Latent Profile Analysis of the Five Facet Mindfulness Questionnaire (FFMQ) Subscales in a Naturalistic Sample of Patients with Past or Present Major Depression: A Replication and Extension Study
Jelle Lubbers, Philip Spinhoven, Mira B. Cladder-Micus, Jan Spijker, Anne E.M. Speckens, Dirk E. M. Geurts · 2024 · 6 citationsRead the paper
Abstract Objectives Mindfulness is a multifaceted concept which might be best captured by person-centered profiles rather than by variable-centered assessment of its subcomponents. Patients with different mindfulness profiles may vary in mental health status and might experience differential clinical outcomes of Mindfulness-Based Cognitive Therapy (MBCT). We therefore aimed to assess the relationship between mindfulness profiles and mental health in patients with major depressive disorder (MDD) and to examine whether mindfulness profiles are predictive for MBCT-induced changes in mental health. Method Latent profile analysis (LPA), a method used to identify hidden subgroups of individuals within a population, was performed on pre-MBCT subscale scores of the Five Facet Mindfulness Questionnaire in patients with current or remitted MDD (n=754). Equality of means of pre-MBCT, post-MBCT, and residualized change scores of several mental health measures across the latent profiles was tested. Results LPA identified four distinct mindfulness profiles. Three profiles were similar to previously identified profiles in another MDD sample (“very low mindfulness,” “high mindfulness,” “non-judgmentally aware”) and one profile was similar to that identified in non-clinical populations (“judgmentally observing”). The “high mindfulness” subgroup scored best, the “very low mindfulness” worst, and the other subgroups intermediate on mental health in terms of depressive symptoms, worry, overall functional impairment, and self-compassion. Mindfulness profiles were not predictive of MBCT-induced changes in mental health. Conclusions Mindfulness profiles were differentially related to mental health, but were not predictive of treatment outcome. Future research would benefit from longitudinal assessment of latent mindfulness profiles to examine whether a patient’s profile changes after MBCT and whether hypothesized change in profile would be related to treatment outcome. Preregistration This study was not preregistered.
What the full text shows
Tested in real conditions- Latent profile analysis identified four distinct mindfulness profiles in patients with MDD: 'high mindfulness', 'very low mindfulness', 'non-judgmentally aware', and 'judgmentally observing'. — 4 profiles (n=754 patients with past or present MDD)
- The 'high mindfulness' subgroup exhibited the best mental health outcomes, while the 'very low mindfulness' subgroup exhibited the worst, across depressive symptoms, worry, functional impairment, and self-compassion. (Baseline and post-MBCT measurements)
- Baseline mindfulness profiles did not predict the rate of improvement (residualized change) in depressive symptoms or secondary mental health outcomes following MBCT. — p > 0.05 for all overall Wald chi-square tests of residualized change (n=500 patients with full pre- and post-MBCT datasets)
- The 'judgmentally observing' profile was the most prevalent subgroup in the sample. — 48.5% of the sample (n=366) (n=754 patients with MDD)
- Setup
- Naturalistic cohort of 754 patients with past or present Major Depressive Disorder (MDD) receiving Mindfulness-Based Cognitive Therapy (MBCT) at Radboud University Medical Centre.
- How
- The researchers used Latent Profile Analysis (LPA) on Five Facet Mindfulness Questionnaire (FFMQ) subscale scores to identify subgroups. They then tested for differences in mental health outcomes and residualized change scores using Wald chi-square tests.
- Limitations
- Low entropy value (0.62), indicating less accurate identification of latent profiles. · Fit indices were inconsistent, with some favoring a 7-profile solution over the selected 4-profile solution. · High attrition rate, with 34% of patients not completing post-MBCT questionnaires. · Use of the 24-item FFMQ-SF instead of the full 39-item FFMQ.
Summarized by Gemma 4 from the open-access full text · 8 Oct 2026. Read it