It's timely that I post this now, given the recent news release supposedly saying that some people may have a genetic vulnerability to suffering from PTSD. Whoever posted this made no mention before posting of the steps necessary to be taken before accepting the post at face value -- That'll be the focus of a further blog post I'll be submitting in a couple of days. Today, I just wish to post some more news, about a LACK of a once proposed genetic vulnerability (as reflected in a single gene) to suffering depression as a result of adverse childhood experiences, by Culverhouse et al, here:
Collaborative meta-analysis -- depression and childhood stress
Remember that SSRIs (selective serotonin re-uptake inhibitors) have been, since 1987, the first choice medication for treating depression, such depression supposedly linked to a "deficit" in serotonin in some people.
Even some people with considerable experience reporting on scientific developments have said, on trying to read the article, that they found it "all Greek". So, I emailed the first author, Prof Rob Culverhouse, for a summary which might be able to be passed on a general audience. He was very generous to do so. Personally, I really liked the article, and especially the authors' ability to identify, and deal with, all the methodological problems in previous analysis.
Here's Prof Culverhouse' summary:
"As you know, selective serotonin reuptake inhibitors are a standard treatment for depression. In 2003, Caspi et al. published an article advancing a plausible hypothesis: that a genetic variant related to serotonin transport (5-HTTLPR) was associated with increased risk of depression in individuals exposed to stress. Specifically, the variant was not associated with risk of depression in individuals who were not exposed to stress. However, if exposed to trauma or stressful life events, carriers were more likely to become depressed than non-carriers (perhaps because their serotonin system could not respond adequately). This hypothesis has remained controversial because no one could seem to achieve a clean replication of the finding, although there were claims of “similar” or “supportive” findings and a great deal of research resources had been used to investigate this topic.
Our goal was to see if, with both a large sample and analyses that compared “apples to apples”, we would either find a setting where the gene-by-environment interaction could be cleanly replicated or we would be able to rule out a major effect on depression from this interaction. We gathered as many groups as possible who had data related to this topic and did multiple meta-analyses wherein we defined a common set of phenotypes and a common set of statistical analyses agreed upon by all the the participating groups. Key strengths of our approach were the large sample, the active participation of the data providers, and minimization of heterogeneity in the analyses. We investigated multiple hypothesized factors that, it the signal were real, might have thwarted previous efforts to replicate it. Our consortium included some who believed the interaction was real and some skeptics. It also included groups that had never published on this topic before as well as those who had.
In the end, our results rule out the possibility that the interaction would have a large or robust impact on depression (we saw no evidence for it in any of the primary or secondary models examined). This contrasts sharply with the statistical evidence for sex & stress, which only got stronger and more consistent as more data were added to the analyses.
We view the results of this study as an instance of science working as it is supposed to: experts had their opinions, but in the end it is evidence from the data that settles scientific disagreements. Our conclusion is that if there is an interaction similar to what was hypothesized, it must be of modest effect size and only seen in very specific situations that were not included in our many analyses. As a consequence, it is unlikely that an interaction between 5-HTTLPR and stress would have much clinical relevance for depression. We hope a consequence of our work will be that the field will direct resources to topics more likely to have clinical relevance for depression.
Best wishes,
Rob Culverhouse, Ph.D.
Assistant Professor of Medicine
Division of General Medical Sciences, Box 8005
Washington University in St. Louis School of Medicine"
The summary is a bit simplistic, in that it specifically looked at, and this is important from an ACEs view, only the effects of childhood stress. Other gene variants, other studies have shown, might link experience of general life stress with depression (gene variant BDNF Val66Met), and Late Life Depression (the collective action of a group of genes). It is important to note that Prof Culverhouse' collaborative collection and data analysis (from over 60,000 participants) only looked at the 5-HTTLPR gene -- this study looked at other gene variants, childhood stress, and cognitive processing anomalies Genes, Cognitive Biases, and Depression
One of my hopes is that this research will not only promote attention more to the now known contributors to depression following ACEs, but also lead to survivors experiencing less self-stigma as a result of not feeling as if there's an inherent "weakness" in them.
I would dearly love to see people provided with some more information and guidance in how to appraise the scientific publications they feel are important to them, and this will be the focus of a future post "Gene studies -- Lies, Damn Lies and *^%## statistics" -- should you have any questions about the methodological issues, such as "What's a meta-analysis, and how do I know it's been done well?"
Prof Culverhouse was very generous in promptly replying to my request for his article, and supplementary data -- so, strictly speaking, he should be your first option in getting his materials, but if you have any need, or further questions, you're welcome to contact me.
Comments (0)