New research from the University of Washington shows that the combination of mother-daughter conflict and low serotonin levels accounts for 64 percent of the difference among adolescent girls who harm themselves and those who do not.
“Most people think in terms of biology or environment rather than biology and environment working together,” said [Theodore Beauchaine, a UW associate professor of psychology.] “Having a low level of serotonin is a biological vulnerability for self-harming behavior and that vulnerability increases remarkably when it is paired with maternal conflict.”
It’s been a while since I’ve taken undergrad psych classes, but I believe that terminology doesn’t mean that if a girl cuts herself ten times, six of those cuts are due to her mean mom and low serotonin levels. I think it means that of one hundred girls who engage in self-destructive behavior, sixty-four of them can chalk it up to their maternal relationship and low levels of serotonin.
Want to be a better mother? Check out some blogs and sites on the topic here, here and here. Want to lower your serotonin levels? Sorry, but that’s a whole ‘nother post.
2 responses so far ↓
1 Grace // Mar 10, 2008 at 4:14 pm
Always go to the source. You can download the full text of the research article at Beauchaine’s research webpage.
http://tbeauchaine.psych.washington.edu/tb/#Publications
The variance analysis is explained.
They do not rely on self-diagnosed blame.
2 Andrea Bartz // Mar 10, 2008 at 7:36 pm
Thank you for your comment, Grace! I must have been unclear in my blog post - I didn’t assume the researchers relied on self-diagnosed blame. I was simply commenting that “xyz accounts for abc percent of the difference” is tricky language, and I had to remind myself of what it means. And I believe I was correct, but maybe my language was unclear. The “64 percent” means that of 100 girls exhibiting this behavior, about 64 of them are influenced by the combination of a poor parental relationship and low serotonin levels. I was making the point that it’s NOT accurate to say that if a girl cuts herself, those two factors are 64 percent to blame, and some other factor is 36 percent behind it.
I read through the study and they explained that they used three forms of analysis: multivariate analysis of variance to test for group differences in family interaction patterns between self-injuring and typical adolescents, correlations between 5-HT and variables picked out from that first analysis, and interactions between peripheral 5-HT and family interaction
quality using multiple linear regression techniques. Beyond that, I’m not sure the paper re-explains percentages of variability, since it’s assumed most colleagues would understand. (Unfortunately, I’m no colleague, and am drawing no knowledge from last winter’s Research Methods in Psychology course.)
I hope you don’t mind the link to your blog! I enjoyed it and thought others might too.
Leave a Comment