Then, argh, weep, grind, gnash my teeth. Making the news rounds today is a story based on an observational (therefore CRAP) study about higher cholesterol being ASSOCIATED with the development of Alzheimer's later. Crap. It's crap. Breathing, you might recall, is associated with dying, in that all who breathe die. Walking is associated with falling; running even more so. If you sleep, you have an association with dying later. Hope that makes you anxious. You might be killing yourself RIGHT NOW what with all that inhaling and exhaling, walking, and sleeping.
From the WSJ Health Blog:
More data this week suggesting that common risk factors for heart disease may also raise the risk of developing Alzheimer’s: A study of nearly 10,000 people found that having high cholesterol during midlife increased the risk of developing Alzheimer’s disease in old age.
The finding was based on cholesterol measurements taken in the 1960s and early 70s, when the patients were in their early 40s, and on Alzheimer’s Disease diagnoses that came decades later.
Not so fast, Cowboy. Schwitzer takes 'em down:
.....in a story sent to me by one of my blog followers because she was bothered by it, CNN reported that:People as young as 40 with borderline or high cholesterol levels are at increased risk for developing Alzheimer's disease or vascular dementia, said a Kaiser Permanente study released Tuesday.
But it took them until the second last line of the story to add an important caveat:Although the Kaiser study does not show proof that lowering cholesterol definitively lowers the risk for Alzheimer's disease as well, many doctors agree that nothing adverse can come of reducing high cholesterol levels.
Wait a minute: the lead says they "ARE AT INCREASED RISK" but the end of the story says there's no proof "THAT LOWERING CHOLESTEROL DEFINITIVELY LOWERS THE RISK."
That's because this was an observational study and the story never addressed the limitations of drawing conclusions from observational studies. It confused association (which is all an observational study can show) with causation (which an observational study can't show). But it didn't stop them from DEFINITIVELY scaring the hell out of cheeseburger-eating-40-year-olds in the lead. Be very afraid: you're at risk and there may be nothing you can do about it! Amazing.
Got that? Even though there was an ASSOCIATION (remember, association does NOT equal causation) between elevated cholesterol (as measured in the 1960s and '70s) and development of dementia later, there was no evidence that treating the cholesterol made any difference.
More fine print: Of about 10,000 individuals OBSERVED, just under 600 developed a vascular or Alzheimer's type dementia. 5% those with "desireable" cholesterol range (under 200) developed dementia. Only 6% of those with borderline cholesterol (200-239) developed a dementia; For cholesterols above 240, 6% developed Alzheimers. So there was a 1% absolute risk reduction for having a lower cholesterol. Hmmmm.....Hardly making me put down my cheeseburger. The vast majority of individuals didn't develop the disease at all (at least as of the conclusion of the study). That's not the way the scare-mongering headlines read, however.
Other confounding issues in this study include that it only observed individuals who were enrolled in Kaiser Permanente at the time of enrollment AND at the conclusion of the study. Patients included with the diagnosis of dementia (either vascular or Alzheimer's type) were captured using ICD-9 codes, a big problem for me. I choose codes for any given encouter NOT using science, but using coding rules, which are bizarre and a world unto themselves. Did physicians code for dementia more for some reason in patients with higher cholesterol? We'll never know. What about the people who dropped out of Kaiser and went to other health plans? Confidence intervals were not optimal. Cases weren't matched, but "modeling" occured, which undermines the data in an unimpressive way. LDL, widely used as a marker for "bad" cholesterol, wasn't included in the cholesterol panels the researchers used in the 60s and 70s. We're not comparing apples to apples, in other words. Nevertheless, if you read the news headlines:
From CNN Health:
"High Cholesterol today might mean Alzheimer's tomorrow"
Mildly High Cholesterol at Midlife Linked to Alzheimer's
High Cholesterol Linked to Alzheimer's