Table 3.

Bayes estimation of BDNF and mortality association mediated by cortisol level among all subjects and by 75th percentile of age using Monte Carlo Markov Chain (MCMC) method^a.

	All subjects (N = 117)		Age ≤ 48 years (N = 90)		Age >48 years (N = 27)
Pathwayb	HR/OR (95% CI)	P-value	HR/OR (95% CI)	P-value	HR/OR (95% CI)	P-value
BDNF to Cortisol (Logistic)	4.12 (1.86–9.24)	0.001	7.48 (3.22–17.92)	<0.001	1.44 (0.43–4.80)	0.185
Cortisol to mortality (Cox)	3.05 (1.70–5.45)	<0.001	2.99 (1.27–7.00)	0.017	3.61 (1.53–8.97)	0.006
BDNF to mortality-Total Effect (Cox)	1.94 (1.19–3.12)	0.011	2.89 (1.42–5.70)	0.004	1.47 (0.73–2.95)	0.221
BDNF to mortality-Direct Effect (Cox)	1.58 (0.90–2.78)	0.106	1.78 (0.72–4.42)	0.185	1.72 (0.78–3.76)	0.159
Mediation percentage (95% CI)c	29.92%		45.45%		−41.10%
	(19.58–40.25%)		(34.05–56.85%)		(−93.52%, 11.33%)

^aAnalysis was performed with weekly average CSF BDNF in ng/mL (≥75th percentile, <75th percentile); cortisol using trajectory analysis grouping (low vs. high); Prior distribution of coefficients for each corresponding pathway (Table 2) was incorporated in MCMC Bayes estimation, which resamples prior distributions to derive a more precise and valid parameter estimate.

^bAll pathways were adjusted for 75th percentile age (<48 vs. ≥48 years old), Glasgow Coma Scale (GCS) and Gene Risk Score (GRS). For “BDNF to mortality-Direct Effect,” cortisol using trajectory analysis grouping (low vs. high) was additionally adjusted.

^cThe mediation percentage was calculated with the following equation: Mediation percentage = {[ln(OR_{Total Effect}) − ln(OR_{Direct Effect})]/ln(OR Total Effect)}^*100.

Bolded p-values indicate statistical significance.