Coffee, Tea, and Dementia Risk

Can your daily coffee or tea help protect your brain? In this episode, we unpack a large long‑term study linking caffeinated drinks with lower dementia risk. With help from a researcher, we explore what the findings really mean, how strong they are, and why food and drink research is often more complex than the headlines suggest.

Based on: 

This Everyday Drink Is Linked to an 18% Lower Risk of Dementia, According to a 43-Year Study

Listener Reflection Questions 

1. Did this episode change how convincing the “coffee reduces dementia risk” headline felt to you — and why?
2. What other factors might explain why people who drink more coffee appeared to have lower dementia risk?
3. What’s one question you could ask next time you see a headline about food and health?

Brought to you by the ISLAND Project and Wicking Dementia Research and Education Centre at the University of Tasmania. To continue the discussion and for more information visit https://island.mooc.utas.edu.au or email island@dementia.utas.edu.au

Transcript

Copy Transcript Copied

Neil Broomfield: 00:01
Welcome back to Brain Matters, the podcast where we slow down health headlines and take a closer look at what the research actually tells us, and just as importantly, what it doesn't. I'm Neil Broomfield recording on Palawa Land. Now, today's topic might be one you're very happy to hear about: coffee and tea. You may have come across headlines recently suggesting that drinking coffee or tea could reduce your risk of dementia or help slow cognitive decline. Well, is that good news we can rely on, or is the story a bit more complicated? To help delve deeper into it, I'm joined by Dr. Emma Lea, senior lecturer at the Wicking Dementia Research and Education Centre. Emma, thanks so much for being with us today.

Emma Lea: 00:48
Thank you for inviting me. It's great to be here.

Neil Broomfield:00:51
Why don't we start with the basics? Emma, the headlines are based on a large, long-running study published in the Journal of the American Medical Association looking at how people's coffee and tea drinking habits relate to their risk of cognitive decline and dementia over time. Can you please elucidate on what the researchers were actually trying to find out?

Emma: 01:14
This was a really large research study that was conducted in the United States with over 130,000 participants, and they were taken from two broader research studies. There was the nurses' health study, and there was also the health professionals follow-up study. These both started back in the 1980s, and it collected data on a whole range of different foods and drinks and other variables. It was quite big. But for this paper, the researchers were looking at whether there was a relationship over the long term, so in this case, around 40 years or so, between drinking tea and coffee and brain health. They collected separate data on caffeinated and decaffeinated coffee. That's a little bit unusual because that's not always done in a lot of research studies looking at this topic. They really wanted to see whether there was any difference between how they might link to brain health. To measure brain health, they primarily looked at the risk of getting dementia. They did also measure a whole range of different measures of cognitive function. For example, they looked at subjective cognitive decline. That's when people perceive whether or not they have some issues with their own memory or attention. They also looked at more objective tests, for example, looking at verbal fluency.

Neil: 02:38
A lot of those tests actually are carried out in the Island Project, I think, aren't they?

Emma: 02:44
Yeah, I think that's right.

Neil: 02:45
Is that study that was done by John by the Medical Association in America, is that what's known as a longitudinal study?

Emma: 02:54
Yes, it's a very long-term study in this case, so longitudinal and it is an observational study as well.

Neil: 03:04
We could say then that essentially they followed a large group of people over many years, tracked what they were drinking and eating, including caffeinated drinks, and looked at how that related to their brain health later on. Now, this is where the headlines come in. The main message we tend to see is people who drank moderate amounts of caffeinated coffee or tea had a lower risk of dementia and sometimes slower cognitive decline. Does that broadly reflect what the study found?

Emma: 03:37
Yes, it does. The researchers found that there was a lower risk of dementia associated with drinking more caffeinated coffee and tea. Thinking of the figures themselves, they found that people who drank more had an 18% lower risk of dementia compared with those who drank them the least. The biggest impact on brain health was when people drank two or three cups of caffeinated coffee, or one or two cups of tea, a day. They didn't actually find a benefit if people drank more than that, which I think is quite interesting. They also did find some other differences in other measures of cognitive outcomes but these were fairly small, so quite modest, really. They also weren't consistent across the participant groups. So remember I said there were two studies there, the nurses' health study and the health professional study. The findings weren't consistent across those groups. 

Neil: 
Were they wildly inconsistent or marginally? 

Emma:
I'm not really sure exactly. For example, they didn't even have consistent measures across both groups. There was one study which didn't use all of the same measures of cognitive status as the other group.

Neil: 04:56
So not like for like.

Emma: 04:58
Yes that's right, looking outside of dementia risk, but at those measures of cognitive outcomes. The main statistically significant difference they found was a higher score on the telephone interview for cognitive status test. That's a more objective test. The findings, the difference, was actually very small. This test can give scores that range between zero and forty-one. The higher the score, the better the score.

Neil: 05:29
That's a curious number. Why 41?

Emma: 05:31
It would just be probably the number of items that are included on that score, and that was the total. They found that people who drank caffeinated coffee had a score that was 0.11 higher than those who drank less, so when you're thinking of zero to forty-one, 0.11 is very small, really. Just to put that into context, people who have uh mild cognitive impairment have been found to have a score that's two or three points lower than those without cognitive impairment on that zero to forty-one scale. So a difference of 0.11 is very small. I think, and the the authors point this out in the paper, that the best way of looking at these findings is that it's a small difference at the population level. When you're looking at an individual and whether or not, there might be a clinically meaningful change for that individual based on caffeinated beverages that they drink,  you really can't guarantee that an individual is going to  have an improvement in their cognition if they drink more caffeinated coffee and tea. It's when we're looking across many individuals at that more population level that there is a small improvement.

Neil: 06:49
Are you able to comment on what at the lowest age was in either cohort of the study? I'm thinking that people,some people, start drinking caffeinated drinks at quite a young age, say even before they're in their teens. Whereas other people think it's horrid until they get to much older and then they begin to enjoy it. Was there a definition or a distinction between people that started earlier in life drinking caffeinated drinks from those that started later or did that not show in the study?

Emma: 07:26
An interesting thing about this study is, because it's the nurses' health study and the health professionals study, these are people who are actually middle-aged when the study began. They don't actually have data looking at people from a much younger age. I think they tended to be in their 40s or so at the beginning of the study, and remember that was back in the 1980s, because we've followed them for around 40 years or so. Yes, that would be something that would be interesting to look at, the age and how that relates. 

Neil: 08:01
Good, thank you for that. Therefore, when we say moderate, we're not talking about extreme amounts of coffee or tea or caffeinated drinks here, are we? It's more that everyday range of a few cups a day. Would that be right?

Emma: 08:14
Yes that's right. The two to three cups of coffee a day and one to two cups of tea is really where the the biggest benefits are found. 

Neil: 08:24
Right and another interesting detail is the protective pattern didn't really show up in the same way for decaffeinated drinks.

Emma: 08:31
That's right. That's another interesting thing about this study, is that they did differentiate between caffeinated and decaffeinated coffee. There are some other studies, that have been conducted, where they haven't actually looked at that difference. I guess that means then, that it's likely to be the caffeine which is neuroprotective and there are a whole range of different reasons for why that might be the case. Things like caffeine can reduce levels of beta-amyloid in the brain, which is a special type of protein in the brain, particularly people living with Alzheimer's disease. Caffeine might also have an effect on things like the plasticity of neurons. It might lower the brain inflammation that might be present, and it might also do things like reduce the risk of type 2 diabetes, and that's a major risk factor for dementia. There are also other bioactive compounds that are present in tea and coffee apart from caffeine, which might also play a role in the link to brain health. Things like polyphenols, which is a phytonutrient or a micronutrient found in plants, and that might have benefits for antioxidants and vascular health.

Neil: 09:50
Can we just backtrack a little here, Emma? What exactly is caffeine? Is it found roughly in the same amounts in similar cups full of tea and coffee? Is it found in any other everyday drinks or food? What about sports drinks or Coca-Cola or things like that?

Emma: 10:09
It's definitely found in other beverages as well. In fact, in this study, they did also look at caffeine intake in general. So they looked at caffeine in coffee specifically, but they also looked at the presence of caffeine in a whole range of different foods and drinks. They do have that information and they did find an improvement looking at caffeine more generally.

Neil: 10:34
This is now the point where Brain Matters hits the brakes a little bit. This was an observational study, and that's really important. Can you, Emma, explain what that means and why it matters?

Emma: 10:47
Observational studies, like this one, are ones where researchers make observations of the effect of something, but they don't actually try to change the environment or who is exposed to a particular thing in the environment. This type of research design can show links or associations, but it can't actually prove causation. It can't prove, for example, that coffee is the reason for the difference. Whereas if we compare observational studies to experimental studies, they are able to show causation because they're actually making those changes to the environment or to those things that people are exposed to. It's very different to observational studies. Observational studies have their own benefits. Particularly when we're looking at things, like food and nutrition and dietary behaviors and so on, observational studies may actually be preferable because they can look at real-world nutrition and dietary behaviors and broader lifestyle patterns.

Neil: 11:56
Let's explore that a little bit further then. What are some of the other reasons we might see this link apart from the drinks themselves? We learnt about so-called co-founders in the context of research with Dr. Hannah Ferrer in last month's Brain Matters episode. Were these taken into account in this study?

Emma: 12:15
The researchers did actually take into account quite a range of different confounders, which might otherwise have explained the results. They looked at things like lifestyle factors, whether or not participants smoked, their physical activity levels, whether or not they lived alone, as well as some demographic factors, like level of education and the socioeconomic status of the neighborhood that they lived in. I didn't see any evidence that they looked at things like personal income, but because of the sample, the nurses and the health professionals,  you'd consider that really there's probably not going to be a massive difference in the income levels of participants. They're probably fairly homogenous. That's something actually that probably needs further research because it means that it's a little bit harder to generalize from this study to broader populations because they're quite well educated in that group in that study sample. There might be other lifestyle factors which might explain the results that they didn't consider in the study. There might be expected to be some differences between people who drink coffee and those who don't, with things like sleep patterns, for example, stress levels or social patterns, things like that, which I didn't see any evidence of them having considered those as possible confounders. I guess another thing to think about is just how drinking and eating are tied up in social and cultural activities. It's not just about the nutritional value of the foods and drinks that you're taking in. There is a whole range of other factors that are really important. For example, when I was reading the study, reading the paper, I was thinking what about those people who are drinking tea and coffee for reasons to maybe have a break in their day or  a time out, perhaps having a chat with family or friends or colleagues if they're at work. That might have an impact on brain health in and of itself. However, because they did find a stronger relationship with caffeinated compared to decaffeinated coffee. That might actually take that into into account. Perhaps it's not a bigger factor, but it's still something to think about. They did also think about the possibility of reverse causation. That's where the assumed cause and effect might be swapped around. They did say that that can't be ruled out of their study. That's things like if someone is experiencing early cognitive changes, that might actually have an influence on the amount and the types of tea and coffee that they drink. It also might have an impact on how accurate their self-report of what they've actually eaten and drunk, whether or not it's accurate or there's some kind of bias in there, they may not have recorded things correctly. The researchers did actually make some adjustments in their analysis to try to take into account that possibility. For example, any participants who had major chronic disease present at the start of the study, they didn't include those people in the analysis. I think that they have made a really good attempt to take into account those different types of confounders.

Neil: 15:51
If I can try and precise just this last part of our chat then. People who regularly drink caffeinated drinks might differ in lots of ways from people who don't. They might have different routines, social habits, overall health behaviours. And those can all play into brain health and its outcomes. Is that right? 

Emma: 16:11
Yes, that's right. 

Neil: 16:12
Good. Thank you. Another thing that headlines don't always tell us is how big the effect actually is. Was this a large reduction in risk or was it something more modest than that?

Emma: 16:22
I think, from my perspective, it was definitely more modest. For example, the researchers found a hazard ratio of 0.82 and that equates to an 18% lower individual risk of dementia.

Neil: 16:37
Can you explain what you mean by hazard ratio?

Emma: 16:40
That's a measure of how often a particular event happens in one group compared to another group over time. If we have a hazard ratio of one, that means that there's no difference between those two groups that are being compared. Whereas, if the hazard ratio is less than or more than one, then it does mean there's a difference between those groups. They found a hazard ratio of 0.82 between those with the highest and the lowest intake of caffeinated tea and coffee. It's it's quite close to one. It's not a huge drop under one. That equates to an 18% lower individual risk of dementia.

Neil: 17:27
It wasn't a ratio between the two disparate groups, but between the amounts of drink that they took. Is that right?

Emma: 17:36
Yes. It's not between the nurses and the health professionals. They actually pulled those groups together and looked at them as one group. It's between those with the highest and the lowest intake of caffeinated tea and coffee. It's good, but it's fairly modest when compared to other dementia risk factors. We already know that there are other dementia risk factors, untreated hearing loss or midlife hypertension, things like this, that actually have a much bigger impact on dementia risk, more so than tea and coffee is suggested just from this paper.

Neil: 18:16
WClearly then, Emma, this isn't something that's going to override everything else we know about brain health. It's, I suppose, more like a small piece of a much bigger puzzle. 

Emma:
Yes, that's right. 

Neil:
Food and drink research often makes the headlines, but it is also one of the hardest areas to study well. Why is that?

Emma: 18:35
That's a great question. It can be really difficult to work out the relationship between food and drink, and particularly specific foods and drinks, or components of food and drinks, and aspects of health, things like brain health or associations with chronic diseases such as dementia. One reason is just because there are so many different variables involved and most people eat and drink a really large variety of foods and beverages. The types of foods and drinks that they consume might change with the seasons, they might change over time for a whole variety of different reasons. Things like preparing food in a different way might have some effect on the components of that food. In this situation, thinking about the caffeine that might be present in the coffee, if you prepare the coffee in a different way, it might actually have an influence on the levels of caffeine in there or other bioactive compound levels. Sometimes researchers will want to study specific components of food. In this case, caffeine, or it might be a particular vitamin or a type of fat, or they might want to look at intake of specific foods like fish or nuts, or they might look at diet as a whole, like a Mediterranean diet. There are complicating factors with all of these different types of approaches. It is quite a complex area. The other thing that can make it a little bit tricky to research food and drink, is that we're often asking people to remember what they've been eating. That can be difficult and it might not be an accurate representation of what they were actually eating. There can be a number of reasons behind that. It might be just that they  haven't remembered accurately what they have consumed, but it also might be a little bit of response bias there. That means they might think, the researcher doesn't want to hear about the amounts of high-fat foods, for example, or those extra cups of coffee that I've been drinking. That might actually influence what they've been recording. It might not be super accurate. In this particular study, participants filled in food frequency questionnaires. They filled these in themselves, and that means they reported how often they consume different foods and drinks, and how much of each.  The researchers did actually compare those food frequency questionnaires with potentially a more accurate way of measuring food and drink intake, and that was against a seven-day dietary record. That's where the participants actually wrote down and recorded everything that they ate and drank over those seven days, and actually to the point of where they're weighing everything as well. So it's quite accurate. They found that there was a strong association between those food frequency questionnaires and the seven day dietary records. That just means that it gives us more confidence in the accuracy of the data that's coming from the food frequency questionnaires.

Neil: 21:53
On the other side of that coin, I'm just thinking, this was an American study where the data is whatever it is. If this were conducted in somewhere completely different, say northeast Russia or China, we might have a completely different set of results out of this, might we not?

Emma: 22:11
We definitely could. Yes. There's definitely more research that can be conducted on other populations.

Neil: 22:20
Even Australia, I dare say, would throw up possibly different results too.

Emma: 22:24
Yes that's right.

Neil: 22:25
Although our diet is probably nearer American than Northeast Russia.

Emma: 22:31
Yes, it's probably a fairly typical Western diet.

Neil: 22:35
But what's become obvious from what you've said is that it's not like testing, say, a medication where you can tightly control everything. T

Emma: 
That's right. 

Neil:
So let's bring this back to real life. If someone listening enjoys their morning coffee or tea, what's the takeaway?

Emma: 22:50
I would say to them that you should continue enjoying your morning coffee, with this study suggesting that when looked at a population level, it might be giving brain health a little bit of a boost. But if you don't already drink coffee or tea, I don't think there's enough evidence in this paper, especially when considering what we know about other ways to reduce dementia risk, to actually start drinking it. Especially, of course, if you don't actually enjoy drinking coffee or tea.

Neil: 23:19
Very good. Okay. So it follows that it might be encouraging, but it's not a reason to suddenly start drinking coffee if you don't already, as you've just said. And it's not a reason to dramatically increase how much you drink either. That is right, I suppose?

Emma: 23:32
Yes, that's right. This paper suggests that there's no benefit to drinking more than those two or three cups of caffeinated coffee or one to two cups of tea a day. The effects are modest enough to not warrant starting to drink tea or coffee if you don't already. We can think of it as moderation being key to this. It's also important to think about all those other lifestyle factors that contribute to dementia risk and brain health, potentially more so than tea or coffee.

Neil: 24:06
Well, this, of course, is all part of the Island study, which is roughly halfway through its 10-year span. When it comes to brain health, we already know there are multiple factors that impinge on this, like physical activity, cardiovascular health, social connection, and so on and so forth, many others besides. Caffeine is just one small part of that bigger picture. Okay then. Before we wrap up, dear listeners, why don't we do a quick brain matters check-in? If you see a headline like coffee reduces dementia risk, what sort of questions should one ask, Emma?

Emma: 24:44
There are a few questions that come to mind, but if I had to narrow it down to just one, I would say to maybe try to find out if the study found an association or a causation. This does tie into other questions, like what type of study was conducted. Thinking back to whether it was an observational study or an experimental study, for example. It also helped you to think more about those other factors which we've talked about, which could uh explain the relationship. Then it would be possible to maybe check to see whether the researchers were taking those factors into account when they undertook their analysis?

Neil: 25:25
Let me try and encapsulate that then. You should ask yourself, is this based on observational research or an experiment? Could there be other explanations? And how big is the effect in reality? Here's something to think about after this episode. Did hearing more about the study change how convincing that headline felt to you? What else might explain the link between coffee and tea or other caffeinated things, and brain health? What's one question you might ask next time you see a food or health claim in the news? Emma, thanks so much for helping us to go deeper into this seemingly everyday matter, which turns out to be far more complex than might have first appeared.

Emma: 26:11
Thank you. It's been great to be here.

Neil: 26:13
Listeners, if you enjoyed coffee or tea, this research might be reassuring, but as always, a single study is just one part of a much bigger evidence base. This has been Brain Matters, where we slow down health headlines and make sense of the research behind them. We'd love to hear your thoughts, and you can join the discussion using the link in the show notes. Thanks so much to you all for listening, and see you again on another topic in this Brain Matters series of talks very soon. Meanwhile, my thanks to you, Emma, again, as well as to Florence Sward for organizing and producing this series. And let's not forget Drew Stansbury for recording this, with his usual technical wizardry. Until soon, goodbye.