American death rate from substance abuse and mental disorders triples since 1980 | Hacker News Reader

American death rate from substance abuse and mental disorders triples since 1980

sciencebulletin.org

130 pointsupen77 comments

There's a really interesting result buried in this.

"about half of US counties saw increases in suicide and violence"

The separation of perceived and actual violent crime rates is a pretty infamous thing. Violent crime has been declining for decades, but at least since ~2000 perceived rates have failed to track that result. People toss around lots of theories, from "the media scares us" to "bigots are scared" to "perception can only go so low".

This implies that we might be missing a better answer: people give national answers using local assumptions. If your county and all of the ones nearby have steadily rising rates of assault and homicide, it looks an awful lot like violence is on the rise (and for you, it is). This is especially true since you'll hear about every murder in your town, but only a few from NYC regardless of how many happen there (or don't happen).

So violent crime rates have dropped steadily, largely because they're power-law effects and the deadliest regions have gotten much safer. But there are a lot of people looking around at rising violence where they and their families live, and that's getting buried by the data from a handful of cities.

In many counties the rate of violence is still low enough that it will be difficult to measure accurately, never mind trying to reason about it anecdotally.

(a typical year around here includes 0 murders and about 20 aggravated assaults, less than 1/1000 individuals)

Media scare tactics around violence focus on urban areas, but the counties with high and increasing violence are the rural ones, especially the severely depopulated and very isolated counties. Of course the flip side is, nobody lives there. The ratios are high but the absolute numbers are not.

I like to use this site to explore this data, although it cannot produce deep links: https://wisqars.cdc.gov:8443/cdcMapFramework/mapModuleInterf...

You're correct, but people in rural areas often dismiss the data as lies (egged on by some sectors of the media) rather than asking for better data. So it's an article of faith among rural and semi-rural conservatives that all government data are lies and that the inner cities are some sort of urban-warfare crime zone which is being covered up in the mainstream media. It suits certain political actors to promote this version of events.

>> the deadliest regions have gotten much safer.

Not so sure I agree with that.

- St. Louis, MO

- Detroit, MI

- New Orleans, LA

- Baltimore, MD

- Chicago, IL

All have had incredibly high murder rates since 2002. That's over a decade of violent crime that hasn't slowed down at all and continues to remain steady to this day.

I'm from St. Louis, so I happen to be familiar with those numbers.

Here are the homicide rates in St. Louis in the early 90s (per 100k people):

And here's 2002-2014:

Homicides here started falling in the late 1990s and remained (relatively) low throughout the 2000s. The rate did jump in 2014 and has remained higher since, but it's still lower than it was in the early 90s.

Which, I think, is to say that your narrative isn't correct, at least regarding St. Louis. It's much messier than that (and mostly good news (relatively)).

Wow. In the mean time, Toronto has a homicide rate of around 2.0 (in 2012 [1]), making it almost 18x lower than St. Louis.

Genuine question, why do you continue to live there? Why wouldn't you move somewhere safer?

[1] https://en.wikipedia.org/wiki/Crime_in_Toronto

Violent crime is almost entirely concentrated in the poorest neighborhoods, and the people living there don't generally have the means/ability to move.

There are (at least) three things to understand about those numbers:

1. The City of St. Louis occupies a very small portion of what any reasonable person would call "St. Louis." And those numbers include only the City portion. The rest of the area is much safer. St. Louis County, which surrounds St. Louis on three sides and contains three times as many people (over 1 million), had 15 homicides in 2014 for a homicide rate of 1.5 (lower than Toronto) [0]. Most people living in the St. Louis area are living in very safe places.

2. Not all cities occupy such a small portion of their region. Toronto, to keep your example, is 243.33 sq mi. St. Louis is 66 sq mi. If St. Louis's borders were extended to match Toronto's reach, then it would include a whole bunch of the safe areas in St. Louis County, which would considerably soften the numbers. (Note that nothing on the ground would actually change.) This is why reasonable regional comparisons should be done at the level of the MSA. Arbitrary borders are, well, arbitrary and produce misleading comparisons.

3. Even when looking only at the City of St. Louis, you have to understand that many neighborhoods have 1 or 2 or 0 homicides each year. [1] Most of the violence is concentrated on the city's Northside and in the Southeast. Extreme violence is not a part of the everyday lived experience for most citizens.

The way outsiders talk about crime in places like St. Louis very sloppy.

(Bonus #4: Crime and safety and law and order aren't the only metric by which to choose where to live.)

[0] This number is almost certainly too low, because of how crime is reported in the County. I got that number here:

http://www.kmov.com/story/29948033/homicides-on-the-rise-in-...

Some other sources suggest it may be around twice that for the entire County, meaning the rate is likely closer to 3.0 per 100,000 (which is still considerably lower than in the City).

[1] https://nextstl.com/2013/01/understanding-st-louis-homicides...

This is a very good list, and I'd like to expand on point four: even among city-life-specific dangers, murder generally ranks fairly low.

Where I live, the annual pedestrian and bicyclist death count is is comparable to the St. Louis murder rate. And, of course, far more evenly distributed across neighborhoods. Add in car crashes and I'm probably at far more risk here than I would be in St. Louis.

I'm at more risk from smog and water pollution just by living in a city, and to fight that I should move not to Toronto but to a ritzy suburb. But dense populations mean good emergency response times, which argues for city again!

The list continues - you're absolutely right that safety isn't the only goal, but even on that basis it's not clear that murder is the statistic to observe.

These statistics don't tell the whole truth, as it can often be members of a subgroup killing eachother while the general population remains relatively safe.

It seems that they are unable or unwilling to separate deaths due to mental disorders from substance abuse. Probably due to the thinking that anyone who dies from a substance abuse issue is dealing with a mental disorder. While this may be true some of the time, I'm not sure that it's clear that it is always the case. And even if it was, substance abuse deaths would just be a subset of deaths from mental disorders, still would seem misleading to lump them together.

Substance abuse is generally regarded as a form of mental illness and is often comorbid with depression or anxiety disorders. It's not totally unfair to lump them, after all it is way easier do die from dangerous drugs than (purely) from a mental illness. That might give the reader some weird impressions or distort the problem to lean more heavily on the typical "War on Drugs" logic.

I think it's significantly better for public health policy making to consider substance abuse a mental health issue than it is to consider it a weird personal problem or a social variable that we have no control over.

That's a valid point, but the number tripled. Is that really down to statistical methods, or is there actually a huge problem which, admittedly, may not be measured perfectly?

How much wider of a net are we casting when we consider what counts as a "death due to mental illness or substance abuse" compared to 25 years ago?

I don't doubt it may have gone up, but I do doubt that these numbers tell us as much as some might think.

You make a good point. While drug abuse is often an attempt at self-medication for untreated mental illness (or illness which responds poorly to pharmaceutical industry offerings), it's a mark of the overall social disregard for both mental illness and drug use that this isn't broken down more carefully. On the mental illness side I'd expect to see things get worse from both years of overseas wars (often fought by people from the lowest socioeconomic quintile) and recessionary suffering and subsequent failure to recover. On the drugs side, fentanyl has turned the calculus of dosage, lethality, and morbidity upside down in pursuit of large profits, as much for corporate suppliers as criminal ones.

expect all this to get significantly worse rather than better if the incoming administration and Congress pursue their stated goal of repealing the Affordable Care Act aka Obamacare without a clear alternative ready to go. Last discussion I heard suggested that they might need 2-3 years to agree on a replacement health coverage system and that the problem was likely to be thrown back in the lap of the states for the time being.

Nobody seems to address the elephant in the room which is that the death rate from substance abuse and mental disorders was possibly underreported in the past

Kinda like how cancer wasn't that big of an issue until we began looking for it.

Not mentioned in the article, but I suspect some of this is due to the closing of most state run mental hospitals, which ramped up starting in 1972.

https://en.wikipedia.org/wiki/Deinstitutionalisation

Edit: More in-depth article on the topic: http://www.pbs.org/wgbh/pages/frontline/shows/asylums/specia...

probably, but keep in mind that those places are extraordinarily expensive to run. even though we are seeing massive increases in deaths as a result of their closure, it may still have been the right decision. there's a value to everything, including life.

Easier to stomach if the defunding of those programs had been in order to redirect that funding toward better-producing programs.

It's easy to put a value on someone else's life. How much should we spend on your life before it's ok to consign you to death?

no, it's not easy or pleasant or anything to put value on someone else's life. but it's a reality.

and to your question: i don't think it's your job to spend anything on my life. and i would certainly choose to die if the cost was bankrupting my family, for example.

The flipside of course, is how much extra are you willing to spend to save every preventable death?

How much are you you willing spend on someone else's life before it's ok for you to consign their death?

Here's the interactive version: http://vizhub.healthdata.org/subnational/usa

You can play with this map to see where the changes are biggest and smallest.

Interestingly, the strip from Maryland to Georgia - which normally stands out in US political maps, due to issues of slavery and race - stands out as being one big area where deaths due to drugs or substance abuse didn't increase much at all, annually, from a relatively low start.

Interesting to see the dramatic rise in deaths related to substance abuse from 1996 onwards, the year Purdue Pharma introduced Oxycontin. It's especially glaring in the most impoverished parts of Appalachia.

Is there any obvious reason for the higher death rate in the "bible belt"?

One thing I could imagine is rural exodus, leaving a higher percentage of old people.

Poverty.

Poverty combined with boredom. And from my personal experience, a complete unwillingness to discuss such problems (i.e, "Not my family").

Is it reasonable to assume that substance abuse deaths are far greater than mental health disorder deaths?

As for the increase: no shit. Alcohol, one of the most damaging drugs, plays a central role in our contemporary marketing economy that can scientifically manipulate into engaging with alcohol on a massive scale. On the pharma side, we shove extraordinarily dangerous drugs like Oxy, Klonopin, SSRIs and Adderall on under the guise of them being medication. Thanks to the industry surrounding all these things, people don't even know they are engaging in dangerous, life-threatening drugs until it's too late.

> Is it reasonable to assume that substance abuse deaths are far greater than mental health disorder deaths?

I wouldn't assume that, especially not based on this article.

It all depends on how you count up the deaths anyway. For example, considering substance abuse, what counts? Do we only take overdoses? What about accidental poisonings? What if I just smoke cigarettes until I am 85 and then die of heart disease that I was genetically predisposed to anyway? What about eating fatty foods to obesity and then dying from a heart attack?

I don't know how US death certificates and coroners work.

But things like death by suicide have tricky stats. In the US a death after self injury with the intent to die is counted as suicide. That intent to die bit is hard to know, so some deaths are going to be counted as something else. (Accidental death, maybe.) And because of the strong stigma and taboo around suicide historical data may undercount it.

Also, in the UK, for a coroner to rule a death as suicide it has to meet a higher burden of proof. The coroner has to be satisfied beyond all reasonable doubt that the death was suicide. For other deaths (but not murder) the burden of proof is the balance of probabilities.

So it's easier (in England) for a coroner to rule something is an accidental death than a death by suicide.

I guess the researchers are aware of all of this, and that they'll have mentioned how the cleaned up the data.

But it's something to be aware of when reading it.

By my count around a third of Americans die from drug misuse and drug-related causes:

http://alexkrupp.typepad.com/sensemaking/2014/05/the-one-sta...

And that was written before this study came out, which estimated that there may be an additional 60,000 annual tobacco-related deaths that are not being counted by the CDC:

http://www.nejm.org/doi/full/10.1056/nejmsa1407211

I think given the way people reason about such things you should probably advertise it as 20% of Americans die from tobacco use and then argue that this should be considered together with other drug related causes.

I'm not arguing that tobacco is anything other than a drug, I'm saying that emphasizing it in things like your first paragraph here will reduce the number of people that come to the wrong conclusion.

Take a look at what Reagan did to national mental health services in the 80s. Pretty hard not to draw some correlation there.

drugs have become the new treatment plan. and instead of fixing the mentally ill, they just sedate them until they reach an early death.

That was part of the problem. The other part is that "One Who Flew Over the Cuckoo's Nest" inspired the left to change laws so that it became hard for Psychs to institutionalize people who need help.

Both parts need to change. The "they aren't hurting anybody so leave them alone" people need to understand that they are hurting themselves and need to have their freedom of choice taken away. There needs to be strong oversight to prevent abuse. There needs to be sufficient funding to humanely take care of the mentally ill and, when possible, make them better. Addiction needs to be treated as a mental illness since, once the individual becomes addicted, it literally is a chemical abnormality in the physical brain mostly beyond the ability of the individual to address without medical help.

Before expanding the process to lock up people I'd have a good look at the current state of me tal health services - see the recent article about UHS about how they're abusing the system and their patient just to milk a bit more profit. Locked wards are fine IMO, but please don't allow this to for-profit institutions - same wrong incentives as for-profit prisons, i.e. current day slavery in the US.

What legislation do you think most contributed to deinstitutionalization?

it became hard for Psychs to institutionalize people who need help.

The prospect of a society in which any psychologist can, on a whim, circumvent due process and imprison anyone against their will indefinitely, is terrifying.

It's much, much better this way.

i agree that liberty is more important in this case, but it is important to acknowledge that the cost of liberty is the well-being of thousands (maybe millions) of mentally-ill would would otherwise be treatable.

>> it became hard for Psychs to institutionalize people who need help.

> The prospect of a society in which any psychologist can,

You've mixed up psychologist and psychiatrist.

It's easy to find ways that psychiatrists can detain people against their will, and force them to undergo medical treatment against their will, that don't rely on a single doctor and that have strong protections against abuse.

Something like two doctors - one of which knows the patient, plus an approved MH professional who's role is to protect the legal rights of the patient, plus free advocacy during all stages of detention to protect the legal rights, with powers of a close relative to ask for the detained person to be released; with a default position that people have capacity to make decisions about themselves, and that any limitations of capacity are limited to a single domain (I might not be able to say what happens to my money, but you should not use that to assume my capacity chose my medication is limited); with a recognition from all involved that this is a significant interference with my human rights and that we only tolerate this because it's in my best interests.

Same problem with psychiatrists. The psychiatrists associated with large-scale direct intake at hospitals are also extremely overworked and often not particularly sophisticated (psychiatrists who keep up with research and are effective practitioners can make more money outside the hospital system).

In the very worst case, which is unfortunately common, you have chains like UHS operating mental health hospitals in numerous states with what appear to be quotas. You don't have to drive yourself to a UHS hospital to end up in one. Most hospitals, even the very good ones, don't do on-site psych; in many of them, if you show up for any psych reason, you'll be held and then transported to an affiliated mental health hospital.

So, yes, there are very significant due process concerns with acute mental health care.

Psychologist no. But psychiatrist yes. Actually there's a few others who have that authority as well.

Involuntary commitment, 72 hour hold. Police can do this and some mental health professionals.

http://www.disabilityrightsca.org/pubs/502401.pdf

I feel you're committing the fallacy of the excluded middle here. The current standard in which a person needs to present a danger to themselves or others also results in everyone within the mental health system being treated like they're one step away from being violent criminals, resulting in significant additional stigmatization. My best friend went through a mental health crisis earlier this spring and was involuntarily hospitalized for several weeks in virtually carceral conditions.

Like I said above "There needs to be strong oversight to prevent abuse". That pretty much rules out whims.

We have people who talk to themselves and freeze to death in the Winter. I disagree when you say "it is better this way." There are real risks in allowing institutionalization; on the other hand, innocent people are dying from treatable mental health issues. Surely there is some middle position to be had?

>The "they aren't hurting anybody so leave them alone" people need to understand that they are hurting themselves and need to have their freedom of choice taken away.

You bring up an interesting point. I wonder who becomes in charge of diagnosing, which ones do we choose, for how long do we lock them away? Did we use shock therapy here in the US?

> Did we use shock therapy here in the US?

You still do! It's a useful treatment for some people with treatment resistant and severe, life threatening, depression.

It's used much less often, and there are some extra bits of bureaucracy around it.

And it's safer than conventional treatment for some people - pregnant women, for example.

To add to this: it's nothing at all today like it was in the 50s. Patients are not strapped down and thrashing: they are sedated and given muscle relaxants so that they don't thrash. The most common side effect is memory loss.

> Both parts need to change. The "they aren't hurting anybody so leave them alone" people need to understand that they are hurting themselves and need to have their freedom of choice taken away.

One of the big functions of involuntary commitment to mental hospitals in pre-1950s America was for parents to put away their "troubled daughters" because of social behavior they did not approve of (promiscuity, friend circle, etc). You get glimpses of this in memoirs from the era (the most recent one I have read that comes to mind is Herbert Huncke's autobiography). This continues today for rich parents who send their children to be involuntarily committed to "residential centers," "boot camps," and certain rehab programs.

This is a slippery slope that social reformers, busybodies, religious fundamentalists, etc have used to label any behavior they do not approve of, but cannot convince legislators to make against the law, as "hurting yourself" and grounds for taking freedom away. You are already most of the way down the slope by advocating for locking away drug addicts. Don't forget that Nixon's administration ramped up the drug prohibition as a way to break the civil rights and anti-war movements: https://harpers.org/archive/2016/04/legalize-it-all/. Today a similar policy will just be used to target leftist collectives and human rights groups, anarchists, the groups that sprang out of the Occupy movement, left-wing hackers, etc.

Involuntary institutionalization starts as a way to help the mentally ill, progresses through to oppression of women, and ends up as a tool of political repression (as it did in the USSR, and currently still does in Belarus).

need to have their freedom of choice taken away

A lot of people undergoing mental health issues also have severe financial stress and the related economic insecurities, which end up in a vicious cycle of poverty and worsening mental health - a situation of which I have first-hand experience.

It's depressing that your go-to strategy is to reduce their freedom of choice (and inevitably, their dignity and autonomy) instead of considering whether their functional needs are adequately met. It's as if you're arguing that people who are not able to participate and compete fully in the market can't be allowed any choices whatsoever.

Whenever I see a headline like this, I think "did reality change, or did the reporting methods?"

also worth thinking: did X become a bigger killer, or did we impede all the other killers so much that X is now coming into fruition more often.

My pet theory is that we are isolated and there are less ways to feel connected to society (suburbs, too much TV) and the over importance of image, wealth, and status. Too many McJobs and lonely suburbs and people start to get depressed and other forms of mental illness. Naturally these people self medicate when their needs are not met anywhere else.

Corporations used to feel more obligated to the society that they were in. Churches used to be more popular. Even if the Catholic church did evil things and controlled people, it did/does provide a place for people to feel like they belonged somewhere.

It would be interesting to see how these morbidity rates correlated with losses in wages and job security, since that magical year 1980. Particularly among middle-aged white males -- which are a pretty large group; and for whom stark rises in suicide rates have been observed by others, e.g.:

http://www.nytimes.com/2015/11/03/insider/more-white-men-die...

Whoah the quantity of subscribe popups, notifications and ads plastered all over the page is a huge turnoff. All this JS interaction seems to have somehow broke the animated gif as well.

Protip for publishers: one prominent entry point into "subscribing" with multiple choices is sufficient. I get that you want me to subscribe, but if I don't want to you're surely getting negative returns by creating such an obnoxious UX.

It's ironic that with all that JS, they couldn't be bothered to come up with something better than a gif to display the actual content.

I do wonder if they accounted for statistical bias.

Death rate is 3x larger, but US population was ~60% of what is was now. That alone means that the death rate is actually only 1.8x times bigger.

Additionally, it's been shown that doctors have been bribed and has caused vast over-diagnosis.

Why was this clickbait, masquerading as a study, linked on HN?

Regarding Kentucky cancer stats: https://en.wikipedia.org/wiki/Paducah_Gaseous_Diffusion_Plan...

Regardless of how you feel about legalization, policy, etc, this should prove, at the very least, that we aren't winning the "War on Drugs."

Yup. The war on drugs is a waste of everything.

Prohibition really has turned out to be a seriously damaging thing to society. Drugs are classified not based on their potential to cause harm or abuse, but instead by some nebulous concept of morality. We, as a society need to embrace a more scientific based approach.

The value judgments are inseparable from the 'harm or abuse' measurements; there is no value-neutral way to assess harm and abuse. You can say you want to be scientific and come up with numbers, but the selection of the metrics and the 'critical values' will be political and ethical decisions.

Having said all that, I am in favor of full legalization of every substance, even though I personally find the use of almost all substances 'distasteful'.

You can scientifically quantify harm both to individuals and society. For example, potential for abuse, potential for individual harm, and potential for social harm are all factors that could be quantified and considered.

Using those metrics alone, the juxtiposition of the classification of marijuana versus alcohol or nicotine is quite interesting. You could perhaps include many other drugs as well such as cocain, mdma, and perhaps even LSD (though this one does pose some concerns socially given that it can cause manic episodes).

My point is that the current "war on drugs" has no scientific basis. And, as such, where we've ended up has been very destructive in terms of the cost to society (imprisonment, surge in use of more harmful but easier to source substances, etc...).

If you are going to do a cost/benefit analysis of drugs, what do you include in the costs? Do you include direct medical care, counselling, crime, and/or insurance? Do you count the financial or emotional costs borne by family members? Do you count lost income? Do you count lost taxes? How do you account for positive outcomes from drug use, such as potential improvements in productivity, or creativity? Do you count benefits to happiness, emotional well being, sense of belonging, or fellow-feeling? How do you count all these? Do you evaluate risks from a straight expected-value or do you use some risk-aversion factor? What is your time preference (i.e. 'interest rate'), and how do you derive it?

I am not aware of any government policy based on science; the ones that reference scientifically derived evidence are mostly justifying intuitive/emotional judgments. It may be a Hume-ian perspective, but I am not aware of any compelling evidence to the contrary.

You can use whatever you'd like. It's comparative analysis, so it doesn't matter too much. E.g. Marijuana versus Alcohol versus Oxycodone versus meth.

> LSD (though this one does pose some concerns socially given that it can cause manic episodes)

There still seems to be a bit of "reefer madness" going on when it comes to psychedelics, even inside the legalization circles.

Some people have reported a link between manic episodes and LSD - just like people have reported links between manic episodes and alcohol, manic episodes and marijuana, etc. That doesn't seem to unbelievable, considering the vast array of things that can trigger mania (even things most people consider innocuous, like visiting another country).

Make drugs illegal. Lock up drug users in prison with other drug users who get drugs into prison because it's not as hard as you think. Teach prisoners no employable skills. Dump them back outside with the most damming scarlet letter in American society. Give them no public assistance because of drug conviction. Reduce life options to selling drugs or becoming full-blown addict to escape the painful reality. Observe drug use or sale. Arrest again. Change nothing.

What a surprise.

I'd be curious to see these numbers separated. Did they both increase proportionally?

Here is the detailed report: http://ghdx.healthdata.org/us-data

While these subjects are connected for some individuals - it's not wise to pile them up together into the single stat number.

Wrong stats.

Which drugs? Top 5-10?

Which mental disorders? Top 5-10?

(I upvoted you.)

For mental disorders it's something like:

1) Anorexia Nervosa (historically had very high death rate, something like 20%)

2) Body Dysmorphic Disorder (most people don't know that this has a high suicide rate)

3) Major depression

4) Psychotic illness (people with psychosis tend to live 20 years less than the general population)

5) ADHD

(I'm not sure of the ordering of the last 3.)

As for drugs, far and away the leading cause of death and harm is going to be alcohol.

So, clearly the anorexic people die because they starve themselves, people with depression commit suicide, but what's killing the people with ADHD? I know ADHD is correlated with greater risk taking, greater likelihood to abuse drugs, but is it severe enough to put it in the top 5 or 10?

Yes

I believe this is connected to stagnant wages (which starts in 1970s for Americans), with a catch.

My hypothesis:

1. Counterintuitively, bad economic circumstances don't lead to chronic stress. People in countries with objectively terrible economic conditions are not more stressed or unhappier. It is simply: 'life'.

2. Caveat: Bad economic conditions do cause chronic stress where the media is projecting an illusion of rising success and wealth for the average person. This is because large numbers of people begin to feel left behind, even though most of their 'successful' peers are probably in debt to their eyeballs. Chronic stress is caused by the constant reminders in the form of fraudulent statistics, advertisements and social conditioning by fictional television shows that cause them to compare their circumstances to largely imaginary ideals.

My solution, which sounds drastic but I believe it to be meritorious, is to ban television (but we can keep the Nature channels). I believe a large number of social ills shall mostly disappear once this is accomplished.

Supporting evidence:

http://news.harvard.edu/gazette/story/2009/03/fijian-girls-s...

Reply on news.ycombinator.com