Skip to content

Sniffing out suicide


Christopher Banks at Bipolar Bear has drawn my attention to a Science Daily article claiming a new sweat sniffing way to spot a suicide that is 97% accurate.
Lars-Håkan Thorell says he has a sort of lie detector or e-meter that can tell if you are ‘hyporeactive’ and therefore likely to kill yourself.

He tested 783 inpatients at German mental hospitals and found :

“Hyporeactivity was most prevalent in the bipolar patients: of 126 patients, 80.2% were affected, compared to 67.3% of the depressed patients and 58.5% of those with other diagnoses. The study also shows that people with recurrent depression run a risk of becoming hyporeactive at some later point in life.”

Oh my God!

There must have been a bloodbath right? Such high rates of correctly predicted suicides among the hundreds of inpatients enrolled in the study.

Well not quite.

German psychiatric hospitals aren’t like that.
Not any more at least.


As medwire News reveals, only 4.6% of Thorell’s subjects committed suicide over the period of the study.

That’s thirty six tragedies. But not upwards of half the 783 guinea pigs dead.

How can they say they’re 97% accurate?

Because they predicted 97% of the suicides.
That’s all except one.

But to do so they had to predict over half the patients would kill themselves.
It’s like backing most of the horses in a race.

It’s pretty hard to predict something as rare as suicide. It’s even rare among psychiatric inpatients. So a fairly accurate testing system will still throw up a lot of false positives.

Thorrell’s team got false positives at over ten times the rate they got true positives.
HIV testing in NSW is about 99.95% ‘accurate’. That makes it 60 times less likely to generate false positives than Thorell is.

But what if they do find an accurate biological test for suicide?
What would that mean?

Well for starters it would revolutionise the field of mental health diagnostics.

‘Mental illnesses’ like bipolar, depression and ‘other diagnoses’ are all currently defined according to their symptoms, most notably by the “psychiatrist’s bible” the Diagnostic and Statistical Manual (DSM).  They’re descriptions of how people behave and what they say.

Psychiatry is not a dogmatic religion and it accommodates many fads and cults that allow varying interpretation of its holy book though. That’s why you’re likely to get as many different mental health diagnoses as you meet diagnosticians.

The holy grail of mental health diagnosis is a diagnostic biomarker. You hear some group claiming to have found it every few months. It can be an armpit sniff or a blood test or an EEG trace or a genetic marker or one of those psychedelic fMRI readouts …

Funny thing is they never pan out.

They haven’t really found any reliable biological diagnostics for ‘mental illness’ since the early 20th Century when they discovered epilepsy was a neurological disease.

That’s because ‘mental illnesses’ aren’t illnesses any more than coughing is an illness. They’re symptoms. They are probably caused by all sorts of stuff.

Some may be signs of physiological illnesses, some might be caused by stress, some might be developmental, some may be due to toxins, some might be because life sucks, some might be genes, some may be environment …

Put it all together and it’s called ‘you’.

But a reliable test for suicide would change everything.

Finally! An objectively confirmable symptom with an objective test for it!

A real mental illness!

‘Suicidal Disorder’ would become the disease with ‘bipolar’, ‘depression’ and ‘other diagnoses’ the prodromal symptoms. The mental health system would become heavily aligned toward suicide prevention. Productivity would be measured according to ‘accurately predicted’ suicides kept alive.

At last! Quantifiable mental health performance indicators!

Of course it could all end tragically. Perhaps both suicide and hyporeactivity are caused by something else. What if the sniff test for suicide is detecting residue from one of the drugs they give psychiatric patients and that’s what’s causing the suicides? It’s not unprecedented you know. Then some drug company might get it’s arse sued off due to the ‘reliable’ evidence linking their product to suicide.

But how about a really radical approach to suicide prediction?
How about we listen to people?

Yeah, some people might give false positives – say that they will kill themselves then not do it.
But maybe those people need some listening to anyway.

Sure some people will deny they’re suicidal right up to the day they kill themselves.
But maybe people so desperate to keep their plan secret should be allowed to carry it out.

It must be hard living in a world where there’s no one to turn to.

From → DSM, hurts

  1. Rexie permalink

    “How about we listen to people?”

    Totally agree. There is something about this attentive communication that a very effective channel is built. (I think attention facilitates a great exchange of energies.)This measure will temporary benefits but is still a very good approach.


  2. Easy answers are usually bunk. Anything where there’s a simple biological test to predict someone’s behavior smacks of faux sciences like phrenology that get completely disproven over time. The flip side of the sweat test is an expensive patented antidote that changes one’s sweat in order to prevent suicide.

    Sometimes there might be a simple solution, like penicillin to kill bacteria, but people are much more complex than bacteria, because we are conscious and participants in civilization, and thus there’s no easy fix for complex social, psychological, and even spiritual problems.


Over to you

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: