In the summer of 2011, a man named Anders Breivik boarded a ferry for Utoya, a small, mostly barren island a few hours 39 kilometers north-west of Oslo, where a bomb he’d planted in the morning had already blown a crater through downtown. He was wearing a fake police uniform and carried enough weapons and casings to take out a small militia. By the time Breivik finished his slow massacre of the trapped teens and young adults who had been attending a political camp on the island, he’d killed 69 (eight, additionally, died in the blast) and left dozens wounded, in an event typically referred to in Norway by its date—July 22.

In a country so small and intricately linked that journalists often remark that one in three Norwegians knew one of Breivik’s victims, the public struggled to make sense of a day so horrific it seemed filched from fairy tales.

But Trond Idaas zeroed in on a largely unexamined group when, three months after the attack, he began a survey of Norwegian journalists who covered the massacre. Idaas, a veteran of Norway’s main morning newspaper, Aftenposten, worked as a journalist for many years, including during the Balkan Wars in the mid 1990s. He noticed that colleagues who covered it returned different, shaken—and the lack of resources available to help them recover from the aftermath of war convinced him to switch careers. He completed a master’s thesis on stress reactions in journalists and now is a special adviser for the Norwegian Journalists Union. Of the 700 journalists who had been involved in the first 24 hours of Utoya coverage, more than 500 responded to Idaas’ survey, which found that the amount of stress journalists suffered from covering Utoya went up if they thought their reporting caused harm.

A few months later, Idaas was presenting at a trauma conference outside of Oslo when he met Klas Backholm, a Finnish journalist and a clinical researcher specializing in trauma. Backholm was at the conference summarizing work from his doctoral thesis, which studied the stress responses of journalists covering the Finnish school shootings in 2007 and 2008. His findings echoed what Idaas was seeing in his own work. The two are now collaborating on analysis and presentation of a subsequent Utoya survey, which they expect to finish later this month.

Classic symptoms of Post Traumatic Stress Disorder, like hyper-vigilance or persistent memories, traditionally correlate with a personal experience of violent, often life-threatening events. But both Idaas and Backholm had found that the intensity of the journalists’ “stress response,” a symptomatic measurement that at strong levels is a risk factor for full-blown PTSD, was most strongly correlated with whether they experienced an ethical or moral dilemma while reporting. In short, the journalists who were most at risk were the ones who felt their work had done more harm than good.

For example, some reporters called the cellphone numbers of students on Utoya to get information on what was happening. They would later learn that Breivik used the ringing to track victims hiding. Those who placed the phone calls spent several days waiting for the names of the dead to find out if their reporting had resulted, literally, in bloodshed. Perhaps needless to say, the phone calls correlated with high stress responses.

Earlier studies of reporters in war zones have suggested already that journalists are at higher risk for stress symptoms than other conflict-adjacent workers such as safety officers or medical technicians. Unlike other first responders, Idaas points out, journalists “are not there to help anyone,” a circumstance that “runs against ethical and moral norms as a human being.” It’s a similar concept to the growing study of “moral injury” in the military, where soldiers seem to experience stress symptoms when ordered to partake in actions that run contrary to their own personal moral code.

“You can’t compare journalists to soldiers, but still it’s something of the same feelings—going beyond your ethical or moral threshold into a gray zone,” says Backholm. “No other occupation group is asking these questions: If I do my work, do I cause more harm?”

In 2008, a group of social scientists released the results of a survey of hundreds of journalists who covered the the 2004 tsunami that killed 280,000 people in Southeast Asia. Many of the fatalities were Norwegians on holiday. Despite the staggering death toll, according to Idaas, journalists who covered the tsunami had less anxiety than journalists who covered Utoya. Nine months after the tsunami, 91 percent of responding journalists qualified their stress symptoms as “low,” with roughly 10 percent at “risk” for developing PTSD or more serious reactions. Nine months after Utoya, about three times as many journalists were still “at risk,” according to Idaas.

Alexis Sobel Fitts is a senior writer at CJR. Follow her on Twitter at @fittsofalexis.