Neurotoxicity and LSD treatment: a follow-up study of 151 patients in Denmark

Larsen frames the study against the historical backdrop of LSD's introduction into psychiatry in the 1950s and its rapid uptake at research and clinical centres in Scandinavia. Early enthusiasm for LSD as an aid to psychotherapy and as a model psychosis coexisted with sporadic reports of serious adverse events, including homicide, suicide and persistent perceptual disturbances. Debate about the long-term safety of clinical LSD use persisted into the 1970s; in Denmark, preserved case material and later political and legal activity culminated in the 1986 LSD Damages Law, which allowed treated patients to apply for compensation under a reversed burden of proof (treatment deemed the cause of harm unless another cause was more likely). This paper sets out to use the case records held in the Danish State Archives to evaluate long-term outcomes after LSD treatment in Denmark. Larsen aimed to describe both potential benefits and harms decades after exposure, and specifically to seek evidence bearing on the alleged neurotoxic potential of LSD by analysing the case material of patients who had applied for compensation under the 1986 law.

The investigators analysed archival case material for 151 of 154 applicants who, under the LSD Damages Law, sought reparatory compensation for harm attributed to LSD treatment administered in the Danish public health system from 1 January 1960. The available files included application forms, medical records and case notes, specialist certificates where provided, and tribunal summaries. Access to the Danish State Archives was granted on condition of confidentiality, and the author reviewed the material twice in 2013. Data extracted in a structured manner included hospital of treatment, patient sex and age at the year of treatment, social outcome (for example, receipt of invalidity pension), documentation of information given and consent, number of LSD sessions and available dosage data. Diagnostic labels were recorded as they appeared in the case material and were mapped where appropriate to modern ICD categories (ICD-8 at the time of compensation applications; ICD-10 terminology used for defining long-term complications during the 2013 review). Long-term complications were defined a priori and recorded from the case material: 1) flashback experiences (mapped to concepts in ICD-10 and DSM-IV hallucinogen persisting perception disorder), 2) psychotic development (schizophrenia or persistent delusional disorder), 3) persistent or recurrent mood disorder (depression or bipolar disorder), 4) anxiety disorders independent of flashbacks/psychosis/depression, and 5) psychosurgery. A given complication was counted as present if it was mentioned by the applicant and confirmed in medical certificates or noted in one or more medical certificates; in a small number of cases the author judged the complication to be present despite incomplete documentation and included these. The LSD tribunal had requested psychiatric specialist certificates for most applicants; such specialist certificates were available for 111/151 patients (74%). Where certificates were absent, the author used other written responses from psychiatric departments or previously submitted specialist reports to inform case classification. The archives did not always report dosages or exact dates, so some analyses were constrained by incomplete documentation.

The 151 analysed applicants were a subset of an estimated nearly 400 patients treated with LSD in Denmark; case material for other treated patients was not preserved. Frederiksberg Hospital accounted for the largest number of treatments, and LSD had been used across psychiatric departments in Denmark. Two psychoanalytic clinics and other hospitals had treated patients with LSD but, according to the preserved records, not all treated patients sought compensation. Demographics showed a preponderance of young female patients among applicants: 25 women were aged 25 or younger at treatment compared with 14 men in the same age category, and two female patients were under 18 at the time of treatment. At the time of application for compensation, 91/151 patients (60%) were receiving an invalidity pension attributed mainly to mental disorder; about 20% of the whole group had severe social problems such as inability to work or long-term sick leave. Duration of prior mental disorder varied: 47/151 cases (31%) received LSD within the first year of their psychiatric treatment, and in 26 cases (17%) LSD had been used as the first treatment choice. At Frederiksberg, early intervention with LSD was used in 22 of 109 patients (20%). Documentation of consent was incomplete by modern standards. For 83/151 patients (55%) the records indicate that the patient consented on the basis of information given; in 65 cases (43%) information on consent is absent or unknown and three patients required persuasion. Administration details show sessions typically started in the morning and were terminated in the afternoon with benzodiazepines or chlorpromazine. Dosage data were fragmentary: dosage information was present in 55/151 cases (36%), from which typical practice could be inferred—initial doses commonly 25–50 micrograms then rapidly increased to a usual maximum of 200–250 micrograms. A minority received very high doses (pockets of cases with doses reported above 400 micrograms, up to 800 micrograms in three patients). Approximately 10% of cases included co-administration of methylphenidate, and other psychotropic drugs (meprobamate, chlorpromazine, amitriptyline, benzodiazepines) were documented as adjuncts or as means to terminate reactions. Acute outcomes, where the case material permitted evaluation (134/151 cases; 17 lacked sufficient information), were: 52/134 (39%) judged improved after LSD treatment, 34/134 (25%) unchanged, and 48/134 (36%) worsened. The records contain illustrative clinical vignettes in which some patients experienced short-term improvements in social functioning or interpersonal disinhibition but later developed severe adverse sequelae such as chronic depression or recurrent, distressing flashbacks. Long-term harms were prominent in the tribunal files. Flashbacks and recurrent perceptual disturbances—sometimes accompanied by anxiety or post-traumatic type symptoms—were frequently documented and often long-lasting. Eight patients subsequently underwent psychosurgery; in most of these the LSD treatment had been a last therapeutic option and often coincided with clinical deterioration, while some patients experienced partial improvement after surgical intervention but remained severely ill. The tribunal awarded compensation to all applicants whose documents were considered; awarded amounts ranged from 50,000 to 510,000 DKK. Twelve applicants received subsidised treatment (nine in Denmark, three abroad), and three cases were noted as still being treated at the time of record handling.

Larsen interprets the findings as indicating substantial long-term harm in this group of LSD-treated psychiatric patients. The study contrasts sharply with earlier short-term follow-ups from the 1960s that reported few complications; possible explanations explored by the author include strong selection bias in the present sample (only patients who applied for compensation were analysed), under-reporting of adverse events in early studies, differences in therapeutic setting and the absence of systematic psychotherapy accompanying drug administration in some centres. The author notes important limitations in attributing causality. The sample is restricted to applicants under a law that assumed treatment causality unless another cause was more likely, the records are often incomplete, and preserved case material represents less than half of the patients known to have been treated historically. Nonetheless, Larsen argues that certain findings—the unusually high frequency and persistence of flashbacks and anxiety—are unlikely to reflect the natural course of psychiatric disorders alone and are consistent with LSD-related persistent perceptual disturbances reported elsewhere. Mechanistically, the discussion highlights LSD’s serotoninergic and dopaminergic receptor interactions (noting action at 5-HT2A and other receptors) as background for both acute and possibly sustained effects, while acknowledging that receptor-level mechanisms and long-term neurobiological consequences remain incompletely understood. Larsen places the results in the context of renewed contemporary interest in psychedelic-assisted therapies, warning that the present findings counsel caution: the author emphasises limited knowledge about long-term harms, the apparent neurotoxic potential suggested by the archival cases, and the need for careful selection criteria and rigorous long-term follow-up if clinical research with LSD proceeds. Finally, the paper situates its conclusions historically: early LSD research seldom used randomised controlled trial designs and routinely under-emphasised adverse outcomes, limiting the ability to generalise safety claims from mid-20th-century studies. The author recommends that modern research consider these historical harms when planning future clinical work with LSD.