Mindfulness trait mediates between schizotypy and hallucinatory experiences

The study investigates whether dispositional mindfulness and perceived stress mediate the association between schizotypy and hallucinatory experiences in non-clinical adults. Prior work links hallucinations in schizophrenia to deficits in self- and reality-monitoring, and shows that schizotypy (positive, negative, disorganized) in healthy individuals is associated with hallucinatory experiences. Evidence suggests mindfulness training can reduce such experiences, and facets of mindfulness (e.g., non-judging, acting with awareness) are inversely related to schizotypy; stress is also implicated via diathesis–stress accounts. The authors hypothesized that mindfulness traits and stress would mediate relationships between schizotypy dimensions and hallucinatory experiences: positive schizotypy would elevate experiences via lower mindfulness (especially non-judging) irrespective of stress, whereas negative/disorganized schizotypy would act via lower mindfulness (especially acting with awareness and non-judging) and higher stress.

The paper reviews: (1) neurocognitive accounts of hallucinations pointing to impaired self- and reality-monitoring (e.g., anterior insula alterations; comparator model) and the complexity of neural substrates; (2) schizotypy as a multidimensional trait (positive, negative, disorganized) associated with hallucinatory experiences and risk for schizophrenia; (3) evidence that brief mindfulness training reduces hallucinatory experiences in high positive schizotypy, and that schizotypy correlates negatively with mindfulness facets (notably non-judging, acting with awareness); (4) diathesis–stress models suggesting stress can trigger psychotic-like experiences, with findings that negative/disorganized schizotypy interacts with stress, and that mindfulness–hallucination links may be mediated by stress. Gaps include whether mindfulness and/or stress fully or partially mediate schizotypy–hallucinatory experience associations and which mindfulness facets are most relevant.

Design and preregistration: Cross-sectional, preregistered online survey (OSF: https://osf.io/ke28h) in a Japanese non-clinical sample. Participants: 715 adults recruited via Lancers (Japanese crowdsourcing). Inclusion: age ≥18, general population. After exclusions for neurological/psychiatric history and failing instructional manipulation checks (IMC), 458 participants were analyzed (245 males, 208 females, 5 unknown; Mean age = 42.56, SD = 10.13, range 20–76). Socioeconomic status not recorded. Ethical approval obtained. Measures:

• Schizotypy: Japanese SPQ (74 yes/no items; subscales for positive/cognitive-perceptual, negative/interpersonal, disorganized; higher scores = higher schizotypy; Cronbach’s α = 0.91; validated with SPS r = 0.70).
• Mindfulness trait: Japanese FFMQ (39 items; 5 facets: observing, describing, acting with awareness, non-judging, non-reactivity; 1–5 Likert; higher = higher mindfulness; α = 0.80 overall; convergent validity with cognitive control, metacognition, emotion regulation).
• Stress: Japanese Perceived Stress Scale (PSS-14; past month; 0–4 Likert; higher = higher perceived stress; α = 0.89; convergent validity with depression r = 0.76).
• Hallucinatory experiences: AHES-17 (17 items; 0–4 Likert; higher = more auditory hallucinatory experiences; α = 0.85; convergent validity r = 0.69 with LSHS). Procedure: Surveys administered via SurveyMonkey on participants’ own computers. Demographics first (age, sex, mother tongue, neurological/psychiatric history). Scale order randomized. IMC items at the end (custom item per Oppenheimer et al., 2009). Participants instructed to respond in a quiet environment. Informed consent obtained. Compensation: 100 yen. Data analysis: Descriptive statistics, Pearson correlations, and structural equation modeling (SEM) were conducted. Fit indices considered: χ² (non-significant preferred), CFI, GFI, AGFI (≥0.90), RMSEA and SRMR (≤0.05). Due to large N sensitivity, non-χ² indices were emphasized. Preregistered analysis tested the hypothesized model with both mediators; exploratory analyses tested stress-only and mindfulness-only mediator models, including a reduced mindfulness model with three facets (observing, acting with awareness, non-judging). SEM implemented with lavaan 0.6-6 in R 4.0.2; other analyses in JASP 0.12.2.

Preregistered model: Mixed fit—some indices acceptable (GFI = 0.973, SRMR = 0.042, CFI = 0.963) but others poor (χ²(6) = 67.700, p < 0.001; AGFI = 0.752; RMSEA = 0.150). Stress did not mediate the relationship between schizotypy and hallucinatory experiences. Exploratory models:

• Stress-only mediation (Fig. 1C): Poor fit overall despite significant paths from schizotypy to stress and stress to hallucinatory experiences. Fit: GFI = 0.926; χ²(3) = 102.222, p < 0.001; CFI = 0.855; RMSEA = 0.269; SRMR = 0.136; AGFI = 0.630.
• Mindfulness trait (all five facets) mediation: Some indices acceptable (CFI = 0.972; SRMR = 0.033; GFI = 0.980) but others poor (χ²(3) = 44.196, p < 0.001; RMSEA = 0.173; AGFI = 0.700). Mediation via mindfulness supported.
• Mindfulness three-facet mediation (observing, acting with awareness, non-judging; Fig. 1D): Similar mixed fit (CFI = 0.964; SRMR = 0.044; GFI = 0.974; χ²(3) = 45.433, p < 0.001; RMSEA = 0.176; AGFI = 0.755). Best relative fit by AIC (19,425.249) compared to hypothesized model (27,348.274) and five-facet model (24,358.496). Descriptive outcomes: n = 458; key scale means (SD): Positive SPQ 8.42 (5.89), Negative SPQ 16.84 (8.08), Disorganized SPQ 4.98 (3.64), AHES-17 24.14 (11.73), PSS 23.14 (6.07). Correlations showed expected associations among schizotypy, mindfulness facets, stress, and hallucinatory experiences (see Table 2). Overall, mindfulness (particularly observing, acting with awareness, and non-judging) mediated schizotypy–hallucinatory experience links, whereas stress did not provide adequate mediation.

Findings indicate that dispositional mindfulness, not chronic perceived stress, primarily accounts for the association between schizotypy and hallucinatory experiences in this non-clinical sample. Specifically: (1) Positive schizotypy appears to increase observing, which relates positively to hallucinatory experiences, potentially by enhancing the vividness of internal/external experiences and biasing source judgments. (2) Positive schizotypy is associated with reduced non-judging, which may exacerbate evaluative engagement with vivid sensations and cognitions, further elevating hallucinatory experiences. (3) Negative and disorganized schizotypy relate to lower acting with awareness, which may impair self-monitoring of one’s actions and internal signals, aligning with accounts of disrupted self-monitoring in auditory hallucinations. Stress showed significant paths but the stress-only mediation model fit poorly; differences between acute experimentally induced stress (prior studies) and self-reported chronic stress (current study) may explain discrepancies. These results refine understanding of cognitive mechanisms in schizotypy-linked hallucinatory experiences and support models emphasizing attention to present-moment experiences and action monitoring.

A preregistered online survey in Japanese adults found that mindfulness traits mediate the relationship between schizotypy and hallucinatory experiences, whereas perceived chronic stress did not adequately mediate this relationship. A model emphasizing three mindfulness facets—observing, acting with awareness, and non-judging—best explained the data by relative fit. The work underscores the role of attention to current experiences and actions, consistent with self-monitoring accounts of hallucinations, and suggests that enhancing mindfulness facets may be a promising avenue for mitigating hallucinatory experiences. Future research should employ experimental and longitudinal designs, examine modality-general versus modality-specific mechanisms, and test generalizability across cultures and age groups.

Key limitations include: (1) Cross-sectional design precludes causal inference; no instrumental variables were used. (2) Hallucinatory experiences were assessed only in the auditory modality, limiting modality generalization. (3) Sample restricted to Japanese adults (mean age ~42), which may limit cultural and age generalizability; different schizotypy factor structures have been reported across populations. (4) Stress was measured as perceived stress over the past month (chronic), which may not capture effects of acute stress implicated in prior experimental work. The SEM models did not achieve complete fit, indicating unmodeled factors may be involved.