When Chris Kyle returned home from Iraq after serving as a Navy Seal, he felt disconnected, unable to talk about his experiences. So he wrote a bestselling book, American Sniper, detailing his four tours of duty, during which he killed more than 150 insurgents. He also mentored other veterans struggling to readjust on "wounded warrior" hunting trips.
Tormented by memories of combat and unable to engage with Americans who have never been to war, veterans often feel guilt or shame at leaving comrades. "There's definitely a lot of hurt from losing my guys or the fact that I got out," Kyle told Texas News Service. "So being able to do this makes me feel like I'm still a part of it and still giving back."
Last weekend Kyle, 38, and friend Chad Littlefield, 35, also a veteran, were shot and killed at a Forth Worth, Texas, gun range. The suspect is 25-year-old Eddie Ray Routh, a former US Marine who served in Iraq.
Kyle had reportedly invited Routh, who was admitted to a mental hospital last year after threatening to kill his family and himself, to go shooting.
It is a grim irony that the killings, the latest in a series of violent crimes involving veterans with mental disorders, may have occurred because the victims feared Routh might kill himself. Routh is in jail charged with two murders. He is on suicide watch.
Last year, as the Iraq and Afghanistan wars began to wind down, releasing tens of thousands of soldiers, the Defence Secretary Leon Panetta voiced concern there was an "epidemic" of suicides. This is contested, but in 2012 more active-duty US soldiers killed themselves than were killed in combat: 177 against 176. The trend was reflected across the wider US military: 349 suicides, 295 combat deaths.
Military suicides have, in fact, been steadily escalating but were masked by high combat deaths, now falling. In 2010 and 2011, 499 and 417 US troops died in action, while military suicides averaged 300 a year.
A 2011 report, Losing the Battle, from the Centre for a New American Security, depicts a worsening crisis, with one suicide every 36 hours between 2005 and 2010. "The numbers continue to grow," says Phillip Carter, director of the CNAS Military, Veterans and Society Programme.
Carter says the task of reducing suicides is compounded by a hazy understanding of what pushes some soldiers, and not others, over the edge. "We know that there is a correlation between combat stress and suicide. But we don't know exactly what factors cause some people to commit suicide. There doesn't seem to be any obvious answer."
Numerous causes are cited, including post-traumatic stress disorder [PTSD], traumatic brain injury [TBI], depression, broken relationships, insomnia, legal issues, substance abuse or reluctance to seek help in a military culture that traditionally sees this as unmanly. One fear is suicides have yet to peak, especially among Iraq and Afghanistan veterans, who are 30 to 200 per cent more likely to kill themselves.
Alarmingly, while most of those who killed themselves in 2011 never served in either theatre, last year a majority had, raising concerns that multiple deployments, which can heighten stress by separating soldiers from unit camaraderie, are a significant factor.
And then there is the challenge of reintegrating into civilian life, given that fewer than 1 per cent of Americans have served in Iraq or Afghanistan. Many veterans are left with a deep sense of isolation, unable to expunge anguish about horrible things they have seen or done.
"There are myriad pathways to suicide. There's never one common thread that links them all together," says David Little, executive secretary with the Action Alliance at the National Suicide Prevention Centre.
A big problem is lack of empirical data. Last month the Veterans Affairs Department - criticised for doing too little, too late - released The Suicide Data Report 2012. On average 22 veterans (as against those still in uniform) kill themselves each day.
Robert Bossarte, who compiled the VA report, told the Washington Post the rise reflects a wider national trend, where suicides rose almost 11 per cent between 2007 and 2010.
Two-thirds of veterans who kill themselves are over 50.
Even so, while suicide numbers among veterans are up, the percentage of veterans who take their own lives remains below the national rate. The VA compiles statistics by extrapolating from the 30 to 40 per cent of veterans who seek care (6500 killed themselves in 2012, up from 3871 in 2008-2009). The 2012 report used death certificates from 21 states. Two of the most populous, California and Texas, are among those yet to supply data. And in suicides of elderly veterans police rarely find out if decades-old combat stress was a factor.
"The VA report is important," says Little. "It's the first time we've had good data to compare suicide rates among veterans to the population in general." But, he says, it raises as many questions as it answers.
Incomplete data means the VA is sometimes unaware if suicides saw combat, what work they did while in uniform, when they left the service, if they called crisis lines, and if they received treatment.
The recent Alabama standoff, during which police killed a Vietnam veteran who had taken a 6-year-old boy hostage, highlights a national crisis. A key question is how to get veterans to seek aid before overcome by despair.
Suicide can be impulsive and many are likely to use guns. In 2010 68 per cent of military suicides shot themselves, against 4 per cent who died of drug overdoses. "There's nothing worse than the abuse of alcohol and the ability to get your hands on a weapon," General Peter Chiarelli told Stars and Stripes.
A 2010 Israeli Defence Force study found suicides dropped 40 per cent since 2006, after troops were forbidden to take weapons on leave. Despite National Rifle Association opposition, last month Congress empowered commanders to inquire if at-risk troops owned private guns.
Other initiatives include phone help lines, a public awareness campaign and more mental health staff. Routh told police he had struggled to cope. How many share his pain?