7 ms·

Main findings: Compared with those who did not drink tea and coffee, drinking 2 to 3 cups of coffee and 2 to 3 cups of tea per day was associated with a 32% (H

Main findings: Compared with those who did not drink tea and coffee, drinking 2 to 3 cups of coffee and 2 to 3 cups of tea per day was associated with a 32% (HR 0.68, 95% CI, 0.59 to 0.79; P < 0.001) lower risk of stroke and a 28% (HR, 0.72, 95% CI, 0.59 to 0.89; P = 0.002) lower risk of dementia. Moreover, the combination of coffee and tea consumption was associated with lower risk of ischemic stroke and vascular dementia. Additionally, the combination of tea and coffee was associated with a lower risk of poststroke dementia, with the lowest risk of incident poststroke dementia at a daily consumption level of 3 to 6 cups of coffee and tea (HR, 0.52, 95% CI, 0.32 to 0.83; P = 0.007).

caffeine 5y ago
The take away here is: there is no obvious evidence in this data that coffee and tea cause stroke or dementia. So if we were worried about that, at least there is no obvious relationship.
- sokoloff 5y ago
  Isn’t there a stronger conclusion evident: that they appear to reduce the risk (rather than not increase it)?
  - headsoup 5y ago
    As is never highlighted, especially in media, is that it's relative risk reduction. Going by their study (and simplifying for argument's sake), there were 10k strokes out of 365k subjects, so <3% risk. A 30% risk reduction means now you're at 2% risk, so figure out whether that's worth adding coffee and tea for.
    - rmbyrro 5y ago
      On an individual level, it might be the difference between having dementia/stroke or not. The decrease from coffee/tea might make you cross the line back into healthy state, perhaps? I don't usually take conclusions from single studies like this, though. It's too easy to fool ourselves with statistics.
      - headsoup 5y ago
        Of course, but there are probably thousands of other decisions we make through life that could increase or decrease our chance 1%. Yeah, statistics and graphs can be a special type of evil
    - ThunderSizzle 5y ago
      A <1% death rate was enough to cause the lockdowns in 2020. If the media ran this endlessly, saying you'll have 30% less chance of stroke or dementia if you drink coffee and tea, people would probably panic on that too.
      - rebuilder 5y ago
        This study was conducted over 10+ years, the COVID numbers are probably per case, or possibly annualised.
      - icoder 5y ago
        Not sure about your country, but here the main problem was (and is again) the risk of an overrun health care system.
        dahfizz 5y ago
        A large uptick in strokes would cause a similar issue, no? You can put that spin on anything. DISCLAIMER: I am not downplaying the severity of Covid. My words mean exactly what they say.
        sokoloff 5y ago
        That seems to depend on the typical duration of stay of a stroke patient in the various departments (emergency, ICU, general in-patient). I'm not familiar with stroke treatment, but from anecdotal exposure in my extended family's experience, they did not result in extended ICU stays.
        icoder 5y ago
        A large enough uptick in strokes at some point will, yes, but I don't think coffee/tea consumption will have that effect, plus our current consumption clearly _doesn't_, while Covid clearly _does_. (At least where I live, ICU's were previously not overrun due to anything) It's not a spin, it's just the daily struggle of the health care system right now. I see no other solution but to take that seriously. That doesn't mean we should take this study as serious just because it discusses something that may lead to something that, by itself, with a theoretical (large enough) uptick would overrun the system too.
    - pwagland 5y ago
      Right, but the point that parent was making is that rather than just not implicating tea/coffee in causing these, the study shows that it reduces the chance of these. That it's a relative risk reduction is not particularly relevant to that. That said, I do agree that changes in relative risk are confusing and/or misleading to many people. A more interesting way of presenting this data, IMO, is the "number needed to treat", that is in order to reduce 1 stroke from occurring, ie, roughly 120 people need to drink 2 cups of coffee/tea per day to reduce 1 stroke, based on the numbers given here. This is obviously a lot more than the naive "30% means roughly 1 in 3 would see improvement" interpretation that the headline number pushes you towards.
      - falcolas 5y ago
        Having seen how dementia and strokes affect people, and their families, I’ll absolutely take a 1% reduction.
  - caffeine 5y ago
    Well, that’s the conclusion the authors want you to draw. IMHO, that’s a higher burden of proof than a correlation study like this can offer. As in, if I didn’t already drink coffee, and I was worried about stroke, I don’t think this study would be enough to confidently prescribe coffee as a stroke-prevention tool. But if I do drink coffee and I am worried about stroke, this study makes me a little less worried, because the answer to “well did coffee-drinkers disproportionately get stroke?” is “no.” The fact that it goes the other way just makes it a stronger “no”. Reasonable people could differ in interpreting this.
- Zigurd 5y ago
  The name checks out
kristopolous 5y ago
Apparently drinking soda can lead to those problems. Could it be that tea and coffee drinkers in general drink less soda? Was that controlled for? Alternatively maybe the soda studies are just proxies for tea and coffee.
- iamstupidsimple 5y ago
  > Was that controlled for? Very little can be controlled for in studies like this. There are just too many factors to slice on, and before long you're just reading into noise.
  - xattt 5y ago
    I think studies like this tend to overlook selection bias. What is typically omitted is that there is some degree of “tolerance” that’s needed to consume moderately caffeinated beverages. Individuals might avoid these drinks if they feel like they get negative effects (ie anxiety, palpitations, headaches) from them. I know this barely scratches the surface of the uncontrolled factors, but I wanted to throw this point out there.
- hajile 5y ago
  I'd love to see stats for caffeine-free people in a study like this. It seems like anybody can tell you that Soda is worse for you than tea/coffee (possible exception of sweet tea).
- hinkley 5y ago
  Also, are they the same mechanism? We know that staying mentally active can delay dementia symptoms for decades. Could it be that people who do mentally taxing work for a living drink more tea and coffee per capita, especially as they approach a certain age, so that coffee and dementia are negatively correlated but not causal?
- ac29 5y ago
  "estimate the associations between coffee/tea consumption and incident stroke and dementia, adjusting for [...], consumption of sugar-sweetened beverages"

caffeine 5y ago

The take away here is: there is no obvious evidence in this data that coffee and tea cause stroke or dementia. So if we were worried about that, at least there is no obvious relationship.

sokoloff 5y ago
Isn’t there a stronger conclusion evident: that they appear to reduce the risk (rather than not increase it)?
- headsoup 5y ago
  As is never highlighted, especially in media, is that it's relative risk reduction. Going by their study (and simplifying for argument's sake), there were 10k strokes out of 365k subjects, so <3% risk. A 30% risk reduction means now you're at 2% risk, so figure out whether that's worth adding coffee and tea for.
  - rmbyrro 5y ago
    On an individual level, it might be the difference between having dementia/stroke or not. The decrease from coffee/tea might make you cross the line back into healthy state, perhaps? I don't usually take conclusions from single studies like this, though. It's too easy to fool ourselves with statistics.
    - headsoup 5y ago
      Of course, but there are probably thousands of other decisions we make through life that could increase or decrease our chance 1%. Yeah, statistics and graphs can be a special type of evil
  - ThunderSizzle 5y ago
    A <1% death rate was enough to cause the lockdowns in 2020. If the media ran this endlessly, saying you'll have 30% less chance of stroke or dementia if you drink coffee and tea, people would probably panic on that too.
    - rebuilder 5y ago
      This study was conducted over 10+ years, the COVID numbers are probably per case, or possibly annualised.
    - icoder 5y ago
      Not sure about your country, but here the main problem was (and is again) the risk of an overrun health care system.
      - dahfizz 5y ago
        A large uptick in strokes would cause a similar issue, no? You can put that spin on anything. DISCLAIMER: I am not downplaying the severity of Covid. My words mean exactly what they say.
        sokoloff 5y ago
        That seems to depend on the typical duration of stay of a stroke patient in the various departments (emergency, ICU, general in-patient). I'm not familiar with stroke treatment, but from anecdotal exposure in my extended family's experience, they did not result in extended ICU stays.
        icoder 5y ago
        A large enough uptick in strokes at some point will, yes, but I don't think coffee/tea consumption will have that effect, plus our current consumption clearly _doesn't_, while Covid clearly _does_. (At least where I live, ICU's were previously not overrun due to anything) It's not a spin, it's just the daily struggle of the health care system right now. I see no other solution but to take that seriously. That doesn't mean we should take this study as serious just because it discusses something that may lead to something that, by itself, with a theoretical (large enough) uptick would overrun the system too.
  - pwagland 5y ago
    Right, but the point that parent was making is that rather than just not implicating tea/coffee in causing these, the study shows that it reduces the chance of these. That it's a relative risk reduction is not particularly relevant to that. That said, I do agree that changes in relative risk are confusing and/or misleading to many people. A more interesting way of presenting this data, IMO, is the "number needed to treat", that is in order to reduce 1 stroke from occurring, ie, roughly 120 people need to drink 2 cups of coffee/tea per day to reduce 1 stroke, based on the numbers given here. This is obviously a lot more than the naive "30% means roughly 1 in 3 would see improvement" interpretation that the headline number pushes you towards.
    - falcolas 5y ago
      Having seen how dementia and strokes affect people, and their families, I’ll absolutely take a 1% reduction.
- caffeine 5y ago
  Well, that’s the conclusion the authors want you to draw. IMHO, that’s a higher burden of proof than a correlation study like this can offer. As in, if I didn’t already drink coffee, and I was worried about stroke, I don’t think this study would be enough to confidently prescribe coffee as a stroke-prevention tool. But if I do drink coffee and I am worried about stroke, this study makes me a little less worried, because the answer to “well did coffee-drinkers disproportionately get stroke?” is “no.” The fact that it goes the other way just makes it a stronger “no”. Reasonable people could differ in interpreting this.
Zigurd 5y ago
The name checks out

kristopolous 5y ago

Apparently drinking soda can lead to those problems. Could it be that tea and coffee drinkers in general drink less soda? Was that controlled for? Alternatively maybe the soda studies are just proxies for tea and coffee.

iamstupidsimple 5y ago
> Was that controlled for? Very little can be controlled for in studies like this. There are just too many factors to slice on, and before long you're just reading into noise.
- xattt 5y ago
  I think studies like this tend to overlook selection bias. What is typically omitted is that there is some degree of “tolerance” that’s needed to consume moderately caffeinated beverages. Individuals might avoid these drinks if they feel like they get negative effects (ie anxiety, palpitations, headaches) from them. I know this barely scratches the surface of the uncontrolled factors, but I wanted to throw this point out there.
hajile 5y ago
I'd love to see stats for caffeine-free people in a study like this. It seems like anybody can tell you that Soda is worse for you than tea/coffee (possible exception of sweet tea).
hinkley 5y ago
Also, are they the same mechanism? We know that staying mentally active can delay dementia symptoms for decades. Could it be that people who do mentally taxing work for a living drink more tea and coffee per capita, especially as they approach a certain age, so that coffee and dementia are negatively correlated but not causal?
ac29 5y ago
"estimate the associations between coffee/tea consumption and incident stroke and dementia, adjusting for [...], consumption of sugar-sweetened beverages"