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.

The WHO is making a mistake about state-owned alcohol retailers

Last week, the World Health Organization (WHO) released a new report:

Nordic alcohol monopolies protect public health

This continues the current WHO anti-alcohol campaign, apparently heading back towards the days of Prohibition, or something like it. Their online announcement will give you an idea of their tone:

• WHO highlights Nordic alcohol monopolies as a comprehensive model for reducing alcohol consumption and harm

WHO has released a new report highlighting the Nordic alcohol monopolies as an effective model for reducing alcohol consumption and harm. Unlike commercial alcohol sales systems, these state-owned monopolies operate with a public health mandate, restricting availability and limiting the influence of private profit interests. Countries like Finland, Iceland, Norway, Sweden, and the Faroe Islands have maintained relatively low alcohol consumption levels, despite being part of a region historically known for heavy drinking and related harms.

So, their basic point is that retail alcohol monopolies can be an effective way to reduce alcohol consumption and harm. By removing the private profit motive of sales, monopolies often have:

• fewer outlets
• shorter hours of sale
• no advertising and promotion.

I should comment on this because I seem to be the only person regularly blogging about the wine industry, and who actually has personal experience of a state-owned Nordic / Scandinavian alcohol retailer. Also, I should compare this with the situation in the USA and Canada.

The WHO claim is that they have compiled some data that support their current agenda against alcohol. One example is shown in the graph below, looking at alcohol per capita consumption (APC) in the various countries. The Nordic countries have smaller APC than most of the other European countries.

However, the WHO have not looked at how the “government monopolies” actually behave, in practice. They may once have behaved the way WHO imagines, but that was decades ago. My own experience is of Systembolaget in Sweden, for the past quarter of a century (I also have experience of Australia), where the Swedes are dedicated to providing excellent customer service.

My own personal experience of Systembolaget includes:

• it is not closely supervised by the government, but acts as an independent retailer that just happens to be owned by the government (eg. it has recently expanded its opening hours)
• if you can import it then they will sell it, although you may have to order it (and wait a few days for delivery) rather than them always stocking it on their shop shelves
• if you can pay for it then they will sell it to you, unless you are clearly intoxicated at the time or are under-age
• they have plenty of easily accessible shops
• only when the importer runs out does supply of a product end.

Systembolaget currently lists as being available nationally: 15,251 wines, 4,067 beers, 5,272 spirits, 419 cider etc, and 170 alcohol-free drinks. This is hardly a limit on supply, for a country of 10.5 million people.

I have written quite a few blog posts on this situation, including the following:

An explanation of the situation:

Why are there wine monopolies in Scandinavia?

Product availability:

Wine monopolies, and the availability of wine
The availability of wines in government-owned retail monopolies
The availability of older wine vintages in Sweden?

The prices of alcohol:

Why is wine often cheaper in Sweden than elsewhere?
Is Scandinavia currently the most attractive wine export market?

Availability of alcohol from elsewhere within the European Union (EU):

Sweden is not actually restricted to a government alcohol retail monopoly

Similarity to alcohol availability elsewhere in the world:

Why does the world have Three-Tier systems for alcohol supply?

Not everything is necessarily perfect, of course:

My annoyances with my alcohol monopoly

While on the point of alcohol sales in the rest of the EU, looking at the above graph, I would ask the obvious question: what about Croatia, Malta, Cyprus, and especially Italy and Greece? It is not like these countries don’t produce lots of alcohol, including wine, beer and spirits, and sell them both nationally and internationally. The locals apparently consume less alcohol per person than the Nordic countries, according to the graph, without government ownership of retail. Also, Denmark should be noted in this list as a Scandinavian country, but their government does not own the alcohol retail.

More to the point, there was a research report published last year (Classifying national drinking patterns in Europe between 2000 and 2019), which looked at drinking patterns within the different countries of Europe four times this century, and the changes in drinking behavior during that time make a mockery of the one-time-only analysis of the WHO.

Finally, this whole issue undervalues the fact that, as alternative examples, both the USA and Canada also have states and provinces with partial or complete government ownership of alcohol retail. This is mentioned in the report as an alternative to Prohibition but not elaborated upon; and it hardly differs from the Nordic countries.

For Canada you can read: Alcoholic drinks in Canada. For the USA, you can read Alcoholic beverage control state, which lists the relevant states:

Alcoholic beverage control states, generally called control states, less often ABC states, are 17 states in the United States that have state monopolies over the wholesaling or retailing of some or all categories of alcoholic beverages, such as beer, wine, and distilled spirits.

Mind you, the United States continues to be the top global market for wine consumption, and for the USA there is pressure on the current retail situation:

Retailer DtC wine shipping: the time has come
Poll: NYers want wine in grocery stores

So, WHO’s anti-alcohol campaign seems to have led to some mis-understanding about the real situation among the countries of both Europe and North America. Unfortunately, they may not need to continue their campaign for long, because APC has been decreasing in many markets worldwide, no matter what the retail structure (What’s driving wine’s structural decline?). Nevertheless, Current medical evidence says that wine is not harmful in small doses.

Wine price variation within the European Union

The International Organisation of Vine & Wine usually publishes their State of the World Vine and Wine Sector report in April each year. The 2023 report was pretty depressing in many ways, showing a decline in global wine consumption starting in 2018 and increasing after 2021. The 2024 report is not expected to show anything much different.

Similarly, the Silicon Valley Bank just released its US Wine Industry Report 2025, which noted that “the wine industry is undergoing a significant reset, marking the first demand-based correction in three decades ... While painful to experience now, these shifts may result in a healthier industry with more accessible bottle prices that will attract new consumers tomorrow.” I hope so.

In addition to these reports, there are the upcoming tariff portents of the new US president. These may have global effects in the not too distant future, although perhaps “tariffs on imported wine and spirits are more likely to harm American businesses than businesses abroad” (How Trump’s tariff threats will affect the wine and spirits industry and consumers).

So, as a change from all of this, now might be a good time to have a look at the past and present prices of wine among the states of the European Union (EU), and a few of its allied countries, rather than the USA. The situation is not quite as simple as we might expect or like. ¹

Let's start with a simple look at recent EU wine consumption, as shown above (Annual wine consumption in the European Union). As you can see, it varies quite a bit, but the general trend has been slightly downwards — the 2024 value was only 90% of the 2016 value. The average across the nine years was 100 million hectoliters (2,200 million gallons.

There was a distinct dip during the main years of the Covid-19 pandemic (2020–2021), suggesting that a lot of the consumption occurs in public places such as bars and restaurants, which were being avoided at that time. The unique rise in 2022 looks like a reaction to being safely allowed back into public drinking again. (Compare the USA: Alcohol consumption rises sharply during pandemic shutdown.)

Moving on to the different countries, we can look at the alcohol prices for 2020, as shown in the second graph (How alcohol prices vary across the EU) — this is simply a standardized measure with the EU average set at 100. Once again, these values vary quite a lot, with the Hungary price being only 38% of that of Finland.

Of the highest countries, the values for Finland and Sweden include a lot of spirits, which may also be true for Ireland. The wine-producing and consuming countries (eg. Portugal, Italy, France, Germany, Spain) are mostly in the centre of the price list.

The three countries to the right of the main graph are members of the European Free Trade Association, which is why there are data for them. The high price data for Norway and Iceland match those for two of the other Nordic countries (Finland and Sweden), but not the fifth one, Denmark (which matches Switzerland, instead!).

Moving on specifically to wine, we have Eurostat data for changes in price since 2015, as shown in the table above (Food price monitoring tool). The data refer to Wine from Grapes, showing a Harmonised Index of Consumer Prices, with 2015 set to 100. The data allow us to calculate the change in price for two periods of 5 years each (2015–2019 and 2020–2024 inclusive).

The EU average is at the top, with the countries then in alphabetical order. Note that there are the same 27 EU countries, plus Norway and Switzerland (but not Iceland), and the UK for the first period (when it was in the EU, before Brexit).

Note that there was a much greater (average) price increase in the past 5 years compared to the first 5. I think we all know that prices increase through time at an increasing rate. However, it is worth noting the two decreases in price during the first 5 years (Ireland and Cyprus) – lucky them!

Note also that the top 6 countries in the second list are more than double the average — this is pretty serious. Indeed, the top 10 countries are in eastern Europe (plus Czechia, which is way down). In particular, the poor situation for Croatia has attracted discussion (Croatia ranks first in Europe for wine price increases), although quality is another matter (How this power couple is trying to make Croatian wine world class), as is historical significance (How a 150-year-old grapevine is helping California producers adapt to climate change).

The countries from northern Europe discussed above (Sweden, Norway, Finland) are actually in the middle of this list — as I noted, these are wine prices only, not spirits prices. Also, Ireland, Switzerland and Italy are doing very well (see this video for Switzerland: Countries are struggling to contain inflation, but not Switzerland). Austria has not done much worse in the second 5 years compared to the first, which is unusual within the list (see: Austria bucks (some) trends; however: Austria’s wine exports to fall back in 2024).

It seems unlikely that things are going to get much better. It is, of course, the younger people who are the future, on all continents, but it has been observed that wine is very traditional. So, it has been noted (Leveraging wine sales in 2025):

Amid falling overall consumption in several countries, there is also ongoing debate about how to attract more Millennial and adult-Gen-Z drinkers to wine. Sustainability cues could be important; IWSR research showed a greater proportion of regular wine drinkers in Gen Z and Millennial generations in the US associated organic wine with high quality.

So, there is cause for optimism, if we can just see the potential areas of growth. However, The Wine Economist (Mike Veseth) once noted Kenneth Boulding’s historical point that when the future eventually rolls around it never matches the predictions, it is always unexpected (The future of wine?). So, we shall see what we see.

---

¹ The European Union currently consist of 27 countries. The United Kingdom was in the European Economic Community from 1973, which became the European Union, but then withdrew from the EU via Brexit in 2020 (along with Greenland, an autonomous province of Denmark, way back in 1985). Norway, Switzerland and Iceland have submitted membership applications in the past, but subsequently frozen or withdrawn them; they are, however, members of the European Free Trade Association. Turkey is a candidate EU country (and a member of the OECD along with North America and Australia).