Statistics is a greatly specialized science. No regular physician can know if a statistical finding is significant until a statistician has gone over the numbers. It’s not something that can be decided on one’s feel or hunch.
That being said, statisticians are human too, prone to making mistakes. Sometimes they act like the proverbial crazy professor, who is so smart that he forgets to do a reality check. I’m not even an amateur statistician but I have some common sense. And different from many reporters, I don’t go blank on any number presented. Let’s think about it with our horse sense. If it doesn’t make sense on a simple level, we know enough!
I learned from Pathophysiology Professor Van Gool (Wilhelmina Gasthuis), that a single case history may teach more about a mechanism of illness than the largest and most expensive statistical studies. After all, statistics can only give a correlation (which could show mere coincidence), not a causation. The rooster indeed may “think” that his crowing makes the sun rise. And is it a pure coincident that that a stalk population decline was followed by falling birth rates? (It is.)
My father, physician and not a statistician, warned that statistics are like bikinis: they hide the parts you’d have liked to see most.
The value of good statistics is that it could hint to a possible correlation between cause and illness. But the real proof of the pudding lies in finding a real causative mechanism. Without that, it’s just a clue.
One of the trickiest things in statistics is that something must be more or less rare compared to other people or people in other situation. There hardly ever will be a perfect control group to compare the patients with. There are so many examples of historical monumental medical mistakes because of a flawed control group. Two examples.
Mothers who spontaneously lost their pregnancy too early and were given DES had a greater chance next times to deliver a baby old enough to survive. However, without DES, mothers who spontaneously lost their pregnancy too early also had a greater chance next times to deliver a baby old enough to survive. (The pharmaceutical company knew and sold the junk anyway – which proved besides not working very harmful – first discovered by a high incidence of a rare cancer in girls born from this.)
Men who had a heart attack between the ages of 50 and 60 were much more often workaholics than men who didn’t. Until they found, 50 years later, that being hyperactive actually improves survival when one gets in cardiac trouble. The overworking doesn’t cause danger but protects.
Another major source of statistical mistakes may come from doing research looking back instead of forward or coming in half-way something in motion already. One needs to figure out what bias has crept in before one starts measuring. Here, hindsight is not 20/20!
Dementia in Holocaust Survivors
The statistics now presented say that survivors have a 1.2 times higher chance to get dementia over a span of 10 years. (It is unclear to me why people living with dementia and who already died while dement, were excluded from these statistics and what this exclusion does with the figures.) That means a 20% higher chance. If between 100 non-exposed people (one report calls them erroneously ‘non-survivors’) 9 get dementia, between survivors, that would be 10-11. In individual cases, that is meaningless. It would be different if the numbers were 9/18.
Second of all, the survivors will have a much greater potential to get older than the control group. This is because the physically and emotionally weak have died decades ago. What’s left are mainly sturdy fighters. Now, the researchers write that they corrected for age but until I see how they did this, I doubt if that was done properly. They have likely measured rather that Holocaust survivors still alive, live longer.
Thirdly, the control group should have only consisted of Europeans. I knew a couple, he Holocaust survivor and dementing and she Moroccan and … dementing too. But no one had noted about her. Unaware racism is everywhere. Ashkenazic patients do get more medical attention. Even her own family didn’t want to know. “She’s fine.” So, most likely, in the control group, dementia was under-diagnosed. Most Israelis are of Mideastern descent. That totally kills the “significant” 20% find.
Four, there’s this tendency to look intently for trauma and trouble in Holocaust survivors. (This also is done to the elderly. Endless lists of “the ailments of old age.” They never talk about what great it is that above 60 the chance to heart attacks diminishes, allergies die out, keeping up appearance takes a nosedive, that above 90, it’s highly unlikely you still get any of the popular killer diseases, most illnesses, etc.) This, on top of the many (Jews included) who see Jews as a pitiful bunch, sufferers.
Five, as mentioned above, until a mechanism has been found by which Holocaust stress would give dementia (on contribute to developing dementia), these correlations are at best hints, not facts. But, causes for dementia are far from known. I wouldn’t hold my breath on this one.
Survivors of the Holocaust still alive probably live longer and better than those who weren’t subjected to genocide. In any case, they should not fear developing dementia more than anyone else unless there is proof of that. And then I don’t mean, shaky hints or marginally elevated risks.