Mostly Harmless? Understanding the Adverse Effects of Psychotherapy

Every treatment powerful enough to heal is powerful enough to harm.

Introduction

Psychotherapy is one of the most widely practiced and empirically supported interventions in mental health care. Its benefits are well established across disorders and modalities, yet its potential to cause harm remains comparatively underexplored. Unlike medicine, which is guided by the principle primum non nocere (first, do no harm), psychotherapy has often been viewed as inherently safe. After all, how could talking be harmful?

Over the past two decades, this assumption has begun to change. Researchers have started to examine negative or adverse effects systematically, developing taxonomies, checklists, and patient-reported instruments to capture deterioration, dependency, and other unwanted consequences of treatment. The emerging literature suggests that psychotherapy, while generally effective, is not harmless. A small but consistent proportion of patients experience worsening symptoms, lasting distress, or social and occupational impairment as a direct or indirect result of therapy.

This essay reviews the conceptual and empirical development of the study of psychotherapy’s adverse effects—from early psychoanalytic observations to contemporary prevalence data—and considers what this shift means for clinical practice. To understand psychotherapy’s risks is not to diminish its value, but to approach it with the same scientific and ethical standards applied to any other form of healthcare.

Early reflections

Recognition that psychotherapy can have negative outcomes is as old as the discipline itself: Sigmund Freud, the founder of psychotherapy, thought about what he called the negative therapeutic reaction, describing patients who deteriorate as analytic insight deepens (1923/1955). He attributed this paradoxical worsening to unconscious guilt and a punitive superego that resists improvement. This implied that the mechanisms responsible for change—interpretation, confrontation, and affective insight—could generate distress or regression under certain conditions.

Karen Horney (1936/2007) questioned Freud’s view that the negative therapeutic reaction results from unconscious guilt. She argued that it often reflects injured pride—the patient’s feeling of “humiliation” when interpretations are “perceived not as help but as an attack” (p. 30). She located the reaction “between two people, one of whom tries to help, and the other who … can experience this help only as humiliation” (p. 32), thereby shifting the focus to the interaction within the analytic encounter.

Quantitative research on psychotherapy began after in 1952 Hans J. Eysenck published an influential but methodologically controversial meta-analysis that questioned the effectiveness of psychotherapeutic treatments, sparking a wave of empirical studies and marked the beginning of evidence-based psychotherapy research.

Ten years later, after reviewing psychotherapy outcome studies, Allen E. Bergin (1963) noted that both improvement and deterioration occurred within the same datasets—effective and ineffective therapists essentially canceled each other out. Bergin argued that control groups often reflected the effects of nonprofessional therapeutic conditions, which could also exist within professional psychotherapy, but were concentrated among a subset of competent therapists. The remainder, he suggested, contributed negative effects that obscured the overall impact of psychotherapy.

Classifying Adverse Effects

Despite this early awareness, systematic attention to psychotherapy’s adverse effects did not follow. More than a century after Freud described the negative therapeutic reaction, empirical research still treated harm as an afterthought. In 2014, Jonsson et al. reviewed 132 randomized controlled trials of psychological interventions and found that only 21% mentioned any form of harm monitoring and fewer than 3% defined how such events were assessed. A parallel audit by Duggan et al. (2014) of 82 NIHR-funded trials reached a similar conclusion: none of the psychological treatment studies reported adverse events in their final publications.

In recent years, efforts have increasingly focused on defining and classifying side effects in such a way that they can be systematically investigated. For example, Hoffmann, Rudolf, and Strauß (2008) proposed a three-part model that distinguishes between different negative outcomes:

• Treatment failure (absence of expected improvement),

• unwanted effects (predictable burdens arising from otherwise appropriate interventions),

• and treatment-induced harm (clinically significant deterioration caused by therapy itself).

This model clarified that negative outcomes do not necessarily imply malpractice, but rather, may represent the inherent risks of interventions that are powerful enough to produce psychological change.

Based on this framework, Linden (2013) developed one of the few structured tools for monitoring side effects in psychotherapy research: the Unwanted Events–Adverse Treatment Reactions Checklist (UE–ATR). His classification system categorizes side effects as:

• Unwanted events (any negative development during treatment),

• adverse treatment reactions (unwanted events caused by correctly delivered therapy),

• and malpractice reactions (harm resulting from incorrect or unethical practice).

It is important to emphasize that the occurrence of side effects does not equate to treatment failure. Rather, it is good practice to recognize and address such effects (Linden & Schermuly-Haupt, 2014).

How Psychotherapy Can Harm

Adverse effects in psychotherapy are best understood as unintended consequences of mechanisms that normally support change. Interventions that engage strong emotion, challenge core beliefs, or alter established behavioral patterns can also provoke destabilization or interpersonal strain when delivered without adequate containment or matched pacing. Dependency, in turn, may develop when therapeutic structure or alliance dynamics unintentionally reinforce reliance on the therapist rather than autonomy.

Empirical research and conceptual reviews suggest that adverse effects cluster into several broad domains (Linden & Schermuly-Haupt, 2014; Ladwig, Rief, & Nestoriuc, 2014; Balder, Linden, & Rose, 2024):

• Symptomatic deterioration and emotional destabilization: Temporary or sustained increases in anxiety, depressive symptoms, or trauma-related distress, often linked to intense exposure or confrontational interventions.

• Demoralization and loss of self-efficacy: Experiences of failure or hopelessness when progress stalls or therapeutic goals feel unattainable.

• Dependency and impaired autonomy: Overreliance on the therapist’s approval or guidance, sometimes persisting beyond treatment.

• Interpersonal and social disruption: Strain in family or workplace relationships, or destabilization of social roles as patients change patterns of communication and behavior.

• Cognitive or memory distortion: Unintended creation of false or exaggerated recollections through suggestive or interpretive methods.

• Boundary violations and ethical misconduct: Rare but serious breaches of professional limits, including emotional or sexual exploitation.

• Stigmatization and secondary consequences: Negative social reactions or self-stigma related to treatment participation or diagnostic labeling.

Group settings may additionally amplify interpersonal conflict or social comparison effects, contributing to dropout or temporary worsening (Schneibel et al., 2017).

Prevalence and Severity of Adverse Effects

Estimates of the frequency and severity of psychotherapy’s adverse effects vary depending on study design, population, and definition. Reviews typically report that between 5% and 20% of patients experience unwanted effects attributable to treatment (Linden & Schermuly-Haupt, 2014). Although the majority of these reactions are mild or transient, a smaller proportion involve sustained deterioration in symptoms, functioning, or well-being.

Population-based and service-level data provide converging estimates. In Austria, Leitner et al. (2012) found roughly ten percent of more than 2,000 former patients reported a clear worsening of symptoms, while about one-third described other burdens such as emotional exhaustion or dependency. Comparable figures appear in outpatient research: Balder, Linden, and Rose (2024) found that one-third of patients experienced at least one side effect, most frequently relational strain or short-term symptom aggravation.

Patient-reported inventories show similar patterns. Ladwig, Rief, and Nestoriuc (2014) captured demoralization, dependency, and stigma as common experiences, while severe violations—such as coercion or sexual misconduct—were rare but clinically significant. A representative German survey by Strauß et al. (2021) found that over half of respondents had experienced at least one negative effect, and thirteen percent reported lasting deterioration in vitality or work ability.

Large-scale service audits in the United Kingdom yielded more conservative but still notable results. In a national survey of over 14,000 participants within the NHS’s Improving Access to Psychological Therapies (IAPT) program, Crawford et al. (2016) reported that 5.2% of patients believed therapy had produced lasting negative effects. A subsequent study in secondary-care services found higher rates, with 14.1% reporting enduring harm (McQuaid et al., 2021).

Severity typically follows a skewed distribution. Most adverse effects are mild or moderate, such as transient increases in anxiety or interpersonal tension, and tend to resolve spontaneously. Persistent deterioration—defined as sustained worsening of symptoms or functioning—appears in roughly 2–6% of patients. Severe events, including suicidality or serious boundary violations, are rare (<1%) but underscore the need for systematic monitoring.

Recent evidence also challenges the assumption that distress necessarily signals progress. Verkooyen et al. (2024) found that negative therapy experiences did not predict better outcomes once covariates were controlled. Temporary discomfort may accompany meaningful change, but distress alone is not a reliable indicator of therapeutic gain.

Across the studies that directly measured the occurrence of any psychotherapy adverse effect, the pooled Bayesian estimate was approximately 45%. Restricting the analysis to adverse effects rated as severe, lasting, or harmful yields a corresponding Bayesian estimate of about 9–10%.

In this context, “Bayesian” sounds impressive, but it’s just a calculation I chose that doesn’t require confidence intervals and still sounds scientific and cool. It also treats the study results as 100% comparable, which they are not. Nevertheless, I think the estimates are useful as a rule of thumb as long as you keep in mind their wide margin of error.

Conclusion

Psychotherapy operates by modifying patterns of thought, emotion, and behavior through guided reflection, practice, and interpersonal feedback. The same processes that facilitate recovery can, depending on patient characteristics or therapeutic context, also contribute to distress or symptom worsening. Almost half of treatments result in an unwanted effect. Severe effects occur in about one out of ten cases.

Psychotherapy is therefore not risk-free: while most patients benefit from treatment, adverse effects occur in a substantial proportion of cases. As with pharmacological interventions, the ethical norm in psychotherapy should be to inform patients about potential risks and adverse effects at the beginning of treatment.

Despite growing awareness of psychotherapy’s potential adverse effects, systematic monitoring remains the exception rather than the rule. Standardized definitions, patient-reported monitoring tools, and transparent reporting procedures are still used too rarely in outcome research. To close this gap, the integration of these methods should become a gold standard for the field—and a prerequisite for publication in peer-reviewed journals.

As psychotherapists, we shouldn’t avoid discussing side affects out of concern for diminishing the public’s trust in psychotherapy. Addressing the issue openly strengthens the field in the long run by fostering transparency and accountability. If we have confidence in the profession, we should also trust that it can withstand critical examination.

Michael Linden (2013, p. 296) noted,

“The recognition of adverse treatment effects is not a sign of bad, but of good clinical practice.”

References

Abeling, N., Nestoriuc, Y., Rief, W., & Ladwig, I. (2017). Was kann in der Psychotherapie schiefgehen? Eine empirische Untersuchung negativer Therapieerfahrungen ehemaliger Patientinnen und Patienten. Verhaltenstherapie, 27(2), 107–118.

Balder, T., Linden, M., & Rose, M. (2024). Side effects in psychodynamic and cognitive behavior therapy. Journal of Contemporary Psychotherapy, 54(3), 235–244.

Berk, M., & Parker, G. (2009). The elephant on the couch: Side effects of psychotherapy. Australian and New Zealand Journal of Psychiatry, 43(9), 787–794.

Crawford, M. J., Thana, L., Farquharson, L., Palmer, L., Hancock, E., Bassett, P., … Glover, G. (2016). Patient experience of negative effects of psychological treatment: Results from a national survey. British Journal of Psychiatry, 208(3), 260–265.

Duggan, C., Parry, G., McMurran, M., Davidson, K., & Dennis, J. (2014). The recording of adverse events from psychological treatments in clinical trials: Evidence from a review of NIHR-funded trials. Trials, 15, 335.

Fazhan, M., Yin, H., Li, Y., Zhou, J., & Zhang, L. (2023). The preliminary development and psychometric properties of the Psychotherapy Side Effects Scale (PSES). Brain and Behavior, 13(5), e2885.

Foulkes, P. (2010). The therapist as a vital factor in side-effects of psychotherapy. Australian and New Zealand Journal of Psychiatry, 44(3), 283.

Freud, S. (1923/1955). The ego and the id (J. Strachey, Trans.). In J. Strachey (Ed.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 19, pp. 1–66). London: Hogarth Press.

Hoffmann, S. O., Rudolf, G., & Strauß, B. (2008). Unerwünschte und schädliche Wirkungen von Psychotherapie: Eine Übersicht mit dem Entwurf eines eigenen Modells. Psychotherapeut, 53(1), 4–16.

Horney, K. (1936/2007). The problem of the negative therapeutic reaction. Psychoanalytic Quarterly, 76(1), 27–42; discussion 43–74.

Jonsson, U., Alaie, I., Parling, T., & Arnberg, F. K. (2014). Reporting of harms in randomized controlled trials of psychological interventions for mental and behavioral disorders: A review of current practice. Contemporary Clinical Trials, 38(1), 1–8.

Ladwig, I., Rief, W., & Nestoriuc, Y. (2014). Welche Risiken und Nebenwirkungen hat Psychotherapie? Entwicklung des Inventars zur Erfassung negativer Effekte von Psychotherapie (INEP). Verhaltenstherapie, 24(4), 252–263.

Leitner, A., Märtens, M., Höfner, C., Koschier, A., Gerlich, K., Hinterwallner, H., Liegl, G., Hinterberger, G., & Schigl, B. (2012). Psychotherapie: Risiken, Nebenwirkungen und Schäden. Donau-Universität Krems, Fakultät Gesundheit und Medizin.

Linden, M. (2013). How to define, find and classify side effects in psychotherapy: From unwanted events to adverse treatment reactions. Clinical Psychology & Psychotherapy, 20(4), 286–296.

Linden, M., & Schermuly-Haupt, M.-L. (2014). Definition, assessment and rate of psychotherapy side effects. World Psychiatry, 13(3), 306–313.

Lorenz, T., Schwarzer, G., & Ladwig, I. (2020). Predictors and impact of psychotherapy side effects in young adults. Counselling and Psychotherapy Research, 20(4), 619–631.

McQuaid, A., Crawford, M. J., Killaspy, H., Bassett, P., & Wykes, T. (2021). Patient experience of lasting negative effects of psychological therapy delivered in secondary care mental health services: A national cross-sectional survey. BMC Psychiatry, 21, 578.

Moritz, S., Schilling, L., Hauschildt, M., & Lüdtke, T. (2019). PANEPS: A measure of positive and negative effects in psychotherapy. Frontiers in Psychology, 10, 347.

Rozental, A., Kottorp, A., Boettcher, J., Andersson, G., & Carlbring, P. (2016). Negative effects of psychological treatments: An exploratory factor analysis of the Negative Effects Questionnaire. PLOS ONE, 11(6), e0157503.

Schimmel, P. (2010). The elephant on the couch: Side-effects of psychotherapy. Australian and New Zealand Journal of Psychiatry, 44(5), 489.

Schneibel, R., Linden, M., & Brähler, E. (2017). Adverse events of group psychotherapy in the in-patient setting: Results of a prospective study. Acta Psychiatrica Scandinavica, 135(6), 541–550.

Strauß, B., Lutz, W., Schöttke, H., Rubel, J., & Deisenhofer, A. K. (2021). Negative effects of psychotherapy: Estimating prevalence in a random national sample. BJPsych Open, 7(6), e186.

Verkooyen, S., et al. (2024). No pain, no gain revisited: The relation between therapy-induced distress and outcome. Frontiers in Psychology, 15, 1378456.