The mental health crisis in forced displacement: 2026 statistics and trends
What the prevalence data actually shows, where services fall short, and how untreated distress compounds the cost of displacement.
TL;DR
Roughly one in five people in conflict-affected populations lives with a mental health condition at any given time, and about one in eleven with a moderate or severe one. Study-level PTSD and depression estimates among refugees commonly land between 20 and No current figure, several times general-population baselines. Service coverage does not track that burden: specialist staffing in displacement settings is minimal and mental health draws a small share of humanitarian health funding. Post-arrival stressors such as legal limbo, unemployment, and family separation predict symptom severity as strongly as war exposure, which means much of the burden is responsive to policy rather than fixed by past trauma.
Why the numbers are harder than they look
Almost every widely quoted refugee mental health figure comes from a systematic review pooling dozens of small studies, and those studies rarely agree. Prevalence estimates for PTSD in displaced populations have been published anywhere from under No current figure to over No current figure. That range is not measurement failure so much as a signal that the underlying populations, instruments, and timing differ enormously.
A screening questionnaire administered in a transit centre three weeks after flight measures something different from a clinical interview conducted in a resettlement country five years later. Both are labelled PTSD prevalence in the resulting headline. Analysts who quote a single number without the accompanying method are, in practice, quoting the study design rather than the population.
What survives across all of it is the direction and magnitude of the gap. Displaced populations carry several times the general-population rate of common mental disorders, and the elevation persists long after physical safety is re-established. That finding is stable enough to plan services around, even when the precise percentage is not.
Post-traumatic stress disorder
Depression and anxiety
Children and adolescents
Post-arrival stressors
Trauma exposure is only half the explanation
The most consequential shift in this field over the past two decades has been the move away from treating displaced mental health as a pure trauma-response problem. Early programming assumed that war exposure caused the symptoms and that trauma-focused therapy was therefore the intervention. Field evidence complicated that quickly.
Studies comparing groups with similar conflict exposure but different post-flight conditions kept finding that the post-flight conditions explained more of the variance in symptom severity. Insecure legal status, prohibition on working, overcrowded or unsafe accommodation, and separation from spouses and children are each independently associated with worse outcomes. They are also chronic rather than historical: they renew the stress response daily.
This reframes what counts as a mental health intervention. Resolving an asylum backlog, granting work rights, or reuniting a family are not adjacent social policies that happen to help. On the available evidence they operate on symptom severity at a scale comparable to clinical treatment, and they reach people who will never enter a clinic.
Where service coverage breaks down
Specialist workforce
Many low-income and conflict-affected countries operate with well under one psychiatrist per No current figure people, and camp settings often have none resident at all. Care therefore has to run through non-specialists.
Share of humanitarian budgets
Mental health and psychosocial support has historically drawn a very small fraction of humanitarian health funding, far below its share of the burden of disease in the same populations.
Continuity of care
Onward movement, camp closure, and shifting shelter arrangements break treatment courses. A six-session protocol assumes a person stays in one place for six weeks, which displacement rarely guarantees.
Language and interpretation
Assessment instruments validated in one language and culture perform unevenly elsewhere. Without trained interpreters, screening either misses cases or manufactures them.
Stigma and help-seeking
In many displaced communities psychological distress is expressed somatically, so people present with pain or fatigue rather than a mental health complaint, and never enter the relevant service pathway.
Aid worker and host-community burden
National staff and host communities absorb the same shocks with even less coverage, and are usually excluded from programme reporting entirely.
Methodology note: coverage descriptions synthesise WHO mental health atlas reporting, UNHCR operational reporting, and IASC MHPSS guidance. Camp-level provision varies widely between and within operations, so these should be read as recurring structural gaps rather than as measured coverage rates for any named site.
What has actually worked at scale
Because specialist staffing is not available and will not become available on any realistic timeline, the interventions with the strongest field evidence are the ones deliverable by trained non-specialists. Brief transdiagnostic protocols such as Problem Management Plus, group interpersonal therapy, and the WHO Self-Help Plus package have all shown measurable symptom reduction when delivered by lay counsellors with supervision.
The common design features matter more than the brand names. Sessions are few, usually five to eight. Content is transdiagnostic rather than tied to a single diagnosis, which avoids the need for diagnostic capacity that does not exist. Delivery is group-based where possible, which multiplies scarce counsellor time and adds a peer-support effect that individual sessions lack. Supervision structures are built in, because unsupervised lay delivery degrades quickly.
The persistent weakness is what happens after the protocol ends. Short interventions are evaluated at three or six months; displacement lasts years. Follow-up data thins out precisely where the policy question lives, and programmes are rarely funded long enough to generate it.
The socioeconomic cost of leaving it untreated
- Untreated depression and PTSD reduce labour-market participation, which slows economic self-reliance and extends dependence on assistance well beyond the point where physical security has been restored.
- Psychological distress in caregivers is a strong predictor of child developmental and educational outcomes, so the cost transmits across generations rather than resolving with the caregiver.
- Somatic presentations of untreated distress consume primary-care consultations and diagnostics repeatedly, which is more expensive over time than treating the underlying condition.
- Return and integration decisions taken under acute distress are less durable, contributing to secondary movement that shows up in displacement statistics without an obvious cause.
The framing that lands with finance ministries is not that mental health care is compassionate but that untreated distress is expensive. It suppresses the labour-market participation that self-reliance strategies depend on, and it does so in exactly the cohort those strategies target. Our displacement severity index weights health-system access for related reasons: severity is not only about how many people moved.
Reading the statistics responsibly
- Check the instrument before comparing rates. PCL-5, HTQ, PHQ-9, and GHQ-12 do not produce interchangeable numbers, and much of the variation between headline figures is instrument choice rather than real difference.
- Distinguish screening positives from clinical diagnoses. Most large refugee studies report screen-positive proportions, which run higher than diagnosed prevalence.
- Note the sampling frame. Camp-based convenience samples over-represent people already in contact with services; household surveys in urban displacement settings are rarer and usually smaller.
- Record time since displacement. Symptom trajectories change substantially over the first two years, so cross-sectional rates are only meaningful alongside displacement duration.
- State the comparison population explicitly. Elevated relative to host-community rates is a different claim from elevated relative to global baselines.
The humanitarian data literacy guide covers the general version of these rules, and the API integration guide sets out how to build a citable longitudinal panel when you need to combine sources.
What to watch through 2026
Three trends are worth tracking. Funding pressure across the humanitarian system is squeezing exactly the programme category that was already underfunded, and mental health support tends to be cut before food or shelter because its effects are slower to show. Protracted displacement continues to lengthen, which shifts the caseload from acute trauma response toward chronic condition management, a service model most operations are not built for. And remote or digital delivery, accelerated during the pandemic years, is now being applied in low-connectivity displacement settings where the evidence base is far thinner than the enthusiasm.
None of these change the underlying prevalence picture. They change whether the gap between burden and coverage widens or narrows, which is the number that actually matters for the people counted in it.
Frequently asked questions
What percentage of refugees have mental health problems?
WHO estimates that around one in five people in conflict-affected populations lives with a mental health condition at any given time, and about one in eleven with a moderate or severe condition. Study-level estimates for PTSD and depression among refugees commonly fall between 20 and No current figure, though figures vary widely with the screening instrument used, the language of assessment, and how long ago people were displaced.
Why are PTSD rates among refugees so much higher than in the general population?
Two mechanisms stack. Pre-flight and flight exposure includes violence, bereavement, and loss of home and livelihood at rates the general population never encounters. Post-arrival conditions then sustain the response: legal uncertainty, unemployment, overcrowded shelter, and separation from family are chronic stressors that prevent recovery. Research consistently finds these daily stressors predict symptom severity as strongly as the original war exposure.
What mental health services exist in refugee camps?
Most camp provision follows a layered model: community and family support at the base, focused non-specialist support such as Problem Management Plus or group interpersonal therapy in the middle, and referral to specialist care at the top. Because specialist staff are scarce, the middle layer delivered by trained lay counsellors carries most of the caseload. Coverage varies enormously between camps, and urban displaced populations outside camps are typically served least well.
Do refugee mental health conditions improve over time?
Partly, and unevenly. Symptoms often decline in the first years after arrival where legal status, housing, and income stabilise. Where those conditions remain unresolved, elevated rates persist for a decade or more. Improvement tracks the resolution of post-arrival stressors more closely than the passage of time on its own.
Which displaced groups carry the highest mental health burden?
Unaccompanied and separated children, survivors of torture and conflict-related sexual violence, older people separated from family support, and people with pre-existing severe mental illness who lose medication continuity. Internally displaced people are frequently worse served than cross-border refugees, because they fall outside the refugee-specific service architecture while remaining inside the country in crisis.
Where can researchers find refugee mental health data?
WHO publishes prevalence estimates for conflict-affected populations, UNHCR reports service coverage through its public operational data portal, and the IASC MHPSS Reference Group maintains assessment and monitoring guidance. Systematic reviews in the epidemiological literature remain the best source of pooled prevalence, since routine humanitarian reporting captures service contacts rather than population rates.
Related reading
- Displacement severity index, how health access is weighted against displacement scale.
- Humanitarian data literacy, interrogating prevalence figures before citing them.
- Understanding displacement, the conceptual background to the populations described here.
- WHO mental health in emergencies, prevalence estimates and the layered service model.
- IASC MHPSS Reference Group, assessment and programme guidance.