Atlas of risk

ronv

Joined Nov 12, 2008
3,770
Pretty interesting.
They should probably show how they interact.
Like low fruits and vegetables with obesity and high cholesterol.
And alcohol and pregnancy. :rolleyes:
 

jpanhalt

Joined Jan 18, 2008
11,087
Such observations related to diet and lifestyle tend to go in cycles. Publication bias is probably a major contributor to that phenomenon and to the skepticism of many who have worked in the field. Such studies get play in the news, because almost everyone wants to prolong his life.

In the very early 1990's, I attended a lecture by Michael S. Brown (Nobel Prize, Physiology or Medicine, 1985) in which he discussed the original design and results of the Framingham Study on cardiovascular disease (http://en.wikipedia.org/wiki/Framingham_Heart_Study ). The original study was designed as a two-tailed study to assess overall death rates in the two groups of subjects. Nether group had detectable cardiovascular disease at entry. One group continued as they pleased; the other group had dietary and lifestyle interventions that were generally considered to be "healthy" (e.g., lower cholesterol foods, exercise, non-smoking, etc.). Overall death rates in both groups were not different. In fact, the intervention group had more suicides and accidental deaths (e.g., getting hit by a car while bicycling) than the control group. Contrary to the objections of some, the study design was changed after data were available to look only at cardiovascular deaths. It is clear that such manipulation of design and data after the fact should have been addressed in the statistical analysis, but it wasn't. Those who objected were ignored.

Thus, while it is hard to see how eating more vegetables and fruits per se would lead to more traffic deaths, it is important to remember that concurrent lifestyle changes might also occur that would negate the advantages of the changed diet. The term "risk compensation" has been applied to such offsetting effects (http://en.wikipedia.org/wiki/Risk_compensation ).

John
 

jpanhalt

Joined Jan 18, 2008
11,087
This sentence brought back memories:
Correction for multiple comparisons should be completed with these datasets, but is often ignored by investigators.
I have seen it done and published in the microbiology literature.

John
 

GopherT

Joined Nov 23, 2012
8,009
I'm sure Bertus will particularly like this MRI study paper and attached statistics.

http://prefrontal.org/files/posters/Bennett-Salmon-2009.pdf
I think the authors should have more clearly defined the degree of dead. The rudimentary nervous system of a fish does allow reflexes for minutes or hours after death.
Catfish, for instance, still exhibit an undulation in the meat after processing into filets.

They should have taken their positive correlation and examined it as a function of time since death. I look forward to more PITA-funded follow-up to this research. ;)
 

studiot

Joined Nov 9, 2007
4,998
Since I was rumbled by a well read gopher, perhaps the new statistical collection of howlers from professor Gary Smith is worth a mention

Standard Deviations
Gary Smith
Duckworth- Overlook 2014

;)
 

JoeJester

Joined Apr 26, 2005
4,390
PETA .... People Eating Tasty Animals ... of course there is an opposition group called PETA ... People for the Ethical Treatment of Animals.

PITA ... a bread that holds a tasty GYRO sandwich or the person who deserves this medal
 

Attachments

Top