Kelly Bailey via Compfight
OK – so the title of this post should have been “DOES Beer Help You Sleep?”. When I read the articles that talked about it, I was sceptical – but hopeful – to say the least…
This is another study I read recently in the PLoS One Journal titled “The Sedative Effect of Non-Alcoholic Beer in Healthy Female Nurses”. The conclusion they reached is this:
“The moderate consumption of non-alcoholic beer will favour night-time rest, due in particular to its hop components, in addition to its other confirmed benefits for the organism.”
I read through the study, the sceptic in me kicked in, and even I can see problems with it from the outset.
How does this “research” get funding? Here’s what my first glance reveals:
- It’s very, very small – 17 nurses in all. Drawing any conclusion from such a small group is suspect. I could find 17 people with red hair that all have toenail fungus. I could then assume that toenail fungus is the cause of red hair (or is it the reverse?).
- They used surveys (self-reporting) as a baseline for entry into the study. Self reporting is not realiable, since it introduces variability that’s very hard to account for. I don’t know how much beer I had last week – much less last month!
- Sleep was measured using an Actiwatch, which is considered a standard for measuring sleep outside of a sleep lab. It’s accelerometer based, and monitors limb movement throughout the night. It cannot tell what “sleep state” you are in – and without that, it’s impossible to tell whether your sleep was restful or not. A better test would be a sleep lab that could actually monitor sleep stages.
The actual results they obtained were “noise” as far as I can determine, with the exception of Time-To-Sleep. But then again, by using an Actiwatch, how can you really tell?? And using yourself as a control isn’t exactly science, either…
Here are a couple of the charts from the actual study. The sleep efficiency chart above is typical of all the others they gathered with the exception of Time-To-Sleep below (20 minutes down to 12 minutes is statistically significant – but the actual onset of sleep would be very hard to measure with an Actiwatch if you’re just lying still in bed – so I’m suspect).
I have a couple of questions for you about this study in particular, but in the whole sleep-study field in general:
- Should we even acknowledge such small scale, self-reporting studies?
- I don’t know anyone with an Actiwatch and don’t know how they get around this “lying very still but awake” problem. Do you? It doesn’t seem all that different from a FitBit.
Leave a comment and tell me what you think!
