The largest mental health study ever conducted on researchers surveyed 138,000 PhD students and postdocs and found that nearly one in three are clinically depressed or anxious, and one in five has had suicidal thoughts
The largest study ever conducted on researcher mental health surveyed 138,000 PhD students and postdocs across 148 studies and found that nearly one in three are clinically distressed. The rates have not improved since the pandemic ended.
There is a particular kind of exhaustion that accumulates in academic research, one that is difficult to name because it does not fit neatly into the categories that medicine has built for suffering. It is not simply overwork, though the hours are long. It is not simply uncertainty, though the career path offers very little of the opposite. It arrives in the gap between how meaningful the work is supposed to feel and how it actually feels at two in the morning, when a grant application is due and the last three papers have been rejected and the position this PhD was supposed to lead to has been filled by someone who published more.
Researchers at the University of Vienna have now conducted the most comprehensive quantitative analysis ever assembled on what that experience is doing to the people inside it. Across 148 studies and four large databases, they synthesized mental health data from 138,446 PhD students and postdoctoral researchers from institutions across dozens of countries. The findings were published in Nature Human Behaviour. What they found is that the mental health of early-career researchers is not a problem that affects a particular type of person in a particular type of program in a particular country. It is a structural condition of academic research itself, and it is getting worse.
The numbers across every outcome
Nearly one in three early-career researchers, 29.9 percent by the study’s pooled estimate, reported elevated psychological distress when measured against validated clinical thresholds. That headline figure holds across a broad spectrum of specific conditions. Elevated depressive symptoms were reported by 29.8 percent. Elevated anxiety symptoms by 29.7 percent. Eating disorder symptoms by 28.3 percent. Alcohol misuse by 22.9 percent. Non-suicidal self-injury by 18.6 percent. Suicidal ideation by 18.8 percent, meaning that roughly one in five PhD students and postdocs reports having had thoughts of ending their life.
To understand what these numbers mean, consider the comparison the researchers make against age-matched peers in the general population. Population-based studies of young adults using equivalent screening thresholds consistently find depression prevalence in the range of 8 to 17 percent and anxiety in the range of 6 to 10 percent. Among early-career researchers, the equivalent figures are approximately 30 percent for both. Depression rates among academic researchers are two to three times higher than in their same-age peers outside academia. Anxiety rates are three to five times higher.
These are not marginal differences. They describe a population whose mental health, by any clinical measure, is severely compromised relative to what would be expected given their age and, in many cases, their level of privilege and opportunity.
The problem does not belong to any particular group
One of the most significant findings in the meta-analysis is what the moderator analyses did not find. The researchers tested whether the mental health burden differed across gender, academic discipline, career stage, country income level, and a range of individual protective and risk factors including social support, supervisor quality, loneliness, and job satisfaction.
Almost none of these variables made a meaningful difference. Depression and anxiety rates were essentially the same whether the researcher was a woman or a man, whether they studied psychology or engineering, whether they were in the first year of a PhD or finishing a postdoctoral fellowship, whether they were at a wealthy research university in the United States or at an institution in a lower-income country.
“The findings suggest that psychological distress among ECRs is not a phenomenon limited to specific groups,” said lead author Aljoscha Dreisoerner. “Rather, we see a broad pattern of mental health problems that is evident across many areas of academia.”
This consistency across groups carries an important implication. When a problem does not cluster around particular demographics or individual circumstances, it is not primarily a personal problem. It is a structural one. The conditions that produce this level of distress are present throughout academic research, regardless of who the researcher is or where they are working.
What those conditions actually are
The paper identifies the structural drivers with precision, not as background context but as the central explanation for why individual-level interventions have consistently failed to move the numbers.
The academic labor market has produced a situation in which the number of PhD graduates far exceeds the number of permanent academic positions available to employ them. Most early-career researchers work on short-term contracts, often renewable only once, with no institutional commitment beyond the current grant cycle. They relocate repeatedly for positions that may last one or two years before the search begins again. The financial insecurity of this period frequently extends into the mid-thirties and beyond, covering the years when most people outside academia have achieved some degree of professional stability.
Into this material precarity, the academic system has layered an increasingly metric-driven evaluation culture. Hiring committees, grant panels, and tenure decisions now rely heavily on quantifiable outputs: citation counts, h-indices, publication rates in high-impact journals, and grant funding totals. These metrics create powerful incentives toward a specific kind of research productivity that is legible to evaluation systems, while making different kinds of intellectual contribution, slower, deeper, or riskier work, structurally disadvantaged. The result is a work culture in which overwork is not merely tolerated but valorized, in which guilt about taking breaks is normalized, and in which the blurring of boundaries between professional and personal life is treated as evidence of commitment rather than as a symptom of dysfunction.
Between 30 and 50 percent of doctoral candidates leave their programs before completing a degree. For those who remain, the attrition of motivation and wellbeing is documented in this study’s numbers. The researchers describe this not as a talent pipeline with unfortunate individual losses but as a system producing widespread psychological harm at scale.
The pandemic made it worse and it did not get better
The temporal data in the meta-analysis adds a dimension that makes the findings more urgent. Before the COVID-19 pandemic, the pooled prevalence of elevated psychological distress among early-career researchers was already high, around 24.3 percent. During the pandemic period from 2020 to 2022, it rose sharply to 34.2 percent as research was disrupted, social isolation increased, and the already precarious career pathways became even more uncertain.
What has happened since is that the prevalence has not returned to pre-pandemic levels. The post-pandemic estimate in the data sits at 33.3 percent, statistically indistinguishable from the pandemic peak. The conditions that produced the increase appear to have become permanent features of the academic environment rather than temporary responses to an acute crisis.
What would actually help
The researchers are explicit that the consistent failure of individual-level factors to moderate distress rates means that individual-level interventions are unlikely to produce population-level change. Workshops on resilience, mindfulness programs, and counseling services are not without value for the individuals who use them. But they address symptoms without addressing causes, and the causes are institutional and structural.
“Structural changes are much needed,” said principal investigator Urs Nater. “These include more stable career paths, greater predictability for mid-level academic staff, and regular monitoring of mental health at universities.”
The specific reforms the researchers identify include expanding the number of permanent academic positions relative to PhD graduates, developing transparent and merit-based promotion criteria that reduce the power of individual gatekeepers over careers, establishing institutional safeguards against power abuse and discrimination, and creating career guidance infrastructure that prepares researchers for employment trajectories outside academia rather than treating non-academic careers as a form of failure.
They also call for regular, systematic monitoring of mental health within institutions, comparable to the workplace health monitoring that exists in other professional sectors, as a mechanism for accountability that would make the current situation harder to sustain without acknowledgment.
The study represents the most statistically robust account yet assembled of what the academic research environment is doing to the people working inside it. Whether that account changes anything will depend on whether the institutions and policy bodies that govern academic research treat it as documentation of an urgent structural failure or as one more entry in a literature they have been declining to act on for decades.
Source
Aljoscha Dreisoerner, Vanessa Goetz, David Frohnmayer, Ulrich S. Tran, Martin Voracek, Urs M. Nater. “Prevalence and severity of mental health problems in early-career researchers: a systematic review and meta-analysis.” Nature Human Behaviour, June 29, 2026.
DOI: 10.1038/s41562-026-02505-5
