Junky Medical Journalism

The Washington Post reported yesterday that Overweight Elderly at Higher Risk for Alzheimer's

Overweight elderly people are more likely than those who stay trim to be stricken by Alzheimer's disease, researchers reported today...The findings add Alzheimer's to the long list of serious ailments caused by being overweight or obese ... For every 1 percent increase in their body mass index (BMI) at age 70, there was 36 percent increased risk for Alzheimer's

[emphasis mine].

By the time the story made it into print, some of the outrageous claims were toned down or corrected:

Study Links Excess Weight To Likelihood of Alzheimer's
Overweight elderly women are more likely than those who stay trim to be stricken by Alzheimer's disease, researchers reported yesterday... The findings add Alzheimer's to the long list of serious ailments associated with being overweight or obese ... For every one-point increase in their body mass index (BMI) at age 70, there was a 36 percent increased risk for Alzheimer's

See if you can tell what changed. (A formula for the body mass index is here. A "1% increase" in BMI implies a 1% increase in weight. A one-point increase in BMI (for a 5'5" person) implies a 6 lb. increase in weight, which, for most women anyway, is a lot more than 1%).

The article does not answer the most obvious questions, which are: What is the probability that a person will develop Alzheimer's at some point during their life -- e.g. 1%, 10%, 70%? And how is this probability affected by one's weight? (keep reading)

The cited report appears in The Archives of Internal Medicine/Vol. 163, July 14, 2003. For non-subscribers there is a $9 charge to download the article. I repost it here for non-commercial Fair Use.

The report does not demonstrate that excess weight causes Alzheimer's, all it shows is that the women in the study (221 women in Göteborg Sweden who turned 70 in 1971-72) who were diagnosed with Alzheimer's between the ages of 79-88 (all 17 of them) were, on average, heavier during their 70s than the women in the study who did not first develop Alzheimer's between ages 79-88. But in fact there could have been several reasons why a woman was in the latter (lighter) group, and not only because she lived to a spry old age of 88 without having developed Alzheimer's, these include: (a) she died at some point between the ages of 70 - 88 without having developed Alzheimer's; (b) she developed Alzheimer's before age 79; (c) she developed a form of dementia other than Alzheimer's sometime before age 88.

Furthermore, the report stressed the following:
* there was no significant difference in body weight (BMI, to be precise) between the men who developed dementia and those who did not.
* there was no significant difference in body weight (BMI) between the women who developed dementia before age 79 and those who did not.
* there was no significant difference in body weight (BMI) between the women who developed non-Alzheimer's types of dementia and those who did not.
* "A higher mean BMI may be protective against death, especially in women". In other words, the prevalence of Alzheimer's increases markedly with age, and heavier women might succumb to Alzheimer's partly because they live longer. The study did not give any detail on the actual longevity of its subjects.
* "physical activity was not assessed in this study. We therefore cannot comment on its potential influence on BMI and dementia risk". In other words, it is just as possible that instead of excess weight causing dementia, they are both exacerbated by the same underlying cause of inactivity. Indeed, it's even possible that the earliest (undetected) stages of dementia cause a slowdown in physical activity that in turn causes weight gain. But we don't know.

Although the research paper may have made some progress towards establishing a connection between weight and Alzheimer's , it's less than fully convincing. (You'll need more than 17 Swedish women to convince me). It's certainly not what I would consider front page news, as did the editors of the Washington Post and the Seattle Times. One would think that there are already enough reasons to watch one's weight without having to confuse the issue with an unproven connection to Alzheimer's. There are certainly enough other health care issues for newspapers to write about. Doing a half-assed job of communicating an inconclusive research report doesn't seem to be a very good way to inform the public about health issues.

Oh, and to answer the unanswered questions, a different study says that prevalence of Alzheimer's is in the ballpark of 15% for those age 65+, and in the neighborhood of 35% for those over 80 (that figure invokes a whole 'nother can of sticky health policy issues for another time). As for the article's claim of a "36 percent increased risk of Alzheimer's" -- that refers to a statistical measure of "risk", and it does not follow that a woman who is six pounds overweight has a 50% chance of developing Alzheimer's in her 80s, or that a woman who is 25 lbs. overweight will have a greater than 100% of developing Alzheimer's... The truth is, nobody really knows the extent to which being overweight increases your odds of developing Alzheimer's, or if it even does.

Oops, and I see that The Age also gets it wrong: "Obesity may lead to Alzheimer's" and

with every unit increase in body mass index at age 70, each woman's chance of developing Alzheimer's increased by 36 per cent.

Posted by Stefan Sharkansky at July 16, 2003 07:00 AM