This is actually a myth.

The data for suicide statistics have been polluted in in some rather peculiar ways. Hospitalizations for any self-harm-related injury, no matter how obviously sublethal, are categorized as suicide attempts.

https://www150.statcan.gc.ca/n1/en/pub/82-003-x/2001002/arti... (pdf)

>Hospitalizations related to suicide attempts were defined as the presence of ICD-9 codes E950 to E959 in the first accident code for a patient discharged alive. This ICD-9 category includes injuries resulting from attempted suicide, as well as self-inflicted injuries specified as intentional, but without suicidal intent.

So that will deceptively inflate the statistics for whichever gender self harms more frequently... which turns out to be girls and women.

https://www.ncbi.nlm.nih.gov/pubmed/25795294

>The goal of the current study was to use meta-analytic techniques to better conceptualize the presence and size of gender differences in the prevalence of NSSI (non-suicidal self injury) The results showed that across studies women were significantly more likely to report a history of NSSI than men.

This next study tries to investigate that issue by breaking down "suicide attempts" by degree of actual suicidal intent. They devised a spectrum ranging from deliberate self harm to suicidal gestures to serious attempts to end one's life. Surprise surprise, the self-harm end of that spectrum is more populated by females and the serious-attempt end is more populated by males.

>*A significant association between suicide intent and gender was found, where ‘Serious Suicide Attempts’ (SSA) were rated significantly more frequently in males than females.*

There is also an undercounting of men who attempt suicide by hospital staff. A recent Australian study showed that the number of men attempting suicide was much greater based on reports of first responders compared to hospital staff, but the latter being the current source of statistics.

https://www.beyondblue.org.au/media/news/news/2019/05/28/sui...

Thanks for commenting; interesting sources.

Yeah, I was thinking about this. probably the definition of "suicide attempt" is too broad then.

How do you come to that conclusion? The CDC data is tracking the number of hospitalizations. If the suicide attempt results in a hospitalization, that sounds pretty serious to me.

Hospitalization is generally mandated for any expression of an immediate intent to kill one's self, even if there isn't an attempt at all. So there is a line-drawing problem; if someone says they tried to kill themselves even if what they tried obviously wouldn't work (e.g. surface-level cuts away from blood vessels) their attempt could "result in hospitalization" even though hospitalization would have theoretically resulted even if no attempt had been made.

This still doesn't sound like an overly broad definition of suicide-attempt to me. Somebody saying they want to die and giving themselves minor cuts seems like a pretty serious situation. Unless the supposition is that boys don't get hospitalized when they do the same sort of thing, number of hospitalizations sounds like a robust metric.

It is serious. Just pointing out a way that someone could think there's room for narrowing the definition. Obviously any suicidal ideation may need help; categories that aren't totally clear can just make it harder to decide what sorts of resources and interventions to prioritize on a systems level.

Serious is not the same as fatal.

People who aren’t being listened to sometimes take an overdose as a cry for help, expecting to be found. They still go to hospital.

That is quite different from shooting yourself in the head.

We're talking about suicide attempts right? So yes, not fatal, obviously.

No - the context from upthread is this:

> Suicide attempts have always been way more frequent among women/girls, whereas men/boys actually commit suicide more often.

Sure... but I was responding to the assertion that the discrepancy in suicide attempts comes from suicide attempts being over-counted. I don't think that holds water.

If you're expecting to be found, it's arguable how much that counts as a suicide attempt.

It shouldn’t, but it does because hospitals only see an overdose and don’t have access to the patients prior beliefs.

how else to explain the fact that number of attempts is higher for women yet number of suicides is lower? it means number of attempts can't predict number of suicides, so it is counting something unrelated.

It's counting suicide attempts, not suicides. Obviously one rate does not predict the other, but does that mean suicide attempts aren't worth tracking? I don't think so.

And I don't know how to explain the discrepancy, but that's what the data says. Being unable to explain the data doesn't mean the data is wrong.