How We Help Others May Reveal Which Psychiatric Symptoms Will Persist

For decades, psychiatrists have searched for reliable signals that predict which patients will struggle with persistent symptoms and which will recover. A new study from Karolinska Institutet in Sweden suggests that one such signal may be hiding in plain sight: the way people help others. Published in Translational Psychiatry, the research followed more than 4,400 adults with psychiatric disorders for a full year and found that distinct flavors of prosocial behavior—helping anonymously, helping publicly, helping when emotionally aroused—were each linked to different long-term symptom trajectories, from problematic alcohol use to substance use to chronic anxiety.

The study’s central insight is that prosociality is not a single, uniform trait. Drawing on the Prosocial Tendencies Model, the researchers distinguished six context-dependent subtypes: altruistic helping, in which a person assists others at a personal cost; anonymous helping, where the helper’s identity is never revealed; compliant helping, performed when asked; dire helping, triggered by emergencies; emotional helping, carried out in moments of affective arousal; and public helping, done in front of an audience. Behavioral and neuroimaging research indicates these subtypes recruit partly dissociable neural circuits spanning the prefrontal cortex, striatum, and temporoparietal regions, meaning that measuring them separately may capture genuinely different psychological processes.

To test whether these subtypes matter clinically, the team enrolled 4,484 Swedish adults—71.1 percent of them women, with a mean age of 35.8 years—who reported at least one current or past psychiatric diagnosis from a mental health care setting. At baseline, participants completed the Prosocial Tendencies Measure, rating each subtype on a five-point scale. They then reported their alcohol use, substance use, anxiety, and depressive symptoms repeatedly over twelve months, using validated instruments including the AUDIT, DUDIT, GAD-7, and PHQ-9. Alcohol and substance use were assessed every second month, while anxiety and depression were tracked monthly through the REDCap online platform.

The analytical approach was deliberately sophisticated. Rather than estimating a single average symptom course for the whole cohort, the researchers used latent class linear mixed models to identify subgroups of participants with distinct longitudinal trajectories—stable high-symptom classes, stable low-symptom classes, and, in the case of anxiety, a class showing improvement over time. Logistic regression then tested whether baseline prosocial subtypes predicted membership in these trajectory classes, with all models adjusting for demographic covariates such as age, gender, and education, as well as for shared variance across psychiatric symptom domains.

Three findings stood out. Higher anonymous prosociality was associated with lower odds of belonging to the stable high alcohol use trajectory, which comprised 68.1 percent of the alcohol-use sample, with an odds ratio of 0.89. In other words, people who readily help others without seeking credit were less likely to remain in the persistently risky drinking group. Conversely, higher public prosociality—helping in front of others—was associated with higher odds of belonging to the stable high substance use trajectory, which included 13.2 percent of participants, with an odds ratio of 1.15. And higher emotional prosociality predicted membership in the stable high anxiety class, which accounted for nearly half of the anxiety sample, with an odds ratio of 1.25, the strongest association in the study.

Notably, no prosocial subtype predicted depressive symptom trajectories, even though baseline measures of prosociality did correlate with depression levels at the study’s start. The authors suggest several explanations. Anxiety and depressive symptoms showed strong reciprocal associations in their models, so the baseline link between anonymous helping and depressive symptoms may partly reflect shared variance with anxiety. Alternatively, depressed mood may color self-reports of social tendencies in a state-dependent way, without representing a stable trait-like influence on the course of depression itself.

The substance use finding is particularly intriguing because public prosociality showed no association with baseline substance use levels. This temporal pattern suggests that a strong orientation toward socially visible helping is not a consequence of drug use but may instead constitute a longitudinal vulnerability marker—possibly reflecting a heightened dependence on social approval and external validation. Previous work in healthy samples had already linked public prosocial behavior to later substance use, and the new results extend that pattern into a clinical population. The researchers also note that public prosociality correlated moderately with altruistic prosociality, raising the possibility that earlier reports of reduced altruism in substance use disorders partly reflected this shared variance with socially visible helping.

The alcohol findings tell a complementary story. At baseline, lower altruistic and lower anonymous prosociality were both independently associated with higher alcohol use, echoing prior experimental work showing reduced prosocial decision-making in alcohol use disorder. Yet only anonymous prosociality showed a prospective association with the alcohol trajectory, and altruism showed no longitudinal association with later symptom development at all. This supports recent longitudinal evidence suggesting that reduced altruism may not be a premorbid risk factor for substance-related problems, but rather a correlate that emerges alongside them. The lower propensity for anonymous helping among risky drinkers may instead reflect greater self-referential motivational processes, consistent with documented antisocial traits and selfish choices in this population.

Emotional prosociality—helping when affectively aroused—emerged as the most robust marker, showing both a strong baseline association with anxiety symptoms and a prospective prediction of the persistent high-anxiety trajectory. The authors interpret this pattern as potentially reflecting heightened emotional responsivity and empathic concern, which can exacerbate anxiety through interpersonal over-involvement. Intriguingly, higher emotional and public prosociality were also initially associated with belonging to the anxiety symptom-improvement class, but these associations disappeared after adjusting for pandemic-related contextual factors, including COVID-19 infection severity and subjective loneliness. The core three associations, however, remained robust across sensitivity analyses, suggesting they capture relatively stable behavioral dispositions rather than situational artifacts of data collection during the pandemic.

The study has limitations worth noting. Prosocial behavior was measured only once, at baseline, so the design cannot establish causal direction or track how helping tendencies change alongside symptoms. Self-report measures are vulnerable to social desirability bias, and the latent class solutions require replication in independent cohorts. Still, the implications are striking: prosociality is neither uniformly beneficial nor uniformly diminished in psychiatric illness. Different subtypes point in opposite directions, a pattern the authors liken to the Aristotelian doctrine of the mean, in which an intermediate level of social engagement represents a healthy balance between withdrawal and over-involvement. If future research validates these markers with objective, computational behavioral tasks, clinicians may one day add a simple question—how, and for whom, do you help?—to the toolkit for predicting which symptoms will linger, and for designing interventions that promote more regulated, sustainable forms of social connection.

Subject of Research: Longitudinal associations between context-dependent prosocial behavior subtypes and persistent psychiatric symptom trajectories

Article Title: Distinct prosocial behaviors associated with persistent symptom trajectories in psychiatric disorders

Article References: Jangard, S., Vieira, J. B., Garke, M. Å., Ojala, O., Månsson, K. N. T., Hedman-Lagerlöf, M., Sörman, K., El Alaoui, S., Rozental, A., Forsström, D., Lundin, J., Shahnavaz, S., Lundgren, T., Olsson, A., & Jayaram-Lindström, N. (2026). Distinct prosocial behaviors associated with persistent symptom trajectories in psychiatric disorders. Translational Psychiatry, 16(1), Article 522. https://doi.org/10.1038/s41398-026-04515-3

Image Credits: AI Generated

DOI: 10.1038/s41398-026-04515-3

Keywords: prosocial behavior, psychiatric disorders, anxiety, alcohol use, substance use, depression, symptom trajectories, longitudinal cohort study, Karolinska Institutet, social dysfunction, translational psychiatry, mental health