A Mixed-Methods Narrative Review of Feasibility, Acceptability, and Mechanism-Level Studies of Mindfulness-Based Interventions in People with Dementia and Cognitive Impairment

Abstract Objectives Mindfulness-based interventions (MBIs) are increasingly proposed as non-pharmacological approaches to support psychological well-being among people experiencing dementia or memory-related decline, yet the evidence base remains fragmented. This mixed-methods narrative review examined the feasibility, acceptability and potential mechanisms of MBIs for individuals with subjective cognitive decline (SCD), mild cognitive impairment (MCI), cognitive impairment (CI) and dementia. Method Searches of MEDLINE, CINAHL and SCOPUS (inception–2024) identified qualitative and quantitative studies evaluating MBSR, MBCT or adapted mindfulness programmes that reported feasibility, acceptability or mechanisms of action. Study quality was assessed using CASP tools, and findings were integrated using a parallel-results convergent synthesis. Twenty-one studies met inclusion criteria. Results Across populations, MBIs demonstrated high session attendance and favourable acceptability, although home-practice adherence decreased with more advanced impairment. Within included studies, evidence for improvements in psychological outcomes (e.g. stress, agitation, affect, worry) was mixed and largely derived from small, underpowered pilot trials. Preliminary mechanistic signals—such as enhanced non-reactivity, attentional regulation and reduced stress markers—were most evident in SCD and MCI groups. Qualitative findings indicated shifts in emotional regulation, acceptance of memory loss and greater perceived agency in daily coping. Conclusions Together, the convergence of quantitative and qualitative evidence suggests that MBIs are feasible and valued by people with memory loss, but robust mechanistic and efficacy claims are premature. Larger, theoretically informed trials with mechanistic measures are needed to determine who benefits, how and under what conditions. Preregistration This study is not preregistered.