Anaclitic LSD therapy (historical)

1. Introduction

Version followed: the London day-hospital and private-practice method of Joyce (A. J.) Martin and Pauline McCririck (c. 1954–late 1960s), as reconstructed from Martin's own papers, Grof's 1970 review, and historian Zoë Dubus (2020, 2023). Verified in this revision from full texts in the MAPS bibliography: Martin's 1957 day-hospital series, her 1962 Acta Psychotherapeutica paper on twelve homosexual men, and her 1961 case report printed in Crocket, Sandison and Walk (1963). These confirm her selection criteria, the session format, the adverse reactions she reported, and the orientation-change aims of her work with homosexual men. None of these three papers describes the bodily "fusion" holding; that account rests on Dubus and Grof. Martin's 1964 paper "L.S.D. analysis" is not used here [uncertain]. No complaint, injury, legal case or regulatory finding against Martin or McCririck is documented in the sources used [uncertain]. Historical: not for use. This tab gives no doses, routes, schedules or administration details.

What it was. "Anaclitic" (literally "leaning on") LSD therapy was a psychoanalytic form of LSD treatment developed by Joyce Martin (1905–1969), a London psychiatrist and Freudian psychoanalyst, with her colleague Pauline McCririck. They worked at the Marlborough Day Hospital, London, and in Martin's private practice in her home. Dubus dates the start of Martin's LSD work to 1954. Martin published the method under the label "LSD analysis" (1964). Martin and McCririck called the bodily holding element "fusion" [source: Dubus 2020; Dubus interview; Grof 1970]. Grof (1970) counted it as one of four main LSD-therapy approaches of the era. He credits it to "A. J. Martin", the name Martin used on her papers.

Stated idea of change. The drug was used to bring on deep regression, in which patients relived early emotional deprivation and showed infantile dependency needs. The therapist was not to stay detached. She used psychoanalytic interpretation and, for selected patients, took on an active maternal role in the session, including physical comfort. The aim was to give a corrective experience of unconditional maternal love. Martin said this would build the transference quickly, strengthen the ego, let previously unbearable feelings be integrated, and relieve conflict and symptoms [source: Grof 1970; Dubus 2020; Dubus interview]. These are the developers' claims, not established mechanisms.

Legal status today (2024–25 sources). LSD is in Schedule I of the 1971 UN Convention, Schedule I in the US (since 1970), Class A in the UK, Schedule III in Canada and S9 (prohibited) in Australia. Switzerland grants limited medical use by special authorisation. In 2024 the FDA gave an LSD product breakthrough-therapy designation for generalised anxiety disorder [uncertain: secondary, unverified for the scheduling list; the FDA designation (March 2024, MindMed's MM120) is from Psychiatric Times 2024]. None of the sources used describe any current regulated protocol that uses anaclitic or fusion holding. Current trial manuals limit touch tightly (see §4) [source: MAPS 2010; McHerron et al. 2025].

2. Framework

The framework below describes how the method was organised, as history. It is not a treatment plan.

Phase A: Day-hospital LSD analysis (all patients).

Patients: chronic neurotic patients who had not responded to other psychiatric treatment. Martin's 1957 series had 50 day-hospital patients: 19 obsessional, 22 chronic tension states, 3 "sexual neurosis" and 6 "psychopaths" [source: Erowid ref 1627, abstract of Martin 1957]. Martin's own six selection criteria were: no previous psychotic episode, a good personality (later "good ego development"), some knowledge of unconscious processes, some insight, high intelligence, and a strong desire to get well [source: Martin 1957 full text; Martin 1962; Martin 1963].
Grof's description of the anaclitic indication: deep neuroses, borderline psychotic disorders and severe childhood emotional deprivation [source: Grof 1970].
Format: individual drug sessions, roughly weekly, within ongoing psychoanalytic therapy. The 1957 series gave between 2 and 20 sessions per patient. Each session ran during the day and was ended medically the same afternoon so the patient could go home [source: Erowid refs 1627, 1783].
Grof's account: sessions were repeated at regular intervals, and "All the therapy is conducted during the sessions themselves" [source: Grof 1970].

Phase B: "Fusion" (a small subgroup only).

Who: patients judged to have suffered a severe lack of maternal love [source: Dubus interview].
Entry criterion as Dubus reports it: several prior sessions, an established trust and alliance, the patient's familiarity with the drug's effects, and the patient's verbal consent. Dubus notes that consent was unusual for the period.
Course: the therapist's physical closeness was increased step by step over sessions. Fusion lasted about 1 to 1.5 hours per session, over 1 to 3 sessions depending on the patient's need. Physical contact then stopped and therapy moved to other stages [source: Dubus interview].
Not specified: criteria for ending fusion and the content of the "other stages" are not described in the sources used [uncertain].

Phase C: Outcome and follow-up (as self-reported).

The 1957 abstract reports 5 "cured", 14 markedly improved, 27 slightly improved and 4 unimproved out of 50.
At two-year follow-up, 9 had relapsed, mostly in the chronic-tension group [source: Erowid ref 1627; Martin 1957 full text]. Martin also reported that 2 of the 50 had violent reactions early on and were judged unsuitable for day treatment, with an underlying psychotic trend suspected. She reported no accidents [source: Martin 1957 full text].
Grof cites Martin as reporting "good and relatively rapid results" [source: Grof 1970].
These were uncontrolled clinician ratings. Edwards (2005) warns that writing of this era often rushed to "messianic conclusions" [source: Edwards 2005]. Martin's 1962 paper reported that seven of twelve homosexual men "became heterosexually orientated" and stayed so at three to six years. This too was uncontrolled and anecdotal [source: Martin 1962].

Scale and end of the programme. Martin is said to have treated 1,000–3,000 patients over about 15 years [source: Dubus interview; Dubus flags this as a reported estimate]. Doctors trained at the hospital. The work was presented at conferences, including the Royal Medico-Psychological Association meeting in London in February 1961, whose proceedings Crocket, Sandison and Walk edited [source: Erowid ref 2156; Martin 1963]. A teaching film was made but is lost [source: Dubus interview]. No source used says exactly why or when Martin's own programme ended [uncertain]. Her 1963 case report shows one practical limit. That patient's drug sessions stopped when the general hospital could no longer give him an overnight bed afterwards, because Martin would not send him home alone [source: Martin 1963]. For the wider British context, Ronald Sandison told Ben Sessa in 2005 why his own LSD work at Powick ended around 1964. He gave five reasons: burnout, a shift to community psychiatry, new consultants with no interest in LSD, bad press over LSD misuse and a 1964 murder trial, and new regulatory bodies (the notes name the CSM, 1966) [source: Sessa 2005, MAPS meeting notes]. This is Sandison's account of his own unit, not of Martin's. Martin died in 1969 [source: Dubus 2020]. More widely, Sandoz stopped supplying psychedelics for research in 1965, regulation of drug trials tightened after the thalidomide disaster, the pharmaceutical industry was unwilling to fund trials, and nonmedical use was banned in 1970 [source: Hall 2021].

3. Guiding principles

Part A gives the historical principles, as published. They are given so they can be recognised, not followed.

Regression is the therapeutic route. In practice: the drug state was used to bring patients back to early deprivation so it could be worked on directly [source: Grof 1970].
Abandon analytic detachment. In practice: the therapist acted as an active, loving presence rather than a neutral interpreter [source: Grof 1970; Dubus 2020].
Support is genuine, not role-play. In practice: Martin held that her maternal support was real rather than artificial. Dubus reports that she set this against the more detached approach of Sandison and Ling [source: Dubus 2020].
Gratify the early need, then move on. In practice: fusion was time-limited (1–3 sessions) and then withdrawn [source: Dubus interview].
Trust and consent first. In practice: fusion was offered only after an alliance had formed and the patient had agreed verbally [source: Dubus interview].
Frightening dissociation can be productive. In practice: Martin reported that "schizophreniform" states with complete dissociation, though frightening, aided improvement. In two cases she linked benefit to a sense of acceptance by a "higher mind" [source: Erowid ref 1783, abstract of Martin 1959]. This is Martin's interpretation and is not established.

Part B gives the principles a clinician applies today. They are drawn from current trial guidance and from the lessons of this history.

No regressive bodily holding of a person in an altered state. In practice: never offer "re-mothering" contact. Touch, if any, is minimal, pre-agreed and protocol-bound [source: MAPS 2010; McHerron et al. 2025; synthesis].
Consent is set before the session and re-checked during it. In practice: touch preferences are agreed in preparation and re-confirmed through verbal and non-verbal check-ins [source: McHerron et al. 2025; MAPS 2010].
The participant can always stop touch. In practice: agree a stop word. Honour it unless touch is needed to prevent injury [source: MAPS 2010].
Never sexual contact. In practice: say this explicitly in preparation [source: MAPS 2010].
Structural safeguards over personal virtue. In practice: use two therapists, a neutral third party for concerns, video with independent review, and supervision [source: McHerron et al. 2025].
Do not cultivate dependency. In practice: frame the relationship as time-limited and plan its ending from the start [synthesis].
Do not reframe alarming reactions as breakthroughs without assessment. In practice: treat psychotic-like or dissociative reactions as clinical events needing medical and psychiatric review [synthesis].

4. Intervention inventory

This inventory has two parts.

Part A describes, in general terms only, each component of the historical method as the sources report it. It is for recognition, not use. Part A gives no therapist language, because the method must not be reproduced, and none of Martin's session wording is given in the sources.
Part B gives what a clinician does today with this knowledge.

Part A: Historical components (not for use)

A1. Patient selection

Purpose and when used: Martin chose chronic neurotic day-patients who had not responded to other treatment [source: Martin 1957 full text]. Grof describes anaclitic "mothering" as aimed at patients with severe early deprivation, deep neurosis or borderline psychotic states [source: Grof 1970]. Fusion was reserved for a small subgroup judged to have lacked maternal love [source: Dubus interview]. The versions differ. Martin's own papers exclude anyone with a previous psychotic episode. Grof's later summary names borderline psychotic states as an indication.
Steps (as reported): patients were referred to the Marlborough Day Hospital after other treatments failed, sometimes from another hospital [source: Martin 1962]. Martin screened them against six criteria: no previous psychotic episode, good personality or ego development, some knowledge of unconscious processes, some insight, high intelligence, and a strong desire to get well [source: Martin 1957 full text; Martin 1962]. She still made exceptions. In the 1961 case she judged the man's ego development poor and noted a risk of paranoid psychosis. She went ahead because he was desperate and saw the treatment as "a last hope" [source: Martin 1963]. None of the three papers opened describes a consent procedure [source: Martin 1957; Martin 1962; Martin 1963].
Example language: none (not for use).
Watch for: the selection targeted the most vulnerable patients, those with deprivation histories and borderline or psychotic features. This is the group most exposed to the harms described in the Harm account [synthesis]. A patient's desperation, framed as a "last hope", was treated as a reason to go ahead despite a recognised psychosis risk [source: Martin 1963].

A2. Day-hospital drug session with psychoanalytic interpretation

Purpose and when used: to bring on regression and work on the material that surfaced. This was the core of every course [source: Grof 1970].
Steps (as reported): roughly weekly day sessions, 2 to 20 per patient in the 1957 series. Interpretation was given during the session, and the session was ended medically the same afternoon [source: Erowid refs 1627, 1783]. Sessions lasted most of the day. Martin held that the therapist should stay with the patient for most of each session to keep the transference relationship [source: Martin 1962; Martin 1963]. Ward staff, such as a nursing sister, were also present at times [source: Martin 1963]. Grof stresses that the work happened within the sessions themselves [source: Grof 1970]. No administration details are given here.
Example language: none (not for use).
Watch for: sending patients home the same day after deep regression. In the 1961 case, a general hospital provided an overnight bed after sessions. The sessions stopped when that bed was no longer available, because Martin would not let the man go home alone [source: Martin 1963]. Martin's 1957 paper acknowledged that the drug "may precipitate a psychosis". It reported that 2 of 50 patients had violent early reactions [source: Martin 1957 full text]. Routine aftercare between sessions is not described in the sources [uncertain].

A3. Graded physical "mothering" and fusion (described generally)

Purpose and when used: to give a regressed, deprived patient an experience of unconditional maternal love, build the transference rapidly and strengthen the ego [source: Dubus 2020; Dubus interview; Grof 1970].
Steps (as historians report, in general terms): offered only after several sessions, trust, familiarity with the drug state and verbal consent. The therapist's proximity increased over successive sessions, from sitting near the patient to prolonged close holding. Comfort such as a warm drink was sometimes offered. Each fusion period lasted about 1 to 1.5 hours, over 1 to 3 sessions, and then physical contact ended [source: Dubus interview]. Grof describes the therapist meeting basic needs for warmth and comfort through close body contact [source: Grof 1970]. Dubus cites Grof and Grof (2010) for the fusion description; that book is not drawn on here [uncertain]. Martin's own papers opened (1957, 1962, 1963) do not describe this holding. The 1961 case mentions only comfort such as extra blankets [source: Martin 1963]. This element is not given as a technique.
Example language: none (not for use).
Watch for: this is the element linked to boundary concerns. See the Harm account.

A4. Patient-to-therapist touch

Purpose and when used: Dubus reports that Martin's clinics also allowed patients to touch the therapist. She calls this unusual for the period and says it needed firm boundaries [source: Dubus interview].
Steps: the sources used give no details [uncertain].
Example language: none (not for use).
Watch for: two-way, open-ended touch while one party is in an altered state removes a clear boundary [synthesis].

A5. Reframing dissociative ("schizophreniform") reactions as therapeutic

Purpose and when used: Martin reported that some patients entered states of complete dissociation. She judged these frightening at the time but beneficial afterwards. In two cases she attributed benefit to a felt acceptance by a "higher mind behind the universe" [source: Erowid ref 1783, abstract of Martin 1959]. This is Martin's interpretation.
Steps: none to give. This was an interpretive stance, not a procedure.
Example language: none (not for use).
Watch for: psychotic-spectrum reactions being labelled breakthroughs instead of being assessed [synthesis].

A6. LSD treatment aimed at changing homosexual men's orientation

Purpose and when used: Martin framed homosexuality as "arrested emotional development". She used LSD analysis with homosexual men who wanted to change. Her 1962 paper in Acta Psychotherapeutica reports twelve men drawn from 100 chronic neurotic day-hospital patients. It excluded "Inverts" and men "accepting their homosexuality without any desire to change" [source: Martin 1962]. The paper's stated procedural aim was a rapid positive transference to resolve early conflicts. Its outcomes, however, were counted by orientation: seven men "became heterosexually orientated" and were reported to remain so at three to six years, with marriage recorded as an outcome [source: Martin 1962]. Her 1961 case (printed 1963) was a 32-year-old man who was depressed, stammering and suicidal, and who "deplored" his homosexuality. Success was reported as his ending homosexual relationships and marrying [source: Martin 1963]. Clancy Cavnar, PsyD (Chacruna, 2018), presents Martin's 1962 study as one of the psychedelic attempts to "cure" homosexuality [source: Cavnar 2018].
Steps (as reported, in general terms): the same day-hospital LSD sessions as A2, with the therapist present for most of each session. Individual or group follow-up continued for up to about two years [source: Martin 1962]. In the 1961 case the sessions brought rage, sadistic fantasies, and sexual and aggressive urges aimed at the therapist and at women staff. Martin treated these as abreaction to be worked through. Afterwards the man was transferred to a male psychiatrist for further drug-assisted abreactions [source: Martin 1963]. Neither paper describes a consent process beyond the patients' wish to change [source: Martin 1962; Martin 1963].
Example language: none (not for use).
Watch for: any drug or regression therapy that treats sexual orientation as a target is conversion practice [synthesis]. Martin's papers report no follow-up for harm, and no account by the men themselves was found [uncertain]. The absence of reported harm is not evidence that there was none.

Part B: What a clinician does today

B1. Recognising and contextualising historical claims

Purpose and when used: clients, trainees or practitioners may cite "anaclitic" or "fusion" therapy as validated precedent for holding clients in altered states. Use this when that claim comes up.
Steps:
1. Name the history accurately: London, c. 1954–1960s, Martin and McCririck, uncontrolled case series [source: Dubus 2020; Erowid ref 1627].
2. Explain that outcome claims were the clinicians' own uncontrolled ratings, from an era Edwards (2005) criticises for "messianic conclusions".
3. Contrast it with current trial touch rules (B3–B6).
4. State that the method is not part of any regulated protocol in the sources reviewed [synthesis; uncertain beyond those sources].
Example language (illustrative): "That method existed in 1950s London, but it was never properly tested. Today's trials deliberately limit touch."
Watch for / working: the person can tell historical precedent apart from evidence and current standards.

B2. Assessing a client who received regressive or holding-based drug therapy

Purpose and when used: a client discloses past or recent drug-assisted therapy that involved holding, "reparenting" or bodily contact. This may be an elderly former patient of 1960s programmes, or, more often, someone treated in an underground setting [synthesis].
Steps: all steps here are [synthesis].
1. Establish safety now: current contact with the practitioner, any ongoing pressure, and any risk to self.
2. Get a factual account at the client's pace: setting, who was present, what was agreed beforehand, what happened, and whether the client could stop it.
3. Screen for PTSD, dissociation, shame, persistent perceptual or psychotic symptoms, and dependency on or grief for the practitioner.
4. Ask how the client understands the experience now. Do not impose the label "abuse" or "healing".
5. Explain the client's reporting options (see Harm account).
Example language (illustrative): "What did you agree to beforehand about being touched? Did it stay within that?" "Were you able to say no once you were under the drug's effects?"
Watch for / working: a coherent narrative, reduced self-blame, and choices about reporting made by the client.

B3. Agreeing touch preferences in preparation (regulated trial teams)

Purpose and when used: to settle any touch before the altered state, when the participant can consent fully. Do this in preparation sessions, before any dosing session [source: MAPS 2010; McHerron et al. 2025].
Steps:
1. Explain the purpose, risks and benefits of touch [source: McHerron et al. 2025].
2. Agree a preference, either a blanket "no touch" or a conditional rule such as "touch only if I initiate" [source: McHerron et al. 2025].
3. Specify which touch is allowed (MAPS examples are holding a hand or a hand on the shoulder or arm), who may initiate it, how close therapists will sit, and how the participant signals no [source: MAPS 2010; McHerron et al. 2025].
4. State plainly that there will never be sexual contact [source: MAPS 2010].
5. Record the agreement on a standard form [source: McHerron et al. 2025].
6. Recommended limit: rule out prolonged full-body holding or lying beside the participant. This is [synthesis] informed by the history in this tab. MAPS 2010 lists being held among touch a participant may request, so this limit is stricter than that manual.
Example language (illustrative): "Some people want a hand to hold, some want no touch at all. Let's decide now, while you're clear-headed." Also, verbatim from MAPS: therapists "at no time will they engage in any form of sexual contact".
Watch for / working: the participant states the preference in their own words, and the record matches what they said.

B4. Consent check-ins and a stop word during sessions

Purpose and when used: consent can shift in an altered state, so it is re-checked throughout the session [source: McHerron et al. 2025].
Steps:
1. Re-confirm the agreed stop word, used as a command (MAPS: the word "stop") [source: MAPS 2010].
2. Offer touch only within the prior agreement, and ask before each instance [source: MAPS 2010].
3. Watch for verbal and non-verbal signs of discomfort with closeness, and adjust distance [source: MAPS 2010; McHerron et al. 2025].
4. Honour "stop" at once. The only exception is touch needed to prevent injury [source: MAPS 2010].
Example language (illustrative): "Would it help if I put my hand on your shoulder, as we agreed? Yes or no is fine."
Watch for / working: the participant declines or adjusts freely and shows no sign of complying to please.

B5. Responding to a request for holding beyond the agreement

Purpose and when used: a regressed participant asks to be held or "mothered". This is the situation anaclitic therapy answered with bodily fusion.
Steps:
1. By default, do not go beyond the agreed touch, even when asked [source: McHerron et al. 2025].
2. Offer non-touch alternatives: grounding, guided imagery, psychodynamic containment such as naming and staying with the need, or comfort objects such as a weighted blanket [source: McHerron et al. 2025].
3. Review the request in integration [synthesis].
Note on the evidence: McHerron et al. note that withholding touch can also cause harm and call for research. Treat this as an open question [uncertain].
Example language (illustrative): "I can hear how much you want to be held. I'm staying right here with you. Let's put the blanket around you."
Watch for / working: distress settles with presence. The participant does not feel rejected, which you check in integration.

B6. Team and structural safeguards

Purpose and when used: to protect participants when they are most vulnerable and suggestible. Apply these as standing arrangements in any drug-assisted therapy service.
Steps:
1. Use two therapists, and keep at least one in the room throughout [source: MAPS 2010].
2. Provide a neutral third party outside the therapist pair for concerns. The male-female pair has itself been linked to misconduct [source: McHerron et al. 2025].
3. Consider video recording with independent review, weighing privacy [source: McHerron et al. 2025].
4. Provide specialised training in consent negotiation and boundary transgressions, plus supervision that covers complex boundary situations [source: McHerron et al. 2025].
5. Make the reporting route known to participants in writing before the first session [synthesis].
Example language (illustrative): "If anything about how either of us behaves worries you, here is someone outside our team you can contact."
Watch for / working: concerns are raised and logged. Video review is actually carried out.

B7. Harm reduction and integration for clients seeking "reparenting" in unregulated settings

Purpose and when used: a client plans, or has attended, an unlicensed retreat or session that promises holding, "re-mothering" or "fusion" under a psychedelic [synthesis].
Steps: all steps here are [synthesis].
1. Do not facilitate, recommend or help obtain illegal drug use.
2. Assess psychiatric risk: psychosis or bipolar history, dissociation, trauma, and current medicines and interactions (in general terms).
3. Discuss the specific boundary risks of touch in altered states. Use the red flags in the Harm account.
4. Help the client set personal limits and a safety plan: a sober contact, the right to leave, and no one-practitioner-alone sessions.
5. Offer integration support afterwards, and refer for regulated care.
Example language (illustrative): "I can't advise on the drug itself. I can help you think about what you're agreeing to physically, and how you'd say no."
Watch for / working: the client can name the red flags and their own limits.

B8. Meeting deprivation needs without drugs or holding

Purpose and when used: a client's longing to be "re-mothered" is real and clinically important. This is the need anaclitic therapy addressed.
Steps: all steps here are [synthesis]; specific models are not sourced in this tab.
1. Name and validate the attachment need.
2. Work with it through the relationship and through verbal and imaginal methods, with clear boundaries.
3. Refer to established attachment- or trauma-focused psychotherapies where indicated.
Example language (illustrative): "That longing makes sense, given what you missed. We can work with it here, in words and in how we are together."
Watch for / working: less idealisation of a rescuer figure and more capacity to self-soothe.

5. Handling problems

All guidance in this section is [synthesis] unless a source is tagged. The sources used do not discuss problem management for this method.

Client idealises anaclitic or fusion therapy and asks you to hold them. Decline clearly and warmly. Explain that the boundary protects them, and explore the need behind the request (B8).
Rupture after you decline touch, or after you hold firm to the agreed limit in a trial (B5). Name the hurt, own its impact without reversing the boundary, and repair the rupture in integration. If the client remains distressed, consult supervision.
Dependency on a former practitioner. Typical signs are grief, loyalty, or difficulty acknowledging harm.
- Do not push the client to condemn the practitioner.
- Work with ambivalence.
- Note that rapid, intense transference was an explicit aim of Martin's method [source: Dubus 2020], so attachment of this kind is expected rather than a sign of the client's weakness.
Client minimises boundary crossings ("I asked for it", "I consented beforehand"). Explain that consent given before an altered state does not license escalation during it. Ongoing consent is the current standard [source: McHerron et al. 2025].
Overwhelm or flashbacks when recounting. Slow down. Use grounding and titrate disclosure. Stabilise before detailed processing.
Persistent perceptual, dissociative or psychotic symptoms after drug sessions. Refer for psychiatric assessment. Do not frame these as "breakthroughs" (contrast A5).
Non-adherence: the client goes ahead with an unregulated "reparenting" retreat. Stay engaged and keep the harm-reduction stance (B7). Agree a check-in afterwards.
You learn that a colleague offers holding in drug sessions. Consult your professional body's guidance and a senior colleague, and follow the steps in the Harm account. Record what you heard and when.
Trial participant reports discomfort with a therapist's touch. Stop the touch at once. Inform the neutral third party and the principal investigator. Document, and review any video [source: McHerron et al. 2025 for the third-party and video measures].
Client who went through a sexual-orientation change attempt (drug-assisted or not). Take an affirming stance; orientation is not the problem to treat. Explore shame, grief and relationships or marriages entered under pressure. Screen for depression and suicide risk by standard practice. Avoid any framing that echoes "arrested development" [synthesis; the historical framing is from Martin 1962].

Harm account

Evidence note. The sources used document no complaint, legal case, regulatory finding, injury or death attributed to Martin or McCririck, and no first-person account by a former patient [uncertain]. The harms below are therefore of three kinds:

adverse reactions and aims that Martin herself reported in her papers: violent reactions, an acknowledged psychosis risk, and orientation-change work with homosexual men [source: Martin 1957; Martin 1962; Martin 1963];
structural risks that historians and current ethics researchers identify in this kind of practice;
one documented modern case of touch-related misconduct in a psychedelic trial, which shows the same vulnerability. That case is not linked to Martin's method.

The tags below keep these apart.

What was actually done

Who and where. From about 1954, Joyce Martin, a London psychiatrist-psychoanalyst, and her colleague Pauline McCririck gave LSD-assisted psychoanalytic therapy at the Marlborough Day Hospital and in Martin's home practice [source: Dubus 2020; Dubus interview].
Selection. Chronic neurotic patients who had failed other treatments were referred. Martin's own criteria excluded anyone with a previous psychotic episode and required insight, intelligence and a strong wish to get well. She still made exceptions for desperate patients despite a recognised psychosis risk [source: Martin 1957 full text; Martin 1963]. Grof later described patients with severe early deprivation, borderline or psychotic features as the main candidates for the anaclitic approach [source: Erowid ref 1627; Grof 1970].
Repeated day sessions. Patients had roughly weekly drug sessions (2 to 20 in the 1957 series). Each lasted most of the day, with the therapist present for most of it and interpreting [source: Erowid refs 1627, 1783; Martin 1962]. The session was ended medically the same afternoon and the patient usually went home. At least one patient was given an overnight hospital bed instead [source: Martin 1957; Martin 1963].
Regression and maternal stance. In the session the therapist moved out of analytic neutrality into a mothering role, responding to the patient's infantile needs for comfort and warmth [source: Grof 1970].
Graded physical closeness ("fusion"). This was used for a subgroup, after several sessions and the patient's verbal consent. Over successive sessions the therapist moved from sitting nearby to lying beside and holding the drugged patient for about 1–1.5 hours, over 1–3 sessions. Physical contact then stopped [source: Dubus interview]. Patients were also allowed to touch the therapist [source: Dubus interview].
Interpretation of severe reactions. Martin reported frightening states of complete dissociation ("schizophreniform" reactions) and judged them beneficial afterwards [source: Erowid ref 1783].
Application to sexuality. From the day-hospital patients, Martin treated twelve homosexual men who wished to change. She excluded those who accepted their orientation, and reported that seven "became heterosexually orientated" [source: Martin 1962]. In the 1961 case, the reported success was that the man stopped having homosexual relationships and married [source: Martin 1963].
Spread. Doctors trained at the hospital, and the work was presented at the 1961 London meeting and in a film, now lost [source: Dubus interview; Erowid ref 2156]. Dubus says that some methods of this era passed to underground therapists without safety evaluation [source: Dubus interview].

How it caused harm

The mechanisms below are [synthesis] unless tagged.

Boundary erosion through holding a drugged, regressed patient.
- The patient's capacity to monitor, refuse or withdraw from bodily contact is reduced in the drug state and in regression.
- Contact escalated step by step across sessions, so each step looked small.
- Two-way touch was permitted [source for practice: Dubus interview].
- Current researchers treat consent in altered states as needing ongoing re-checking because it can shift [source: McHerron et al. 2025].
Consent problems.
- Verbal consent was obtained [source: Dubus interview], but it was given within an intense, deliberately accelerated transference [source: Dubus 2020] to the person who would do the touching.
- No independent party, written agreement or stop procedure is reported in the sources used [uncertain].
Dependency and transference. The method aimed to build transference rapidly and to meet infantile dependency needs directly [source: Dubus 2020; Grof 1970]. Martin wrote that the drug built a positive transference faster than years of psychoanalysis [source: Martin 1962]. This creates risks of dependency, painful endings when contact is withdrawn after 1–3 sessions, and exploitation by a less scrupulous practitioner. No documented dependency case from Martin's patients was found [uncertain].
Missed psychiatric deterioration. Martin acknowledged that LSD "may precipitate a psychosis". She reported two violent early reactions among 50 patients, both judged unsuitable for day treatment [source: Martin 1957 full text]. She still went ahead with a suicidal patient whose ego development she judged poor, accepting a risk of paranoid psychosis [source: Martin 1963]. Reframing dissociative or psychotic-like reactions as beneficial (Martin 1959) risks normalising decompensation. Sending patients home the same day after deep regression risks unsupported distress between sessions. Martin herself stopped one man's sessions when no overnight bed was available [source: Martin 1963].
Pathologising sexual orientation (documented). Martin framed homosexuality as arrested emotional development and selected only men who wanted to change. She counted heterosexual orientation and marriage as success: seven of twelve men in 1962 [source: Martin 1962; Martin 1963]. Cavnar (2018) groups this work with other psychedelic attempts to "cure" homosexuality [source: Cavnar 2018]. Treating sexual orientation as a target is now regarded as conversion practice [synthesis]. One of the men was suicidal before treatment [source: Martin 1963]. The men's outcomes beyond Martin's own follow-up are unknown, and no first-person account was found [uncertain].
Overstated results. Outcomes were uncontrolled clinician ratings [source: Erowid ref 1627]. Martin's 1962 paper is internally inconsistent: the detailed patient's age differs between sections, and the category counts do not match the table [source: Martin 1962]. Edwards (2005) criticises the era's "messianic conclusions".
Modern documented parallel (not Martin's method).
- In a MAPS-sponsored Phase II MDMA trial for PTSD in British Columbia, begun in 2015, participant Meaghan Buisson alleged repeated sexual assault by co-therapist Richard Yensen. His wife, psychiatrist Donna Dryer, was the other co-therapist.
- MAPS acknowledged an "unethical" sexual relationship that Dryer knew of and did not report, cut ties with both, paid for therapy on a compassionate basis, and formalised a code of ethics on dual relationships.
- The civil claim settled on undisclosed terms. The police investigation and the medical college complaint were unresolved at the time of reporting.
- The provincial psychology college could not act because Yensen was not registered [source: CBC].
- These are allegations and organisational acknowledgments, not court findings.
- McHerron et al. (2025) note that the male-female dyad structure has been linked to such misconduct [source: McHerron et al. 2025].
Who is most at risk [synthesis]:
- people with early deprivation or abuse;
- people with borderline, dissociative or psychotic vulnerability;
- people with strong rescue fantasies;
- anyone treated by a single practitioner, a closed couple, or an unregulated practitioner.

Recognising it today

Red flags in a colleague's, retreat's or group's practice. All items are [synthesis] unless tagged.

Offers of "re-mothering", "reparenting", "holding", "fusion", "corrective maternal experience" or "cuddle work" during psychedelic, MDMA or ketamine sessions.
Touch not agreed before the session, or "we'll see what you need on the day".
Touch that escalates across sessions, full-body holding or lying beside the client, and therapist-initiated touch.
A single practitioner alone with a drugged client, or a couple team with no outside reporting route [source for the couple-team concern: CBC; McHerron et al. 2025].
No video, no supervision, and secrecy or loyalty demands.
Framing distress or dissociation as "breakthrough", and cultivating dependency ("only I can give you this").
Citing 1950s–60s anaclitic therapy as proof of safety or efficacy.
Targeting sexual orientation or gender identity. One pattern is recruiting only clients who "want to change" and counting heterosexual relationships or marriage as outcomes, as Martin did [source for the historical pattern: Martin 1962].
Unregistered practitioners outside any regulator's reach [source: CBC].

How legitimate practice differs [source: MAPS 2010; McHerron et al. 2025]:

written touch preferences agreed before dosing;
minimal touch, such as a hand or shoulder, offered only on request or with permission;
an explicit stop word, and an explicit no-sexual-contact rule;
two therapists, a neutral third party, and video with independent review;
specialised training and supervision;
non-touch alternatives for regressive needs.

What to do instead and how to help those harmed

Instead.
- For research or treatment with psychedelics, use only regulated programmes with published touch policies (see B3–B6).
- For deprivation and attachment needs, use non-drug, boundaried psychotherapies and refer appropriately [synthesis].
- The anaclitic or fusion method itself has no place in current practice.
Assessment and treatment of affected clients. Use B2 and section 5: establish safety, take a paced narrative, screen for PTSD, dissociation, psychotic symptoms and shame, work with ambivalent attachment to the practitioner, and use trauma-focused therapy after stabilisation [synthesis]. Former 1950s–60s patients will now be elderly; for them, focus on meaning-making and any lasting effects [synthesis]. For anyone put through an orientation-change attempt, take an affirming stance. Do not treat orientation as the problem. Explore shame, lost years, and relationships entered under pressure. Assess depression and suicide risk by standard practice, and follow your professional body's position on conversion practices [synthesis].
Ethical, reporting and legal duties. All items are [synthesis]; check local law and your code.
- Sexual contact with a client is prohibited by all major professional codes.
- Support the client to report, if they wish, to:
  - the practitioner's licensing body;
  - the trial sponsor, ethics committee or IRB, and regulator, where it happened in a trial;
  - the police, where assault is alleged.
- Know that unregistered practitioners may fall outside any regulator's reach [source: CBC].
- Follow mandatory-reporting rules for colleague sexual misconduct, minors or vulnerable adults where your jurisdiction imposes them.
- Document disclosures factually, and respect the client's choices where no mandatory duty applies.

Sources

Pages consulted

Books and papers cited [uncertain where relied on beyond the full texts listed above]

Martin, A. J. (1957). LSD treatment of chronic psychoneurotic patients under day-hospital conditions. International Journal of Social Psychiatry, 3(3), 188–195. (Full text: see Pages consulted.)
Martin, J. (1963). A case of psychopathic personality with homosexuality treated by LSD. In Crocket, R., Sandison, R. A. & Walk, A. (eds), Hallucinogenic Drugs and their Psychotherapeutic Use, pp. 112–115. London: H. K. Lewis. (Full text: see Pages consulted.)
Grof, S. & Grof, C. (2010). Holotropic Breathwork. SUNY Press. (Cited by Dubus for fusion.)

Sandison, R. (2001), cited by Cavnar (2018) for Martin's 1962 outcomes.

← Previous · VII-035LSD treatment of alcoholism (historical)Next · VII-037 →Hypnodelic therapy (historical)