3 ms·

I'm not saying this to be mean, but I think it's important to point out that Dave Asprey is a quack, and his "Upgraded Coffee" is snake oil. Asprey is happy to

I'm not saying this to be mean, but I think it's important to point out that Dave Asprey is a quack, and his "Upgraded Coffee" is snake oil. Asprey is happy to cherry-pick studies to support his products, but the fact is that there is simply no good reason to believe that mycotoxin levels in ordinary coffee are high enough to have any measurable effect on humans who drink it. (And if you look carefully at his citations, you'll notice that he does not bridge that crucial gap between "mycotoxins are bad for you" and "my coffee has lower mycotoxin content"). This guy's combination of scientific literacy and alarmingly poor reasoning really makes me nauseous, because I can't simply write it off as mere incompetence. For instance, look at this infographic: http://www.bulletproofexec.com/wp-content/uploads/2012/04/Butter-Infographic1.jpg http://www.bulletproofexec.com/wp-content/uploads/2012/04/Bu... . Specifically, look at the part comparing 1910 and 2000. That's not just a case of spurious correlation. It's like something I would make up as a wacky example of spurious correlation.

beagle3 14y ago
Well, you may call him what you want, and yes, he is not right about everything. > you'll notice that he does not bridge that crucial gap between "mycotoxins are bad for you" and "my coffee has lower mycotoxin content" I'm not sure which gap you think needs to be bridged here? > Specifically, look at the part comparing 1910 and 2000. That's not just a case of spurious correlation. It's like something I would make up as a wacky example of spurious correlation. I'm not sure what you're reading into this. When I look at it, I don't think "butter protects against heart problems" (which seems to be your interpretation, and I agree cannot be concluded from this data, not even remotely), but rather "butter can't be the evil it is made to be, as this statistic shows". Which is perfectly reasonable. The thing about the QS movement, and people like Asprey, Roberts and Ferris, is that they offer very cheap experiments with easily measurable outcome, so you don't have to wait for a 30-year double blinded test that is blessed by anyone. I've tried several things they recommend, and the vast majority of the practice checks out (regardless of the theory behind it). I haven't tried his coffee, but I did start putting butter in my Lavazza coffee, and as a result lost weight and am feeling much better. You don't have to be a "believer" to try and see if things work for you. I don't think he's right about everything, but a lot of the counterintuitive things him (and Ferris, and Roberts) suggest actually work.
- mistercow 14y ago
  >I'm not sure which gap you think needs to be bridged here? I said it explicitly: "there is simply no good reason to believe that mycotoxin levels in ordinary coffee are high enough to have any measurable effect on humans who drink it." >When I look at it, I don't think "butter protects against heart problems" (which seems to be your interpretation, and I agree cannot be concluded from this data, not even remotely), but rather "butter can't be the evil it is made to be, as this statistic shows". Which is perfectly reasonable. It isn't perfectly reasonable. Life expectancy in 1910 was 48 for men, and 51 for women. The average age for heart attacks today is 66 for and 70 for women. Some, but not all of that is infant mortality, yes, but the fact is that if people in 1910 had less heart disease than people today, it is likely largely because people in 1910 often didn't long enough for heart disease to become a problem. And that's just one factor that the comparison fails to control for. The way people lived and ate in 1910 was completely different from the way people lived and ate in 2000, and it is quite likely that those changes simply overwhelm the change with respect to one particular kind of saturated fat. That doesn't mean that butter fat is harmless. It just means that you can't see the harm if you also change everything else. As a side note, where are these statistics about heart disease in 1910 coming from? We get the top-level URL for dietheartpublishing.com (which is low on data, high on sensationalist paleo-quackery), but no link to specific data. Are we relying on what doctors were able to diagnose in 1910? Because that's a whole other level of flawed. >I've tried several things they recommend, and the vast majority of the practice checks out (regardless of the theory behind it). What you think you've experienced through personal trials is not significant evidence.
  - beagle3 14y ago
    > "there is simply no good reason to believe that mycotoxin levels in ordinary coffee are high enough to have any measurable effect on humans who drink it." I don't keep the links handy, but a one second google brings up this: http://www.ncbi.nlm.nih.gov/pubmed/16284846 http://www.ncbi.nlm.nih.gov/pubmed/16284846 and a tens of other references. Coffee was not studied as extensively as far as I know, but if you have reason to believe that it is better protected than wheat and maize, please let me know why. The only non-asprey link about coffee toxin studies I could find is a single article from the 1980 that says "I looked, I didn't find, I stopped looking". > Some, but not all of that is infant mortality, yes, but the fact is that if people in 1910 had less heart disease than people today, it is likely largely because people in 1910 often didn't long enough for heart disease to become a problem. Well, in 1860-1910, US white infant mortality rate was coming down from 21% to 10% (with black mortality rate, when documented, being 50% higher than the white one). I don't know if that data is included in the "1910 life expectancy", but it is likely that it is. In Italy, it was going down from 25% to 13% over the same period. That's enough to make the "48/51" lifespan you quote meaningless (sources I found say 54, but whatever). http://eh.net/encyclopedia/article/haines.demography http://eh.net/encyclopedia/article/haines.demography If you consider "life expectancy of anyone aged > 4", then all of a sudden 50 years jumps to 62 years; and if you consider "life expectancy of anyone aged > 20", then it will surely increase the 1910 number to 70 or more, while hardly nudging today's numbers. I agree that his statistic is not conclusive in any way, but your interpretation is not valid either. > That doesn't mean that butter fat is harmless. It just means that you can't see the harm if you also change everything else. That is true. And yet, there is virtually no direct evidence of harm from butter than I can find. (Neither is there from cholesterol intake, for that matter - or salt intake, with the exception of a well defined 20% of the population - that is, if you're not in those 20%, there is no evidence of salt harm). > Are we relying on what doctors were able to diagnose in 1910? Because that's a whole other level of flawed. Yes, that's true. But that throws away any study done before the early 1990s - which is also flawed. We're damned if we do, damned if we don't, and everyone gets to pick their own favourite data and interpretation. > What you think you've experienced through personal trials is not significant evidence. What I discovered is significant for my daily life, is significant for my daily life (duh, it's a tautology). I'm not saying you should believe it because it worked for me. I'm saying you should try and figure what works for you. FWIW, my ex was a medical doctor and I regularly read BMJ and NEMJ for a while. That was before Ionnadis's paper, but I remember thinking back then that 70% of what was published was statistically invalid. Much of what is considered significant evidence in medicine and health today, is not significant evidence. If the experiment is simple and cheap enough, you should do it yourself.

beagle3 14y ago

Well, you may call him what you want, and yes, he is not right about everything. > you'll notice that he does not bridge that crucial gap between "mycotoxins are bad for you" and "my coffee has lower mycotoxin content" I'm not sure which gap you think needs to be bridged here? > Specifically, look at the part comparing 1910 and 2000. That's not just a case of spurious correlation. It's like something I would make up as a wacky example of spurious correlation. I'm not sure what you're reading into this. When I look at it, I don't think "butter protects against heart problems" (which seems to be your interpretation, and I agree cannot be concluded from this data, not even remotely), but rather "butter can't be the evil it is made to be, as this statistic shows". Which is perfectly reasonable. The thing about the QS movement, and people like Asprey, Roberts and Ferris, is that they offer very cheap experiments with easily measurable outcome, so you don't have to wait for a 30-year double blinded test that is blessed by anyone. I've tried several things they recommend, and the vast majority of the practice checks out (regardless of the theory behind it). I haven't tried his coffee, but I did start putting butter in my Lavazza coffee, and as a result lost weight and am feeling much better. You don't have to be a "believer" to try and see if things work for you. I don't think he's right about everything, but a lot of the counterintuitive things him (and Ferris, and Roberts) suggest actually work.

mistercow 14y ago
>I'm not sure which gap you think needs to be bridged here? I said it explicitly: "there is simply no good reason to believe that mycotoxin levels in ordinary coffee are high enough to have any measurable effect on humans who drink it." >When I look at it, I don't think "butter protects against heart problems" (which seems to be your interpretation, and I agree cannot be concluded from this data, not even remotely), but rather "butter can't be the evil it is made to be, as this statistic shows". Which is perfectly reasonable. It isn't perfectly reasonable. Life expectancy in 1910 was 48 for men, and 51 for women. The average age for heart attacks today is 66 for and 70 for women. Some, but not all of that is infant mortality, yes, but the fact is that if people in 1910 had less heart disease than people today, it is likely largely because people in 1910 often didn't long enough for heart disease to become a problem. And that's just one factor that the comparison fails to control for. The way people lived and ate in 1910 was completely different from the way people lived and ate in 2000, and it is quite likely that those changes simply overwhelm the change with respect to one particular kind of saturated fat. That doesn't mean that butter fat is harmless. It just means that you can't see the harm if you also change everything else. As a side note, where are these statistics about heart disease in 1910 coming from? We get the top-level URL for dietheartpublishing.com (which is low on data, high on sensationalist paleo-quackery), but no link to specific data. Are we relying on what doctors were able to diagnose in 1910? Because that's a whole other level of flawed. >I've tried several things they recommend, and the vast majority of the practice checks out (regardless of the theory behind it). What you think you've experienced through personal trials is not significant evidence.
- beagle3 14y ago
  > "there is simply no good reason to believe that mycotoxin levels in ordinary coffee are high enough to have any measurable effect on humans who drink it." I don't keep the links handy, but a one second google brings up this: http://www.ncbi.nlm.nih.gov/pubmed/16284846 http://www.ncbi.nlm.nih.gov/pubmed/16284846 and a tens of other references. Coffee was not studied as extensively as far as I know, but if you have reason to believe that it is better protected than wheat and maize, please let me know why. The only non-asprey link about coffee toxin studies I could find is a single article from the 1980 that says "I looked, I didn't find, I stopped looking". > Some, but not all of that is infant mortality, yes, but the fact is that if people in 1910 had less heart disease than people today, it is likely largely because people in 1910 often didn't long enough for heart disease to become a problem. Well, in 1860-1910, US white infant mortality rate was coming down from 21% to 10% (with black mortality rate, when documented, being 50% higher than the white one). I don't know if that data is included in the "1910 life expectancy", but it is likely that it is. In Italy, it was going down from 25% to 13% over the same period. That's enough to make the "48/51" lifespan you quote meaningless (sources I found say 54, but whatever). http://eh.net/encyclopedia/article/haines.demography http://eh.net/encyclopedia/article/haines.demography If you consider "life expectancy of anyone aged > 4", then all of a sudden 50 years jumps to 62 years; and if you consider "life expectancy of anyone aged > 20", then it will surely increase the 1910 number to 70 or more, while hardly nudging today's numbers. I agree that his statistic is not conclusive in any way, but your interpretation is not valid either. > That doesn't mean that butter fat is harmless. It just means that you can't see the harm if you also change everything else. That is true. And yet, there is virtually no direct evidence of harm from butter than I can find. (Neither is there from cholesterol intake, for that matter - or salt intake, with the exception of a well defined 20% of the population - that is, if you're not in those 20%, there is no evidence of salt harm). > Are we relying on what doctors were able to diagnose in 1910? Because that's a whole other level of flawed. Yes, that's true. But that throws away any study done before the early 1990s - which is also flawed. We're damned if we do, damned if we don't, and everyone gets to pick their own favourite data and interpretation. > What you think you've experienced through personal trials is not significant evidence. What I discovered is significant for my daily life, is significant for my daily life (duh, it's a tautology). I'm not saying you should believe it because it worked for me. I'm saying you should try and figure what works for you. FWIW, my ex was a medical doctor and I regularly read BMJ and NEMJ for a while. That was before Ionnadis's paper, but I remember thinking back then that 70% of what was published was statistically invalid. Much of what is considered significant evidence in medicine and health today, is not significant evidence. If the experiment is simple and cheap enough, you should do it yourself.