The Mental Health of Lonely Marijuana Users

Mr. Lonely ¹

Does Smoking Pot Offer Relief to the Lonely?  A new paper by the original Tylenol and social pain researchers claims that it does (Deckman et al., 2013). Let's take a closer look.

Comfortably Numb: Marijuana Use Reduces Social Pain, Research Finds

Marijuana use buffers people from experiencing social pain, according to research published online on May 14 in Social Psychological and Personality Science.

"Prior work has shown that the analgesic acetaminophen, which acts indirectly through CB1 receptors, reduces the pain of social exclusion," Timothy Deckman of the University of Kentucky and his colleagues wrote in the study. "The current research provides the first evidence that marijuana also dampens the negative emotional consequences of social exclusion on negative emotional outcomes."

You could be forgiven if you thought, as I initially did, that the University of Kentucky IRB must hold a liberal view on the administration of controlled substances to undergrads participating in psychology experiments. But that's not what happened here... the data are entirely correlational, based on self-report, and largely problematic (in my view).


Marijuana Lowers Self-Worth and Worsens Mental Health in Those Who Are Not Lonely

That's my interpretation of the article, which is SO clunky compared to the fun and breezy query, Can Marijuana Reduce Social Pain? ²

The paper begins with the premise that "Social and physical pain share common overlap at linguistic, behavioral, and neural levels" (Deckman et al., 2013). So let's give a pain reliever to reduce the sting of rejection!  A critique of the original work asked why the authors chose Tylenol, as opposed to an NSAID like aspirin, ibuprofen, or naproxen. In the current study they tried to develop a mechanistic account of why acetaminophen might reduce social pain:

Prior research has shown that acetaminophen—an analgesic medication that acts indirectly through cannabinoid 1 receptors—reduces the social pain associated with exclusion. Yet, no work has examined if other drugs that act on similar receptors, such as marijuana, also reduce social pain.

The problem is that acetaminophen's mechanism of action is surprisingly unclear (Toussaint et al., 2010). One prominent hypothesis claims that Tylenol might exert its analgesic effects through descending serotonergic pathways at the level of the spinal cord. In fact, the paper that Deckman et al. cited in favor of cannabinoid 1 (CB₁) receptors describes a very complex pathway that includes indirect involvement of CB₁, with actual pain suppression occurring in the spinal cord. ³

An even more basic question: if acetaminophen acts through CB₁ receptors, then why isn't it a potential drug of abuse, or known by experienced pharmanauts for its psychoactive properties?  The drug experience vault Erowid says:

Acetaminophen is a non-salicylate analgesic and antipyretic (pain killer and fever reducer). It is a common over-the-counter pain medication found in hundreds of products around the world. At higher doses it is known to cause liver-damage and has a low therapeutic index (ratio of effective dose to toxic dose), making it dangerous when included in recreationally used pharmaceuticals [e.g., Tylenol with codeine]. It is not known to be psychoactive.

On the other hand, we all know that cannabis is psychoactive. The design of the cannabis study included cross-sectional national survey data, a two year longitudinal survey of 400 high school students, and a Mechanical Turk-implemented version of cyberball, an online game to simulate social exclusion. In all cases, participants reported their marijuana use, and this was related to the variables of interest.

I'll focus on the national survey data in this post, which comprised Study 1 (Marijuana Use Buffers Lonely People From Lower Self-Worth and Self-Rated Mental Health) and Study 2 (Marijuana Use Predicts Fewer Major Depressive Episodes Among the Lonely).

Study 1 used data from the National Comorbidity Survey: Baseline (NCS-1), 1990-1992 (ICPSR 6693), which you can download for yourself. The survey recruited 8,098 individuals from the ages of 15 to 54 living in the U.S., and included over 4,000 variables. Only four variables were chosen for the present study: self-reported loneliness (1= often, 4 = never), marijuana use (0 = none, 1 = daily, 8 = once or twice a year), self-worth (1 = high, 4 = low), and overall mental health (1 = excellent, 5 = poor).

Loneliness was used as a proxy for social pain. Contrary to what the headlines suggested, the impact of pot smoking on social pain was not directly examined. Instead, the study assessed the effects of loneliness (high, low), marijuana use (high, low) and their interaction on self-worth and mental health.

Loneliness and pot smoking interacted to predict feelings of self-worth [B = 0.03, t(5609) = 2.20, p = .03]. Given the huge number of participants, this level of statistical significance is not very impressive.

Fig. 1 (modified from Deckman et al., 2013). Study 1: Marijuana use moderates the relationship between loneliness and self-reported feelings of self-worth. [NOTE: items were reverse-scored for display purposes.]


For lonely people, the amount of pot smoked didn't make too much of a difference in their self-worth (see red arrow above).  For socially connected people, greater marijuana use resulted in lower self-worth, although it's not clear this was significant (pairwise statistical tests were not reported).

I also question how the High Marijuana Use and Low Marijuana Use groups were determined, because over 5,000 participants did not smoke pot at all in the last 12 months. Does the heavy use group combine those who smoke 6 joints a year with those who smoke daily?

Table depicting the mean level of loneliness (1=often to 4=never) for participants at 9 levels of pot smoking (0=none, 1=daily, 8=once or twice a year). Unlike the figure above, the values were not reverse-scored. Data from the National Comorbidity Survey: Baseline (NCS-1), 1990-1992 (ICPSR 6693).


In the lonely group, the frequency of marijuana use had even less of an impact on self-rated mental health. In contrast, heavy pot use resulted in worse mental health among the socially connected. A modest loneliness by marijuana use interaction was observed for mental health [B = 0.03, t(5609) = 2.07, p = .04], similar to what was seen for self-worth.

Fig. 2 (modified from Deckman et al., 2013). Study 1: Marijuana use moderates the relationship between loneliness and self-reported mental health. [NOTE: items were reverse-scored for display purposes.]


Looking at Fig. 2 above, it's clear that marijuana use does not buffer the lonely from the negative consequences of social pain: the black circle and gray square are overlapping. But the authors interpret this result differently:

Marijuana use buffered the lonely from both negative self-worth and poor mental health. This evidence suggests that at relatively high levels of social pain, marijuana use lessens negative consequences of social pain.

As part of the six sentence Discussion of Study 1, they point out one weakness to motivate Study 2:

This study contained some limitations. First, it only assessed self-ratings of both self-worth and mental health. If marijuana use weakens the relationship between social pain and self-reported psychological well-being, then there should also be a lower rate of validated clinical diagnoses of poor psychological well-being.
. . .
To address the limitation of Study 1, Study 2 sought to show that marijuana buffered lonely participants from experiencing a standardized diagnosis of poor psychological well-being. Study 2 used a different nationally representative sample from Study 1 to test this hypothesis.

HOWEVER, the dataset used in Study 1 has extensive information on DSM-III-R diagnoses (including depression) for the majority of participants, so I'm not sure why this wasn't included. Study 2 used data from the National Comorbidity Survey Replication (NCS-R; Kessler & Merikangas, 2004), a different national sample of 10,000 respondents.

Speaking of replication, Deckman et al. should have been able to completely replicate the pot × loneliness analyses for self-worth, self-rated mental health, and DSM depression in both National Comorbidity Samples. I'm not sure why they didn't.

For Study 2, non-users were excluded (unlike in Study 1). The final sample included 537 participants with info on loneliness, marijuana use, and whether they experienced a major depressive episode during the past year. Once again, the results demonstrated that if you're lonely, smoking a lot vs. a little pot will not affect whether you'll experience a major depressive event (red arrow below). If you're not lonely, heavy marijuana use increases the risk of major depression.

Fig. 3 (modified from Deckman et al., 2013). Study 2: Marijuana use moderates the relationship between loneliness and a having a DSM-IV major depressive event in the past 12 months.


Study 3, a survey of 400 high school students, was the most puzzling of all. At Time 1 the students were asked about loneliness, lifetime marijuana use, and depression. Two years later, they were asked again about depression, using the Behavior Assessment System for Children (second edition), but not about marijuana use and loneliness (which could have changed drastically in 2 years).

At any rate, lonely heavy pot users were the least depressed of all at Time 2. I'm not sure how to interpret this; the pattern differs from what was seen in adults. Maybe the lonely heavy pot users at Time 1 bonded with their peer group over two years and were no longer lonely at Time 2.

Fig. 4 (modified from Deckman et al., 2013). Study 3: Marijuana use moderates the relationship between loneliness and depression over 2 years in adolescents. 


Conflicting earlier studies in adolescents have suggested that lonely high school students are more likely (Page, 2000) and less likely (Grunbaum et al., 2000) to use marijuana. A recent study indicated that heavy marijuana users are more likely to engage in self-injury (Giletta et al., 2012), but this was true only for Americans and not for Dutch and Italian students. I imagine there's a huge literature on these issues, but it wasn't addressed at all in the present paper.

Overall, I don't think the authors have demonstrated that marijuana reduces social pain, at least not in adults. They used a very select set of questions from huge, comprehensive national surveys and then called this a limitation of the study:

Another potential limitation to some of the above studies lies in how social pain was measured. In Studies 1–3, single-item measures of loneliness were used as a proxy for social pain. These studies use large community sample data sets and thus our ability to include numerous measures was limited.

There were many other questions that could have assessed social pain in NCS-1 and NCS-R, including a series of questions about friendships, e.g. "How much do your friends really care about you--a lot, some, a little, or not at all?"

Has this paper advanced the agenda of the social pain/physical pain isomorphists? We already knew that opiates were good at alleviating both types of pain. And it's a truism to say that people turn to alcohol and all sorts of recreational drugs to dull the pain of a lonely existence. For the most part, we assume this isn't a healthy way to cope. Some studies suggests that depression is decreased in heavy marijuana users (Denson & Earleywine, 2006) but others find an increase (Pacek et al., 2013).

In sum, Deckman et al., (2013) presented evidence that heavy marijuana use is detrimental to the mental health of socially connected individuals and not especially effective in buffering lonely users from social pain.


Footnotes

¹ However, please note that highly reliable source TMZ claims "Akon doesn't drink ... Akon doesn't smoke ... but Akon was pretty damn surprised when he found out his pal Justin Bieber might be doin' both."

² University press offices!! I'm sure you'd love to hire me to write your press releases. Price quotes are available upon request, please leave a comment.

³ You might also want to know something about the distribution of CB₁ receptors in the anterior cingulate cortex, the purported locale of physical/social pain overlap.

⁴ Survey questions were:

LONELY - During the past 30 days how often did you feel lonely?
POT - On the average, how often in the past 12 months have you used marijuana or hashish? Just
WORTHY - I feel I am a person of worth, at least equal with others.
MENTAL HEALTH - How would you rate your overall mental health? Is it excellent, very good, good, fair, or poor?

Data for these four questions were available from 5,631 participants. Ratings were standardized, reverse-scored, and analyzed using weighted least squares regression.


References

Deckman, T., DeWall, C., Way, B., Gilman, R., & Richman, S. (2013). Can Marijuana Reduce Social Pain? Social Psychological and Personality Science. DOI: 10.1177/1948550613488949

Dewall CN, Macdonald G, Webster GD, Masten CL, Baumeister RF, Powell C, Combs D, Schurtz DR, Stillman TF, Tice DM, Eisenberger NI. (2010). Acetaminophen reduces social pain: behavioral and neural evidence. Psychol Sci. 21:931-7.

Toussaint K, Yang X, Zielinski M, Reigle K, Sacavage S, Nagar S, Raffa R. (2010). What do we (not) know about how paracetamol (acetaminophen) works? Journal of Clinical Pharmacy and Therapeutics 35 (6), 617-638.