Of the 9,087 participants, 397 cases of dementia were recorded over an average follow-up period of 23 years.
Admissions for alcohol related chronic diseases and cases of dementia from 1991, and the role of cardiometabolic disease were then identified from hospital records. Long term abstainers and those who reported a decrease in alcohol consumption also appeared to have an increased risk.
'People with a history of heavy drinking who abstain for health reasons and those who under-report their drinking also makes it hard to draw any firm conclusions'. The chances of losing your marbles are higher for those that didn't drink a drop of alcohol compared to people who consume about 1-14 units of booze per week.
Scientists have determined that those who fully refuse alcohol increases the risk of dementia - brain damage leading to dementia in varying degrees.
They noted that with every increase of seven units per week there was a significant 17 per cent increase in dementia risk.
Perhaps the most important contribution of the recently published research, however, was the discovery that doing the exact opposite of excessive drinking - abstinence from alcohol consumption in middle-age - is a factor for developing the illness as well.
In the United Kingdom, 14 units of alcohol a week is now the recommended maximum limit for both men and women, but many countries still use a much higher threshold to define harmful drinking.
'These results suggest that abstention and excessive alcohol consumption are associated with an increased risk of dementia, although the underlying mechanisms are likely to be different in the two groups, ' the authors wrote.
What is slightly odd, however, is that the higher risk of dementia for people who abstained from alcohol was only apparent if they didn't drink wine.
'The study tells us little about how drinking above low risk guidance beyond the of age of 55 affects the development of dementia.
Dr Doug Brown, chief policy and research officer at Alzheimer's Society, said: "By finding evidence that drinking lots of alcohol, and also drinking no alcohol at all both increase dementia risk, this study supports other work that continues to question whether drinking up to the equivalent of six glasses of wine per week might have a protective effect against dementia".
The authors say while the study is important to fill gaps in knowledge, "we should remain cautious and not change current recommendations on alcohol use based exclusively on epidemiological studies".