[REUTERS] -- In the search for an explanation of why a U.S. soldier left his base in Afghanistan at night and killed 16 civilians in their homes, some experts have raised the possibility that mental illness or a brain injury played a role in the massacre.

We're going to look into all of that, General John Allen, who commands U.S. and NATO troops in Afghanistan, told CNN on Monday, declining to comment further on the mental state of the soldier suspected in Sunday's attack. A U.S. official told Reuters that the staff sergeant had suffered a traumatic brain injury in a vehicle rollover in 2010 in Iraq, and was treated and returned to duty.

Experts caution against jumping to conclusions, but two facts are known. This was the sergeant's fourth deployment. And the risk of mental illnesses such as post-traumatic stress disorder (PTSD), depression, and anxiety-related disorders is generally higher during subsequent deployments than during a soldier's first.

The more exposure there is to trauma the worse it's going to be, said Dr. David Reiss, a psychiatrist in private practice who has treated patients with PTSD. Especially if someone is deployed repetitively, then the whole issue of expecting to go home, not going home, just amplifies it.

As with most aspects of mental illness, however, the link between deployments and mental illness is not straightforward. For instance, the incidence of PTSD during a third deployment is 2.7 percent among male members of the armed forces, according to a study released last September by the Armed Forces Health Surveillance Center in Silver Spring, Maryland.

But the incidence of PTSD for GIs serving a fourth deployment is lower: 1.9 percent, found the study, which analyzed medical records of 1.3 million U.S. military members who had served in Afghanistan or Iraq.

Even among GIs aged 30 or older - the sergeant is reportedly 38 - that pattern persists. The rate of PTSD among these older soldiers is 2.4 percent during a third deployment but 1.8 percent during a fourth. By comparison, the rate is 1.2 percent during the first deployment.

The rate of other mental illnesses also peaks during the third deployment. About 1.9 percent of male GIs over 30 are diagnosed with an anxiety disorder during their first deployment, compared to 2.5 percent during their third and 2 percent during their fourth.

'DWELL TIME'

Another factor influencing the chance of developing a mental illness while serving in the military is the length of time between deployments, or dwell time. The longer the dwell time, found the 2011 study, the larger the percentages diagnosed with the conditions after the deployments.

During the period of the study - members of the armed forces who returned from Iraq or Afghanistan between October 1, 2001, and December 31, 2010 - 42 percent of men deployed twice, 13 percent deployed three times, almost 4 percent deployed four times, and just over 1 percent deployed five times.

One reason for the decrease in the rate of mental illness after the third tour of duty is that service members who develop a mental illness during a deployment are less likely than their counterparts to deploy again, found the 2011 study, as a result, repeat deployers may be more psychologically 'resilient' than their never or less frequently deployed counterparts.

Why time between deployments should raise the risk of mental illness is even less clear. But the 2011 study suggested that with long dwell times between repeat deployments, a soldier might be more successful at transitioning back to non-combat life - and that, ironically, might make it more difficult to readjust yet again to 'warrior' status. As a result, short dwell times may be less psychologically traumatic.

The Army has not released information on how long the sergeant, assigned to a unit at Joint Base Lewis-McChord near Tacoma, Washington, waited between his return from Iraq, his previous deployment, and his deployment to Afghanistan last December.

The link between PTSD or traumatic brain injury and aggression or violence is only beginning to be unraveled. Overall, mental illness does not increase the risk of committing a violent act, particularly a criminal one. But when someone with a mental illness is violent, said Reiss, it's going to be less focused, more irrational and it tends to be more dangerous in that way.

In particular, PTSD is marked by sleep problems, irritability, difficulty concentrating, jumpiness, and feeling constantly 'on guard,' noted Eric Elbogen, an associate professor of psychiatry at the University of North Carolina School of Medicine, and colleagues in a 2010 paper. Those symptoms, in turn, are associated with anger and hostility in soldiers returning from Iraq and Afghanistan, they concluded from interviews with 676 veterans. Violence was also linked to traumatic brain injury, they reported in The American Journal of Psychiatry.

PTSD AND VIOLENCE

The PTSD-violence connection is believed to arise from feeling threatened even in safe situations, research shows. As a result, Some may act on impulse or go to extremes to protect themselves, says the Department of Veterans Affairs National Center for PTSD. Aggressive behavior is more common in those with PTSD than those without PTSD.

The flashbacks that characterize PTSD can make the person feel he is re-living the trauma, triggering an aggressive act meant to ward off the threat. People with PTSD are also more likely to perceive, or imagine, threats. Research has found that their amygdala, the brain structure that perceives threat, is overactive, while the hippocampus, which puts information into context (loud noise = car backfiring, not sniper fire, for instance), is often shrunken in people with PTSD, says neuroscientist Richard Davidson of the University of Wisconsin, Madison.

PTSD can also be marked by emotional numbing, studies show. As a result, someone with the illness may lack empathy and be unable to grasp the gravity of an act of violence.

The link between traumatic brain injury and violence has received even less study, with the result that estimates of the connection vary wildly: published analyses have found the frequency of aggressive behaviors after TBI to be as low as 11 percent and as high as 96 percent. A 2009 study found that people who had suffered TBI were 66 percent more likely to commit a violent act than were people who had not been injured.

Although it is not clear why TBI should raise the risk of violence, one possibility is that damage to the prefrontal cortex, located just behind the forehead, impairs judgment and impulse control, said Ricardo Jorge, assistant professor of psychiatry at the University of Iowa, who has studied veterans with traumatic brain injury. In particular, damage to the prefrontal cortex may silent the messages that region sends the violence-prone amygdala.

But these are only generalities, which are extremely difficult to apply to any given soldier. We have never done very well predicting aggression or violence at the individual level, said Dr. Thomas Insel, director of the National Institute of Mental Health.

(Additional reporting by David Alexander; Editing by Eric Beech)