Recent study shows that alcohol does have an important benefit in older age

This post is about alcohol, dementia and old age. There is a new research report that concludes that the occurrence of dementia is the lowest among people with moderate alcohol consumption, and the risk is higher among those who never drink or who drink excessively.

But, to set the scene, my wife spent a Monday recently in a hospital, being examined some of the time, but mostly just sitting there waiting (I kept her company). I then did exactly the same thing myself on the Tuesday (and she accompanied me). In my wife’s case they said: “It will be okay”, and so far they have been right. However, in my case they gave me some pills to take 3 times per day, every day for the rest of my life. These are the outcomes of getting older — back in the old days we didn’t make it, to get any older, but these days we slowly undergo physical and mental degeneration for quite a long while.

In this regard, old age is acknowledged to be the most important predictor of the syndrome known as dementia (a general decline in cognitive abilities affecting one's ability to perform everyday activities). Therefore the number of people living with this condition is expected to grow, along with the number of old adults. It should be obvious, then, that detecting this as far ahead as possible will help us (family, friends, colleagues, community, society) cope with the situation as best we can.

To this end, a recent report appeared:

Identifying early predictors of cognitive impairment and dementia in a large

nationally representative U.S. sample. Rand Research Report 2024 RR-A3207-1

It’s relevance to us in this blog should be obvious: alcohol is often implicated in cases of dementia. Well, this report makes it clear that alcohol does indeed play a potential role, but not in the way you may have suspected — there is good news aplenty.

Let’s start with what this study is, since we know that it is not a scientific experiment. It is simply a survey in which people answer questions and are measured in various ways. So, it is the quality of the surveying that matters, in terms of its size and how well it represents the intended population. According to the authors:

The data for this study come from the Health and Retirement Study (HRS), a United States nationally representative, biennial longitudinal survey of adults over age 50. The first wave was conducted in 1992 with a target population of the cohorts born in 1931 through 1941. Older cohorts and a younger cohort (born in 1942 through 1947) were added in 1998 [cohort = wave], so that the 1998 study represented the population born in 1947 or earlier. Since 1998, refresher cohorts of 51- to 56-year-olds have been added to the HRS every six years to maintain a population representation of adults ages 51 and older. The sample includes about 20,000 individuals per wave; more than 45,000 individuals have participated in the HRS since its inception.

This sounds pretty darned good to me — better than most such surveys. Their objective is clear:

In this report, we aim to identify predictors of dementia and cognitive impairment for individuals in the United States up to 20 years in advance using the cognition and dementia measures from the Health and Retirement Study (HRS).

So, the HRS already has within it “a validated probabilistic measure of dementia and cognitive impairment that was developed in prior research”, so that each person has been evaluated for cognitive ability. The HRS then:

allows us to evaluate the predictive power of many potential dementia risk factors, such as demographics, socioeconomic status (SES), labor-market measures, lifestyle and health behaviors (such as exercising and smoking), subjectively reported and objectively measured health, genes, parental health, cognitive abilities, and psychosocial factors (such as personality traits, social activities, and loneliness). We estimate how these factors predict cognitive impairment and dementia of individuals two, four, and twenty years after age 60.

So, that idea should be clear, from this here figure. The authors used their fancy univariate and multivariate analyses to try to work out how each of 181 potential risk factors, measured now, relates to developing dementia at a later age: “dementia prevalence among persons approximately age 80 according to their observed characteristics when they were about age 60”.

The report itself is quite readable, and you should consult it if you are interested in learning about all of the relevant factors they detected. For example:

In terms of explained variation, an individual’s baseline cognitive abilities, health, and functional limitations were the strongest predictors of dementia, whereas parental health, family size, marital history, and demographics were the weakest ones.

However, here we are interested in alcohol, which does make a suitable appearance:

Alcohol consumption is also predictive of dementia incidence; the incidence and prevalence of dementia are the lowest among those with moderate alcohol consumption, and the risk is elevated among those who never drink or who drink excessively. We found the same patterns in both sex groups and in all three dementia models that we considered.

This remarkable observation is shown in the next graph. This graph shows the males in the left pair of graphs and the females in the right pair. The top pair shows the effect of moderate exercise, for comparison, and the bottom pair shows the effect of alcohol intake. Each individual graph shows five possible situations (horizontally), and the rate of dementia associated with it (vertically).

So, to take exercise for men, as an example you should be familiar with: as the incidence of exercise decreases from left to right, the incidence of dementia increases. Or put around the other way, as exercise increases dementia decreases. You do not have all be told that exercise is good for you, but here it is one clear example.

Now look at alcohol. For both men and women, the highest incidence of dementia is among people who do not drink alcohol at all. For both men and women, the lowest incidence of dementia is among people who do have only one drink (or a couple of drinks) at a time. However, further increasing alcohol intake beyond this does increase the incidence of dementia — remember, five drinks at one time is a whole bottle of wine!

So, what people have been saying for years, that a little bit of alcohol is good for you, actually turns out to be true in this study.

The authors are moderate but clear:

Alcohol consumption is predictive of dementia incidence; the incidence and prevalence of dementia are the lowest among those with moderate alcohol consumption, and the risk is elevated among those who never drink or who drink excessively. We found the same patterns in both sex groups and in all three dementia models that we considered ... The lifestyle and health behavior predictors are noteworthy because they are modifiable risk factors and can be controlled by the individual.

Based on their whole analysis, they note:

We found strong associations between the outcomes and several modifiable risk factors. Therefore, our results suggest that there might be scope for slowing cognitive decline and dementia among at-risk people through behavioral changes and interventions. Such lifestyle modifications could be achieved by individuals taking the initiative to make such necessary changes, and public policy could also play an important role.

Our results suggest that it might be beneficial for maintaining cognitive health to exercise at least sometimes, even if it is only light physical activity, such as walking. Consuming alcohol in moderation, working longer, and engaging in hobbies and novel information activities after retirement are also associated with a lower risk of developing dementia. Similarly, maintaining good physical health is associated with reduced dementia incidence, which suggests that adopting a healthy lifestyle might be beneficial not only for general health but also for brain health ...

All these findings point toward the importance for policymakers and other stakeholders to promote healthy behaviors in the population and to strengthen individuals’ access to quality health care.

To this latter end, getting the bureaucrats’ alcohol-consumption ideas right is important. It seems that the World Health Organisation, and a number of other national and international bodies, have not yet got it right, as they have recently come out strongly against alcohol. The basic issue is that we cannot do proper scientific experiments (where we manipulate people’s behavior) to study the relevant social situations critically. All we can do is look at large databases, of which the one discussed here seems to be an important one.