Z-score neurofeedback, heart rate variability biofeedback, and brain coaching for older adults with memory concerns

Figures

• 

  Fig. 1. Study Participants Flowchart. Number of participants is shown for each phase and assessment point of the trial. Reasons for exclusion of participants and summaries of the most common reasons for withdrawal/loss to follow-up are shown on the right.

• 

  Fig. 2. Summary of the Study and the Memory Boot Camp Program. A summary of each phase of the study is displayed on the left. The right panel is a summarized description of the intervention phase of this trial: The Memory Boot Camp program. MoCA: Montreal Cognitive Assessment; BAI: Beck Anxiety Inventory; BDI-ii: Beck Depression Inventory-II; ESS: Epworth Sleepiness Scale; ISI: Insomnia Severity Index; PSQI: Pittsburgh Sleep Quality Index; WSAS: Work and Social Adjustment Scale.

• 

  Fig. 3. Change in Neurocognitive and Self-Report Measures across Timepoints. LS mean values for neurocognitive and self-report questionnaire scores at each timepoint: -3 months (baseline assessment), 0 months (pre-treatment assessment), 3 months (post-treatment assessment) and 9 months (follow-up assessment). Pairwise LS mean change values between timepoints are summarized in the top of each graph (values from Table 2). For all neurocognitive and self-report measures except for the NeuroTrax, there was a score change observed after the active treatment period (gray shading) in a direction indicating improvement. Error bars represent the standard error. MoCA: Montreal Cognitive Assessment; NeuroTrax GCS: Global Cognitive Score; BAI: Beck Anxiety Inventory; BDI-ii: Beck Depression Inventory-II; ESS: Epworth Sleepiness Scale; ISI: Insomnia Severity Index; PSQI: Pittsburgh Sleep Quality Index; WSAS: Work and Social Adjustment Scale. ns: p > 0.0125, change not significant at the Bonferroni-corrected significance level. * p≤0.0125. ** p≤0.008. *** p≤0.001. **** p≤0.0001.

• 

  Fig. 4. Change in HRV Physiological Measures across Timepoints. LS mean values for HRV parameters and breathing rate at each timepoint: -3 months (baseline assessment), 0 months (pre-treatment assessment), 3 months (post-treatment assessment) and 9 months (follow-up assessment). Pairwise LS mean change values between timepoints are summarized in the top of each graph (values from Table 3). Respiration rate decreased during the active treatment period (gray shading). Ln(SDNN), ln(RMSSD), and ln(HF) LS mean values increased over the control waiting period. Although not specifically labeled, there was no change over the waiting period for ln(VLF) or ln(LF). Error bars represent the standard error. VLF: power in the very low frequency domain; LF: power in the low frequency domain; HF: power in the high frequency domain. SDNN: standard deviation normal to normal beat intervals; RMSSD: root mean square of successive normal to normal beat interval differences. aThese LS mean values have been back-transformed from the natural log values displayed in the top graph. Error bars are not included because they are not appropriate for back-transformed data. Significance markers are not included on…

• 

  Fig. 5. Change in QEEG Metrics across Timepoints. For participants with SOIs at trained sites, LS mean |z-score| values at each timepoint: -3 months (baseline assessment), 0 months (pre-treatment assessment), 3 months (post-treatment assessment) and 9 months (follow-up assessment) are shown for the QEEG parameters absolute power and relative power percent for the frequency bands delta, theta, beta, and high beta. Pairwise LS mean change values between timepoints are summarized in the top of each graph (values from Table 4). There was no change in |z-score| over the control or follow-up period for any parameter. For absolute power of delta, absolute power of theta, relative power of theta, absolute power of beta, absolute power of high beta, and relative power of high beta, there was a significant decrease (improvement) in |z-score| over the treatment period (gray shading). Error bars represent the standard error. ns: p > 0.0125, change not significant at the Bonferroni-corrected significance level. * p≤0.0125. ** p≤0.008. *** p≤0.001. **** p≤0.0001.

• 

  Fig. 6. Correlation between Changes in MoCA Score and Physiological Metrics over the Treatment Period. Data points for participants’ change from pre-treatment to post-treatment (post minus pre) on MoCA score (Y-axis) versus each physiological metric (X-axis) are graphed. Reference lines are drawn at X = 0 and Y = 0, which divide the graph into quadrants. Points that demonstrate improvement in both graphed measures (where possible) are shown in black, while those that demonstrate no change or decline are shown in gray. For 12 of the 15 graphs, points in a single quadrant are black and points in the other three quadrants are gray, because an improvement was defined for both the X-axis and Y-axis measures. For change in MoCA score, an improvement was an increase in score. For change in heart rate, respiration rate, and all eight QEEG metrics, an improvement was a decrease. For the HRV metrics ln(SDNN) and ln(RMSSD), an improvement was an increase. For the three graphs representing change in ln(VLF power), ln(LF power), and ln(HF power), two quadrants (top left and top right) contain black points and two quadrants (bottom left and bottom right) contain gray dots because an improvement in the…