Table 2.

Linear mixed model estimates of within- and between-arm changes in log_e plasma NfL (observed data analysis).

		Estimate [95 % CI]
		Model 1	Model 2	Model 3	Model 4
Within-arm estimated change
HSMP	V2 – V1	−0.07 [−0.14 to 0.00]	−0.07 [−0.14 to 0.00]	−0.07 [−0.14 to 0.00]	−0.06 [−0.14 to 0.01]
	V3 – V1	−0.11 [−0.18 to −0.04]	−0.11 [−0.18 to −0.04]	−0.10 [−0.18 to −0.03]	−0.10 [−0.18 to −0.02]
CMBAS	V2 – V1	−0.04 [−0.11 to 0.02]	−0.04 [−0.11 to 0.02]	−0.05 [−0.11 to 0.02]	−0.05 [−0.12 to 0.02]
	V3 – V1	−0.02 [−0.09 to 0.04]	−0.02 [−0.09 to 0.04]	−0.03 [−0.10 to 0.04]	−0.03 [−0.10 to 0.04]
Between-arm estimated change
HSMP – CMBAS	V2 – V1	−0.02 [−0.12 to 0.07] p = 0.64	−0.02 [−0.12 to 0.07] p = 0.63	−0.02 [−0.12 to 0.08] p = 0.65	−0.02 [−0.12 to 0.09] p = 0.77
	V3 – V1	−0.08 [−0.18 to 0.02] p = 0.11	−0.08 [−0.18 to 0.02] p = 0.10	−0.08 [−0.18 to 0.02] p = 0.13	−0.07 [−0.17 to 0.04] p = 0.21
Data included in models
Participants		137	137	132*	124†
Observations		370	370	360	336

The unadjusted model (Model 1) included random participant intercepts (all other covariates were fixed effects), as well as trial site, visit, trial arm, and the visit × arm interaction. Model 2 adjusted Model 1 for age at baseline, sex, and education level. Model 3 adjusted Model 2 for renal insufficiency, as well as baseline BMI and PASE. Model 4 adjusted Model 3 for APOE. Continuous covariates were mean-centered. CI = confidence interval. CMBAS = caring mindfulness-based approach for seniors. HSMP = health self-management program. NfL = neurofilament light. PASE = physical activity scale for the elderly. V1 = baseline visit. V2 = post-intervention visit. V3 = follow-up visit. APOE = apolipoprotein E gene. P-tau-181 = tau phosphorylated at threonine 181.

^⁎Model 3 included 5 fewer participants versus Models 1–2 due to missingness in the PASE covariate.

^†Model 4 included 8 fewer participants versus Model 3 due to missingness in the APOE (n = 6) and P-tau-181 (n = 2) covariates.