Biological Measures &
PTSD Treatment Efficacy

by David Baldwin, PhD

This is an amalgam of two separate posts to the traumatic-stress list, responding to separate questions by Dennis Grant and by W. Jake Jacobs, regarding biological and endocrinological measures of treatment efficacy for PTSD.

It is always good to read [Dennis Grant's] posts, but sadly this does not make the questions easy to answer...

Mulling this over, I keep coming back to Seymour Antelman's "time-dependent sensitization" work (TDS). In a number of studies, mostly (but not only) in rats, he and his crew have demonstrated that a brief single stressor (sometimes psychological, sometimes a drug vaccine) causes a delayed but long-term change in sensitivity to this or similar stressors. Similar if not identical to kindling, I suppose. At first, I pessimistically thought of this as a problem for your question -- that the veterans most affected by combat traumas would be those who had experienced childhood traumas that had already heightened their biological responsivity to stress -- and so would be most difficult to "debrief" out of PTSD following combat. But now I've thought of another angle on this.

Importantly, these TDS differences are not apparent immediately (an hour or a day, say) but appear after some delay and may strengthen over time. Lenny Rosenblum, for example, found something like this in differential responsiveness to anxiety probes in bonnet monkeys at young adulthood that depended on how predictably their mothers had had to spend time away foraging for food when these subjects were infants (see 1994, Biological Psychiatry 35, 221-227).

My point here is that it seems possible that effective psychotherapy may deal with some aspects of PTSD symptoms -- nightmares, say -- causing a feeling of relief, and starting a positive chain of events (less perceived threat) that might later include some reduction in the biological or neuroendocrine sensitivity (etc.) that is characteristic of PTSD. Maybe, like the reverse of Antelman's TDS, this psychophysical change would not be apparent immediately, but instead might appear only after some delay.

Without having done a thorough search on this (yet), my guess is that those few studies that did use psychophysical measures to assess therapeutic efficacy would take the last of these measures at the final therapy session, not some weeks or months later. If so, its possible that we may not have data addressing this possibility because no one thought to look for a delayed biological response to treatment.

I decided to post this to the entire list because I wonder if anyone knows of any studies that looked for such a delayed biological or psychophysical return toward normalcy well after the end of self- reported effective therapy (among clients who weren't mugged or something in the interim!). As usual, comments are welcome.

Sometime later, Jake Jacobs wrote:

Putnam's data are indeed impressive. But I haven't an answer to the following question: Does EMDR bring about acute (during the procedure) changes in cortisol, ACTH, or other hormonal indicators of stress? - W. Jake Jacobs

I am unaware of any studies on any psychotherapy treatment for PTSD that looked for changes in any neuroendocrine measures (cortisol, ACTH, etc.). I've found eight treatment outcome studies (with PTSD) that used some sort of psychophysical measure (most often HR). Generally, these reports show the clearest evidence of treatment efficacy with self-report measures tied closely to disturbing trauma symptoms, followed by more general symptom scales (anxiety, depression, the SCL-90-R's GSI), with the least sensitive measures being any psychophysical ones. If anybody has some references on this, please post or email them. Really, I've _looked_ for this stuff.

So psychophysical measures appear to be the least sensitive to therapeutic change among PTSD patients -- with the possible exception of hormonal or neuroendocrine measures that weren't even tested (so far as I know). Certainly measurement problems are a likely scapegoat for this dearth.

Here are some other possible reasons, IMHO:

  1. Maybe the PTSD patients had different values on these measures before they were traumatized (perhaps even predicting PTSD given the trauma exposure), and the measures were relatively unaffected (at least not promptly) by treatment -- I've found almost nothing on pre-trauma (baseline) measures of such patients.
  2. Effective treatments (e.g., EMDR; whatever) seem most rapidly effective with the intrusive symptom cluster (flashbacks, nightmares); as these disturbing symptoms are minimized, and people report feeling much better, it may set the stage for slower and subsequent recovery in other clusters (avoidance, hyperarousal) as reflected in neuroendocrine or psychophysical measures. But --
  3. I have yet to find any PTSD treatment outcome study that looked for delayed effects in non-self-report measures (e.g., psychophysical) -- the follow-up measures are generally phone or brief in-person interviews with the former patients.

If anybody's got something more on this, I'm real curious to hear about it. This is an extremely important issue, methinks...