A study of 10,000 veterans asked whether years of standing next to blasts leaves a mark that later reads as anger.
When a veteran comes home angry, the people around him tend to file it under personality. Short fuse. Always been that way. This study asked a different question. What if some of that anger is dose-related, built up over a career spent a few feet from things designed to explode?
The Setup
The blasts that don't knock you down.
Most blast research has focused on the big events, the IED that ends a deployment. This study looked at the small ones. The thousands of low-level blasts a service member absorbs over a career from firing heavy weapons, breaching doors, and standing near artillery. Researchers call it military occupational blast and impulse exposure. Investigative reporters have raised concerns it may be tied to later harm, to self and others.
Who lives closest to the blasts
Researchers used job assignment as the marker of exposure. These are the roles that eat the most blast over a career.
They matched 5,000 veterans from high-blast jobs against 5,000 with the same age, sex, and race who served elsewhere. Then they turned a language model loose on every clinical note in both groups, hunting for documented anger, aggression, or violence, and checked its work by hand.
What Happened
The gap was there in the raw numbers.
Veterans from blast-heavy jobs were more likely to have anger documented in their records. Not a little more. Almost half again as likely, before any adjustment.
17.2%
of blast-job veterans had documented anger, aggression, or violence
12.0%
of matched veterans from other jobs
The Number That Tells the Story
The gap shrinks. It doesn't close.
Here's the honest part. Blast-heavy jobs come bundled with other things that stoke anger: combat, brain injury, PTSD. So researchers kept stripping those out, one layer at a time, to see if the blast signal was really blast or just everything that rides along with it. The effect got smaller with each layer. It never reached even.
Odds of documented anger in blast-job veterans, as more is accounted for
Bars show odds above even. The 1.0 line means no difference. Every bar stays right of it.
Blast jobs alone
1.53x
+ age, sex, race
1.52x
+ combat & service
1.26x
+ brain injury & illness
1.22x
+ PTSD
1.16x
Even after accounting for combat, brain injury, and PTSD, blast-heavy work still carried about 16% higher odds of documented anger. The signal is real but weak, and it survives everything thrown at it.
The Tangle
These don't compete. They travel together.
The reason the blast effect is hard to isolate is that the jobs nearest the blasts are also the jobs nearest combat, brain injury, and trauma. They arrive as a package. PTSD matters most of all here: anger is a core symptom of it, and 70% of the anger-flagged veterans carried a PTSD diagnosis, against 35% of the rest.
1.47x
Traumatic brain injury
The strongest single risk factor for documented anger in the model.
1.32x
Combat exposure
Serving in a combat theater raised the odds on its own.
0.33x
Female sex
The single largest protective factor. Roughly a third the odds.
0.67x
College degree
More education tracked with lower documented anger.
Read It Straight
What this study doesn't say.
The honest edges, because a finding like this gets sensationalized fast.
They used occupation as a stand-in for exposure, not a measured dose per person. An infantryman and an artillery instructor get sorted the same way even if their actual blast histories differ wildly. A real per-person measure would be sharper. Occupation is a blunt instrument.
The study's own clinicians note that anger gets underreported by veterans and under-documented by providers. So the true gap is more likely bigger than what the records show, not smaller. These are conservative estimates.
The records show an association between blast-heavy work and documented anger. They cannot prove the blasts caused it. The authors are careful about this, and so should anyone repeating the finding be.
The model agreed with human reviewers 96% of the time and three separate methods were cross-checked against each other. That's strong for reading subtle language at this scale, but it is still a machine inferring a hard-to-pin-down thing from clinical text.
The Takeaway
Here is why it matters that the signal held. A firefight can't be un-fought. A brain injury can't be undone. But blast exposure is something the military can dial down, and already is, with stand-off distances, overpressure limits on the heavy guns, and tracking who absorbs how much over a career. If anger is partly a dose a man accumulates, then it is partly a dose we can cap.
A blast injury and a bad temper can look identical from the outside. They are not the same thing, and they do not ask the same thing of us.