Hacker News Reader

Comment by Estragon | Hacker News Reader

  > a recent study estimated that these drugs were killing 
  > 300,000 - 500,000 Americans each year

Looks like sloppy epidemiology to me. The only way to nail a claim like that down would be to establish a control group with very similar life stressors, something you couldn't do from medical records. People take sleeping pills because they're under stress bad enough to keep them from sleeping, and that has an impact on survival rates.

  > what is clear is that the longterm use of them is terrible
  > for you and that they have zero longterm efficacy in the
  > first place.)

This is interesting. What's the evidence?

View on HN

"This is interesting. What's the evidence?"

http://www.madinamerica.com/2011/11/anxiety/

The actual book (Anatomy of an Epidemic) explains it better, but those are just some quick links. They're also vastly more addictive than heroin. While the worst of heroin withdrawals are over in less than a week, benzo withdrawals can literally last over a year, with people unable to sleep and in severe pain that entire time:

http://en.wikipedia.org/wiki/Benzodiazepine_withdrawal_syndr...

This cannot be stressed enough. Benzodiazepines are insidious because people get used to taking them when they have trouble sleeping, and when they try to stop, they cannot sleep properly for a very long time. I have seen this happen with two of my friends, and it is painful just to watch. Not being able to sleep is a crippling enough horror that people go right back to consuming. To paraphrase Fight Club: NO, what you NEED is some exercise, good food, and to relax.

The Z drugs are not benzos.

To claim that the Z drugs are "vastly more addictive than heroin" is, frankly, ridiculous.

Rebound insomnia is nasty; and the Z drugs should be treated with respect; and sleep hygiene should be tried first; but it's stupid to scare people with nonsense like "addictive as heroin".

... and chew valerian root :) Fight Club is full of useful information, by the way (the book, too).

> This is interesting. What's the evidence?

I'd be interested to see the evidence for "terrible for you".

Hypnotics just make you sleep - they don't treat the cause of the insomnia. (For some people they're great at kicking you back into a routine.) There's evidence that long term use can create rebound insomnia.

Here's a guide for general public stating the need for using the meds short term:-

(http://www.nice.org.uk/nicemedia/pdf/TA077publicinfoenglish....)

Here's a slightly more technical document with links to the evidence that they used:

(http://www.nice.org.uk/nicemedia/pdf/TA077fullguidance.pdf)

Here's a guide aimed at clinicians with many links to good evidence:

(http://www.cks.nhs.uk/insomnia/evidence)

Epidemiology specifically does not and can not nail down claims, but rather provides probabilities of possible causal relationships.

It is a collection of statistical methods based on interpretation of available data, usually from public health records.

Anyone who complains that it doesn't prove anything outright has no idea of what it is, how it works or what it tries to achieve.

Reply on news.ycombinator.com