The contagious nature of bacterial or viral infections like strep throat or influenza is well understood. You’re at risk of catching the flu, for example, if someone near you has it, as the virus can be spread by way of droplets in the air, among other modes of transmission. But what about a person’s mental health? Can depression be contagious?
A JAMA Psychiatry paper published earlier this year seemed to suggest so. Researchers reported finding “an association between having peers diagnosed with a mental disorder during adolescence and an increased risk of receiving a mental disorder diagnosis later in life.” They suggested that, among adolescents, mental health disorders could be “socially transmitted,” though their observational study could not establish any direct cause.
It makes some intuitive sense. Psychologists have studied how moods and emotions can spread from person to person. Someone howling with laughter might be contagious in the sense that it makes you laugh, too. Similarly, seeing a friend in emotional pain can evoke feelings of despair — a phenomenon termed emotional contagion.
For more than three decades, researchers have investigated whether mental health disorders, too, may be induced by our social environment. Studies have found mixed results on the extent to which friends’, peers’, and families’ mental health can impact an individual’s mental health in turn.
The JAMA Psychiatry study — conducted by researchers at the University of Helsinki and other institutions — analyzed nationwide registry data on 713,809 Finnish citizens born between 1985 to 1997. The research team identified individuals from schools across Finland who had been diagnosed with a mental disorder by the time they were in ninth grade. They followed the rest of the cohort to record later diagnoses, up until the end of 2019.
The study found that ninth grade students who had more than one classmate diagnosed with a mental health disorder had a 5 percent higher risk for developing a mental illness in subsequent years than students without any classmates with diagnoses. The risk was particularly high in the immediate year after exposure: Students with one diagnosed classmate were 9 percent more likely to receive a mental health diagnosis, while students with more than one diagnosed classmate were 18 percent more likely to receive a diagnosis. The risk was greatest for mood, anxiety, and eating disorders. Increased risk was observed after adjusting for an array of possible parental, school- and regional level confounders like parental mental health, class size, and area-level unemployment rates.
These results might seem like compelling evidence for social transmission of mental health disorders, but other researchers — such as Eiko Fried, an associate professor of clinical psychologist at Leiden University — have suggested that the Finnish team may not have controlled for all relevant confounders. Fried brought up living in a poor neighborhood, which increases depression risk, as an example of a confounder in an email to Undark. “These kids end up in the same schools, and you get an aggregation of depression in those schools. This now looks like social contagion, until the confounder — neighborhood — is taken into account.”
Seeing a friend in emotional pain can evoke feelings of despair — a phenomenon termed emotional contagion.
The researchers did control for neighborhood employment rates and educational levels, but it’s possible they still didn’t account for other influential contextual factors. To the extent that these shared factors are insufficiently measured, estimates of correlated outcomes risk pinning causality on the wrong variable. In a post on X (formerly Twitter), Fried said it might be more plausible that hidden confounders explain what’s going on, rather than social contagion.
In response to an email query laying out critiques of potentially confounding variables, the lead author of the Finnish study, Jussi Alho, underscored the utility of using classrooms as a point of reference by pointing to another potential influence: The tendency for people to seek out or be attracted to those who are similar to themselves. “In our study, we mitigated this self-selection bias by using school classes as proxies for social networks,” he explained. “As institutionally imposed social networks, school classes are well suited to research, as they are typically not formed endogenously by individuals selecting similar others as classmates. Moreover, school classes are arguably among the most significant peer networks during childhood and adolescence, given the substantial time spent together with classmates.”
By Alho and his co-authors’ reckoning, as they write in the paper, the strength of the Finnish study lies in the fact that the social networks being investigated were not chosen independently by the research subjects. At the same time, Alho allowed that critics have a point: “We cannot fully rule out residual confounding,” he wrote in an email to Undark, “due to unmeasured or inaccurately measured covariates in our study.”
These confounders are a persistent problem dogging this line of research. A 2013 study published in the journal Health Economics, for example, examined the mental health status of college student roommates over their first year, testing for possible “contagion among people who are placed together largely by chance.” The authors described the study a natural experiment, which they argued that would be able to produce, in their words, “unbiased estimates” of “causal effect.”
Are the mental health issues spreading between persons in social networks? Or are some other unknown factors merely creating that impression?
The researchers found “no significant overall contagion of mental health and no more than small contagion effects for specific mental health measures” like general psychological distress, depression, and anxiety. Even in this case, though, the mild contagion effect could be attributable to unmeasured factors, such as the students sharing comparable social environments and upbringings. After all, they’re attending a school they might have selected for based on similar academic interests or extracurricular skills.
All these possible influences make it hard to know what’s driving what. Are the mental health issues spreading between persons in social networks? Or are some other unknown factors merely creating that impression?
Whatever the answer, such personal exposures may be driving a different sort of contagion: public awareness. Generalized anxiety disorder, for example, first appeared as a diagnosis in the third edition of the Diagnostic and Statistical Manual of Mental Disorders in 1980. The condition causes “excessive, frequent and unrealistic worry about everyday things,” according to the Cleveland Clinic’s Health Library. By the time a fourth edition of the DSM and its updated GAD diagnostic criteria came out in 1994, GAD had “morphed from a rarely diagnosed condition into a disorder with a lifetime prevalence reaching up to 5 percent in a community sample,” according to a 2017 paper on the history of the diagnosis. Data from a 2016 Agency for Healthcare Research and Quality report on anxiety in children indicates that childhood anxiety occurs in approximately one in four children ages 13 to 18, while the lifetime prevalence of severe anxiety disorder in that age group is 5.9 percent.
What’s causing these rates is potentially better awareness among both patients and clinicians. Or, it could result from an umbrella of other factors like evolving diagnostic criteria and improved access to treatment. But as Alho and his colleagues suggest in their paper, it’s possibly also driven by knowledge and acceptance of mental health disorders gained through social networks. After all, being exposed to a peer with a mental disorder, the researchers noted in their study, may well aid in “normalization of mental disorders through increased awareness and receptivity to diagnosis and treatment.”
UPDATE: A previous version of this piece stated that a study in JAMA Psychiatry found that ninth grade students who had one or more classmates diagnosed with a mental health disorder had a 5 percent higher risk for developing a mental illness in subsequent years than students without any classmates with diagnoses. The finding was for more than one classmate.