3 ms·

So you can reduce your dementia incidence risk from Q1 -> Q5 by a whopping 0.08%-points. But in media you surely will read about a 40% reduction. *edited: %-po

So you can reduce your dementia incidence risk from Q1 -> Q5 by a whopping 0.08%-points. But in media you surely will read about a 40% reduction. *edited: %-points instead of %

KempyKolibri 8mo ago
Sure, because both are true (although that 0.08% is only over 8 years of known omega 3 consumption - as timescales increase the absolute risk moves towards the relative risk). That 0.08% reduction would mean approximately 28,000 fewer EOD cases - not to be sniffed at!
- AlecSchueler 8mo ago
  > That 0.08% reduction would mean approximately 28,000 fewer EOD cases - not to be sniffed at! What would it mean for salmon stocks and increased environmental damage?
  - KempyKolibri 8mo ago
    Depending on where you source your omegas from, potentially zero impact! To be clear my preference would be to source n3s from algal supplements and, once food safety testing for humans is complete, n3s from GM rapeseed. In time I hope we end up with lab meat/plant-based meat alternatives that use these n3s so we can get the benefits of fish without the environmental and ethical concerns of getting n3s from fish.
    - susiecambria 8mo ago
      If from menhaden, there's a raging debate on the one hand about trout, the Chesapeake Bay (Maryland and Virginia), the ecology and environment more broadly, and the other hand a Canadian company based in a small rural county in Virginia (Omega Protein, which, BTW, does not provide year-round benefits to all of its employees which creates a drain on already super limited services and supports. Omega Protein is not alone in this.). I don't know enough about any of this to have an informed opinion, but I do understand that menhaden put Reedville, VA on the map.
llm_nerd 8mo ago
This is talking about early onset, which is a particularly terrifying outcome. And yes, 1 in 1000 for a horrible outcome sounds much better than 2 in 1000, doesn't it? And to be clear, many things that people worry about is less likely than that. Homicides (over an 8 year period about about 0.04 per 1000 people), terrorism (vanishingly small), and on and on. None of this means that people should stock up on omega-3s, and as likely the study is actually finding a correlation with something else (e.g. wealthier people enjoy more fish rich diets and are less exposed to toxins, or something else), but halving something terrifying that isn't that uncommon is legitimately newsworthy.
ricardobeat 8mo ago
The 40% (66%?) is the number that matters. Same way you wearing a helmet reduces your changes of brain damage in a motorcycle accident by 90%, yet you’re not on a motorcycle most of the time.
- djoldman 8mo ago
  When it comes time to decide whether or not to take action and what that action should be, I'd say that the total potential risk reduction is more important. One should weigh the cost of the proposed intervention in time/money/other_expense against the potential benefit. The potential benefit is the total reduction in risk * the magnitude of the unwanted outcome. 40% is less relevant.
  - KempyKolibri 8mo ago
    The thing is, the 0.08% doesn't capture the total potential risk reduction - only the risk during the timeframe of the study (8 years in this case). Where we're talking about exposures and outcomes that stack over time (exposure to LDL and heart disease being a classic one) the absolute risk is, in my opinion, more misleading than the relative risk. For example you see this oft-quoted stat about "statins only increase lifespan by 3 days" based off relatively short RCTs, but this doesn't capture the effect of statin use over decades, which is where we see much, much bigger gains. It seems to me that both RR and AR are things to take into consideration and we have to be mindful of the shortcomings of each.
mrob 8mo ago
The reduction in risk is 0.08 percentage points, not 0.08 percent. The "%" symbol always means "percent", not "percentage points". The 0.08 percentage point reduction is a 40% reduction. EDIT: don't assume the causality
- baxtr 8mo ago
  Maybe they wanted to say "down to" instead of "by"?
- sieabahlpark 8mo ago
  [dead]
fulafel 8mo ago
... over 8 years. Order of magnitude difference if it extrapolates to lifetime.

So you can reduce your dementia incidence risk from Q1 -> Q5 by a whopping 0.08%-points. But in media you surely will read about a 40% reduction. *edited: %-points instead of %

KempyKolibri 8mo ago

Sure, because both are true (although that 0.08% is only over 8 years of known omega 3 consumption - as timescales increase the absolute risk moves towards the relative risk). That 0.08% reduction would mean approximately 28,000 fewer EOD cases - not to be sniffed at!

AlecSchueler 8mo ago
> That 0.08% reduction would mean approximately 28,000 fewer EOD cases - not to be sniffed at! What would it mean for salmon stocks and increased environmental damage?
- KempyKolibri 8mo ago
  Depending on where you source your omegas from, potentially zero impact! To be clear my preference would be to source n3s from algal supplements and, once food safety testing for humans is complete, n3s from GM rapeseed. In time I hope we end up with lab meat/plant-based meat alternatives that use these n3s so we can get the benefits of fish without the environmental and ethical concerns of getting n3s from fish.
  - susiecambria 8mo ago
    If from menhaden, there's a raging debate on the one hand about trout, the Chesapeake Bay (Maryland and Virginia), the ecology and environment more broadly, and the other hand a Canadian company based in a small rural county in Virginia (Omega Protein, which, BTW, does not provide year-round benefits to all of its employees which creates a drain on already super limited services and supports. Omega Protein is not alone in this.). I don't know enough about any of this to have an informed opinion, but I do understand that menhaden put Reedville, VA on the map.

llm_nerd 8mo ago

This is talking about early onset, which is a particularly terrifying outcome. And yes, 1 in 1000 for a horrible outcome sounds much better than 2 in 1000, doesn't it? And to be clear, many things that people worry about is less likely than that. Homicides (over an 8 year period about about 0.04 per 1000 people), terrorism (vanishingly small), and on and on. None of this means that people should stock up on omega-3s, and as likely the study is actually finding a correlation with something else (e.g. wealthier people enjoy more fish rich diets and are less exposed to toxins, or something else), but halving something terrifying that isn't that uncommon is legitimately newsworthy.

ricardobeat 8mo ago

The 40% (66%?) is the number that matters. Same way you wearing a helmet reduces your changes of brain damage in a motorcycle accident by 90%, yet you’re not on a motorcycle most of the time.

djoldman 8mo ago
When it comes time to decide whether or not to take action and what that action should be, I'd say that the total potential risk reduction is more important. One should weigh the cost of the proposed intervention in time/money/other_expense against the potential benefit. The potential benefit is the total reduction in risk * the magnitude of the unwanted outcome. 40% is less relevant.
- KempyKolibri 8mo ago
  The thing is, the 0.08% doesn't capture the total potential risk reduction - only the risk during the timeframe of the study (8 years in this case). Where we're talking about exposures and outcomes that stack over time (exposure to LDL and heart disease being a classic one) the absolute risk is, in my opinion, more misleading than the relative risk. For example you see this oft-quoted stat about "statins only increase lifespan by 3 days" based off relatively short RCTs, but this doesn't capture the effect of statin use over decades, which is where we see much, much bigger gains. It seems to me that both RR and AR are things to take into consideration and we have to be mindful of the shortcomings of each.

mrob 8mo ago

The reduction in risk is 0.08 percentage points, not 0.08 percent. The "%" symbol always means "percent", not "percentage points". The 0.08 percentage point reduction is a 40% reduction. EDIT: don't assume the causality

baxtr 8mo ago
Maybe they wanted to say "down to" instead of "by"?
sieabahlpark 8mo ago
[dead]

fulafel 8mo ago

... over 8 years. Order of magnitude difference if it extrapolates to lifetime.