of subjects who could not be personally interviewed, systematic assessments (follow-up 1: 482; follow-up 2: 336) focusing particularly on dementia could be obtained from GPs, relatives or caregivers. Within three years, follow-up assessments were unavailable for only 49 subjects (1.5%). Proxy information could be obtained for 98.0% (n=295) of the 301 patients who had died in the interim. Since dementia is associated with a higher mortality rate, proxy information is particularly important in order to avoid underestimation of incident dementia cases.

At baseline there were 3,202 persons without dementia. Alcohol consumption information was available for 3,180 subjects:

* 50.0% were abstinent

* 24.8% consumed less than one drink (10 grams of alcohol) per day

* 12.8% consumed 10-19 grams of alcohol per day

* 12.4% consumed 20 or more grams per day

* A small subgroup of 25 participants fulfilled the criteria of harmful drinking (>60 grams of alcohol per day for men, respectively >40 grams for women)

* One man (>120 grams of alcohol per day) and one woman (>80 grams of alcohol per day) reported an extremely high consumption of alcohol

* Among the consumers of alcohol almost half (48.6%) drank wine only

* 29.0% drank beer only

* 22.4% drank mixed alcohol beverages (wine, beer, or spirits)

Alcohol consumption was significantly associated with male gender, younger age, higher level of education, not living alone, and not being depressed.

The calculation of incident cases of dementia is based on 3,202 subjects who had no dementia at baseline. Within the follow-up period of three years:

* 217 cases of dementia (6.8%) were diagnosed, whereby 111 subjects (3.5%) suffered from Alzheimer dementia. Due to the relatively small numbers, other subgroups of dementia (vascular dementia: n=42; other specific dementia, e.g. dementia in Parkinson's disease, Lewy body dementia, alcohol dementia: n=14; dementia with unknown aetiology: n=50) were not considered in the following analyses.

Univariate and multivariate analyses revealed that alcohol consumption was significantly associated with a lower incidence of overall dementia and Alzheimer dementia. In line with a large-scale study also based on GP attenders aged 75 years and older, the study found that light-to-moderate alcohol consumption was associated with relatively good physical and mental health. This three-year follow-up study included, at baseline, only those subjects 75 years of age and older, the mean age was 80.2 years, much higher than that in most other studies.