Reader Comments

Post a new comment on this article

Regression to the mean?

Posted by oldcolonial on 31 May 2012 at 12:48 GMT

Does the study look for this effect? It is unclear to me that the study indicating that AR is not related to prior health status will ensure that the observed effect is nothing more than the regression to the mean.
Just a thought.

No competing interests declared.

RE: Regression to the mean?

trobc replied to oldcolonial on 05 Jun 2012 at 01:08 GMT

I posted a similar comment before seeing yours. I agree that regression to the mean needs to be addressed to estimate the true magnitude of the negative effect, if any.

No competing interests declared.