Effects of a mindfulness-based versus a health self-management intervention on objective cognitive performance in older adults with subjective cognitive decline (SCD): a secondary analysis of the SCD-Well randomized controlled trial

Abstract

Background: Older individuals with subjective cognitive decline (SCD) perceive that their cognition has declined but do not show objective impairment on neuropsychological tests. Individuals with SCD are at elevated risk of objective cognitive decline and incident dementia. Non-pharmacological interventions (including mindfulness-based and health self-management approaches) are a potential strategy to maintain or improve cognition in SCD, which may ultimately reduce dementia risk.

Methods: This study utilized data from the SCD-Well randomized controlled trial. One hundred forty-seven older adults with SCD (M_Age = 72.7 years; 64% female) were recruited from memory clinics in four European countries and randomized to one of two group-based, 8-week interventions: a Caring Mindfulness-based Approach for Seniors (CMBAS) or a health self-management program (HSMP). Participants were assessed at baseline, post-intervention (week 8), and at 6-month follow-up (week 24) using a range of cognitive tests. From these tests, three composites were derived-an "abridged" Preclinical Alzheimer's Cognitive Composite 5 (PACC5_Abridged), an attention composite, and an executive function composite. Both per-protocol and intention-to-treat analyses were performed. Linear mixed models evaluated the change in outcomes between and within arms and adjusted for covariates and cognitive retest effects. Sensitivity models repeated the per-protocol analyses for participants who attended ≥ 4 intervention sessions.

Results: Across all cognitive composites, there were no significant time-by-trial arm interactions and no measurable cognitive retest effects; sensitivity analyses supported these results. Improvements, however, were observed within both trial arms on the PACC5_Abridged from baseline to follow-up (Δ [95% confidence interval]: CMBAS = 0.34 [0.19, 0.48]; HSMP = 0.30 [0.15, 0.44]). There was weaker evidence of an improvement in attention but no effects on executive function.

Conclusions: Two non-pharmacological interventions conferred small, non-differing improvements to a global cognitive composite sensitive to amyloid-beta-related decline. There was weaker evidence of an effect on attention, and no evidence of an effect on executive function. Importantly, observed improvements were maintained beyond the end of the interventions. Improving cognition is an important step toward dementia prevention, and future research is needed to delineate the mechanisms of action of these interventions and to utilize clinical endpoints (i.e., progression to mild cognitive impairment or dementia).

Trial registration: ClinicalTrials.gov, NCT03005652.

Keywords: Cognition; Compassion; Mindfulness; Randomized controlled trial; Subjective cognitive decline.

Fig. 1

Consort flow diagram of enrollment and randomization to CMBAS and HSMP. The ns analyzed and excluded reflect the PP analyses. “Analyzed” participants were those with ≥ 2 observations for the respective measure (i.e., used to estimate the change in the outcome). While the LMMs also included participants who had baseline data only, these data were used solely for the estimation of intercepts (see Additional file 1: Table S2 for ns with non-missing baseline observations). CMBAS, Caring Mindfulness-Based Approach for Seniors; HSMP, Health Self-Management Program; V2, post-intervention; V3, follow-up; PP, per-protocol; PACC5_Abridged, Abridged Preclinical Alzheimer Cognitive Composite 5; Att. Comp., attention composite; Exec. Comp, executive composite; DRS-2, Mattis Dementia Rating Scale-2; RAVLT, Rey Auditory Verbal Learning Test; Coding, Wechsler Adult Intelligence Scale-IV Coding; Cat. fluency, category fluency; Lett. fluency, letter fluency; TMT, Trail-Making Test; Stroop interfer., Stroop interference; Stroop incongr., Stroop incongruent; MST Recog., Mnemonic Similarities Task Recognition

Fig. 2

Estimated change in cognitive composite scores for each trial arm. The graphs visualize the trajectories modeled using the PP linear time LMMs. The cognitive retest effect parameters were omitted from the graphed models, as these resulted in discontinuous trajectories. The time-by-arm interaction was not significant for any composite (ps > 0.29), although PACC5_Abridged scores increased in both arms during the trial (p < 0.001). In order to aid interpretability, the graphed data are for a “prototypical” female participant with sample grand mean values for age, education, state anxiety, and depressive symptoms, at the Barcelona site. Shaded areas are 95% confidence intervals for the fixed effects. Abbreviations: PACC5_Abridged, Abridged Preclinical Alzheimer Cognitive Composite 5; CMBAS, Caring Mindfulness-Based Approach for Seniors; HSMP, Health Self-Management Program; LMM, linear mixed model; PP, per-protocol