3 ms·

There is always the problem of over time people not dying of other causes so that more get to live to the ages where they can get diseases that used to be rarer

by nanis 6y ago

There is always the problem of over time people not dying of other causes so that more get to live to the ages where they can get diseases that used to be rarer. Also, AFAIK, improvements have been made in the diagnosis of Parkinson's. > in the US, the number of people with Parkinson’s has increased 35% the last 10 years 10 years ago, the oldest U.S. baby boomers were reaching 65. Now, they are reaching 75[1]. Given the fact that birth rate went from 19/1000 to 27/1000 between 1935 and 1946 and the fact that people born in 1946 faced a much friendlier environment to grow up in, the numbers are not surprising. The fact is, industry also grew at the same time, so it will always be possible to find positive correlations between prevalence of some disease and some industrial chemical. It is not impossible that a causal relationship exists, but that needs to be established using something other than simple correlations. [1]: https://en.wikipedia.org/wiki/Mid-20th_century_baby_boom#North_America https://en.wikipedia.org/wiki/Mid-20th_century_baby_boom#Nor...

hutzlibu 6y ago
This is a very important, often overlooked point. Also food malnutrition is way less common, than it was before. But we also are exposed to many new chemicals, not occuring in nature, so we could not adopt to them. So ... I guess there is no definite, simple answer.
ddeck 6y ago
> > in the US, the number of people with Parkinson’s has increased 35% the last 10 years > 10 years ago, the oldest U.S. baby boomers were reaching 65. Now, they are reaching 75[1]. The numbers do match up rather well. The US population over the age of 65 has grown around 33% in the last 10 years. I see conflicting information regarding the average/median age at diagnosis, but it seems to be around there. https://fred.stlouisfed.org/series/POPTOTUSA647NWDB https://fred.stlouisfed.org/series/POPTOTUSA647NWDB https://fred.stlouisfed.org/series/SPPOP65UPTOZSUSA https://fred.stlouisfed.org/series/SPPOP65UPTOZSUSA
- alcover 6y ago
  > population over the age of 65 has grown around 33% in the last 10 years Which strikingly echoes GP's "birth rate went from 19/1000 to 27/1000 between 1935 and 1946". But 10 years too late ? I'm confused maybe.
  - ddeck 6y ago
    The data was intended to back up the GP's point, pointing out the growth in the underlying population of those diagnosed very closely mirrors the growth in number diagnosed.
    - alcover 6y ago
      I know. I meant the echoing wavefront seems 10y late : 1935+65 = 2000
      - nanis 6y ago
        > I know. I meant the echoing wavefront seems 10y late : 1935+65 = 2000 Good point. I believe I was thinking about 75. 65 crept in because I was also going to mention that at the beginning of the US Social Security program (1935), the assumption was no one lived much beyond 65. Here is a table that might cast light on the diagnosis of Parkinson's by age: https://academic.oup.com/view-large/1251178 https://academic.oup.com/view-large/1251178 Note population pyramids for 2000[1], 2010[2], and 2020[3]. [1]: https://www.populationpyramid.net/united-states-of-america/2000/ https://www.populationpyramid.net/united-states-of-america/2... [2]: https://www.populationpyramid.net/united-states-of-america/2010/ https://www.populationpyramid.net/united-states-of-america/2... [3]: https://www.populationpyramid.net/united-states-of-america/2020/ https://www.populationpyramid.net/united-states-of-america/2...
        alcover 6y ago
        Thank you. Your population links are fascinating. Going [1]->[2]->[3] shows that the real boom after all is the cohort born around 1960.
nabla9 6y ago
A six-fold increase in the risk in individuals exposed in the workplace to trichloroethylene is more like a natural experiment than "simple correlation".
- bobthepanda 6y ago
  It’s not the only thing they were exposed to though in the natural experiment, so it’s entirely possible that third factor C is what’s driving it. There are more studies according to the article that make the solid case, but one instance of correlation is not necessarily a smoking gun.
mrow84 6y ago
"A 2008 peer-reviewed study in the Annals of Neurology, for example, found that TCE is “a risk factor for parkinsonism.” And a 2011 study echoed those results, finding “a six-fold increase in the risk of developing Parkinson’s in individuals exposed in the workplace to trichloroethylene (TCE).”" It doesn't strike me as fair to dismiss studies like the ones quoted (from the article) as simple correlations.
- nanis 6y ago
  > It doesn't strike me as fair to dismiss studies like the ones quoted (from the article) as simple correlations. I didn't see the particular studies and I am not dismissing them out of hand, but I've seen a lot of studies, lost a lot of battles against all sorts of unsound statistical analysis. When I taught Intro Stats, I had no trouble finding examples of what not to do to illustrate in class: Just look at the front page of the campus newspaper which seemed intent on filling pages with bogus "studies". Again, I am not calling this one[1] bogus, but I am going to point out that study seems to consist solely of finding people who have Parkinson's and have worked around the chemical. Causality requires (not p) => (not q). [1]: https://hero.epa.gov/hero/index.cfm/reference/details/reference_id/700905 https://hero.epa.gov/hero/index.cfm/reference/details/refere...
chemeng 6y ago
While this is always something to consider, generally, 6-fold increase in risk due to exposure is likely not just a population demographic artifact. TCE and its relation to Parkinson's has been studied for at least the last 15 years. Though I haven't come across anything definitive on the mechanism of action, there seems to be indication that a combination of TCE or its analogues combined with other risk factors (assumed to be genetic) generates substantial progressive dopaminergic neuron loss, a hallmark of Parkinson's disease progression.
- nanis 6y ago
  > 6-fold increase in risk due to exposure Interestingly, the Guardian article does not cite a source for that claim, so it is impossible to evaluate it soundly. However, just above, it states: >> the number of people with Parkinson’s has increased 35% the last 10 years which is completely in line with the increase in at risk population + improved probability of diagnosis. That doesn't mean there is no link.
bjornjajayaja 6y ago
Honestly what good comes out of any “industrial scale” chemical? I think people are screwed because this is _one_ chemical; what about all of the synthetic materials in peoples houses these days (the carpet itself)? We need to go back to basics here instead of trying to over-science things. Bust out that bottle of vinegar and keep things simple in my opinion.
- allannienhuis 6y ago
  The whole world runs on industrial scale chemicals. Virtually everything in our modern society relies on them. They are base inputs to almost every chain of goods from foodstuffs to textiles to manufactured items of every sort. Even your vinegar example is primarily produced as an industrial scale chemical.
robocat 6y ago
A comparison can still be made between the same age cohorts i.e. compare rates of Alzheimer’s for the cohort 60 to 65 year olds. There is always a potential for confounding factors, but if the rates are up 35% for all cohorts then that says something. Also the beautiful flow graph https://images2.minutemediacdn.com/image/upload/c_fit,f_auto,fl_lossy,q_auto,w_728/v1555998897/shape/mentalfloss/women.png?itok=KtvGSMWm https://images2.minutemediacdn.com/image/upload/c_fit,f_auto... (graph for US Women I think) does show that after about 70 “mental” starts being a significant percentage of deaths, although Parkinson’s is only one component of that I think. Source for graph is: https://flowingdata.com/2016/01/05/causes-of-death/ https://flowingdata.com/2016/01/05/causes-of-death/
gnramires 6y ago
This can be done by 'Controlling for age' in statistics. If your cohort gets older, you get more prevalence of old age diseases of course. But by isolating by sufficiently small age groups (assuming the age distribution is approximately constant, or flat, within a group[1]), we can tell if the rate per given age is increasing or not. In this case simple multivariate statistics to judge if the changes are significant. [1] The variation in distribution among age groups can be controlled by more sophisticated methods, but it may be sufficient to make the groups small and multivariate analysis (although results might be weakened by assumptions of independent errors if groups are too small).

by nanis 6y ago

hutzlibu 6y ago

This is a very important, often overlooked point. Also food malnutrition is way less common, than it was before. But we also are exposed to many new chemicals, not occuring in nature, so we could not adopt to them. So ... I guess there is no definite, simple answer.

ddeck 6y ago

> > in the US, the number of people with Parkinson’s has increased 35% the last 10 years > 10 years ago, the oldest U.S. baby boomers were reaching 65. Now, they are reaching 75[1]. The numbers do match up rather well. The US population over the age of 65 has grown around 33% in the last 10 years. I see conflicting information regarding the average/median age at diagnosis, but it seems to be around there. https://fred.stlouisfed.org/series/POPTOTUSA647NWDB https://fred.stlouisfed.org/series/POPTOTUSA647NWDB https://fred.stlouisfed.org/series/SPPOP65UPTOZSUSA https://fred.stlouisfed.org/series/SPPOP65UPTOZSUSA

alcover 6y ago
> population over the age of 65 has grown around 33% in the last 10 years Which strikingly echoes GP's "birth rate went from 19/1000 to 27/1000 between 1935 and 1946". But 10 years too late ? I'm confused maybe.
- ddeck 6y ago
  The data was intended to back up the GP's point, pointing out the growth in the underlying population of those diagnosed very closely mirrors the growth in number diagnosed.
  - alcover 6y ago
    I know. I meant the echoing wavefront seems 10y late : 1935+65 = 2000
    - nanis 6y ago
      > I know. I meant the echoing wavefront seems 10y late : 1935+65 = 2000 Good point. I believe I was thinking about 75. 65 crept in because I was also going to mention that at the beginning of the US Social Security program (1935), the assumption was no one lived much beyond 65. Here is a table that might cast light on the diagnosis of Parkinson's by age: https://academic.oup.com/view-large/1251178 https://academic.oup.com/view-large/1251178 Note population pyramids for 2000[1], 2010[2], and 2020[3]. [1]: https://www.populationpyramid.net/united-states-of-america/2000/ https://www.populationpyramid.net/united-states-of-america/2... [2]: https://www.populationpyramid.net/united-states-of-america/2010/ https://www.populationpyramid.net/united-states-of-america/2... [3]: https://www.populationpyramid.net/united-states-of-america/2020/ https://www.populationpyramid.net/united-states-of-america/2...
      - alcover 6y ago
        Thank you. Your population links are fascinating. Going [1]->[2]->[3] shows that the real boom after all is the cohort born around 1960.

nabla9 6y ago

A six-fold increase in the risk in individuals exposed in the workplace to trichloroethylene is more like a natural experiment than "simple correlation".

bobthepanda 6y ago
It’s not the only thing they were exposed to though in the natural experiment, so it’s entirely possible that third factor C is what’s driving it. There are more studies according to the article that make the solid case, but one instance of correlation is not necessarily a smoking gun.

mrow84 6y ago

"A 2008 peer-reviewed study in the Annals of Neurology, for example, found that TCE is “a risk factor for parkinsonism.” And a 2011 study echoed those results, finding “a six-fold increase in the risk of developing Parkinson’s in individuals exposed in the workplace to trichloroethylene (TCE).”" It doesn't strike me as fair to dismiss studies like the ones quoted (from the article) as simple correlations.

nanis 6y ago
> It doesn't strike me as fair to dismiss studies like the ones quoted (from the article) as simple correlations. I didn't see the particular studies and I am not dismissing them out of hand, but I've seen a lot of studies, lost a lot of battles against all sorts of unsound statistical analysis. When I taught Intro Stats, I had no trouble finding examples of what not to do to illustrate in class: Just look at the front page of the campus newspaper which seemed intent on filling pages with bogus "studies". Again, I am not calling this one[1] bogus, but I am going to point out that study seems to consist solely of finding people who have Parkinson's and have worked around the chemical. Causality requires (not p) => (not q). [1]: https://hero.epa.gov/hero/index.cfm/reference/details/reference_id/700905 https://hero.epa.gov/hero/index.cfm/reference/details/refere...

chemeng 6y ago

While this is always something to consider, generally, 6-fold increase in risk due to exposure is likely not just a population demographic artifact. TCE and its relation to Parkinson's has been studied for at least the last 15 years. Though I haven't come across anything definitive on the mechanism of action, there seems to be indication that a combination of TCE or its analogues combined with other risk factors (assumed to be genetic) generates substantial progressive dopaminergic neuron loss, a hallmark of Parkinson's disease progression.

nanis 6y ago
> 6-fold increase in risk due to exposure Interestingly, the Guardian article does not cite a source for that claim, so it is impossible to evaluate it soundly. However, just above, it states: >> the number of people with Parkinson’s has increased 35% the last 10 years which is completely in line with the increase in at risk population + improved probability of diagnosis. That doesn't mean there is no link.

bjornjajayaja 6y ago

Honestly what good comes out of any “industrial scale” chemical? I think people are screwed because this is _one_ chemical; what about all of the synthetic materials in peoples houses these days (the carpet itself)? We need to go back to basics here instead of trying to over-science things. Bust out that bottle of vinegar and keep things simple in my opinion.

allannienhuis 6y ago
The whole world runs on industrial scale chemicals. Virtually everything in our modern society relies on them. They are base inputs to almost every chain of goods from foodstuffs to textiles to manufactured items of every sort. Even your vinegar example is primarily produced as an industrial scale chemical.

robocat 6y ago

A comparison can still be made between the same age cohorts i.e. compare rates of Alzheimer’s for the cohort 60 to 65 year olds. There is always a potential for confounding factors, but if the rates are up 35% for all cohorts then that says something. Also the beautiful flow graph https://images2.minutemediacdn.com/image/upload/c_fit,f_auto,fl_lossy,q_auto,w_728/v1555998897/shape/mentalfloss/women.png?itok=KtvGSMWm https://images2.minutemediacdn.com/image/upload/c_fit,f_auto... (graph for US Women I think) does show that after about 70 “mental” starts being a significant percentage of deaths, although Parkinson’s is only one component of that I think. Source for graph is: https://flowingdata.com/2016/01/05/causes-of-death/ https://flowingdata.com/2016/01/05/causes-of-death/

gnramires 6y ago

This can be done by 'Controlling for age' in statistics. If your cohort gets older, you get more prevalence of old age diseases of course. But by isolating by sufficiently small age groups (assuming the age distribution is approximately constant, or flat, within a group[1]), we can tell if the rate per given age is increasing or not. In this case simple multivariate statistics to judge if the changes are significant. [1] The variation in distribution among age groups can be controlled by more sophisticated methods, but it may be sufficient to make the groups small and multivariate analysis (although results might be weakened by assumptions of independent errors if groups are too small).