Of course Matt is right and the typical mother is being sold a lie in terms of thinking their kid is a potential killer. But also, monitoring kids when using psychoactive drugs (especially when changing drugs or dose levels, and in making sure they are not changing dose or failing to take doses themselves) can go a long way.

Here's the Mother Jones database of the types of shootings we’re talking about: motherjones.com/politic…

So, the database tracks mental health involvement in only the cases we're talking about, public shootings. Notably, you can't know if there is NO mental health issue, you can only know if there is one reported or not one reported. The stats on that database break down in a few categories... "Yes", "No" (no known or reported mental health component", "TBD" and no data. The list shows 50%+ "yes". And to be honest I spot checked some of the "no"s and they are very possibly mental health cases. One of the No's was Luby's, where people reported behavior out of the guy beforehand that would qualify as indicative of mental health problems, and his autopsy results were "No alcohol or illegal drugs". Another "NO" was an 11-year old who pulled a fire alarm and shot kids coming out of his middle school. No prior mental illness specifically reported, but was it there?

Here's my hypothesis, stated: there's a very low rate of amok violence from the mentally ill and there is a higher, but still only one in many million, rate of mass homicide as a side effect from certain psychoactive drugs. Antipsychotics are probably worse in terms of the rate, and antidepressants have risk as well but probably a lower rate (but more widespread use). Some chance that it's the combination effect of the drugs and the fact that we reduced institutionalizations of people for mental health reasons by >75%. Meaning the drugs might not only cause it, it might also be that when we keep people at liberty rather than institutionalized under care because we had the drugs available to do that, we no longer kept them away from the public so we allow more attacks now.

Data consistent with the hypothesis:

-incidence of mass attacks like these in US begin shortly after the approval of the first drug, diazepam

-first shooter, famously from the U Texas bell tower, was known to be on diazepam (and others since have been as well, including Vegas shooter).

-Almost every shooter of this type where we have data was on psychoactive drugs (except terrorists). There are many we don't have data for, though.

-Known cases of shooters on psychoactive drugs: UT, Columbine, Vegas, VA Tech, Stoneman Douglas, Newtown, many others

-incidence of mass shootings increases after SSRIs are introduced and start to be used more widely

-strongly delayed incidence of mass attacks in Europe correlates with the slower penetration of the same drugs in patients there

(remember when Europe thought this only happened in the US?)

-all countries have mass attacks, and death rates trend more with gun laws than attack rates do. Rates of death in Europe and in Commonwealth countries from these types of things are now competitive with rates in the USA.

-In China and Asia in general, there are a ton of school knifings, usually tied to mental illness, it's just that not as many people die due to lack of guns en.wikipedia.org/wiki/L…

-strong correlation of increased violence towards others with all of the relevant drugs: journals.plos.org/ploso…

-SSRIs strongly associated in a cohort study with criminal violence in ages 15-24 journals.plos.org/plosm…

-homicidal ideation is a stated rare adverse event of some of the drugs, and anecdotally reported for many others... listed "rare adverse event" for venlafaxine

-five years before homicidal ideation was stated on the PI for venlafaxine, Andrea Yates drowned her 5 children while on it. She doesn't show up on the gun homicide lists. Lots of non-gun cases like this.

-obviously suicide is a big issue for these drugs as well, but importantly was denied for decades before being categorically proven (and it has been put on the warnings lists, black box warnings)

-extremely low incidence of the side effect would make it very hard to detect and prove, so our situation more likely is “we lack enough data” rather than “we can conclude no connection because no proof yet”

Data irrelevant to the hypothesis:

“NRA talks about mental illness as an issue, therefore we should doubt it.” They could do that to distract unnecessarily, but they would also do that if the hypothesis were true. In fact the politicization of the issue likely frustrates clarity of understanding of it because of knee jerk reactions. But after Mother Jones did their database, even they started to strongly talk about mental illness and the need to consider that.

I haven't seen any data that doesn't line up with the hypothesis yet. There's a strong case for reducing medical privacy rights especially for people who have committed felony attacks, or at least requiring anonymous submission to a central database.

What worries me most about ignoring mental illness as a potential cause is that bombs do a lot more damage then guns, so if we don't address the underlying causes alongside greater gun control, then the death rates could go UP. Mass bombings are up to 3.5x death toll and even more on the injury side I think. The higher potential is offset by the greater difficulty level of execution of course.

One thing I like about the hypothesis is that, if true, it means people aren't total monsters, just people experiencing a very rare side effect of an overall useful drug. I do think that even if this is true, the drugs can easily be doing more overall good than harm.