Loneliness as a Bridge Between Anxiety, Depression, and Suicidal Thinking

Anxiety and depression both raise the risk of suicidal ideation. That much has been established. What has been less clear is the pathway - what, mechanically, explains why someone whose thinking is saturated with worry or hopelessness moves toward thoughts of suicide. A study of more than 62,000 participants in the All of Us Research Program suggests loneliness is part of that explanation.

The study, published in JAMA Network Open, analyzed data from 62,685 participants and asked a straightforward mediation question: does loneliness help explain the association between anxiety symptoms and suicidal ideation, and between depressive symptoms and suicidal ideation? The answer, in both cases, was yes - partially.

What partial mediation means

Mediation analysis tries to answer whether one variable helps explain the relationship between two others. Here, the researchers found that loneliness partially mediated both relationships: some of the elevated suicidal ideation associated with anxiety symptoms, and some of that associated with depressive symptoms, appeared to operate through loneliness rather than directly.

"Partially" is an important qualifier. Loneliness does not fully explain the link between mood disorders and suicidal thinking. Anxiety and depression carry direct associations with suicidal ideation even after accounting for loneliness. But the partial mediation is meaningful, because it identifies a variable that may be targetable independently of the underlying mood disorder.

The case for a transdiagnostic target

Major depression and anxiety disorders are different diagnoses, treated with different first-line interventions - different medications, sometimes different therapeutic approaches. What the study suggests is that loneliness may sit upstream of suicidal ideation in both conditions, regardless of which diagnosis applies. That makes it what researchers call a transdiagnostic target: an intervention point that could potentially reduce suicide risk without requiring precise diagnosis first.

That framing has practical implications. Access to diagnosis and specialized mental health treatment is limited in many communities. Interventions that address loneliness directly - social connection programs, community outreach, befriending services - are already available and do not require psychiatric diagnosis to implement. If loneliness genuinely mediates part of the pathway to suicidal ideation across diagnostic categories, those interventions could reach people who never access formal mental health care.

What cross-sectional data can and cannot tell us

The All of Us Research Program is a national cohort that collects diverse health and social data from hundreds of thousands of participants, making it a rich resource for this kind of analysis. The 62,685 participants in this study represent substantial statistical power.

The fundamental limitation is the study design. Cross-sectional data captures a snapshot in time. It shows that loneliness, anxiety, depression, and suicidal ideation co-occur in a pattern consistent with mediation - but it cannot establish causation or temporal sequence. It is possible, for example, that suicidal ideation causes social withdrawal and loneliness rather than the reverse, or that a third unmeasured variable drives all three. Longitudinal studies tracking participants over time would be needed to determine whether reducing loneliness actually reduces the progression toward suicidal ideation in people with anxiety or depressive symptoms.

The finding is also correlational in nature. The mediation model explains statistical patterns in a large dataset, but whether intervening on loneliness would change clinical outcomes is a question for clinical trials, not observational research.

Why the question matters now

Rates of loneliness, particularly among young adults, rose sharply during the COVID-19 pandemic and have not fully recovered. Rates of anxiety and depression in the same populations also remain elevated. Suicidal ideation is a leading concern in adolescent and young adult mental health. These trends intersect in ways that make understanding the relationships between them practically urgent.

If loneliness is a partially independent contributor to suicidal risk in people with mood symptoms - not merely a byproduct of depression, but a variable with its own pathway - then addressing it is not just a quality-of-life intervention but a potentially protective one. The corresponding author, Katherine Musacchio Schafer, PhD, can be reached at katherine.m.schafer@vumc.org.