Examining the Factor Structure of the 39-Item and 15-Item Versions of the Five Facet Mindfulness Questionnaire Before and After Mindfulness-Based Cognitive Therapy for People With Recurrent Depression

The 39-item and 15-item versions of the Five Facet Mindfulness Questionnaire (FFMQ) show different factor structures before and after mindfulness-based cognitive therapy (MBCT) in adults with remitted recurrent depression. Before MBCT, a four-factor model (excluding the Observing facet) best fit the data; after MBCT, a five-factor model fit best. Configural invariance across time was not supported, meaning the questionnaire's structure changes with intervention, which may invalidate direct pre-to-post comparisons. Both versions had adequate internal consistency and sensitivity to change and did not differ in convergent validity. Researchers should consider omitting the Observing subscale when comparing scores before and after mindfulness interventions. The FFMQ-15 is a suitable brief alternative.

Study at a glance

Abstract

Research into the effectiveness and mechanisms of mindfulness-based interventions (MBIs) requires reliable and valid measures of mindfulness. The 39-item Five Facet Mindfulness Questionnaire (FFMQ-39) is a measure of mindfulness commonly used to assess change before and after MBIs. However, the stability and invariance of the FFMQ factor structure have not yet been tested before and after an MBI; pre to post comparisons may not be valid if the structure changes over this period. Our primary aim was to examine the factor structure of the FFMQ-39 before and after mindfulness-based cognitive therapy (MBCT) in adults with recurrent depression in remission using confirmatory factor analysis (CFA). Additionally, we examined whether the factor structure of the 15-item version (FFMQ-15) was consistent with that of the FFMQ-39, and whether it was stable over MBCT. Our secondary aim was to assess the general psychometric properties of both versions. CFAs showed that pre-MBCT, a 4-factor hierarchical model (excluding the “observing” facet) best fit the FFMQ-39 and FFMQ-15 data, whereas post-MBCT, a 5-factor hierarchical model best fit the data for both versions. Configural invariance across the time points was not supported for both versions. Internal consistency and sensitivity to change were adequate for both versions. Both FFMQ versions did not differ significantly from each other in terms of convergent validity. Researchers should consider excluding the Observing subscale from comparisons of total scale/subscale scores before and after mindfulness interventions. Current findings support the use of the FFMQ-15 as an alternative measure in research where briefer forms are needed.