Are Marijuana Use and Suicide Linked?

A review of the data show there’s cause for alarm

by Libby Stuyt, MD

Correlation does not mean causation, but that is why we do research — to follow correlations in an effort to determine causation. Data linking marijuana use to people with suicidal ideation, attempts, and completed suicides are steadily increasing.

Many states, including Colorado, have made post-traumatic stress disorder (PTSD) an approved condition for medical marijuana. As a psychiatrist having treated many people with PTSD, I know that marijuana is not the answer, similarly to why benzodiazepines or alcohol are not the answer to treat PTSD.

Sure, these addictive substances work in terms of numbing the person so that they do not experience the symptoms. However, to keep the symptoms at bay, the person must use every day, sometimes all day long. This sets them up for developing cannabis or other substance use disorder and the possibility of psychotic symptoms or worsening psychotic symptoms from their PTSD.

A Review of the Data

study of 3,233 veterans in a cross-sectional, multi-site study by the VA found that cannabis use disorder (CUD) was significantly associated with both current suicidal ideation and lifetime suicide attempts compared to veterans with no lifetime history of CUD. The significance persisted even after adjusting for sex, PTSD, depression, alcohol use disorder, non-cannabis drug use disorder, history of childhood sexual abuse, and combat exposure.

An observational study of 2,276 veterans treated in VA PTSD treatment programs around the country found that those who never used marijuana had significantly lower symptom severity 4 months after treatment, those who stopped using marijuana had even lower levels of PTSD symptoms, and those who started using marijuana had the highest levels of violent behavior and PTSD symptoms 4 months after treatment.

The 2020 National Veteran Suicide Prevention Report indicated that Colorado’s rate for veteran suicides is significantly higher than the national rate. There were 217 veteran suicide deaths in Colorado in 2019 — an all-time high and a 25% increase over 2018.

The general population in Colorado has also seen a gradual increase in suicides over the years:

  • 795 people in 2004
  • 919 people in 2009
  • 1,021 people in 2012
  • 1,063 people in 2014
  • 1,242 people in 2018

This trend seems to mirror the increased commercialization of marijuana in Colorado. Available toxicology shows that marijuana was present in the following percentage of suicides:

  • 5.5 percent in 2004
  • 7.1 percent in 2009
  • 11.8 percent in 2012
  • 14.9 percent in 2014
  • 19.1 percent in 2015
  • 22.8 percent in 2018  

Another concerning trend is that not all suicides have toxicology reported. In 2004, 92% of suicide cases had toxicology reports, while by 2018, that number fell to 70.6%.

Even more alarming is the correlation with marijuana and suicide found among young people, especially adolescents. A large systematic review and meta-analysis of 11 studies and 23,317 adolescents found that suicidal ideation and suicide attempts were significantly higher in adolescent cannabis users than in non-users, with an OR of 3.5 (95% CI 1.53-7.84) for suicide attempts in those using cannabis.

The 2019 National Youth Risk Behavior Survey, a cross-sectional, school-based national survey of 13,677 high-school age students found that 50% of those who used marijuana (ever or currently) reported feeling sad or hopeless. It also found that using marijuana (ever) could increase the risk of mental health challenges or suicidal behavior, and trying marijuana before the age of 13 could increase the risk of attempting suicide.

In Colorado, marijuana is the most common drug found in toxicology of teens who die by suicide — even greater than alcohol. In 2018, there were 69 teens ages 15 to 19 who died by suicide. Marijuana was present in 36.7% of the cases, but only 49 of the 69 had toxicology information available. There has been a significant increase in the number of teen suicides in Colorado in the last 5 years, up to 80 in 2019, along with a significant increase in the number with marijuana found in their system. The fact that marijuana is the number one drug found when toxicology is reported correlates with the increased THC potency and availability, and use of concentrates and increase in vaping of marijuana by teens in Colorado.

A recent analysis of 204,780 youths (ages 10 to 24) with the diagnosis of mood disorders, based on Ohio Medicaid claims data linked with death certificate data from 2010 to 2017, found that 10.3% received the diagnosis of CUD. This rate is significantly higher than that reported in the general population. In addition, those with CUD were significantly more likely to engage in non-fatal self-harm and to die. Unintentional overdoses, suicide, and homicide were the three most frequent causes of death.

A much larger recent study looked at commercial and Medicare Advantage claims for 75,395,344 individuals throughout the country from 2003 to 2017. It found recreational cannabis laws permitting dispensaries and lacking dose-related restrictions were associated with significant increases in assaults among people younger than 21 years and increases in self-harm for men ages 21 to 39 years, compared with states with no medical or recreational cannabis laws.

Causation or Correlation?

While marijuana does not cause overdose deaths, this association with violence needs to be investigated thoroughly. There is increasing research demonstrating that regular use of marijuana with THC greater than 10% can result in development of psychotic symptoms. Cannabis-induced psychosis can become permanent, even after cessation of use. A study demonstrated that the highest conversion rate (47.4%) to diagnoses of bipolar disorder or schizophrenia occurs with cannabis-induced psychosis, compared to that from amphetamines, hallucinogens, opioids, or alcohol. Younger age is associated with higher risk of conversion.

There have been multiple studies documenting the association of cannabis use by persons with psychotic disorders and an increasing risk of violence. This can include violence toward one’s self as well as others. A study of 1,136 patients recently discharged from acute psychiatric facilities examined the relationship between continued use of cannabis and violence and found that continuing to use cannabis is a moderate predictor of subsequent violent behavior — even more so than alcohol or cocaine. A recent large, longitudinal study of 2,994 people with a psychotic disorder found 11.4% were using cannabis. Cannabis users were found to report a lower quality of life, being less satisfied with their family relations and financial situation, and to show more aggressive and disruptive behavior and self-harm than non-users.

What’s Driving the Correlation?

There are multiple theories as to why there is such a strong correlation between marijuana use and violence, including suicide. Dellazizzo et al. found a moderate association between cannabis use and physical violence in youths and emerging adults, with a potential dose-response association. They offer an excellent overview of potential mechanisms explaining violent behavior, which could help explain suicidal behavior. These include the fact that cannabis use can cause or exacerbate psychotic symptoms such as delusions; alter a person’s sense of reality by causing perceptual distortions; cause a person to feel anxious, panicky, and paranoid about their surroundings and others; impair executive functioning, creating problems with impulse control and decision-making; and cannabis withdrawal can cause people to feel irritable, restless, and anxious. It is logical that any of these could contribute to suicidal thoughts in someone whose mood is dysregulated by cannabis.

We need more people to be aware of this strong association between cannabis use and suicide, and to conduct more research to determine its cause.

Libby Stuyt, MD, is an addiction psychiatrist in Colorado.

Leave a Reply

Your email address will not be published. Required fields are marked *