Does torture yield reliable information?

Despite the claims of the Bush administration and the recent report by the Department of Justice, there is no convincing evidence to suggest that torture is any more effective than other interrogation techniques.  Additionally, there are risks with torture (which involves long periods of stress and sleep deprivation) of obtaining confessions that contain false information or where the suspect actually begins to believe or repeat what he or she is being accused of (i.e. confirming whatever the interrogators want).  Part of the reason for this is obvious, in that, under enough duress, a person might just say anything to make the torture stop.  But there is also a wealth of evidence from psychology and neuroscience labs around the world that demonstrate adverse effects of stress and sleep deprivation on processes of memory and decision making.  And a recent review article in Trends in Cognitive Sciences (here, or 1 of these 2 for popular press stories about the article) goes over much of this evidence to make the case that torture, as a means to obtain truthful, actionable intelligence, is unreliable.
This is of particular interest for me (though I am upset at myself for not making the connection) as my area of study pertains to the effects of steroid hormones on neural plasticity (where plasticity is a catchall term for the processes that underlie learning and memory).  We have known for a while now that there are only 2 areas in the brain that seem to be any good at adding new neurons in adulthood, and one of these is the hippocampus, which, as I've mentioned before, is critical for memory formation, retention, and recovery.  As it turns out, the ability of this part of the brain to generate new cells, and incorporate them into circuits is critical to processes of remembering.  Similarly, the ability of these new cells to survive appears to be another critical factor in memory retention and possibly recall.  Of the many types of steroid hormones out there, the ones we typically associate with reproduction (e.g. testosterone, progesterone, and estradiol) seem to have the most positive effects on cell proliferation and survival, and can improve learning and memory.  On the other hand, the steroids that are more commonly associated with, and produced in response to, stress (called corticosteroids because they are produced by the cortex of the adrenal gland), have been shown to be detrimental to the processes of cell proliferation and cell survival in the hippocampus, and thus also detrimental to our ability to remember.  This is of importance to the debate over whether or not torture is effective because, while there is merely a lack of evidence to support claims that torture is more effective than other methods of interrogation (and the absence of evidence is not evidence of absence), there is a wealth of evidence that shows how torture can actually impair the memory of the individual being tortured, and may thus result in distorted or false information. 

I will insert a link to the article when it becomes available on the cell press website, but for now, if you have library access, here's the citation:
“Torturing the Brain: On the folk psychology and folk neurobiology motivating ‘enhanced and coercive interrogation techniques.” By Shane O’Mara. Trends in Cognitive Sciences, Vol. 13 Issue 10, September 21, 2009.