A 2025 randomised, waitlist-controlled trial from Royal Holloway, University of London, found that a two-week programme of daily Stoic exercises produced significantly lower depression symptoms and experiential avoidance than no intervention at all, with anxiety improving only at a one-month follow-up rather than immediately. The trial, conducted by the psychologist Georgina King as a doctoral thesis, is described in its own abstract as the first methodologically rigorous evaluation of Stoic training, distinguishing it from the much larger body of earlier Stoicism research that lacked a control group.

We are writers, not clinicians. What follows is a reading of this research, not mental health advice, and it should not be used as a substitute for treatment or a conversation with a doctor or therapist.

King’s trial split 119 participants from a non-clinical population into a Stoic training group of 56 people and a waitlist control group of 63, with the training group practising three daily Stoic exercises and reading Stoic material every other day for two weeks. Assessments were taken before the programme, immediately after, and at a one-month follow-up.

What the trial actually found, in specific numbers

The training group showed significantly lower depression symptoms and experiential avoidance, a term describing the tendency to avoid uncomfortable thoughts and feelings rather than process them, compared with the waitlist group, both immediately after the programme and a month later. Anxiety did not show a significant group difference right after the intervention, only appearing at the one-month follow-up. Using a standard measure of clinically significant improvement, 38 percent of the training group showed a clinically meaningful drop in depression symptoms immediately after the programme, and 34 percent still showed it a month later. For anxiety, 25 percent showed clinically significant improvement right after training, rising to 32 percent by the follow-up.

This is one thesis, not yet published as a peer-reviewed journal article, and the sample, while controlled, is still modest by the standards eventually needed to generalise confidently. The population studied was non-clinical, meaning participants were not selected for diagnosed depression or anxiety disorders, so the findings describe change in a general population sample rather than a clinical treatment outcome. The thesis itself calls for future research with longer interventions and an active control group, rather than treating its own findings as settled.

A much larger, but much less rigorous, body of supporting data

The King trial sits alongside a far larger, longer-running project called Stoic Week, run since 2013 by researchers including Tim LeBon, in collaboration with academic philosophers at the University of Exeter. Tens of thousands of participants worldwide have taken part in a structured week of Stoic practices, with wellbeing measures collected before and after. Reports from recent iterations describe double-digit percentage improvements in measures like life satisfaction, resilience, and reduced negative emotion.

These figures come from uncontrolled, self-selected cohorts rather than a randomised design. Participants signed up because they were already interested in Stoicism, there was no comparison group doing something else during the same week, and reported improvements in this kind of pre-post design are consistently vulnerable to simple attention and expectation effects, quite apart from anything Stoic content specifically contributes. The Stoic Week data is useful as a description of what a large, motivated population reports experiencing, not as evidence with the same weight as a controlled trial.

Why any of this looks familiar to anyone who has done therapy

Part of why formal researchers became interested in testing Stoicism this way is a documented historical connection to an already evidence-based therapy. Albert Ellis, who developed rational emotive behaviour therapy in the 1950s, a direct precursor to modern cognitive behavioural therapy, repeatedly credited the Stoic philosophers Epictetus and Marcus Aurelius as a philosophical foundation for his approach. Ellis frequently cited Epictetus’s line that people are disturbed not by events themselves but by the judgments they form about those events, which maps closely onto cognitive behavioural therapy’s central technique of identifying and examining the specific thought that produces a difficult emotion, rather than treating the triggering event as the direct cause.

This lineage is a plausible explanation for why Stoic exercises show any measurable effect at all in the King trial. It is not itself proof that ancient Stoic texts anticipated a validated therapy, and cognitive behavioural therapy’s evidence base rests on its own large separate body of clinical trials, not on its philosophical ancestry.

Where the “empathy and boundaries” framing gets ahead of the evidence

It is worth being precise about a popular framing this research does not directly test. Modern self-help writing on Stoicism often describes it as a system for balancing empathy toward others with firm personal boundaries. Neither the King trial nor the Stoic Week data specifically measured empathy or interpersonal boundary-setting as outcomes. The trial measured depression, anxiety, experiential avoidance, and rumination. Ancient Stoic ethics, meanwhile, centred on a concept of universal rational community and duty to others quite different from the modern therapeutic idea of a personal boundary, which did not exist as a concept in the texts these exercises are drawn from.

None of this means the empathy-and-boundaries framing is baseless as a personal interpretation of Stoic ideas. It means that framing is an extension made by contemporary writers, not a claim tested by the research described here, and the two should not be presented as though one confirms the other.

What this does not prove

The King trial is one doctoral thesis awaiting peer review and replication, using a non-clinical sample and a brief two-week intervention without an active control condition to rule out the effect of simply doing a structured daily practice of any kind. The Stoic Week data, while large in volume, cannot rule out expectation effects given its uncontrolled design. Anyone experiencing persistent depression, anxiety, or distress is better served by a conversation with a doctor or a licensed therapist than by adopting an ancient philosophical exercise programme on the basis of one unpublished thesis. What the research does support, read plainly, is that a specific, brief Stoic training programme produced a measurable improvement over no intervention in a small controlled trial, a finding consistent with, though not proof of, Stoicism’s documented historical relationship to an already well-established form of psychotherapy.