select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to long term abstinence and heavy drinking associated with dementia

A new prospective cohort study, published in The BMJ, examines the association between alcohol consumption and risk of dementia.


Dr David Vickers, Research Associate in Theoretical Immunology, Imperial College London, said:

The press release sensibly states that the evidence from this study is far from conclusive. However, I would disagree that the strength of the associations are as robust as is indicated. My attention was drawn to the data the authors present in their Figure 2. The wide confidence intervals indicate that the results are compatible with a large difference in dementia outcome. For example, the Hazard Ratios for participants aged 70 years, indicate an increased risk of dementia as the number of alcohol units per week increases—in favour of previously observed trends. However, the 95% confidence intervals were as large as 0.5 to 6.0 (at consuming 70 units of alcohol per week), which is consistent with halving or hextupling the hazard of dementia. Other examples can be found in Tables 2 and 3 of the paper.

“Further, I have trouble inferring the strength of an association from analyses on risk groups with such disproportionate sizes. Abstainers and those who consume > 14 units per week make up such a small fraction of the cohort (14% and 25%, respectively), compared to those who consume 1-14 units per week (61%–based on Table 2 in the article). To me, the results observed in Abstainers and those who consume > 14 units per week is not overly indicative of different mechanism underlying the epidemiology. Instead, I think the main lesson is that extreme outcomes (higher rates of dementia) are more likely to be found in smaller populations. This is all there is to the story, and I agree with Yasar’s commentary that we should remain cautious of these results and not change current recommendations on alcohol use.”


Dr Sara Imarisio, Head of Research at Alzheimer’s Research UK, said:

“Past studies of alcohol and dementia risk have tended to record how much people drink at a single point in time, but the strength of this study is that the researchers have been able to track changes in people’s drinking over a number of years during midlife.

“As this study only looked at people’s drinking in midlife, we don’t know about their drinking habits earlier in adulthood, and it is possible that this may contribute to their later life dementia risk. People who completely abstain from alcohol may have a history of heavy drinking and this can make it difficult to interpret the links between drinking and health. Future research will need to examine drinking habits across a whole lifetime, and this will help to shed more light on the relationship between alcohol and dementia.

“The researchers relied on participants reporting their own drinking levels, meaning people may have underestimated how much alcohol they regularly drink.  While the Whitehall II Study has provided valuable insights into health and ageing, we don’t know the extent to which findings from this group of civil servants apply to the population as a whole.

“Current alcohol guidelines recommend not regularly drinking more than 14 units a week for both men and women, and drinking over this amount impacts on a number of health conditions, not only dementia. We know that a healthy lifestyle, including cutting down on too much alcohol, can improve health and reduce dementia risk, and a good motto tends to be, what is good for your heart is good for your brain. Not smoking, eating a healthy balanced diet, staying mentally and physically active and keeping blood pressure and cholesterol in check are all ways to support healthy brain ageing.”


Prof Clive Ballard, Professor of Age-Related Diseases at the University of Exeter Medical School, said:

“The study is very helpful in quantifying alcohol consumption and identifying a clear increased risk of dementia at a level of more than 14 units a week. It builds on another recent study showing very substantial increases in dementia risk in people with higher levels of alcohol intake and in problem drinking.  This highlights the importance of focussing on alcohol consumption as part of our public health efforts to reduce the risk of dementia and it’s very helpful as part of providing clear guidance regarding safe levels of alcohol consumption as part an overall healthy lifestyle.

“The increased risk in people who are abstinent from alcohol intake compared to those with modest intake is very consistent with previous research, but needs to be interpreted with caution as it may be more likely to reflect differences in other health conditions and differences in cultural backgrounds in people who are abstinent from alcohol.”


Dr Katy Jones, Assistant Professor in Applied Psychology, Institute of Mental Health Centre for Dementia, University of Nottingham, said:

“Given the importance of long term brain health, the power of messages received through the media about alcohol, and the potential subjectivity of ‘moderate drinking’ we should be careful in how these results are framed. An important consideration is that people in better health are more likely to be drinking moderately. Articles with similar findings often conclude researchers aren’t prepared to recommend that people take up alcohol to improve their health. If researchers aren’t prepared to recommend this, we should be treating these findings with caution.

“The sample size seems sound and typical for studies in the area. I notice the data was collected during 1985 and 1993 which was a time of the introduction of unit guidelines in the UK (they were higher then than they are now for men). The data seems to support the conclusions.

“The study supports adhering to current UK unit guidelines and fits with what we call a ‘J-shaped curve’ (people who abstain from alcohol or drink heavily have poorer health outcomes than people in the middle).

“We are sometimes cautious about the J-shaped curve because of three things: people who are in poor health are more likely to quit alcohol, recent evidence suggests alcohol consumption is an indicator rather than a cause of good health, and people find the units system confusing which results in inaccurate estimations of their alcohol consumption.”


Prof David Spiegelhalter, Winton Professor for the Public Understanding of Risk, University of Cambridge, said:

“This study is yet another demonstration of the increased risks faced by alcohol abstainers compared to moderate drinkers.  They had a 50% increased risk of dementia, while another recent review showed a 20% increase in annual mortality risk.    Of course abstainers may be different in many ways other than not drinking alcohol, and it seems important to understand the reasons for these robust and important associations.”


* ‘Alcohol consumption and risk of dementia: 23 year follow-up of Whitehall II cohort study’ by Sabia et al. published in The BMJ on Wednesday 1st August.


Declared interests

Dr David Vickers: “I have no competing interests to declare.”

Dr Sara Imarisio: “Nothing to declare”

Prof Clive Ballard: “No declarations of interest”

Dr Katy Jones: None received

Prof David Spiegelhalter: “No conflicts.”

in this section

filter RoundUps by year

search by tag