Researchers are just now beginning to discover how different biological malfunctions can give rise to symptoms of post-traumatic stress disorder (PTSD)—insight that might one day lead to more targeted treatments. In the meantime they are also exploring the use of biomarkers—hallmark variations in hormones, genes, enzymes and brain function—to apply existing therapies more effectively.

“Trauma exposure can result in enduring biological changes that depend on an individual’s life history, age, gender and a host of other factors,” says Rachel Yehuda, a neuroscientist at Mount Sinai Hospital in New York City. “We must capitalize on this heterogeneity in the service of individualizing treatment approaches rather than insisting that one size fits all.”

Indeed, not all patients get well by way of the most popular forms of therapy. One widely recommended treatment, cognitive behavioral therapy (CBT), typically helps only half of the patients who try it. In 2008 Richard Bryant, a professor of psychology at the University of New South Wales in Australia, and his colleagues attempted to identify that half up front. Before CBT they took brain scans using functional MRI of 14 subjects while showing them photographs of frightening faces.

Seven people—the same who later failed to improve—showed greater than normal activity in brain regions associated with experiencing fear: the amygdala and the ventral anterior cingulate cortex. In another study Bryant found that the people who did benefit from CBT began treatment with larger rostral anterior cingulate cortices. Both animal and human studies have linked this brain area to “extinction,” the psychological process by which we unlearn conditioned responses, including fear.

Extinction is the goal of another common intervention, called prolonged exposure therapy, in which patients actively revisit traumatic memories. It is highly effective for some people but as with CBT, about 50 percent do not respond at all. Yehuda and her colleagues have found that a single enzyme can help distinguish them. In 2009 they evaluated 28 survivors of the 2001 World Trade Center attacks, all of whom were seeking therapy for PTSD. The scientists recorded levels of cortisol and its metabolites, including 5-alpha reductase, before and after prolonged exposure therapy and at three-month follow-up appointments.

They found that those who did not get better had shown significantly lower levels of 5-alpha reductase before treatment. Moreover, lower levels of this enzyme correlated with higher avoidance behaviors or how much the subjects said they steered clear of anything reminding them of 9/11. The researchers speculate that having less 5-alpha reductase—which may be a lifelong trait—can disrupt cortisol levels and with it the body’s normal capacity for handling stress.

More recently, Yehuda and her colleagues conducted a similar study, collecting blood, urine and saliva samples from 37 combat veterans before treatment, after treatment and three months later. Therapy entailed 12 weeks of either prolonged exposure therapy or 30-minute weekly phone calls to discuss symptoms. Overall they spotted several patterns among those who got better: the responders were, on average, 42 years old (a decade younger than the average age of nonresponders) and had typically experienced their first trauma as young adults; the nonresponders were more often in their teens or younger and did not have the social support that the responders did.

In addition, a range of biomarkers flagged those who improved, no matter what treatment they received. Responders were more likely to carry specific genes that rendered them less sensitive to glucocorticoids after treatment. These hormones, including cortisol, serve to shut down runaway inflammation, and greater glucocorticoid sensitivity has been previously linked to an increased risk for PTSD. Vets who recovered also had lower DHEA (dehydroepiandrosterone)/cortisol ratios—indicative of an imbalance between these two opposing hormones that keep a range of immune and stress responses in check. They also had higher blood plasma levels of neuropeptide Y, a neurotransmitter implicated in reducing stress and anxiety. Of particular interest, neuropeptide Y levels tracked with changes in symptoms during treatment, reflecting the patients’ progress in real time.

“I don’t believe that PTSD is an illness that cannot be treated or substantially improved in everyone,” Yehuda adds, “but I do believe that what might really work for one patient will categorically not work for someone else. Biological measures hold the key to helping us understand this.”