Extract:
http://www.ncbi.nlm.nih.gov/pubmed/20870457
The title of the original article is "Mortality related to actigraphic long and short sleep."
Here are a few points:
- The results are based on data collected from 1995-1999. "Actual sleep" was determined by an actigraph, sort of like a pedometer worn on the wrist.
- The results are based on estimated survival, not actual morality rates, because many women could not be contacted in the follow-up.
- From the article, "Subsample recruitment was deliberately structured to include as many older women as possible to increase the power of mortality analyses."
- The age range of the women in the study was 50-81. We know that older people sleep less than younger people.
- The authors do not make causal claims in the original article. It is typical for the popular media to spin it that way, though.
- Unfortunately the data are not fully consistent with their predicted u-shaped conclusion, because the average estimated survival rates for those women who slept less than 4 hours and also for those who slept more than 7.5 hours are nearly the same (86%) as those who slept 6-6.5h (90%). The lowest survival estimates were in the 4.5-5h (54%) and 7-7.5h (58%) sleep duration categories.
- There is huge variability in the risk ratio of the two "worst" sleep duration categories: in the 7-7.5 and 4.5-5 hour rages. In my opinion, this helps explain why the results from these categories are so different from all the other categories.
- Reported sleep is almost always longer than actual sleep, so the women who actually slept 6-6.5 hours probably reported about 7 hours of sleep.
It's analogous to the confusion over cholesterol: people who haven't damaged their bodies and thus require higher levels of cholesterol for repair will have lower levels of cholesterol, but this by no means implies that lowering the level of the substance that repairs said damage- cholesterol- will make you healthier.
LMAO! It was all older people. That explains a lot. And you are right, ridiculous irresponsible spin it ubiquitous in the media. There really should be some kinds of standard and rules about that kind of thing! I have heard many researchers complain about it themselves. – Eva