NOTE-036: Lindahl et al. 2017, varieties of contemplative experience

Contribution

The paper sets out the method and first results of the Varieties of Contemplative Experience (VCE) study, a qualitative interview study aimed at under-reported meditation-related experiences, especially challenging, distressing or impairing ones. It delivers two codebooks: 59 phenomenology categories in 7 domains (Table 4) and 26 influencing-factor categories in 4 domains (Table 6). It also documents that phenomenologically similar experiences were appraised anywhere from very positive to very negative. Before this paper, the adverse-effects literature consisted mostly of case reports and small, theory-driven or single-community studies (Introduction, pp. 3–5).

Key insight

Whether a meditation-related experience counts as “progress” or “adverse” is largely a matter of appraisal. Practitioners, teachers, texts and clinicians often disagree about it. So the study codes what happened (phenomenology) separately from what it was taken to mean, what caused it, and what helped (influencing factors) (Methods, “Influencing factors coding”, p. 12). The taxonomy is valence-neutral on purpose.

Assumptions

These are premises of the design, not formal assumptions:

• Purposive “deviant case” sampling (p. 7). Inclusion required a challenging, difficult or impairing meditation-related experience. The target was 10 men and 10 women in each of three traditions (Theravāda, Zen, Tibetan). Recruitment went through teachers, then snowball sampling.
• Exclusions (p. 7): prior unusual experiences closely resembling the reported ones; a mixed or non-Buddhist practice history; an alternative cause that could “wholly” account for the symptoms.
• Attribution-theory framing (Taves) (p. 6). “Meditation-related experiences” are the effects participants attributed to meditation.
• Retrospective self-report. Interviews ran from Dec 2010 to Mar 2016, typically 45–120 minutes. A follow-up questionnaire was sent after all interviews were done; 53/60 (88%) completed it, and for the other 7 researchers extracted the answers from transcripts (p. 13).
• Saturation. It was judged after 30 and after 60 interviews (pp. 7–8).

Key results

• Sample (Table 1, pp. 13–14). 73 interviewed, 13 excluded: 5 for mixed practice, 4 for reporting no challenge, 4 for alternative causes. That leaves n = 60: 43% female, mean age 48.9 (SD 13.1), 85% US, 94% White (n = 59). 60% were meditation teachers, and 43% had more than 10,000 lifetime practice hours. There were 32 experts (40% Theravāda, 25% Zen, 19% Tibetan, 16% clinical), 11 of whom also gave practitioner interviews.

• Onset context (Table 3, p. 15).

  • Practice before onset: mean 7.1 years (SD 8.0), range 1 day to 25 years. 12% reported challenges within their first 10 days and 18% within their first 50 hours.
  • 72% had onset on or right after retreat and 28% in daily practice.
  • At onset, 25% practised less than 1 h/day, 34% 1–9 h/day and 41% 10 or more h/day.

• Phenomenology (Table 4, pp. 15–16).

  • Participants reported a mean of 6.1 domains (SD 1.0) and 19.6 categories (SD 7.2, range 7–40). Each category was reported by an average of 20 practitioners, ranging from 2 (scrupulosity) to 49 (fear).
  • Top categories: fear/anxiety/panic/paranoia 82%; changes in motivation or goal 78%; positive affect 75%; somatic energy 63%; sleep changes 62%; depression/dysphoria/grief 57%; changes in self–other boundaries 53%.
  • Experts reported 56 of the 59 categories.

• Causality (Table 5, p. 21). Mean score 4.2 ± 1.1 of 6. Subjective attribution 93%, challenge 85%, consistency 88%, de-challenge 60%, re-challenge 61%, exacerbation 58%.

• Impact (p. 21). 88% said effects extended into daily life. Median symptom duration was 1–3 years (days to more than 10 years). 10% reported minimal impairment. 73% reported moderate-to-severe impairment in at least one domain, 17% suicidality and 17% inpatient hospitalisation.

• Influencing factors (Table 6, p. 22). The most-cited categories, with practitioner % and expert % in that order:

  • relationship to teacher 97 / 97
  • worldview or explanatory frameworks 97 / 85
  • amount, intensity or consistency of practice 93 / 94
  • practice approach 93 / 94

  The authors say explicitly that these are participants’ views, “not … a theory or hypothesis about risk factors … put forth by the authors” (p. 22).

Claims

id	claim	strength	support
C1	Meditation-related experiences reported by Western Buddhists span 59 categories in 7 domains	moderate	Thematic coding of 60 practitioner interviews and 32 expert interviews, with saturation checks and a double-coded 15% subset (κ = .70)
C2	The same phenomenology is appraised anywhere from very positive to very negative	moderate	Qualitative (Discussion). No quantified valence data are presented
C3	73% moderate-to-severe impairment, 17% suicidality, 17% hospitalisation, median duration 1–3 years	weak as population figures; descriptive within the sample	Self-report severity items. Sample selected for difficulty (Limitations: “artifact of sampling”)
C4	The experiences were caused by meditation	weak	Six self-reported causality items. The ≥2 score was an inclusion criterion. No control group, no base rates (Limitations concede an RCT is needed)
C5	The results “challenge” the assumption that difficulties occur only in people with pre-existing conditions, on intensive retreat, or poorly supervised	weak–moderate	Existence cases: 12% within 10 days of starting, 25% under 1 h/day, 28% in daily practice. This rebuts “only”, not relative risk
C6	Influencing factors (teacher relationship, worldview, practice intensity) shape valence and duration	assertion by participants	Table 6. The authors disclaim it as their own hypothesis (p. 22)
C7	Findings are relevant to mindfulness-based interventions	informal argument	Some participants had onset under MBI-like conditions. The Limitations section says the results “cannot be generalized … to mindfulness-based interventions”

Concepts

• Meditation-related experience. An effect a practitioner attributed to meditation (attribution-theory sense, p. 6).
• Phenomenology vs influencing factors. Two separate codebooks: what was experienced, versus interpretation, causal attribution and remedy (pp. 12–13).
• Deviant-case sampling. Sampling from the extreme of a distribution in order to study rare phenomena (p. 7).
• Causality score. The number of six endorsed criteria. ≥2 means “possibly related” (p. 10).
• Severity grades (p. 11). Mild: less than 25% reduction in functioning. Moderate: 25–50% reduction, or outpatient treatment. Severe: 50% or more reduction, persistent, or inpatient treatment.
• Influencing factor. Anything reported as a risk factor or a remedy. The two are coded together because the same factor, such as the teacher, is reported as either (p. 13).

Connections

The paper builds on Kornfield (1979), Shapiro (1992: 63% of retreatants reported at least one adverse effect and 7.4% stopped meditating), VanderKooi (1997) and Lomas et al. (2014). It criticises the Chen et al. (2011) and Full et al. (2013) studies for being theory-driven and all-Asian (pp. 4–5). The frame comes from religious studies (Taves; Sharf) and cultural psychiatry (Kirmayer). The companion papers it points to are Lindahl et al. (2014) on light experiences and Britton et al. (2014) on sleep. It announces a VCE-based questionnaire piloted in an MBCT trial (NCT01831362).

In this record, the “sense of self” domain (loss of agency 25%, loss of basic self 25%, self–other boundaries 53%) connects to the consciousness cluster. LIT-080’s “entropic brain” and REBUS frame for psychedelic states has an obvious phenomenological parallel, but neither paper cites the other, and the link is my inference.

Bearing on the record

• It is a citable, carefully hedged source for the claim that meditation has non-trivial adverse effects in some practitioners. It is not a source for rates.
• Its “sense of self” categories are fine-grained operational definitions: agency, ownership, basic self, embodiment, narrative self. They are usable by any consciousness note that needs them.
• Nothing here bears on ML practice. No ANTH- document is implicated.

Limitations

• Frequencies are within a purposively selected, highly educated, 94%-White, mostly US sample. The authors call the 100% rate of challenge a sampling artefact (p. 27).
• Retrospective interviews, some many years after onset. The authors note that more distant interviewees under-report phenomenology (p. 28).
• Category frequencies depend on how broadly each code is drawn and on what was salient to the interviewee, because most categories were not prompted (p. 27).
• The causality evidence is self-report and, in part, built into inclusion (see corrections).
• The expert sample overlaps the practitioner sample (11 of 32). The authors report that removing them does not change the similarities, but give no numbers.
• Reliability was assessed on only 15% of transcripts, with κ of .70 and .56.
• Minor internal inconsistencies: the education figure (73% vs 67%) and the causality cutoff (“>2” vs “≥2”).

Open questions

• What are the actual prevalence and incidence in unselected meditators and in MBI participants? The authors propose a structured questionnaire based on the codebook (piloted in NCT01831362).
• Do any influencing factors (practice intensity, trauma history, teacher relationship) predict duration or severity under controlled comparison? The paper supplies only a list of candidates.
• Would the planned comparisons (Abrahamic traditions, European and Indian samples) change the taxonomy itself, or only the frequencies?

Corrections to the seeded skim

• The dossier omits the two most severe outcomes. 17% reported suicidality and 17% required inpatient hospitalisation (Results, “Duration, severity and associated impairment”, p. 21).
• The causality assessment is weaker than “11 of 13 regulatory criteria” sounds, and the dossier does not look at it.
  • Only six criteria were scored per person, all by self-report or transcript: subjective attribution, challenge, exacerbation, consistency, de-challenge, re-challenge. The score is the sum of endorsements, with ≥2 (“possibly related”) as the cutoff.
  • A score of ≥2 was an inclusion condition, so “100% of respondents met causality assessment criteria” (p. 13) is guaranteed by the design.
  • “Inter-subjective consistency” is satisfied by each category having been reported by several people (mean 20). “Expert judgment” is satisfied by the expert interviews. “Prior published reports” is a literature list.
  • The paper also states the cutoff inconsistently: “score>2” on p. 13 against “score of two or greater” on p. 10 and a reported range of 2–6 on p. 20.
  • The Limitations section concedes the needed design is a prospective RCT with multi-year follow-up. Even so, the paper says the results “provided causality-related evidence for all 11 criteria” (p. 28). That is literally true, but the evidence is thin.
• An internal inconsistency on education. The Limitations section says “73% earning graduate or doctoral degrees” (p. 27). Table 1 gives 42% Master’s plus 25% doctoral, which is 67%.
• The exclusion criterion is narrower than the dossier states. The dossier says “people whose symptoms predated meditation” were excluded. The actual criterion is a prior history of unusual experiences that “closely resembled” the meditation-associated ones. So 32% of the retained sample had a psychiatric history and 43% a trauma history (Table 3).
• The inter-rater figures need their kappas. Phenomenology coding showed 99.5% average agreement but weighted κ = .70. Influencing-factors coding showed 97.5% agreement but κ = .56 (moderate). Percentage agreement is inflated by rare codes.