An evolutionary reorientation of the mental health sciences

More than one billion people worldwide — roughly one in seven — live with a mental health condition. The global economic cost is estimated at $2.5 trillion per year and is projected to reach $16 trillion by 2030. In low- and middle-income countries, where urbanisation and social disruption are accelerating most rapidly, more than 75% of those affected receive no treatment at all. Among young people, prevalence is rising sharply across every region for which data exist.

This report examines the state of mental health research, treatment, and policy across eleven sections. Its central argument is that dominant paradigms — including neuroscience, genetics, and pharmacology — have not, despite enormous investment, delivered the breakthroughs that were expected; and that evolutionary biology offers a powerful complementary framework that has been largely absent from mental health science, education, and clinical practice worldwide. Evolutionary thinking does not replace existing approaches — it provides the explanatory context within which their findings make sense, and points towards new directions for research into prevention and treatment that could be applicable across cultures and healthcare systems. Evolutionary theory has revolutionised our understanding of biology over the last century. It could do the same for the understanding and treatment of mental health.

The report is structured in three parts. Part I (Sections 1–2) sets out the diagnosis: the scale of the crisis and the disciplinary gap in education. Part II (Sections 3–7) assesses the state of the science of many stalled areas — psychotherapy, pharmacology, neuroscience, genetics, and epidemiology. Part III (Sections 8–11) considers applications and the future; workplace, school and community interventions, and institutional infrastructure. Each section reports the current state of affairs and follows by providing specific recommendations which arise from taking an evolutionary perspective on mental health problems.

The Foundation for Evolution and Mental Health

The Foundation for Evolution and Mental Health (FEMH) was registered as a charity in England and Wales on 10 December 2024 — the first charitable organisation in the world dedicated to advancing evolutionary approaches to mental health. Led by field-leading clinicians and researchers, with an international advisory board spanning evolutionary biology, anthropology, psychiatry, psychology, neuroscience, genetics, and public health, FEMH aims to help build the institutional infrastructure described in this report.

FEMH’s charitable objects are global in scope. The Foundation is structured to fund education, research, and clinical innovation anywhere in the world — from cross-cultural fieldwork across multiple continents to training materials adaptable for any national curriculum to open-access resources for mental health professionals everywhere. The mental health crisis is global. The evolutionary perspective — because it addresses universal human biology — can help provide solutions that are global too.

Evolution is the foundational theory of biology — and strangely, it has been almost entirely overlooked in the science and treatment of mental health. Applying it could open unprecedented advances in understanding and relieving mental suffering. The Foundation was established to fast-track that progress, and this inaugural report is our roadmap.

Dr Adam HuntUniversity of Cambridge · Founding Chair, Foundation for Evolution and Mental Health

The Mental Health Crisis — A Global Emergency

Across the world — in rich countries and poor, in cities and villages, among old and young — more people than ever before are living with diagnosable mental health conditions. The numbers are staggering: over one billion people worldwide, roughly one in every seven human beings alive today, could warrant a diagnosis. The Global Burden of Disease Study shows that the number affected rose from 655 million in 1990 to 970 million in 2019 — a 48% increase in under three decades — and the Lancet Commission on Global Mental Health estimated the global cost between 2010 and 2030 will be a staggering $16 trillion.

There is an uncomfortable fact regarding this growing crisis which justifies the calls for action threaded throughout this report: despite decades of research, hundreds of billions in investment, and the dedication of hundreds of thousands of clinicians and researchers, the crisis is not improving. Response and remission rates in usual care are often low, with most patients not responding, and therapies not improving over time. Psychiatric medication has not become much more effective than when most classes of drugs were first developed in the 1950s. Suicide rates in the United Kingdom recently rose to their highest level since 1999. The gap between the scale of the problem and our ability to address it is widening, not closing. This has been called the treatment-prevalence paradox: the amount of people in treatment for mental health problems has dramatically increased, but rather than associating with a drop in population-wide illness as would be expected when rolling out therapies more widely, at the same time there has been an increase in mental health problems.

When a problem of this magnitude resists solution for this long, it is worth asking whether we are missing something fundamental. This report highlights a single oversight more than any other: the organising principle of all life sciences — evolution by natural selection — has been almost entirely absent from how we understand, research, and treat mental health.

The Scale of the Crisis

The problem of mental health and disorder is global. In the United States, 23.4% of adults (61.5 million people) and 32.2% of young adults ages 18–25 experienced mental illness in 2024. In the United Kingdom, 22.6% of adults meet the criteria for at least one common mental health condition — up from 17.6% in 2007 — with 25.8% of 16–24-year-olds affected. In Europe, approximately 140 million people live with a mental health condition; in Australia, 21.5% of the population (4.3 million people). In low- and middle-income countries, where over 80% of affected people live, treatment gaps exceed 75%. In Brazil, mental and behavioural problems were the third leading cause of sick leave between 2007 and 2017, with mood disorders accounting for around half of all claims. In South Africa, approximately 26% of adults screen positive for probable depression, and mental health-related absenteeism costs the economy billions annually — yet only 15% of South Africans with mental illness receive treatment. In Turkey, 17% of the population face mental health issues, 3.2 million people suffer from depression, and antidepressant consumption increased by 56% over five years. In China, the age-standardised prevalence of mental disorders among children and adolescents reached 8.9% in 2021, representing over 30.8 million cases and 2.8 million disability-adjusted life years.

The economic burden is correspondingly vast. In England, the Centre for Mental Health estimates total societal costs — including lost productivity (£110bn / $149bn), reduced quality of life (£130bn / $176bn), and health and care spending (£60bn / $81bn) — at £300 billion ($405 billion) per year. A 2024 macroeconomic analysis measuring lost output through reduced consumption, constrained careers, and lower investment, put the cost to the US economy at $282 billion (1.7% of GDP). The economic burden is undeniably immense.

The convergence of these figures across very different countries is itself significant. From an evolutionary perspective, this universality is a clue: if human beings across every culture and continent are vulnerable to mental suffering at this scale, the explanation must reach past any particular society’s failures, into something more fundamental about the human mind — shaped over millions of years of evolution — and how it manifests in the modern environments in which it now operates.

Children and Young People

The youth data on mental health is perhaps the most alarming dimension of the crisis. In England, 20.3% of 8–16-year-olds had a probable mental disorder in 2023 — up from 12.5% in 2017 — and among 17–19-year-olds the figure has more than doubled to 23.3%. In the US in 2021, 29% of adolescent girls experienced a major depressive episode. As of March 2025, 255,000 children in England were on waiting lists for mental health services. A YoungMinds survey found that 26% of young people reporting they had attempted suicide while waiting.

The fact that mental disorders are so early-onset is particularly paradoxical amongst illnesses; evolutionary forces push most health problems and bodily deterioration to arise after reproductive years, when they have less impact on fertility. That young people are so regularly affected implies vulnerability arises downstream of functional processes, potentially manifesting maladaptively due to modern environments.

Recognition, Technology, and Overdiagnosis

Much of the increase undoubtedly reflects wider recognition. Diagnostic criteria have broadened, stigma has decreased, and clinicians are better trained to identify groups historically missed — girls and women for ADHD and autism, whose criteria were derived from predominantly male samples and often overlook less overt female presentations, and men for depression and anxiety, which in men more often surface as irritability, risk-taking or substance misuse than as classic low mood. Both gaps have narrowed in recent years, contributing to rising measured prevalence. But recognition alone cannot explain the speed of the rise: depressive symptoms among young people were stable from the 1960s to around 2010, then rose sharply. The role of smartphones and social media has been widely debated, but the evidence is more complex than headlines suggest. Researchers have found that digital technology use explains at most 0.4% of the variation in adolescent well-being.

An evolutionary lens reframes the question. Rather than merely asking “is social media harmful?” it asks more broadly: what evolved psychological needs does modern life meet, and which does it fail to provide? The answer inevitably goes beyond technology to raise questions of the structure of schools, workplaces, cities, and social life. It also explains why so many people meet diagnostic criteria in the first place: many of the states that psychiatry labels as disorders — anxiety, low mood, hypervigilance — are rooted in evolved responses that served adaptive functions for most of human history.

Medicalisation and simplistic diagnostic thresholds inspired by a germ-disease model of medicine have furnished science and society with inappropriate conceptualisations of the roots of mental health problems. Understanding which symptoms represent normal responses, which are over-activated or inappropriate for our environments, and which represent genuine dysfunction, is one of the most important contributions evolutionary psychiatry can make — a subject of longstanding recognition in the philosophy of medicine and psychiatry, and explored for its practical implications in advancing science in Sections 3–7.

The Missing Piece

Despite enormous investment, outcomes have plateaued. Psychotherapy recovery rates remain flat (Section 3) and few fundamentally new classes of psychiatric medication have emerged in seven decades (Section 4). The life sciences, by contrast, have been making steady progress with an organising principle since 1859: evolution by natural selection. It is the framework that makes sense of genetics, physiology, and medicine — the reason we understand antibiotic resistance, cancer, and the human immune system. As the geneticist Theodosius Dobzhansky observed, “nothing in biology makes sense except in the light of evolution.” Yet this framework has been almost entirely absent from psychiatry and psychotherapy.

This report sets out the case that integrating evolutionary thinking into mental health could be transformative, providing the foundational framework that gives existing approaches coherence, directs them more effectively, and opens avenues that current approaches cannot reach. Medicine and psychiatry should be more than biopsychosocial; they should be evobiopsychosocial. The scale of this ongoing crisis calls for a solution. The solution could be in the same theory which makes sense of the rest of biology.

The Mental Health Sciences — Trained in Half of Biology

The Educational Gap

Having laid out the scale of the crisis, this section reflects on the gap which could explain the lack of progress. Part of the problem is that the people trained to address the crisis are missing a fundamental piece of their own discipline. The organising principle of the life sciences — evolution by natural selection — is almost entirely absent from the training of mental health clinicians or scientists. The MRCPsych examination syllabus (Member of the Royal College of Psychiatrists) does not include evolutionary science, though genetics is listed as core knowledge. The US ACGME residency requirements (Accreditation Council for Graduate Medical Education) are no different. Clinical psychology training programmes teach cognitive models and neuropsychology but almost never mention the evolutionary framework that explains why humans have the emotional architecture they do. As Abed and St John-Smith observed, “most psychiatrists remain largely unaware of the relevance of evolution to mental disorder and dysfunction” — and the same is true across the mental health professions and researchers more broadly.

Evolutionary medicine, by contrast, has built substantial momentum as a research field. Since 1991 it has produced nearly 900 published articles, 181 dedicated courses across 120 US universities, and a dedicated journal under the International Society for Evolution, Medicine, and Public Health (of which FEMH’s Founding Chair Adam Hunt has served as Communications Chair since 2020). A 2026 scoping review by Brar and colleagues synthesises every education-focused paper on evolutionary medicine and evolutionary psychiatry between 1995 and 2023. Yet integration into formal medical training remains, in their words, “limited, uneven, and frequently peripheral”.

Evolutionary psychiatry is at an earlier stage of the same trajectory. The formal psychiatric case has been made: Abed and colleagues (2019) argued in the British Journal of Psychiatry that evolutionary biology is an “essential basic science for the training of the next generation of psychiatrists”. In 2017 and 2024 Randolph Nesse authored an evolutionary psychiatry chapter in the core Kaplan & Sadock’s Comprehensive Textbook of Psychiatry and in 2024 also produced a landmark review in World Psychiatry; both world-leading authoritative resources. The UK’s Evolutionary Psychiatry Special Interest Group (EPSIG) at the Royal College of Psychiatrists has grown to nearly 4,000 members, with a YouTube channel of 60+ lectures and half a million views. The intellectual foundations are in place. But — as in evolutionary medicine — none of this has yet reached the formal curricula through which mental health clinicians are trained and examined.

The Paradox at the Heart of the Mental Health Sciences

Without an evolutionary framework, clinicians lack the tools to address paradoxes that are fundamental to their daily work:

• Why are harmful mental disorders so early onset, common and heritable? Natural selection should eliminate harmful genes, yet mental disorders are both prevalent and substantially heritable. Evolutionary genetics and the inevitability of trade-offs explains why: most disorders are polygenic, shaped by hundreds of variants that may confer benefits in other traits, contexts or dosages.
• Why are many disorders associated with strengths? Autism-spectrum traits include exceptional systematising ability; hypomania confers creativity and drive; ADHD traits involve novelty-seeking and rapid decision-making. These are plausibly not simple “disorders” with incidental side effects but expressions of evolved variation — the clinical endpoint is one tail of a distribution that confers adaptive advantages across the population.
• Why are anxiety disorders so common? The smoke detector principle: natural selection calibrates defences to over-fire, because a false alarm is far less costly than missing a real threat. Anxiety over-activating is an evolved bias toward caution, rather than a design flaw.
• Why do so many disorders relate to evolutionarily novel situations? Eating disorders, addictions, gambling disorder, body dysmorphia — these involve evolved motivational systems activated by stimuli for which they were never designed. They are predictable consequences of evolved brains interacting with radically novel environments.

A well-established pedagogical scaffold for working through these questions systematically is Tinbergen’s four questions — separating mechanism, ontogeny, function, and phylogeny — which Bateson and Laland (2013) argue is particularly effective at integrating proximate and evolutionary explanations in clinical reasoning. Graves and colleagues (2016) make the complementary case that evolutionary medicine should be taught as a way of thinking about cases — a reasoning method that promotes higher-order clinical thinking — an argument that applies with particular force to the mental health sciences, whose clinical problems turn disproportionately on the suitability of behaviour and emotion.

Beyond the many implications for directing research and treatment covered later in this report, explanatory principles derived from evolutionary theory could in themselves affect clinical interactions. A clinician who understands anxiety as an evolved defence will approach an anxious patient differently from one who presumes a malfunctioning brain. A psychologist who recognises ADHD traits as part of an adaptive spectrum will frame the conversation — and the treatment — in fundamentally different terms.

Proven Pedagogical Potential

After decades of evolution-informed mental health clinicians advocating for and providing anecdotal data of the benefits of adopting an evolutionary lens, the case for integration has now been made empirically. Published in the British Journal of Psychiatry the same week as this report, a multi-site, cluster-randomised trial led by two FEMH trustees — Adam Hunt and Tom Carpenter — randomised 171 practising mental health clinicians across 19 sessions in the UK and Ireland to receive a 30-minute presentation framing anxiety through either an evolutionary or a genetic lens. Clinicians who received the evolutionary framing rated it as five times more useful for patients (OR 5.05, P_E>G = 100%) and three times more useful for clinicians (OR 3.10, P_E>G = 99.96%) compared with genetics education. They also showed greater optimism about recovery, greater anticipated willingness among patients to seek help and share diagnoses, and increased belief in the functional usefulness of negative emotions. Crucially, these effects were driven by both the positive impact of evolutionary education and the negative impact of genetic education — the standard biogenetic framing that dominates current training actively worsened several measures. These effect sizes matched or exceeded established anti-stigma interventions, which is exceptional for a single 30 minute education session. These findings also converge with a 2023 RCT by Schroder and colleagues (N = 877), which found that reframing depression as an evolved functional signal reduced self-stigma and increased patients’ belief in their capacity to recover.

Potential Pathways in Undergraduate and Postgraduate Education

A UK psychiatrist’s specialty training spans six years; we estimate it includes roughly 500 hours of formal didactic teaching across core and higher training; a US psychiatry residency runs four years with roughly 1,000 hours of protected didactic time. A clinical psychology doctorate involves roughly 850 hours. The MRCPsych written papers comprise 300 examination questions. Upstream of all this, undergraduate medical training is itself a major teaching enterprise: a UK MBBS involves an estimated ~2,500 hours of formal didactic teaching across five to six years — UCL’s MBBS BSc, for example, totals 7,600 hours of undergraduate medical training when clinical placements are included — and a US four-year MD programme is broadly comparable.

FEMH’s trustees, who include senior clinicians and researchers in the field, recommend that as few as 5–10 hours of dedicated evolutionary teaching in the foundation years could cover the core concepts of the field. Beyond this, many insights require almost no dedicated time at all: a single slide woven into an existing lecture on anxiety disorders, eating disorders, or depression — taking a few minutes to provide the evolutionary context. Across the combined undergraduate and postgraduate training pipeline of any future psychiatrist — totalling at least ~3,000 hours of formal teaching — this represents well under one half of one percent of total didactic time. For a future GP, surgeon or other non-psychiatric clinician, 5–10 hours is roughly 0.2–0.4% of their MBBS or MD alone.

Most or all of these 5–10 hours could productively be delivered at the undergraduate medical level rather than reserved for psychiatric specialty training. There is reason to be optimistic that space could be found within existing curricula. Some of the more mechanistic biological content currently taught — detailed neurochemistry and genetics for example — does not strongly inform day-to-day clinical decisions, is rarely useful to convey to patients, and likely includes many findings that have not consistently replicated and are awaiting improvements in scientific understanding (see Section 4 for further discussion). Even modestly reducing this material would create the few hours needed for evolutionary content, whilst providing the critical framing which puts other findings in context of the explanatory theory of evolution.

The vast majority of medical students will not become psychiatrists, but almost all will encounter patients with mental illness throughout their careers. Mental disorders are among the most common chronic conditions worldwide, and their interaction with physical illness is extensive: a European primary care study found that 42.5% of patients consulting their GP for any reason met criteria for a threshold or subthreshold psychiatric disorder, and worldwide estimates of mental disorder prevalence in primary care range from 25% to 60%. Yet stigma towards mental illness remains pervasive among non-psychiatric clinicians, and medical training is not associated with reduction in stigmatisation.

The undergraduate years are also a natural home for this content on intellectual grounds: medical students at the start of their training are about to absorb an enormous amount of biology — physiology, genetics, immunology, pharmacology — and evolutionary theory is the framework that ties these disciplines together. Evolution is already widely argued to be a foundational basic science for medicine, yet only 37% of UK medical schools include it in any form — and in the core curriculum at fewer than one in eight. Where evolution is taught, the content focuses overwhelmingly on infectious disease, antimicrobial resistance, cancer, ageing, diet and reproductive biology; mental health and psychiatric disorder are essentially absent. A practical entry point would be a Student Selected Module (SSM) toolkit that interested clinicians or lecturers at any medical school could adapt and deploy. FEMH could develop a ready-to-deploy package, designed from the outset to feed into a centralised research programme.

New Resources for Scaling Evolutionary Education

There is increasing reason for optimism about the immediate scalability and penetration of evolutionary education. Clinicians and the public have ever-more access to advanced AI tools, and FEMH trustees have experience testing these tools for their ability to educate on evolutionary explanations of specific disorders. Recent AI models (at time of writing, GPT 5.5; Claude Opus 4.7) do a reasonable job of summarising the evolutionary literature — offering fairly nuanced, evidence-based accounts of why particular conditions exist, their adaptive logic, and their potential mismatch with modern environments.

At present, most clinicians do not know to ask these questions, because they have never been taught that evolutionary perspectives on mental health exist. A basic grounding in the core principles — the kind that 5–10 hours of foundational teaching would provide — would change this. This means that evolutionary understanding could spread through the profession far more rapidly than has ever previously been possible: the foundational curriculum provides the knowledge of what to ask, and widely available AI tools provide the depth and personalised responses which matter for specialists and trainees at every stage. Scaling evolutionary education to clinicians around the world no longer requires training multiple generations of specialist teachers to take up positions at physical institutions; it requires basic education and encouragement to ask the right questions of pre-existing globally available AI tools.

Equally important is the access the public has to evolutionary explanations of mental health. At present, first-line public resources — NHS Every Mind Matters, Mind UK, the US National Institute of Mental Health patient pages, the Mayo Clinic explainers and their international equivalents — frame common mental disorders almost entirely in terms of biological dysfunction: brain chemistry, altered brain circuits, genetic risk, environmental stressors. Evolutionary framings of why anxiety, low mood, hypervigilance, grief, rumination and disordered eating exist — drawn from the same primary literature this report cites — are almost entirely absent from the pages through which patients first learn what is happening to them.

It is worth reflecting on how unusual the current position of the mental health sciences is within psychology more broadly. Psychologists in general accept that emotions and many aspects of human behaviour have evolutionary origins, and developmental psychology has long drawn on evolutionary thinking — most notably through John Bowlby, whose attachment theory was explicitly grounded in evolutionary biology and ethology. Yet the mental health disciplines — clinical psychology, psychiatry, counselling, psychotherapy, social work and mental health nursing — have largely held themselves apart from this background. The future FEMH is working towards is one in which evolutionary thinking is as naturally interwoven into mental health training, research and public communication as it already is in the wider psychological sciences — a shared foundation for the discipline which doesn’t replace the content, but gives it additional important context.

Psychotherapy — Reviving The Dodo

Apparent Progress, Stagnant Outcomes

Psychotherapy has become ever more widespread. In the United States, over one in five adults report receiving any mental health treatment in a given year; in England, the NHS Talking Therapies programme receives 1.8 million referrals annually. A standard 12-session course of therapy costs an average of £1,550 ($2,060) in the UK — up 34% since 2022 — and between $1,200 and $3,000 in the United States. By any measure, more people are in therapy now than at any previous point in history.

Given this expansion, one might expect substantial improvements in outcomes. The evidence suggests otherwise. Cuijpers and colleagues’ 2024 meta-analysis — the most comprehensive to date, spanning 441 trials and over 33,000 patients across eight major mental disorders — found that response rates remain modest across all approaches: 42% for major depression, 38% for PTSD and OCD, 36% for generalised anxiety, 32% for social anxiety. The field has not been short of apparent innovation — the rise of CBT in the 1990s, trauma-informed care, third-wave therapies such as mindfulness-based cognitive therapy, acceptance and commitment therapy, and dialectical behaviour therapy — yet when rigorously tested, most approaches produce outcomes that are remarkably similar. CBT’s dominance reflects the volume of RCT evidence behind it, making it easier for funders and insurers to approve; demonstrated superiority in outcomes has been harder to show. Trauma-informed care, although becoming more popular in recent years, has not been shown in rigorous RCTs to improve outcomes beyond standard care.

This equivalence is sometimes called the “Dodo Bird verdict,” after the Dodo in Lewis Carroll’s Alice in Wonderland who declares after a race that “everybody has won, and all must have prizes.” It is one of the most replicated findings in psychotherapy research. The most widely accepted partial explanation is that the majority of therapeutic benefit comes not from the specific technique but from common factors: the therapeutic alliance, empathy, and the experience of being understood. Wampold’s 2015 review suggests that specific techniques account for less than 1% of outcome variance, while the therapeutic alliance produces a moderate-to-large effect (d = 0.57). For the most prevalent conditions, it often matters less which therapy a patient receives than whether they have a competent, empathic clinician providing a structured space to talk. This is valuable — we are social primates who regulate emotion through connection and collaboration — but it is not precision medicine. As this section will argue, it is possible that the Dodo Bird’s verdict comes downstream of a matching problem — the wrong intervention for the wrong patient — more than a genuine ceiling on what psychotherapy can achieve.

One Qualitative Exception — What Precision Looks Like

There are striking exceptions to the Dodo Bird pattern. Exposure therapy for specific phobias — building on systematic desensitisation pioneered by Joseph Wolpe in the 1950s — achieves response rates of 80–90%, with 90% retaining gains at four years. Exposure and response prevention (ERP) for OCD achieves improvement in two-thirds of patients, with intensive programmes reaching response rates of up to 94%. These are qualitatively different from the ~38–42% response rates of general psychotherapy.

This stark difference can be made sense of through the lens of evolutionary theory. Fear is an evolved response, and every organism needs to learn which parts of its environment are dangerous and which are not. This is one of the most ancient and conserved learning mechanisms in biology. Exposure therapy works by directly engaging this mechanism: it teaches the brain, through repeated safe experience, that a stimulus is not dangerous. It targets a simple, well-understood evolved system with a specific intervention — and the results are powerful.

This is the template the rest of psychotherapy lacks. The Dodo Bird verdict is often taken as a fact of nature — as though no treatment could ever do better than another. Exposure therapy proves otherwise. Where the underlying problem is well understood and the intervention is matched precisely to the evolved mechanism driving it, response rates approach those of the best physical medicine. The central proposition of this section — and one of the central aspirations of FEMH — is that this template can be extended. If we better understood the evolved architecture of the human mind, we could develop a whole generation of therapies that work with that architecture as precisely as exposure therapy works with the fear system.

Two Distinct Contributions of Evolutionary Thinking

An evolutionary perspective, properly embedded in clinical practice, could transform psychotherapy along two complementary axes. The first is general, reaching every therapeutic encounter: giving clinicians a richer framework for understanding the emotions and psychological patterns they treat. The second is specific: enabling precise triage of each patient to the intervention most likely to help them. Both could be helpful. The first could affect the conversation and relationship within every therapy. The second addresses the Dodo Bird problem at its root — by optimising beyond generalist therapies to precisely match interventions to evolved drivers.

Contribution 1 — Enriching the Therapeutic Encounter

Every therapist in every modality stands to gain from a better answer to the patients’ question: why is this happening to me? A clinician equipped with an evolutionary framework can give functional answers framed in the human story, rather than pathologising ones aimed at an individual’s brokenness. The dominant biogenetic framing (“your depression is caused by a chemical imbalance”) has been shown to increase self-stigma and reduce patients’ belief in their capacity to recover. Evolutionary framing could avoid these harms. Schroder’s 2023 RCT (N = 877) found that framing depression as an evolved signal led to less self-stigma and more adaptive beliefs. The 2026 Hunt and Carpenter trial (171 clinicians, 19 UK/Ireland sessions) found evolutionary explanations were rated as five times more useful for patients and three times more useful for clinicians than genetic explanations.

The parallel with the chemical imbalance theory is instructive. Moncrieff’s 2022 umbrella review found the serotonin hypothesis had never been convincingly demonstrated — yet this scientifically questionable explanation was extraordinarily successful at encouraging medication uptake, because the explanatory framework drove the behaviour. Evolutionary explanations, if more widely known, could have an equal or greater effect on therapy engagement and behavioural change.

A patient struggling with social anxiety can be told, truthfully, that anxiety is a defence system calibrated for ancestral social stakes and that their system is over-firing because modern social pressures are not what it was designed for. A new mother with postnatal depression can be told that her distress is an alarm signal firing in response to conditions of isolation that human mothers never evolved to face. A neurodivergent patient can be told that their mind is differently configured, carrying real strengths alongside the challenges. In each case, the evolutionary lens replaces the oversimplistic “what is wrong with you?” with “what does your evolved psychology need that it isn’t getting?” and “how could you make the most of the functional aspects of your situation and overcome the dysfunctional?”

Contribution 2 — Precision: the Right Intervention for the Right Person

The second contribution evolutionary thinking offers to psychotherapy is more ambitious, with more direct potential for improving outcomes. Consider three patients who present with similar moderate-to-severe depression. Patient A is socially isolated: she has recently moved cities, works remotely, and no longer has close friends. Patient B has lost a spouse; six months on, the acute shock has not subsided into integrated grief. Patient C is neither isolated nor bereaved, but cannot stop ruminating about a missed opportunity, replaying the same mental loop for weeks. Under current arrangements, all three are likely to be referred via a single system and likely end up at the same psychotherapist. A reasonable proportion will improve, because the common factors of any decent therapy help most people. But it is improbable that the same intervention is the best match for all three.

What each patient actually needs is plausibly quite different. Patient A’s depression is a signal that an evolved need — for close, reciprocal, in-person connection — is not being met. Her problem is more ecological than cognitive. The most effective response may be a group programme, a community-based intervention, or a structured social-prescribing referral, rather than individual therapy. Meta-analyses of interventions for loneliness find that both group-based approaches and interventions targeting maladaptive social cognition produce meaningful effects. Patient B could potentially benefit from Complicated Grief Treatment, a specific protocol built on attachment theory with roots in exposure work; in head-to-head trials it is more effective for complicated grief than interpersonal therapy (51% vs 28% response). Patient C would likely benefit from Rumination-Focused CBT, which targets the process of repetitive thinking rather than the content of negative thoughts, with response rates of 81% and remission in 62% of residually depressed patients.

Current services often lack sophisticated systematic pipelines for triaging patients to the intervention whose underlying model best matches their presentation (with notable exceptions of services which direct towards exposure therapy). It is plausible that the Dodo Bird verdict, so often presented as a theoretical ceiling on psychotherapy, is in fact an artefact of this imprecision: bona fide therapies look equivalent on average because strong fits and poor fits are silently averaged together. Supporting this view, a 2021 JAMA Psychiatry RCT found that evidence-based matching of patients to therapists — without changing what those therapists actually did — produced markedly better outcomes than usual assignment (d = 0.75 on general impairment).

An evolutionary framework is uniquely suited to systematically developing this triage task because it provides a theory and framework for interpreting the core psychological needs of humans and directing towards what distressed patients’ evolved systems are responding to: social disconnection, status threat, resource scarcity, ecological mismatch, romantic needs etc. Many problems have a plausible intervention already in existence — sometimes outside mainstream psychotherapy services (e.g. if community based), or requiring a specific approach which not all therapists offer.

Existing Evolution-Informed Tools

Although direct research into the efficacy improvements to be gained by this strategy of evolution-informed triaging awaits formal testing, two existing assessment tools illustrate the questions evolution-informed clinicians can consider first when assessing an individual’s psychological distress.

The SOCIAL model and EFS are first glimpses of a future in which a patient’s initial assessment (potentially by self-implemented questionnaire) better directs the service to what kind of problem they are presenting with — a diagnosis of which evolutionary domain has been disrupted, beyond the symptom profile alone — and quickly routes them to the intervention most likely to address it. Some will go to individual therapy. Others will go to a bereavement specialist, a community group, a sleep or physical-activity intervention, a caregiving-support programme, structured rumination training, and so on. Triage of this kind would carry over the impact of therapeutic relationship, but more precisely identify the psychological and social gaps which a given person needs.

The practical gains could be substantial. Patients would more often receive interventions that match their actual problem, closing the gap between the ~38–42% response rates of generalist therapy and the 80–90% rates achieved where matching is already precise. Psychotherapy services, chronically oversubscribed, would be relieved of the patients whose problems are better addressed elsewhere. The field as a whole would finally have a coherent basis for precision therapy.

Pharmaceuticals — The 70-Year Plateau

A Toolkit Frozen in the 1950s

Most widely used psychiatric drug classes originate from discoveries made in the 1950s. While some more recent developments (such as glutamatergic agents including ketamine and esketamine) represent important advances, overall progress in identifying fundamentally new mechanisms has been limited, and usually a case of accidental discovery. Chlorpromazine, the first antipsychotic, was developed as an adjunct to surgical anaesthesia. Imipramine, the first tricyclic antidepressant, was a failed antihistamine. Lithium’s mood-stabilising properties were noticed during unrelated animal experiments. Benzodiazepines emerged from old dye chemistry revisited decades later in a search for tranquillisers. In each case, researchers stumbled on a compound that changed behaviour and then worked backwards to understand which receptors it affected.

Seventy years later, the field has not moved far from these origins. Modern antidepressants show no meaningful efficacy advantage over imipramine, first tested in 1957; their main improvement has been tolerability. The majority of antipsychotics still work via dopamine D2 receptor blockade — the same mechanism as chlorpromazine. A 2025 analysis in JAMA Network Open found that of 16 novel psychiatric drugs approved by the FDA between 2013 and 2024, 11 (69%) still targeted serotonin, dopamine, or norepinephrine — the same neurotransmitter systems identified seven decades ago.

Cipriani and colleagues’ 2018 network meta-analysis — 522 trials, 116,477 participants — confirmed that all 21 antidepressants studied were more effective than placebo. But Kirsch’s analysis of FDA submission data found the mean drug–placebo difference was just 1.8 points on the Hamilton Depression Rating Scale (a scale running 0–52), falling below the NICE (National Institute for Health and Care Excellence) threshold for clinical significance. For mild and moderate depression — the conditions for which antidepressants are most widely prescribed — approximately 82% of the drug response was duplicated by placebo. The serotonin hypothesis itself, the theoretical basis for SSRIs, was found by Moncrieff’s 2022 umbrella review to have no convincing evidential support.

The Costs

Against modest efficacy, psychiatric medications carry a side-effect burden that is often underplayed. Sexual dysfunction affects 40–65% of patients on SSRIs and SNRIs (serotonin–norepinephrine reuptake inhibitors) — for some, persisting after discontinuation. Emotional blunting — a flattening of both positive and negative emotions — is reported by 60% of antidepressant users, with similar proportions reporting they “didn’t feel like themselves.” Withdrawal effects affect 56% of patients on discontinuation, with 46% describing those symptoms as severe. For benzodiazepines, 20–45% of long-term users experience significant withdrawal, and physical dependence can develop over months. Antipsychotics carry their own distinct burden: tardive dyskinesia (involuntary, often irreversible, movement disorders) affects approximately 30% of patients on first-generation agents. Nearly 60% of psychiatric outpatient visits in the US involve prescriptions for two or more psychotropic medications, with limited evidence supporting most combinations.

Prescribing has expanded enormously: in England, antidepressant prescriptions more than tripled between 1998 and 2018 (18.4 million to 70.9 million per year). This has been driven not by more people starting treatment but by increasing duration of use — patients staying on antidepressants for years, often indefinitely, despite clinical trial evidence usually only extending to 6–12 weeks. The majority of prescriptions are written by GPs, and the infrastructure for regular medication reviews is inconsistent. Patients may begin usage during a crisis, stabilise, and continue indefinitely because no one revisits the decision.

The Industry’s Retreat

Although psychiatric drugs are more widely used than ever before, the pharmaceutical industry has largely abandoned the search for improved treatments. Over the past fifteen years, the world’s largest drug companies have systematically stopped psychiatric drug development: GlaxoSmithKline closed psychiatric labs (2010), Novartis shuttered its Basel neuroscience facility (2011), AstraZeneca discontinued and then fully exited (2012–2024), Pfizer pulled out entirely, cutting 300 research positions (2018), and Amgen ended its neuroscience programmes (2019). Although psychiatric drugs represent a global market of around $50 billion, this is now largely in generics. The science of novel drug development kept failing, reducing incentives for innovation. The probability of a psychiatric compound progressing from Phase I to regulatory approval is 6.2% — among the lowest of any major therapeutic area. The most dramatic recent illustration came in 2024, when AbbVie’s emraclidine — acquired through an $8.7 billion purchase of Cerevel Therapeutics — failed its Phase II trial. AbbVie took a $3.5 billion impairment charge and lost $40 billion in market capitalisation in a single day.

There are recent exceptions — esketamine (2019) targeting NMDA glutamate receptors for treatment-resistant depression (although ketamine has again been in use since the 1960s — this should be recognised as a newly approved psychiatric application of older pharmacology), and Cobenfy (2024), the first antipsychotic in over fifty years to work through a non-dopamine mechanism — but these remain isolated breakthroughs rather than evidence of a new paradigm.

Signal or Malfunction?

From an evolutionary perspective, the pharmaceutical approach to mental health has a fundamental problem: it has largely assumed that symptoms are expressions of dysfunction, when many can also be evolved functional responses carrying information. The real challenge is to distinguish functional signals from genuine dysfunction — and to target treatment accordingly. At present, both are too often treated as the same problem. When a patient presents with anxiety or low mood, the standard pharmacological response is to treat the symptom. But where those emotional states are evolved functional responses carrying information — anxiety signalling threat, low mood signalling that a life strategy is failing — treating the response without also reading what it is signalling is the equivalent of silencing a smoke alarm without ever checking whether there is a fire.

The analogy with pain is instructive. Pain is an evolved signal, and we have excellent painkillers. Nobody would argue that painkillers are never appropriate — they are among the most valuable tools in medicine. But no competent physician would prescribe painkillers instead of investigating the source of the pain. A patient with a broken leg needs both pain relief and a splint. A patient with appendicitis needs both pain management and surgery. Painkillers that masked a serious injury and allowed the patient to keep walking on a fracture would not be considered good medicine — they would be making the underlying problem worse by suppressing the signal that would otherwise prevent further damage. Psychiatric medication, when used without an understanding of what the symptoms are for, risks doing exactly this: relieving the signal while leaving the cause untouched, and sometimes allowing the cause to worsen. The patient feels better in the short term, but the conditions that triggered the evolved response — perhaps social isolation, a mismatched environment, an exploitative relationship — remain unchanged. In some cases, the medication may actively prevent the adaptive response: a depression that would otherwise have motivated someone to leave an exploitative job or a harmful relationship is suppressed, and the person remains in the situation causing their suffering.

This does not mean that psychiatric medication is never appropriate — just as the existence of pain does not mean we should never use painkillers. For severe depression, psychosis, and acute crises, pharmacological intervention can be life-saving. The evolutionary critique is not anti-medication; it is anti-indiscriminate-medication. The question is not “should we ever use these drugs?” but “do we understand what we are suppressing and confirming suppression is the right response?” At present, for the vast majority of prescriptions, the answer is no. But shifting the perspective on the problems of psychiatry to one informed by evolutionary biology offers two transformative contributions that could change this: better phenotyping, and better outcome measures. Each deserves detailed attention.

Cutting Through the Heterogeneity: Evolutionary Phenotyping

One of the central problems in psychiatric pharmacology is heterogeneity. A patient who presents with low mood, sleep disruption, loss of appetite, and social withdrawal receives a diagnosis of major depressive disorder. But this diagnosis tells the prescriber almost nothing about why this person is depressed — and without knowing why, there is no principled basis for choosing a treatment. The same symptom profile could reflect grief after bereavement, chronic social entrapment, inflammation-driven sickness behaviour, seasonal changes in light exposure, or genuine neurobiological dysfunction. These are categorically different problems with categorically different causes. Yet under the current diagnostic system, they all receive the same label and, in most cases, the same prescription.

Depression, from an evolutionary perspective, is not one disease. It is a generic response — much like pain — in which the organism reduces activity, withdraws from engagement, and conserves resources. Just as pain can signal a broken bone, a burn, a bacterial infection, or a torn muscle, depression can be triggered by fundamentally different circumstances involving different evolved pathways. Rantala and colleagues’ 2018 paper suggested twelve distinct subtypes of depressive episodes based on their proximate triggers and evolutionary functions, including infection-related depression, long-term stress, loneliness, traumatic experience, grief, romantic rejection, hierarchy conflict, postpartum hormonal changes, seasonal variation, chemical exposure, somatic disease, and starvation. Hagen’s critical review of evolutionary theories proposes that some forms of depression may function as social bargaining strategies in cooperative relationships, while others represent social navigation mechanisms that emerge when an individual’s major life strategies are blocked.

The same principle applies to anxiety. Social anxiety and panic disorder may both produce avoidance behaviour, but they involve fundamentally different evolved systems. Social anxiety relates to our evolved sensitivity to social hierarchy and group acceptance: humans are a deeply social species, and being excluded from the group was, for most of our evolutionary history, a death sentence. Social anxiety arises from the system that monitors the risk of exclusion and motivates appeasement, status-seeking, or withdrawal when that risk is high. Panic disorder, by contrast, seems to activate the ancient predator-defence system — the freeze–flight–fight cascade that evolved to respond to immediate physical danger. Generalised anxiety may reflect a threat-detection threshold set too low — the smoke detector principle in overdrive — where the system that evolved to err on the side of caution in genuinely dangerous environments is firing constantly in a modern world that triggers it far more often than the one in which it was calibrated. These are different systems, shaped by different selection pressures, operating through different neural circuits.

The implications for pharmacology are profound. A drug that dampens the predator-defence arousal underlying panic may be entirely wrong for someone whose social anxiety reflects a genuine mismatch between their evolved need for group belonging and their actual social circumstances. A person whose depression is triggered by chronic social entrapment — an exploitative job, a controlling relationship — does not have a simple neurochemical problem that a pill alone can fix; they have a life situation that needs to change. Medicating this person may actually prolong their suffering if it suppresses the signal that would otherwise motivate them to act.

If we could distinguish between these groups, targeting could allow drug efficacy to improve. The modest average effect sizes reported in the literature — the 1.8-point HAM-D difference, the number needed to treat of 7 — might reflect the lumping together of patients who genuinely benefit from medication with patients for whom medication is irrelevant or counterproductive. Separate and test these groups as appropriately distinct and effect sizes comparable to the 80–90% response rates of exposure therapy for phobias (Section 3) might be achievable. Recognising when complementary life changes are required would also allow a more efficient blending of therapy and pharmacy.

Precision psychiatry has been discussed for over a decade, but previous approaches have focused on genetics, neuroimaging, and biomarkers — none of them have used an evolutionary framework to guide phenotyping. The results have been disappointing, precisely because biological markers alone cannot tell you what a symptom is for. Evolutionary sub-typing offers something that genomics and brain scans cannot: a principled, theoretically grounded basis for distinguishing between patients whose suffering is a signal, patients whose defences are miscalibrated, and patients whose neurobiology is genuinely dysfunctional. This could transform psychiatric pharmacology from a one-size-fits-all enterprise into something approaching personalised medicine.

Beyond Symptom Reduction: What Should Drugs Actually Optimise For?

The second major contribution of evolutionary thinking to pharmacology concerns what we measure when we judge whether a drug works. Currently, the regulatory system for psychiatric drugs — from clinical trials to FDA and MHRA (Medicines and Healthcare products Regulatory Agency) approval — is built almost entirely around symptom reduction. A drug for depression is judged by whether it reduces scores on the Hamilton Depression Rating Scale. A drug for schizophrenia is judged by whether it reduces the frequency of hallucinations or delusions. A drug for autism is assessed by whether it diminishes “repetitive behaviours” or improves “social interaction” as measured against neurotypical norms. The assumption is that symptoms are the disease; reduce the symptoms, and you have treated the patient.

An evolutionary perspective calls this assumption into question. If many psychiatric symptoms function as evolved responses rather than simple malfunctions — or, in the case of neurodevelopmental conditions like autism, expressions of natural human cognitive variation — then symptom reduction is an inadequate and sometimes misleading measure of whether a patient is actually doing better. The real question is not “are the symptoms reduced?” but “is this person living a meaningful, functional, and fulfilling life?”

There is an irony here. The DSM-5 itself requires, for virtually every psychiatric diagnosis, that symptoms cause “clinically significant distress or impairment in social, occupational, or other important areas of functioning.” Functional impairment is baked into the very definition of a psychiatric disorder. Yet in practice — and especially in the design of drug trials — this criterion is largely ignored. Trials measure whether symptoms decrease on a rating scale, not whether the patient’s life actually improves. A patient whose HAM-D score drops from 22 to 15 is counted as a responder, regardless of whether they are now able to work or maintain relationships. The diagnostic system acknowledges that what matters is functioning; the treatment system measures only symptoms. This disconnect — between how disorders are defined and how drugs are tested — is one of the most consequential oversights in modern psychiatry.

The evidence that these are different questions is substantial. Research on recovery in psychosis has drawn a careful distinction between clinical recovery (symptom remission and functional improvement as assessed by clinicians) and personal recovery (the person’s own sense of living a meaningful life despite their condition). Leamy and colleagues’ systematic review identified five core processes in personal recovery: Connectedness, Hope, Identity, Meaning, and Empowerment (the CHIME framework). A meta-analysis of the relationship between clinical and personal recovery in schizophrenia spectrum disorders found that the two are only weakly correlated: many patients who achieve symptom remission do not experience personal recovery, and — crucially — some patients who retain residual symptoms report living rich, purposeful, and meaningful lives.

The implications for autism are equally striking. Quality-of-life research consistently finds that autistic people’s wellbeing is predicted not by symptom severity but by employment, relationships, and social support. A capabilities approach — which evaluates whether a person has the opportunities and support to live the life they value, rather than how closely they conform to neurotypical norms — has been proposed as a fundamentally more appropriate framework for understanding autistic adulthood.

Concentrating on functional outcomes rather than symptoms themselves aligns with the evolutionary perspective that many symptoms and spectrums of mental disorder are related to functional traits or systems which occasionally cause suffering or disability. It also has direct consequences for drug development and regulation. If the only endpoints regulators accept are symptom-reduction scores, then the entire pharmaceutical pipeline is optimised for producing drugs that suppress symptoms — regardless of whether symptom suppression actually makes patients’ lives better. A drug that reduced hallucination frequency by 30% but did nothing for occupational engagement, social connection, or subjective wellbeing would pass regulatory scrutiny. A drug that produced modest symptom changes but dramatically improved patients’ capacity to hold a job, maintain relationships, and experience meaning in their lives would struggle to demonstrate efficacy under current criteria.

The current system which equates symptoms with dysfunction constrains innovation. It channels billions of dollars of pharmaceutical research toward a narrow definition of success that is neither scientifically nor ethically justified. The assumption that psychiatric symptoms are the disorder — that reducing them is synonymous with helping the patient — is precisely the kind of unexamined premise that evolutionary thinking can help challenge. Very few psychiatric conditions are simple diseases in the “broken brain” sense. Most involve complex interactions between evolved psychological systems and environments that those systems were not designed for. Treating them as though symptom scores are the whole story of the disease has potentially held back the potential for innovation for decades.

This kind of reform is the direction regulators are already moving in for other specialities. The FDA’s Patient-Focused Drug Development (PFDD) programme — launched in 2012 and formalised in the 21st Century Cures Act of 2016 — requires drug developers to systematically incorporate patient-experience data into every phase of drug development, and commits the agency to issuing specific guidance on how that data should be captured and used. Alongside it, the FDA’s guidance on Patient-Reported Outcome measures and the Oncology Center of Excellence’s core patient-reported outcomes set for cancer trials have made PRO endpoints — covering pain, fatigue, nausea, physical functioning and overall quality of life — a standard component of modern oncology drug development. The regulatory machinery for centring patient-reported life outcomes already exists; it has simply not been sufficiently extended into psychiatry.

Mental health is perhaps the single strongest case for such extension. In oncology, tumour burden has a biological meaning that is largely independent of the patient’s social context. In mental health, the symptoms themselves — anxiety, low mood, rumination, hypervigilance, intrusive thoughts — can be evolved functional signals that exist across the whole population, and which do not reliably indicate dysfunction. The clinical question is never simply “how severe is this symptom?” but “is this symptom causing distress and disability for this person, in their specific social and occupational context?”. A drug that lowers symptom scores while leaving functioning unchanged does not constitute appropriate treatment.

Neuroscience — $20 Billion and Counting

The Investment

The fact that the brain is where the mind is housed has justified one of the largest research programmes in the history of science: the project to understand the human brain at the molecular and circuit level, on the premise that doing so would eventually explain and cure mental disorders. Under Thomas Insel (2002–2015), the US National Institute of Mental Health (NIMH) spent approximately $20 billion. A 2012 analysis found that roughly 72% of funds were directed at basic neuroscience, brain physiology, and genetics, and 20% at treatment research. By 2020, that imbalance had deepened further, with basic science absorbing 75% of the budget and treatment research falling to just 14%. The US BRAIN Initiative has invested more than $3 billion since 2014. The EU’s Human Brain Project spent €600 million ($700 million) in an attempt to simulate the human brain computationally.

These investments continue to this day. The Transmitter’s analysis of NIH’s own Research, Condition, and Disease Categorization (RCDC) system shows that the total amount awarded by the NIH for neuroscience-tagged grants — across all of the agency’s institutes and centres, not just the NIMH — reached $10.5 billion in 2024, more than doubling from $4.2 billion in 2008 and now representing 28.4% of all NIH extramural funding. When EU, UK, Chinese, Japanese, and other national neuroscience programmes are included, global spending on neuroscience research likely exceeds $15 billion annually.

Some of this investment has been highly productive. More is known about neural circuits, neurotransmitter systems, and brain structure than at any previous point in human history. Decades of work in rodents and primates have mapped pathways with extraordinary precision. The fear-learning circuitry that makes exposure therapy work (Section 3), the dopamine systems relevant to reward and motivation, the molecular machinery of synaptic transmission — all represent genuine scientific achievements. But when it comes to understanding common mental disorders in humans, the explanatory power and clinical returns have been weak. Insel himself provided the most candid assessment:

In his 2022 book Healing, Insel went further, arguing that mental disorders are categorically different from the conditions that conventional biomedical research was designed to solve, calling them a medical problem that requires a social solution.

The Replication Crisis in Human Neuroimaging

The promise of brain scanning technologies such as fMRI was that they would reveal the neural signatures of depression, schizophrenia, anxiety, and other conditions — that we would eventually be able to see the disorder in the brain. This has not happened. Decades of neuroimaging research comparing patients with psychiatric diagnoses to neurotypical controls have not identified reliable, disorder-specific brain pathology for any common mental health condition. There is no brain scan that can diagnose depression, schizophrenia, or anxiety disorder. The differences that are found are complex, diffuse, overlapping between conditions, and highly variable between individuals.

A landmark 2022 study in Nature by Marek and colleagues — using three of the largest neuroimaging datasets, totalling approximately 50,000 individuals — demonstrated that reproducing brain–behaviour associations requires sample sizes of thousands of participants: typically 1,500 to 3,900 depending on the phenotype. The median neuroimaging study has a sample of approximately 25. This means that the vast majority of published fMRI findings linking brain patterns to psychiatric conditions are statistically underpowered, producing inflated and unreliable effect sizes. Botvinik-Nezer and colleagues showed that when 70 independent teams analysed the same fMRI dataset, they reached substantially different conclusions depending on their analytical pipeline.

The Research Domain Criteria (RDoC) framework, launched by Insel at the NIMH in 2010 to move beyond DSM categories and organise research around neuroscience-defined dimensions, has produced few widely adopted clinical tools or diagnostic innovations that have reached patients after more than a decade. The project of finding simple brain breakages — a malfunctioning serotonin receptor, a clearly damaged circuit — that would explain common mental disorders the way a tumour explains cancer has not succeeded. Whatever is happening in the brains of people who suffer from depression, anxiety, or psychosis, it is not a straightforward malfunction. The differences are complex, interwoven with normal brain function, and variable between individuals. At this point, it has become clear that no simple brain disease will be found.

How Evolutionary Thinking Could Transform Neuroscience

Neuroscience has produced extraordinary basic science, and the brain is obviously where the mind is housed. But if we want to understand the brain in relationship to mental disorder more effectively, we need to understand what brains are for and how they evolved. Nesse and colleagues argued in a landmark 2010 paper in PNAS that evolutionary biology should be taught as a basic medical science. There have already been calls from within the field for neuroscience to move beyond its excessively mechanistic orientation — Gómez-Carrillo and Kirmayer et al. argued in The Lancet Psychiatry that it needs to go “beyond the prevailing, reductive, brain-centric models to include an understanding of the brain in context.” Hunt, St John-Smith, and Abed evobiopsychosocial model takes this further, showing how the evolutionary perspective of Tinbergen’s four questions (mechanism, development, function and phylogeny) can be used to understand and integrate our understanding of the biological, psychological, and social levels which impact mental health.

Evolutionary phenotyping in the scanner. As Sections 3 and 4 argued, DSM categories lump together fundamentally different conditions. Neuroimaging studies that compare “depressed” patients with controls are comparing two heterogeneous groups. A brain that is depressed under grief is almost certainly doing something different from a brain that is depressed under social entrapment or chronic inflammation. Again, evolutionary sub-typing could allow researchers to put the right people in the scanner — distinguishing individuals by their life events and the evolved pathways likely to be involved in their current state — and this could be precisely what is needed to produce the reproducible findings that have so far eluded the field.

Drawing the line within the spectrum at the point of dysfunction. Furthermore, rather than comparing neurotypical controls to people suffering from the most serious dysfunction, a more important and revealing comparison could seek to compare individuals ‘on the spectrum’ or experiencing functional manifestations of a set of symptoms from the individuals experiencing the more severe disability (see box below).

Better animal models through phylogenetics. An evolutionary lens adds unique usefulness to the question of which animal models are genuinely informative. Animal models are useful precisely because and to the extent that we share the system with the animal in question. Anxiety and depression-like responses are conserved across a wide range of mammalian species: the fight–flight–freeze system, the behavioural withdrawal of low mood, and the social defeat response are recognisable in rodents, primates, and humans, because these are ancient evolved responses with a long phylogenetic history. Animal models of anxiety and depression therefore have a strong evolutionary rationale. Conditions like schizophrenia and autism are fundamentally different. There is little good evolutionary reason to expect that language-dependent delusions, theory-of-mind disruptions, or the complex social-cognitive differences involved in autism would be present in mice. Despite this, substantial resources have been directed at mouse models of these conditions — with no fundamental advances in schizophrenia treatment since the 1980s, and still no FDA-approved drugs for the core features of autism. An evolutionary perspective suggests that finding the appropriately close relatives in the phylogenetic tree would be a more propitious route for animal models, and that this will vary between disorders.

There is an instructive parallel from the history of science. Before Darwin, biology was essentially taxonomy and mechanistic anatomy: cataloguing species, describing the structures of animals in exquisite detail, but without any framework for understanding why those structures existed or how they related to one another. The introduction of evolutionary theory did not discard any of that painstaking work — it suddenly helped it make sense. It transformed biology from a descriptive enterprise into an explanatory science. Neuroscience today is in a remarkably similar position. We have mapped many neural circuits, neurotransmitter systems, and brain structures in extraordinary detail. What we lack is the theoretical framework that explains why the brain is built the way it is, what its responses are for, and when those responses cross the line from adaptive to dysfunctional. Evolutionary theory is that framework.

Neuroscience is not short of paradigm proposals. The Research Domain Criteria sought to reorganise the field around brain circuits. Personalised medicine has called for biomarker-driven treatment. Rust calls for embracing complex systems thinking. Each captures something important. But FEMH suggests that the best reframing is not an entirely new paradigm — it is the introduction of the paradigm that already explains all of life. Evolutionary theory is the foundational paradigm of biology, yet neuroscience has been almost entirely disconnected from it, focusing overwhelmingly on the mechanistic at the expense of the functional and phylogenetic. We believe that if neuroscientists were to take seriously the fact that the brain is an evolved organ — shaped by natural selection, full of trade-offs and vulnerabilities, with adaptive responses on spectrums that have both functional and dysfunctional extremes — then there is enormous potential waiting to be unlocked.

Genetics — Thousands of Variants, Few Answers

The Promise and the Reality

The US Decade of the Brain (1990–2000) and the Human Genome Project (completed 2003) together generated extraordinary optimism that the biological causes of mental illness would soon be identified and cured. President Clinton promised that the genome would “revolutionize the diagnosis, prevention, and treatment of most, if not all, human diseases.” Francis Collins, director of the project, stated that the diagnosis and treatment of mental illness would be “transformed.” Mental health conditions had long been known to be run in families — twin studies consistently showed heritability estimates of 40–80% — and the hope was that sequencing the genome would reveal specific breakages: the psychiatric equivalents of the BRCA mutations which would be discovered for breast cancer or the CFTR gene in cystic fibrosis.

For rare conditions and de novo mutations, this has borne fruit. Gene therapy targets have been identified for conditions like Rett syndrome and certain rare forms of epilepsy, and this is a genuine achievement of modern genomics. But for the most common conditions challenging psychiatry — depression, anxiety, schizophrenia, autism, OCD — the germ model of genetic disease has not panned out. What GWAS has found instead implies that these conditions are driven by thousands of common alleles, each of tiny effect, widely distributed across the entire population. The Psychiatric Genomics Consortium has identified: 287 genomic loci associated with schizophrenia, 697 associations at 635 independent loci for major depression (from a trans-ancestry meta-analysis of nearly 700,000 cases), 298 genome-wide significant loci for bipolar disorder (many overlapping with schizophrenia), and 30 independent loci for OCD.

These findings were unexpected. Everyone carries some autism-associated alleles. Everyone carries some depression-associated alleles. The field has identified an extraordinary number of these genetic variants. What it has not done is explain why those variants persist in the human genome. It has produced a map with thousands of flags marking “risk”; what it lacks is a framework for interpreting what those flags mean — and whether the persistence of the “risk” is due to a trade-off of “reward”. Evolutionary theory provides that framework.

Why Do These Variants Persist?

Evolutionary genetics can ask the question that psychiatric genetics usually overlooks: why are these variants still here? If schizophrenia substantially reduces reproductive fitness, the genetic variants contributing to it should have been eliminated by natural selection long ago. The fact that they persist, across every human population, at stable frequencies, demands an evolutionary explanation.

The cliff-edge fitness model, formalised mathematically by Mitteroecker and Merola (2024), offers one of the most compelling explanations. Polygenic traits — traits influenced by many genes, explained in plain language as characteristics shaped not by a single gene but by the combined effect of thousands of genetic variants — like cognitive ability are under positive selection. Across the population, more of the trait is generally better. But the distribution has a threshold — a cliff edge — beyond which it becomes dysfunctional. Fitness increases steadily for 99% of the population; schizophrenia emerges in the 1% whose polygenic loading pushes them past the threshold. The variants persist because they confer advantages in the vast majority of carriers and produce dysfunction only at the extreme. Similarly, functional differences in personality or cognitive styles which represent ‘specialised minds’ may lead to psychopathology in specific individuals as a cost of having an extreme manifestation of a trait. Both these adaptive processes for functional traits would allow for the persistence of detrimental but rare phenotypes at the population level.

If this framework is correct, we should find that the same genetic variants associated with psychiatric disorders are also associated with positive traits in the broader population. Current data seems to tentatively go in this direction. Schizophrenia and bipolar risk alleles are associated with creativity and membership of artistic professions; a 2025 review concludes that schizophrenia is intimately related to specifically human brain evolution. Autism polygenic risk is positively correlated with cognitive ability and there is substantial genetic overlap between autism risk and educational attainment. This suggests that classifying these risk loci as purely associated to a disorder is simplistic. They are genes for cognitive and social capacities that happen to also produce vulnerability to more extreme and potentially harmful manifestations.

The conditions that could push genes past the threshold are varied: chronic stress, social isolation, evolutionary mismatch, prenatal infection, and even stochastic developmental processes — the random variation in gene expression and neural wiring that occurs during biological development. A key evolutionary insight is to conceive causation at the population level: genes need only to be positive across the whole population to be selected; they do not need to help every single individual. A person with high polygenic load exposed to developmental perturbation may cross the threshold into dysfunction, while another with identical genetic risk but different developmental luck can remain adaptive.

The Phenotyping Bottleneck

A genome-wide association study is only as good as the phenotyping — the characterisation of the people being studied — that defines its groups. Here, the science is less sophisticated. Most large-scale psychiatric GWAS rely on minimal phenotyping: brief questionnaires, self-reported diagnoses, or imprecise DSM categories. Cai and colleagues showed that a large proportion of the genetic signal from minimal phenotyping is not attributable to the specific disorder at all, and that a large genetic component unique to each disorder remains inaccessible to minimal phenotyping strategies. Major depressive disorder can manifest in over 10,000 symptom combinations as classified by the DSM; when Nguyen and colleagues tested depression subtypes, they found significant differences in SNP-heritability between them — direct evidence that the genetic architecture changes depending on how you define the condition. The field is correlating highly complex genetic data with inappropriately simple questionnaires, and the signal is being diluted accordingly.

Crucially, an evolutionary framework brings individual differences in personality into the genetic picture. There is a growing consensus among evolutionary researchers — including Hunt and Jaeggi, Troisi, and Del Giudice, and Nettle — that psychiatric variation is partly explained by selection for specialised cognitive and social strategies. Personality traits represent different adaptive strategies: some people are highly agreeable, cooperative, and conflict-averse; others are disagreeable, competitive, and assertive. A 2024 genome-wide analysis confirmed substantial genetic overlap between these personality traits and psychiatric conditions. The highly agreeable individual may be more vulnerable to social exploitation and subordination — a pathway to “social defeat” depression. The highly disagreeable individual may provoke conflict and exclusion — a different pathway to a different form of depression. These are not the same conditions. But current GWAS treats them identically.

An evolutionary framework can sharpen phenotyping by emphasising the task of separating stable individual differences from context-sensitive responses. Psychiatric symptoms can be widely functional reactions to environmental conditions — anxiety in the face of threat, for example. Genetic tests will not detect such universal responses. The same symptoms can also arise due to individual differences, however, which may align with the functional responses — neuroticism is a personality trait with anxious tendencies which may be an adaptive individual strategy underwritten by genetic differences. For genetic discovery, this conflation can be highly costly: it can mix individuals expressing transient, environmentally appropriate states with those exhibiting persistent, cross-context dispositions. One benefit of evolutionarily informed phenotyping is therefore to distinguish baseline tendencies from situational activation — for example: does elevated anxiety persist in the absence of identifiable threat? Are depressive symptoms tightly coupled to specific life events, or do they recur independently? Are antisocial behaviours contingent on local ecological conditions (e.g. instability, resource scarcity), or expressed across stable environments? By stratifying individuals along this axis — context-dependent versus context-independent expression — GWAS could more cleanly isolate the heritable components of underlying traits rather than the by-products of circumstance.

The data infrastructure has already been built out — hundreds of thousands of genotyped individuals in the UK Biobank, the Million Veteran Program, and All of Us, as well as large national datasets associated with health services. What has lacked precision is the associated phenotyping: for example, questions about early life events, social context, personality, environmental triggers, and functional impairment. If evolution-informed phenotyping reveals that depression, anxiety or psychosis subtypes have genuinely different genetic architectures, then polygenic scores could become more accurate, and more useful.

Environment — The Mismatch Between Modern Life and Evolved Minds

The Scale of the Problem

The previous sections have examined how neuroscience and genetics have fallen short of their promise to solve mental disorders. More so than their overlooking of the fundamental explanatory theory of biology, the biological sciences have been long criticised for a starker oversight: the impact of environments in which people actually live. Kirkbride and colleagues’ comprehensive 2024 review in World Psychiatry concluded that there is now “compelling evidence that the risk of developing any mental health condition is inextricably linked to our life circumstances, meaning that a higher burden of population-level psychiatric morbidity is disproportionately experienced by those closer to the margins of our societies.”

The evidence implicating specific environmental factors is extensive:

Poverty and inequality. Children in the lowest income quintile are 4.5 times more likely to experience severe mental health problems than those in the highest. The relationship is not simply about poverty: Marmot’s landmark Whitehall studies demonstrated that health follows a continuous social gradient — a slope that runs from top to bottom of the social hierarchy, with each step down associated with worse health outcomes. Wilkinson and Pickett have shown that rates of mental illness are two to four times higher in more unequal societies, regardless of overall wealth — and that this relationship holds across both rich and poor nations. Mind UK reports that 82% of people experiencing homelessness have mental health problems.

Childhood adversity. Adverse Childhood Experiences (ACEs) — abuse, neglect, household dysfunction — are among the strongest predictors of adult mental health outcomes. People with four or more ACEs are 3.7 times more likely to experience depression and 30 times more likely to attempt suicide. Approximately one-third of population-level psychosis is attributable to childhood adversity.

Loneliness and social isolation. Holt-Lunstad and colleagues’ landmark meta-analysis of 148 prospective studies found that strong social relationships were associated with a 50% increased likelihood of survival, and social isolation was as deadly as smoking 15 cigarettes per day. Loneliness raises the risk of depression by two to three times. In 2023, the US Surgeon General declared loneliness a public health crisis.

Urbanisation. A meta-analysis of 37 studies found that urban upbringing increased the risk of psychotic disorders by approximately 2.4 times compared with rural upbringing. Recent UK research found that growing up in dense, urban areas is associated with significant increases in non-affective psychosis, after controlling for other socioeconomic factors. Urban living is associated with elevated stress responses, reduced exposure to nature, and disrupted sleep — all biologically plausible contributors.

Physical inactivity and nature deficit. A 2023 umbrella review in the British Journal of Sports Medicine found that physical activity reduces depression by approximately 25–30%. Yet 27% of the global population is insufficiently active. Conversely, the 2-hour rule — spending at least 2 hours per week in nature — is associated with significantly better health and wellbeing.

Sleep and circadian disruption. Modern life is characterised by chronic sleep restriction and circadian misalignment — the result of artificial lighting, screen exposure, shift work, and prolonged work hours. Insufficient sleep is associated with a 2-fold increased risk of depression. A 2024 study in Nature Mental Health found that consistent bedtimes and adequate sleep duration were among the strongest behavioural predictors of mental wellbeing.

Diet and the gut microbiome. A 2024 BMJ meta-analysis of 218 RCTs found that exercise is at least as effective as antidepressants in treating depression. Mediterranean and traditional diet patterns are associated with 30% lower depression risk. Conversely, ultra-processed food consumption is associated with a 10% higher hazard of incident depressive symptoms, and the paper’s meta-analysis of six prospective cohort studies found that high versus low ultra-processed food exposure was associated with a 32% higher hazard of depressive outcomes.

Why Current Approaches Fall Short

Public health research has accumulated extensive evidence on environmental risk factors for mental illness. What it has lacked is a coherent theoretical framework explaining why these specific environmental factors have the effects they do. Why is loneliness so devastating? Why does urban upbringing raise psychosis risk? Why are highly unequal societies so much sicker than equal ones? Without an evolutionary framework, these are simply observations — empirical correlations awaiting mechanistic explanation. Public health interventions designed without an understanding of why these factors matter risk being generic, untargeted, and easy to deprioritise when budgets tighten.

The biopsychosocial model, articulated by George Engel in 1977, has long called for the integration of biological, psychological, and social levels of analysis. It has been the official framing of psychiatry and clinical psychology for decades. But in practice, the biological has dominated — pharmacology and neuroscience have absorbed the lion’s share of research funding (Sections 4–6), while the social has often been treated as a peripheral concern. The evobiopsychosocial model — proposed by Hunt, Abed, and St John-Smith — adds a fourth dimension: the evolutionary. It is the dimension that explains why each of the biological, psychological, and social levels matter the way they do, why they interact, and why specific environmental conditions cause specific kinds of mental suffering.

The Evolutionary Lens: Why These Environments Make Us Ill

An evolutionary perspective transforms the catalogue of environmental risk factors into something more coherent: a list of mismatches between the environments human psychology evolved to expect and the environments modern life actually provides. Each of the documented risk factors corresponds to a specific evolutionary mismatch — a way in which modern conditions depart from the conditions in which the human mind was shaped.

Loneliness reflects a mismatch with our evolved social ecology. Humans evolved in tightly bonded groups of 50–150 individuals, where every face was familiar and reciprocal cooperation was essential to survival. We are obligately social: our ancestors who failed to maintain group bonds did not survive. The modern atomised lifestyle — single-person households, weak community ties, transactional rather than reciprocal social interactions — is a profound departure from this baseline. The evolved systems that monitor social belonging register isolation as a crisis, because in ancestral conditions it was one. This is why loneliness is as deadly as smoking 15 cigarettes per day.

Inequality reflects a mismatch with our evolved sensitivity to social rank. Humans evolved in relatively egalitarian foraging groups; gross inequalities in resources or status were rare. The evolved systems that monitor social rank — described by Paul Gilbert’s social rank theory — operate on the assumption that low rank is genuinely dangerous, because in ancestral conditions it often was. Steep modern hierarchies, especially in highly unequal societies, chronically activate the threat system, producing depression, anxiety, and the documented health effects of low socioeconomic status. Wilkinson and Pickett’s findings on inequality are not merely a sociological observation; they are an evolutionary prediction.

Childhood adversity reflects a mismatch with our evolved expectations about caregiving. Children’s developing brains calibrate their threat systems, attachment systems, and stress responses to the environment they expect to encounter. In ancestral conditions, a stable, multi-generational, alloparental network of caregivers was the norm. Modern conditions of family fragmentation, isolated parenting, abuse and neglect activate developmental responses that are functional in genuinely dangerous environments — hypervigilance, withdrawal, distrust — but pathological in ostensibly safe ones. This is why ACEs have such powerful long-term effects on mental health.

Urban upbringing reflects multiple mismatches. The modern city departs from ancestral conditions in many simultaneous ways: reduced exposure to nature, reduced physical activity, reduced sunlight, increased exposure to strangers, increased noise, reduced sleep quality, weaker community ties. Each of these is an evolutionary mismatch in its own right; in combination they may explain the 2.4-fold increase in psychosis risk.

Physical inactivity, sleep disruption, and dietary change all reflect mismatches with the lifestyle conditions human bodies evolved under. Hunter-gatherers walked 6–16 km per day, slept in synchrony with circadian cycles, and ate diverse, minimally processed foods. The modern sedentary, indoor, ultra-processed lifestyle violates each of these conditions. The remarkable finding that exercise is as effective as antidepressants for depression is not surprising from an evolutionary perspective: it is a return to baseline conditions for the human body.

Bringing the Sciences Together

This section completes the scientific foundation of the report. Sections 3–6 examined how pharmacology, neuroscience, and genetics have each fallen short of their promise to improve mental health outcomes. This section has shown that the environmental and social determinants of mental health — the conditions in which people actually live — are among the most powerful predictors of who suffers and who does not, yet they remain poorly integrated into the biomedical sciences.

The pattern across all of these fields is the same: each has been developing with very little fundamental understanding of biological function and evolution. Pharmacology asks what neurotransmitter to modulate, but not what the neurotransmitter system evolved to do. Neuroscience maps brain circuits, but not why they are built the way they are. Genetics identifies risk variants, but not why they persist. Public health documents environmental risk factors, but not why those specific environments have the effects they do. Each field has been working with mechanism and correlation, but without the interpretive framework that would connect them.

The evobiopsychosocial model proposed by Hunt, Abed, and St-John Smith provides a medical formalisation of the evolutionary framework for understanding causation which has been so impactful in the rest of the biological sciences. By integrating Tinbergen’s four questions — function, phylogeny, mechanism, and development — with the biological, psychological, and social levels of analysis, evolutionary theory provides the bedrock principle that can combine each of the sciences and make clear where each is useful.

Workplace Mental Health — The Modern Tribe

The Crisis in Careers

For most working adults in industrialised societies, the workplace is where they spend the majority of their productive waking hours, where they form their primary social relationships outside the family, and where their sense of status, purpose, and belonging is most directly shaped. It is, in evolutionary terms, analogous to the small cooperative groups of interdependent adults in which humans spent most of their evolutionary history — working together on a daily basis to provide for shared needs. It also makes an extraordinary number of people ill.

Globally, the WHO estimates that 12 billion working days are lost annually to depression and anxiety, costing approximately $1 trillion per year in lost productivity. In the UK, the picture is stark. Poor mental health costs employers an estimated £51 billion ($69 billion) per year — comprising £24 billion ($32 billion) in presenteeism, £20 billion ($27 billion) in staff turnover, and £7 billion ($9.5 billion) in absenteeism. In 2024/25, 964,000 workers suffered from work-related stress, depression, or anxiety, resulting in 22.1 million working days lost. Sickness absence has reached a 15-year high of 9.4 days per employee per year, with mental health now the leading cause of long-term absence. 63% of UK employees now show signs of burnout — up from 51% two years ago. One in five UK employees need to take time off work due to poor mental health. The top external stressors were poor sleep (affecting 59%) and money worries (48%).

The business case for investment is clear. Deloitte’s analysis of 26 studies found that for every £1 ($1.35) spent on workplace mental health, employers receive £4.70 ($6.35) back in improved productivity — rising to £6.30 ($8.50) for universal prevention programmes. Yet most workplace mental health initiatives remain reactive (employee assistance programmes, counselling referrals) rather than addressing the conditions that produce distress in the first place. Evolutionary thinking offers both an explanation for this crisis and a practical framework for addressing it: design workplaces that work with human nature, and the returns — in reduced suffering, improved productivity, and lower costs — could be transformative.

The Workplace Through an Evolutionary Lens

An evolutionary perspective reframes workplace mental health from a problem of individual vulnerability to a problem of environmental design. For most of human history, people lived and worked in small cooperative groups with specific structural features: shared goals, face-to-face interaction, relatively flat hierarchies, reciprocity, autonomy over daily tasks, daily feedback of success and downtime with the community. The modern workplace violates many of these conditions, with predictable consequences for mental health.

Social rank and status threat. Paul Gilbert’s social rank theory, grounded in evolutionary biology, proposes that humans possess evolved motivational systems for navigating social hierarchies — systems that involve the interplay of threat, drive, and soothing responses. In workplaces with steep hierarchies, arbitrary authority, and performance surveillance, the threat system can be chronically activated: employees experience anxiety, submission behaviours, and depression because their evolved social-rank monitoring system is responding to genuine signals of low status and limited control. The finding that presenteeism costs more than absenteeism (£24 billion / $32 billion vs £7 billion / $9.5 billion) is consistent with this: many employees are too stressed to function effectively.

Control and autonomy. Karasek’s demand–control model — one of the most replicated findings in occupational psychology — shows that the combination of high job demands and low control produces the worst mental health outcomes. From an evolutionary perspective, this makes sense. Our ancestors had substantial autonomy over their daily activities: when to forage, when to rest, organising cooperative tasks as a process of shared deliberation. The removal of this agency — through micromanagement, rigid schedules, surveillance, and performance metrics — violates a fundamental evolutionary expectation and triggers concerns over critical monitoring. The stress response that follows is the predictable consequence.

Group size and belonging. Robin Dunbar’s work on the relationship between neocortical size and group size suggests that humans can maintain stable social relationships with approximately 150 individuals — the size of a typical hunter-gatherer band. Within this, meaningful cooperative relationships operate in layers of roughly 5, 15 and 35. Large corporations, by contrast, place employees in organisations of thousands or tens of thousands, where most colleagues are strangers and the sense of belonging that characterises natural human groups is absent. There is a comparable lack of bonding and familiarity. Teams that are too large, too fluid, or too anonymous undermine the reciprocity and trust that human cooperation depends on.

Remote work and isolation. The post-pandemic shift to remote work has introduced a different mismatch. More than half of remote workers always or sometimes feel isolated and lacking in companionship. Lonely workers miss more than 5 additional workdays per year and are 5 times more likely to miss work due to stress. The paradox of remote work is that it restores autonomy (which suits our evolved psychology) while removing social connection (which it also needs). An evolutionary framework suggests that the optimal arrangement combines autonomy with regular, meaningful, face-to-face interaction in groups small enough to build genuine reciprocal relationships.

What Would an Evolution-Informed Workplace Look Like?

Workplace interventions for mental health have thus far been informed primarily by occupational psychology and generic wellbeing programmes — resilience training, mindfulness apps, employee assistance helplines. These have shown some success, but they generally treat the individual rather than the environment. An evolutionary approach implies that insights from anthropology and evolutionary psychiatry could be substantially more effective, because they address what human beings actually need from their social groups.

Anthropologists have studied the dynamics of natural human cooperative groups — groups who must work together daily to survive — for decades. This research reveals how cooperation, conflict resolution, leadership, and social bonding actually work in the environments for which our psychology evolved: small-scale, face-to-face groups with shared goals, egalitarian norms, and immediate feedback. Translating these insights into workplace design suggests several principles:

Smaller, stable teams that match the scale of natural human cooperative groups (5–15 for close working relationships, up to ~50 for a meaningful organisational unit).

Genuine autonomy over how work is done. A structural commitment to giving employees control over their daily activities, with expectations but not excessively stifling or rigid demands.

Flatter hierarchies and reciprocal leadership. In hunter-gatherer societies, leadership is earned through competence and contribution, not imposed through formal organisational structures. Workplaces could reduce conflicts and increase the sense of mutual cooperation through transparent decision-making, shared recognition, and leaders who contribute visibly.

Biophilic design: natural light, greenery, access to outdoor spaces, and reduced noise. Research shows that workplaces with strong biophilic features produce lower stress, improved cognitive performance, and 15% higher creativity and wellbeing and 6% higher productivity compared to conventional offices. Employees in biophilic workplaces take over two less sick days a year against an average of eight days. These are corrections of the sensory mismatch between evolved expectations and modern office environments.

Shared food. For hundreds of thousands of years, human groups cooperated on a daily basis because they needed to share food. Communal eating should not be conceived as a perk or a social nicety; it is one of the most ancient and universal mechanisms for building trust, reinforcing reciprocity, and signalling that everyone belongs to the same group. Shared food remains central to community life across most cultures worldwide — yet in most modern workplaces it has often been entirely eliminated, replaced by solitary lunches at desks or hurried takeaways, which evaporates the sense of bonding or direct reward. Reinstating shared meals — such as regular team lunches, and occasional business-wide feasts — would tap into one of the deepest evolved mechanisms for social bonding and could do more for team cohesion and morale and even productivity than any comparable intervention.

Workplace rituals and regular gatherings. Ancestral groups were held together by shared rituals as well as practical cooperation: regular communal activities — singing, dancing, storytelling, seasonal celebrations — that signalled group identity and reinforced the bonds of belonging. Religious communities have preserved many of these functions for centuries: weekly gatherings, shared songs, collective meals, mutual aid. Even though surveys have found a strong sense of belonging is associated with a 56% increase in job performance, 50% drop in turnover risk and 75% decrease in sick days, the secular modern workplace has almost nothing comparable. Meetings are transactional; social events are infrequent and optional. An evolutionary perspective suggests that regular, non-transactional group activities — team rituals, celebrations, creative collaborations — could provide the psychological infrastructure of belonging that human beings evolved to need and that the modern workplace conspicuously lacks. With sufficient evolutionary insight, people could come to feel that their workplace really is their tribe.

An evolutionary reframing of workplace design provides clear implementations which should be tested and expected to improve wellbeing and productivity, but should also matter for uptake. Shared meals, team events, and regular gatherings are currently treated as optional perks — the first line-items to be cut when budgets tighten. An evolutionary framing should change this calculus: these are not luxuries but necessities, more fundamental and ancient components of sustaining a cooperative human group than safe working conditions or fair pay. They may not be legally binding, but they are psychologically binding — and organisations that strip them away are dismantling the social infrastructure their people evolved to expect.

Schools and Young People — Education Against Nature

The Youth Mental Health Emergency

Across the developed world, the mental health of children and young people is deteriorating. The WHO estimates that one in seven adolescents aged 10–19 experiences a mental disorder globally. In the United States, diagnosed mental health conditions among adolescents increased by 35% between 2016 and 2023, with anxiety diagnoses rising 61%. Over 7 million US children (11.4%) now have an ADHD diagnosis — an increase of 1 million since 2016. The CDC’s 2023 Youth Risk Behavior Survey found that 40% of US high schoolers report persistent feelings of sadness or hopelessness, rising to 53% for girls. Australia’s national mental health survey found that the prevalence of mental disorders in 16–24-year-olds rose from 26% in 2007 to 39% in 2020–22 — a ~50% relative increase. In England, 20.3% of 8–16-year-olds and 23.3% of 17–19-year-olds had a probable mental disorder in 2023. Suicide is the leading cause of death for those aged 20 to 34 in the UK, accounting for almost a quarter of deaths in that age group.

The causes of this crisis are debated. Social media and screen time have attracted significant public and political attention, and there are legitimate concerns about their effects on young people. However, the scientific evidence remains contested: Orben and Przybylski’s large-scale analysis found that digital technology use explains at most 0.4% of the variation in adolescent wellbeing, and the field has not established a clear causal relationship, though some researchers argue that the effects on heavy users and adolescent girls may be more substantial.

What receives less public attention is the evidence on academic pressure and the structure of modern schooling. A YoungMinds survey found that 63% of 15–18-year-olds struggled to cope in the lead-up to GCSEs and A-levels, with 13% experiencing suicidal thoughts and 13% self-harming during exam periods. Among children aged 10–11, 56% said that SATs were the first time they really worried about their abilities. A Stanford study of 4,317 high school students found that 56% considered homework their primary source of stress, with many reporting sleep deprivation, headaches, exhaustion, and weight loss; students in high-achieving schools average over three hours of homework per night. The causes of the youth mental health crisis are almost certainly multiple and interacting — social media, academic pressure, reduced play, economic insecurity, family stress — but the contribution of the educational environment itself deserves far more scrutiny than it currently receives.

Crucially, the burden of school-related distress is not evenly distributed. An estimated one in seven children is neurodivergent — including those with ADHD (affecting 5–11% of school-age children depending on diagnostic criteria), dyslexia (5–17% of children in English-speaking populations), and autism — and these children disproportionately struggle within the structure of mainstream schooling. Nearly 60% of dyslexic children meet the criteria for at least one other mental disorder such as anxiety, and 88% of parents report their dyslexic child has poor self-esteem because of their condition. Research on school distress has found that 92.1% of affected children were neurodivergent and 83.4% were autistic, with clinically significant anxiety present in 92.5% of cases. Children with ADHD are at substantially higher risk of academic failure, bullying, and school exclusion. Pupils with SEN support or an EHC plan accounted for 52% of all permanent exclusions in 2023/24 despite being a minority of the school population, and are five times more likely to be excluded than children without special educational needs. While many children navigate school without serious difficulty, a significant minority — concentrated among the neurodivergent — experience the modern educational environment as profoundly hostile.

The system designed to help children and adolescents is being overwhelmed. In England, CAMHS referrals increased by 53% between 2019 and 2022 (from 812,000 to 1.2 million). In 2023/24, 78,577 young people waited over a year for treatment, with 34,191 waiting over two years. In the US, comparable bottlenecks exist, with families routinely reporting waits of many months to see a child psychiatrist.

School as Unnatural

Children’s evolutionary history was not one of sitting still in rows, absorbing abstract information for six hours a day, and being assessed on their ability to reproduce it under timed conditions. Throughout human evolutionary history, children have learned through free play, physical movement, social learning, exploration, and gradual apprenticeship within multi-age peer groups. In hunter-gatherer societies — the conditions under which human childhood development evolved — children spend large portions of the day in unsupervised mixed-age playgroups, where learning and play are inseparable. As psychologist Peter Gray has documented, even as recently as the 1950s the school year was shorter than today, and elementary school children had substantially more recess time than they do now. Since then, children’s free play has been in steady decline: Gray further documents that rates of anxiety and depression in young people are five to eight times what they were in the 1950s, and argues that this increase tracks directly with the decline in self-directed play and the expansion of adult-controlled time.

Modern schools demand the opposite lifestyle: sustained attention to abstract content, physical stillness, deferred gratification, compliance with fixed arbitrary schedules, and constant social comparison through grading and testing. American children spend an average of 6 hours per day in school; the US ranks sixth among OECD and partner countries for total compulsory instructional hours across primary and lower secondary. The increased academic pressure on children — earlier formal instruction, more standardised testing, longer school days, reduced break times — represents a deepening mismatch with the developmental environment that human children are adapted for. In England, break times have been reduced by an average of 45–65 minutes per week since 1995. In the US, weekly recess time has fallen by roughly 50 minutes since the No Child Left Behind Act (2001), with elementary students now averaging just 25 minutes per day; only a handful of states require schools to mandate daily recess, and 86% of teachers have at some point removed recess as a punishment for behaviour. In England, children begin formal schooling at age 5 — earlier than almost anywhere else in the world; Finland, which consistently outperforms the UK on international assessments, does not begin formal instruction until age 7. The time allocated for the one thing that most closely resembles ancestral childhood learning — unstructured social play — is being systematically cut.

The more children engage in unstructured play, the more rapidly they develop the capacity for self-regulation, social negotiation, and impulse control. Reducing play to increase academic instruction is, from an evolutionary perspective, removing a key feature of the developmental experience that builds the cognitive capacities the curriculum then demands.

Meanwhile, school-based mental health interventions have largely followed the same pattern as adult treatments: providing therapies to the most struggling individuals or targeting symptoms without addressing the structural conditions that generate them. The MYRIAD trial — the largest RCT of school-based mindfulness ever conducted (8,376 adolescents, 85 UK schools) — found that mindfulness training was no more effective than standard provision, and in students already at risk, it was associated with deteriorated wellbeing compared with controls at one-year follow-up. Teaching children relaxation techniques while maintaining the environmental conditions that generate their distress is not a solution.

Alternatives Already Exist

The recognition that mainstream schooling is poorly suited to many children is not new. For over a century, alternative educational models have arrived at strikingly similar conclusions about what children need. Montessori schools emphasise self-directed learning, mixed-age groups, hands-on materials, and freedom of movement. A meta-analysis of 32 studies found that Montessori students performed approximately one-third of a standard deviation higher than traditionally educated students on non-academic outcomes including executive function, social skills, and creativity. Steiner (Waldorf) education delays formal academic instruction, prioritises creative and practical work, and structures the day around rhythm and nature. Forest schools move learning outdoors entirely: children attending 12 weekly sessions reported feeling calmer, happier, and less bored than classroom controls. Each of these approaches arrived at similar insights that an evolutionary perspective recommends: children learn and are happiest when they can move, explore, choose, create, and interact in small groups.

For neurodivergent children, these alternatives can be transformative. Montessori environments benefit children with ADHD and autism through self-pacing, sensory-rich materials, and multi-age social learning. Forest schools provide the reduced sensory overwhelm and physical freedom that many neurodivergent children need. Homeschooling has grown substantially, driven in significant part by parents withdrawing neurodivergent children from mainstream schools that were causing harm. At a systemic level, the school choice movement in the US has accelerated rapidly: approximately 1.3 million US students now participate in private school choice programmes, a 25% rise in a single year, and several states have implemented universal school choice policies.

In the UK, the academies programme has given 83% of secondary schools nominal independence from local authorities, with theoretical freedom to depart from the National Curriculum. In practice, however, this freedom is heavily constrained (and at the time of release of this report, provisions in a bill before parliament would require them to again teach the National Curriculum). Academies are not legally required to enter students for GCSEs, but they are judged by Ofsted, ranked in league tables, and measured by ‘Progress 8’ — an accountability metric based on GCSE and equivalent results. Schools that deviate from the exam-driven curriculum risk poor inspection outcomes and reputational damage. The result is that most academies teach much the same content in much the same way as maintained schools, because the accountability system rewards exam performance above all else.

These movements nonetheless represent the largest diversification of education in either country’s recent history — driven by a widespread, if often unarticulated, recognition that one-size-fits-all schooling does not work for all children. What they lack is a coherent framework for what the alternative should look like.

The Evolutionary Case for Change

An evolutionary perspective provides exactly that framework. Education designed around the cognitive and developmental needs of evolved human children would emphasise: free play, particularly outdoors and in mixed-age groups; physical movement throughout the day; multi-age learning communities; project-based and exploratory learning over rote memorisation; later starts to formal academic instruction; reduced testing burden; closer integration with the natural environment. None of these is a radical novelty — each appears in successful alternative models or in the schooling systems of countries that consistently outperform the UK and US.

The question is whether the political and institutional will exists to redesign mainstream schooling along these lines. The evidence is increasingly hard to ignore. The youth mental health crisis is severe and worsening. CAMHS is overwhelmed. Neurodivergent children are being failed at scale. Mindfulness apps and resilience training have not solved the problem. The structural conditions that generate distress — relentless testing, constrained play, sedentary classrooms, age-segregated peer groups — are precisely the conditions that an evolutionary analysis would predict to harm children’s wellbeing. The case for redesign is no longer a matter of fringe educational philosophy; it is a matter of public health.

Community, Social Prescribing, and Evolutionary Approaches to Therapy — Reconnecting with the Environment

Community Interventions: The Evidence

Group-based, community-embedded interventions are among the most effective tools available for improving mental health. Exercise. A 2024 network meta-analysis of 218 RCTs found that walking, jogging, yoga, and strength training produce effects on depression that are at least as large as psychotherapy and antidepressants. The most effective interventions were those of moderate-to-vigorous intensity. Even a relatively modest dose — 30 minutes of walking per day — produces measurable mental health benefits.

Group drumming. A study of group drumming among mental-health service users found significant reductions in depression and anxiety, with shifts in immune markers from a pro-inflammatory to an anti-inflammatory profile — a striking example of how social-musical activity can produce measurable biological changes.

Religious and communal participation. Cohort research on weekly religious attendance shows roughly 29% lower odds of depression and a fivefold reduction in suicide. The benefits flow primarily through the communal experience — shared meals, shared songs, mutual aid, regular contact with people across generations — rather than through religious belief itself. Secular communities that recreate these social structures (volunteer groups, community choirs, sports clubs, fraternal organisations) appear to produce comparable benefits.

Nature exposure. Spending time in nature reduces stress, improves mood, and enhances cognitive performance. The 2-hour rule finds that at least two hours per week in nature is associated with significantly better mental and physical wellbeing. Forest bathing, gardening, and outdoor exercise produce overlapping benefits.

Social Prescribing: A Movement Growing at Pace

Social prescribing — referring patients from primary care to non-medical community-based interventions, supported by trained link workers — has expanded rapidly. The UK now has approximately 1.3 million referrals per year through the NHS, with over 4,000 trained link workers. Cost-effectiveness analyses suggest that social prescribing returns £2.14 to £8.56 per £1 invested, with the variation reflecting which interventions and populations are studied. The UK’s National Academy for Social Prescribing coordinates research and training, while regional networks support local implementation.

Social prescribing is now a global movement. Over 38 countries have national or regional social prescribing programmes, including Canada, Australia, Singapore, the Netherlands, Germany, Portugal, and South Korea. The WHO has formally recognised social prescribing as a tool for addressing the social determinants of health.

Beyond the UK: A Global Picture

The international growth of social prescribing reflects a broader recognition that biomedical interventions alone are insufficient for treating mental health conditions, and that community-based, social, and environmental interventions can be highly effective. Countries that have invested in social prescribing report broadly comparable benefits: reduced GP attendance, reduced antidepressant prescribing, reduced loneliness, and improved patient-reported wellbeing. The cost-effectiveness data is consistent across very different healthcare systems.

The Evolutionary Framework: Why These Interventions Work

Evolutionary science supplies the missing theoretical framework for why community interventions and social prescribing work. They partially reverse the mismatches catalogued in Section 7 — the gaps between evolved expectations and modern conditions — by restoring elements of human ecology that industrialised life has removed.

Social connection. Group activities restore the dense, repeated social contact that human psychology evolved to expect and modern atomised life has stripped away. Singing in a choir, attending a weekly running club, or volunteering at a community centre delivers something the modern social environment often does not: regular interaction with the same people, in shared activity, over time.

Nature and movement. Our stress-response systems, attention, and mood regulation evolved in natural environments. Green prescribing works because reconnection with nature is a return to baseline conditions for the human nervous system. Exercise interventions work because our bodies were built for daily physical activity — hunter-gatherers typically walked 6–16 kilometres per day. The mismatch between evolved physiology and modern sedentary, indoor life is one of the most well-documented contributors to poor mental health.

Ritual, rhythm, and synchrony. The evidence for drumming, singing, and dancing is not coincidental. Robin Dunbar’s research at Oxford has shown that active musical performance — singing, drumming, and dancing — triggers endorphin release (indexed by elevated pain thresholds), while merely listening to music does not. Crucially, synchronised exertive activity enhances this effect: it is the doing together that matters. Weinstein, Launay, Pearce, and Dunbar demonstrated that group singing elevates pain thresholds and social bonding in both small choirs and “megachoirs” of over 200 people — and that larger groups experienced a greater change in social closeness, suggesting that music may have evolved specifically to enable bonding at scales beyond face-to-face grooming. These are rhythmic, synchronised, group activities — the kind of coordinated social behaviour that activates the endogenous opioid system and Paul Gilbert’s “soothing system” of affect regulation.

Purpose and contribution. Many community activities involve contributing to something beyond oneself — volunteering, conservation, community projects. In ancestral environments, every individual’s contribution was visible and valued. The loss of meaningful social roles in modern life is itself a mismatch.

Not all novel features of modern life deepen these mismatches. A paper by Katiyar, Hunt, Orben, Chaudhary, and Jaeggi in Psychological Review (2025) argues that digital technologies can simultaneously deepen and reverse the mismatches created by industrialised life. Instant messaging platforms partially replace the casual social contact lost in atomised societies; online gaming provides shared goals and teamwork that mirror cooperative dynamics absent from many modern settings. The net effect depends on how the technology is used and in what context. The evolutionary framework does not lead to a blanket condemnation of modernity; it leads to a precise analysis of which features of modern life diverge from evolved needs, and which partially restore them.

Without this framework, social prescribing risks being a grab-bag of well-intentioned community activities — vulnerable to cuts when budgets tighten and commissioners demand evidence of mechanism. With it, social prescribing becomes a scientifically grounded programme of mismatch reduction, where every intervention is designed to restore a specific aspect of the environment for which human psychology evolved. The evolutionary perspective also lends precise predictions about the most critical elements of community activity, as well as precision about what a particular individual is most lacking (see Section 3), allowing further development and roll-out of the most effective interventions at scale.

Reducing Mismatch: Informing Existing Therapies and Inspiring New Ones

The implications of evolutionary thinking for mental health treatment extend far beyond community interventions. Every clinician who prescribes exercise for depression, recommends social connection for loneliness, or advises nature exposure for stress is implicitly practising mismatch reduction — even if they do not use the term. The argument is not that we need to replace cognitive behavioural therapy, psychodynamic therapy, or medication with something called “evolutionary therapy.” As discussed in sections 3 and 4, the proposal here is that every existing therapy could be more effective if it were informed by an understanding of why human psychology is the way it is — and that specifically designed mismatch-reduction approaches could complement them.

Randolph Nesse, the founding figure of evolutionary psychiatry, has argued throughout his career that evolutionary thinking provides “sensible explanations that support all kinds of therapy.” His 2019 book Good Reasons for Bad Feelings makes the case that understanding the evolutionary functions of negative emotions — anxiety as an evolved threat-detection system, low mood as a disengagement response to unproductive situations, guilt as a mechanism for maintaining cooperative relationships — transforms how clinicians approach these emotions. A therapist who understands that their patient’s anxiety is an evolved defence mechanism, not a disease, will look for what the response is reacting to and whether its intensity is proportionate to the actual threat — rather than treating suppression as the only goal.

There are already early examples of this being applied in practice. Paul Gilbert’s Compassion Focused Therapy (CFT) is explicitly grounded in evolutionary theory. CFT identifies three evolved affect regulation systems — threat, drive, and soothing — and argues that many mental health problems arise because modern environments over-activate the threat system while under-activating the soothing system that evolved through affiliative, caring relationships. CFT enriches cognitive and behavioural approaches by embedding them in an evolutionary understanding of how emotional regulation works.

Cezar Giosan’s Cognitive Evolutionary Therapy (CET) takes a complementary approach — one that is, in essence, an attempt to solve the mismatch problem by asking the right questions within the consulting room. Giosan, at the University of Bucharest, developed a modified form of CBT that begins with an assessment of a patient’s “evolutionary fitness” across domains that would have been relevant in ancestral environments — social status, mating, kinship, health, and group belonging. Therapy then targets the specific domains where the patient is struggling, using evolutionary insights to guide interventions. In a 2020 randomised controlled trial, CET was significantly superior to standard CBT at increasing engagement in social and enjoyable activities (d = 0.83) and reducing behavioural avoidance (d = 0.62). The evolutionary framing changed what patients did — which is potentially what matters most for long-term recovery. Giosan’s approach is effectively mismatch reduction by another name: identify which evolved needs are unmet, and help the patient meet them.

Basile and colleagues proposed in a 2021 paper in Evolution, Medicine, and Public Health that simply explaining the evolutionary mismatch narrative to patients — telling them why their body and mind respond as they do in modern environments — could itself improve treatment adherence and health behaviours. The authors frame this as a testable hypothesis, citing evidence that narrative-based patient education tends to produce greater intentions for behavioural change than factual information alone.

The Evolutionary Mismatch Lifestyle Scale (EMLS), validated in 2024, provides an empirical tool for this clinical approach. The EMLS is a 36-item scale measuring how “mismatched” a person’s lifestyle is across seven domains including diet, physical activity, social connectedness, and screen use. Across 1,901 participants, individuals with higher mismatch scores were significantly more likely to report poor physical and mental wellbeing. Gurjot Brar’s proposal for Mismatch Reduction Therapy (MRT), published in the EPSIG Substack in 2026, represents one attempt to formalise this into a structured clinical framework targeting the key domains of mismatch — sleep, exercise, social connection, and nature exposure.

The broader point is this: evolutionary thinking does not need to create a single new therapy to transform mental health treatment. It can inform every therapy that already exists. It can guide GPs in explaining to patients why exercise matters. It can help CBT therapists identify which of a patient’s thought patterns are evolved responses to genuine threats and which are misfiring in modern environments. It can help social prescribers design programmes that target the most impactful mismatches. And it can inspire specifically designed mismatch-reduction approaches — whether delivered through social prescribing, primary care, or standalone programmes — that address the root environmental causes of distress.

Understanding Causation: The Cross-Cultural Evidence Gap

If evolutionary mismatch is driving the mental health crisis, the most powerful test is to study human populations who have not yet experienced that mismatch. If depression, anxiety, and eating disorders are genuinely “diseases of modernity,” they should be rare or absent in populations whose lifestyles more closely resemble the conditions under which human psychology evolved. If they are present at similar rates, the mismatch hypothesis needs revising. If we want to know what to prescribe socially, communally, and environmentally, we need to understand what baseline human psychological wellbeing actually looks like.

The need for cross-cultural research to understand human psychology and behaviour is widely recognised, but the data is shockingly thin. In 2010, Henrich, Heine, and Norenzayan demonstrated that 96% of psychological research samples came from Western, educated, industrialised, rich, and democratic (WEIRD) societies — populations that represent just 12% of the world’s population. The United States alone provided nearly 70% of all participants. Even cross-cultural psychology, which aims to address this gap, has overwhelmingly studied urban populations in non-Western countries — people living in cities in China, India, or Brazil — rather than the small-scale, non-industrialised societies that most closely resemble the conditions of human evolutionary history. The field of anthropology, which works most closely with such populations, has done little work in systematically assessing the presentation or prevalence of mental disorders. Ethnographers have rich qualitative accounts of distress, healing, and social life in forager and horticultural societies, but very few epidemiological studies exist. For most non-industrialised societies, we simply do not know how common depression, anxiety, or psychosis is — because nobody has measured it with culturally appropriate tools.

With rapid industrialisation and incorporation of ever more humans into the global economy, the window of opportunity for studying mental health in traditional societies is rapidly closing. This evidence, once it is gone, will never be accessible, because we cannot design experimental conditions which replicate past human societies. This should urge a call for action.

For as long as humanity survives, and reflects upon our origins and nature — potentially for thousands of years into the future — the records we collect now of human life pre-industrialisation will serve as the only direct evidence. The sooner this evidence is captured, the more valuable it will be; and the window for capturing it is rapidly closing. The billions spent on brain science could be spent in any year of the future, utilising the more up-to-date tools and technologies that will inevitably be available. It will soon be impossible to gather data on traditional human societies, and once gone, it will never be possible to collect. From an evolutionary perspective on mental health, few areas of scientific investment are so vital.

Building the Field — From Movement to Discipline

A Movement Taking Shape

The previous ten sections of this report have described a global mental health crisis of extraordinary scale — over one billion people affected, trillions of dollars in economic costs, a treatment gap that leaves the majority of those suffering without adequate help — and have argued that the dominant paradigms of the past several decades, despite enormous investment, have not delivered the breakthroughs that were expected. They have also set out the case that evolutionary biology offers a powerful complementary framework: one that can explain why the human mind is vulnerable to disorder, identify the environmental mismatches that are driving rising prevalence, and point towards new and more precise approaches to prevention and treatment.

Translating the strong theoretical framework of the evolutionary sciences into a functioning discipline for improving explanation, prevention and treatment for mental health problems, however, requires people, institutions, and infrastructure. This section reports on the growing community that has formed around evolutionary approaches to mental health: from the early intellectual contributions of the 1990s to a global network which reaches nearly 4,000 clinicians and researchers, the institutions they have built, and the infrastructure that remains to be developed.

Intellectual Foundations

The application of evolutionary thinking to mental health has developed gradually over three decades. In 1994, the psychiatrist Randolph Nesse and the evolutionary biologist George C. Williams published Why We Get Sick: The New Science of Darwinian Medicine, the book that launched evolutionary medicine as a discipline and first outlined what an evolutionary approach to psychiatry might look like. For years the idea attracted interest but remained at the margins, pursued by a relatively small number of researchers without dedicated institutional platforms.

Gradually, the intellectual foundations deepened. Troisi and McGuire’s Darwinian Psychiatry (1998) was followed by Martin Brüne’s Textbook of Evolutionary Psychiatry and Psychosomatic Medicine (2008, 2nd edition 2016) which provided the first systematic clinical textbook, showing how evolutionary principles could be applied across the full range of psychiatric disorders. Nesse’s Good Reasons for Bad Feelings (2019) brought the ideas to a mainstream audience. Abed and St John-Smith’s Evolutionary Psychiatry: Current Perspectives (2022), published by Cambridge University Press in collaboration with the Royal College of Psychiatrists, assembled the most comprehensive multi-author volume in the field.

In the mid-2020s, the field’s intellectual credentials and standing as a serious scientific discipline have grown beyond dispute. Nesse’s landmark review in World Psychiatry (2023) — one of the highest-impact journals in psychiatry — reflected on its successes, limitations and directions for growth, whilst convincingly arguing that evolutionary biology should be considered a basic science for psychiatry alongside neuroscience, genetics, and psychology. In 2024, Nesse contributed a dedicated chapter on evolutionary psychiatry to the 11th edition of Kaplan and Sadock’s Comprehensive Textbook of Psychiatry — the standard reference work used by psychiatrists worldwide. The field has also matured methodologically and empirically: Hunt and Jaeggi’s DCIDE framework, published in Biological Reviews (2025), set out a systematic method for evaluating evolutionary hypotheses about mental disorders, while Hunt and Carpenter’s study in the British Journal of Psychiatry (2026) — a Q1 journal with an impact factor of 7.6 — tested how clinicians themselves respond to evolutionary explanations, opening an entirely new line of empirical inquiry.

At this point, the intellectual case for evolutionary psychiatry is well established. The challenge that remains is institutional: building the organisations, training pathways, and research infrastructure that a mature discipline requires.

Building the Community

The most significant institutional development in the field has been the growth of the Evolutionary Psychiatry Special Interest Group (EPSIG) at the Royal College of Psychiatrists. Founded in January 2016 by FEMH trustees Dr Riadh Abed and Dr Paul St John-Smith, EPSIG is the first active evolutionary psychiatry group of a national psychiatric association anywhere in the world. In its first decade, it has grown into by far the largest grouping of evolutionary psychiatrists internationally, with nearly 4,000 members. It has organised eight full-day international symposia and numerous smaller events, published 40 quarterly newsletters (accessible through www.epsig.org), and built a YouTube channel with more than 60 open-access lectures by leading evolutionary speakers from around the world — making the field accessible to a global audience far beyond formal membership.

The community has also developed internationally. The World Psychiatric Association have recognised evolutionary psychiatry as a formal section, organising symposia at world congresses and publishing an overview of the section’s activities in World Psychiatry in 2025. Irish colleagues formed their own EPSIG at the College of Psychiatrists of Ireland in 2021, producing a notable cluster of publications in the Irish Journal of Psychological Medicine and launching the Evolution and Psychiatry Substack — an informal but widely read channel. A dedicated trainee subgroup, led by FEMH trustee Dr Tom Carpenter, organised face-to-face events and focus groups for the next generation of psychiatrists — and proved a productive platform for original research, including the Hunt and Carpenter study that drew 171 clinician participants from 19 sessions across the UK and Ireland. All of this has been accomplished with next to no private or public funding, driven by goodwill and the dedication of individuals who see the promise of applying evolutionary theory to mental health problems.

The Next Generation

Alongside the established community, a younger cohort of researchers is now contributing to the field — developing new methods, pursuing new empirical questions, and bringing evolutionary thinking into areas from digital mental health to cross-cultural psychiatry. A notable concentration of this work is based at the University of Cambridge, though the network extends well beyond it.

From Early Movement to Established Discipline

The community at Cambridge and connected via the networks built by FEMH trustees over decades has the ideas, the early evidence, the clinical experience, and international outreach to develop a fully fledged programme building an evolution-informed paradigm in mental health research and treatment. What is missing is some of the institutional infrastructure and financial backing that would allow evolutionary approaches to mental health to be taught systematically, researched at scale, and translated into clinical practice worldwide. Several key gaps remain.

No dedicated academic journal. Evolutionary psychiatry papers are scattered across general psychiatry journals (World Psychiatry, British Journal of Psychiatry, Irish Journal of Psychological Medicine), evolutionary biology journals, and psychology journals. Unlike evolutionary psychology — which has several established journals — evolutionary psychiatry has no dedicated peer-reviewed outlet.

No university departments. There is no department of evolutionary psychiatry, or even a named research centre, at any university in the world. A few individual researchers pursue evolutionary questions within broader departments of psychiatry, psychology, or anthropology, but there is nowhere for undergraduate or graduate students to apply to study evolutionary psychiatry as a discipline.

No dedicated funding streams. The major mental health research funders — NIMH in the United States, the Wellcome Trust, the MRC, and the NIHR in the UK — have no specific programmes for funding evolutionary approaches. Grant review panels are typically composed of established professors in neuroscience, genetics, and pharmacology, and emerging fields that do not fit neatly within existing review structures face a well-documented structural disadvantage. The need for initial funding from private rather than mainstream funders is a pattern seen across the history of science. From its founding in 1953, and especially from the 1960s and 1970s onwards, the Howard Hughes Medical Institute became a major private funder of molecular biology at a time when the field was still treated with caution by mainstream grant bodies — work that proved foundational for modern genetics and immunology. The Simons Foundation launched its Autism Research Initiative (SFARI) in 2006 to direct strategic investment into a field that had struggled to attract sustained, large-scale support through conventional channels; SFARI’s strategic investment transformed the field globally. In each case, private philanthropy provided the activation energy that traditional funding structures could not.

The broader funding context underlines the challenge. Global spending on mental health research reached $18.5 billion across 345 funders between 2015 and 2019, but is generally underinvested in considering its impact: being only 6.1% of UK health funding despite being 21% of the UK disease burden. UK mental health research spending was approximately £124 million ($167 million) per year between 2014 and 2017 — just £9 ($12) per person affected, compared with £228 ($308) per person for cancer. Within this already underfunded landscape, evolutionary approaches receive no identifiable share at all.

FEMH trustees and advisory board members are regularly approached by graduate students who express interest in developing evolution-informed lines of research and careers in mental health. Currently there are no specific funding pools dedicated to supporting this interest. Two forms of funding would be suitable here. Firstly, establishment of a small grant fund for single projects of graduate students (between $1000 and $5000) could allow existing masters and PhD students who are enrolled on relevant programs (e.g. clinical psychology, anthropology or public health) with limited access to research funding (a common situation in the UK and Europe) to pursue evolution-informed projects. These grants could have a very high ROI if they allow students to pick existing low-hanging fruit and set them up for a career in evolution-informed research. A second relevant funding stream is of career grants for PhD and postdoc positions (on the order of $100,000–200,000), fully training the next generation of researchers in preparation for attaining permanent research positions and dedicating their careers to developing the evolutionary science and treatment of mental health conditions.

No systematic training of clinicians. Evolutionary science is not included in the MRCPsych syllabus in the UK, the ACGME requirements in the United States, or the curricula of clinical psychology training programmes anywhere. The Hunt et al. (2026) data make the opportunity clear: clinicians who are exposed to evolutionary explanations rate them as far more useful than the genetic material that is in the curriculum — but they are never exposed to them in their formal training. Pilot implementations in medical education demonstrate that the content travels: Rühli and colleagues (2016) delivered short evolutionary modules within existing Australian and Swiss medical-school curricula with measurable gains in understanding and perceived relevance, and Furtwängler and colleagues’ 2023 efforts to develop digital resources show that scalable delivery can be achieved at near-zero marginal cost. The scoping-review evidence synthesised by Brar and colleagues (2026) confirms that the barrier is adoption rather than lack of content.

Clinical trials waiting to happen. As discussed in Section 10, evolutionary-informed therapies have shown promising results in initial trials — Giosan’s Cognitive Evolutionary Therapy, Gilbert’s Compassion Focused Therapy, and mismatch reduction approaches have all produced encouraging early evidence. What is missing is the infrastructure and funding to conduct the large-scale randomised controlled trials that would be needed for NICE, WHO, or equivalent international guideline recognition. Many hypotheses are ready to test; trial infrastructure is not yet in place.