. 2022 Jul 29;145(12):4398–4408. doi: 10.1093/brain/awac253

Table 2.

Brain regions and NfL correlations in MSA

Cross-sectional cases (n = 50)	Correlation		Multivariate regression
Brain region	rho	P	Adj. R²	P
Whole brain	−0.26	0.06	0.15	0.04^a
Pons	−0.17	0.214	0.11	0.05^a
Striatum	−0.56	<0.001	0.4	<0.001^a
Caudate subregion	−0.5	<0.001	0.23	<0.001^a
Putamen subregion	−0.5	<0.001	0.23	0.001^a
Cerebellum	−0.15	0.28	0.16	0.01^a
MCP width	−0.31	0.03	0.02	0.24
Longitudinal cases (n = 14)			Standardized β
Brain region			Standardized β
Whole brain	−0.390	0.027	−0.225	0.122
Pons	−0.560	<0.001^a	−0.484	0.004^a
Striatum	−0.536	0.002^a	−0.425	0.006^a
Caudate subregion	−0.53	0.001^a	−47.91	0.002^a
Putamen subregion	−0.45	0.007^a	−30.01	0.018^a
Cerebellum	−0.535	0.002^a	−0.444	0.009^a
MCP width	−0.594	<0.001^a	−0.555	0.002^a
LMEM rate of change	Est. FE	SE	R²	P
Brain region	Est. FE	SE	R²	P
Whole brain	1277.8	3013.2	0.15	0.67
Pons	184.3	134.3	0.05	0.2
Striatum	95.28	92.83	0.39	0.32
Caudate subregion	80.23	60.83	0.59	0.02^a
Putamen subregion	12.759	50.91	0.24	0.39
Cerebellum	737.7	1044.09	0.24	0.48
MCP width	0.34	0.16	0.23	0.04^a

Correlation analysis was performed by Spearman rank correlation, from which we obtained P and rho for the association between these brain areas and NfL levels in all MSA cases at baseline. The multivariate regression analysis was performed adjusting for disease duration, UMSARS scores and age at sample to obtain β (the slope), the correlation value with adjusted R² and P-value. LMEM analysis in the longitudinal cohort (n = 14, total number of scans = 33) was performed adjusting for disease duration, UMSARS and age at sample collection. Statistical significance was set at P < 0.05 and Bonferroni correction for multiple testing was used. An LMEM was also performed to determine the relationship between longitudinal brain volume rate of change and the extracted rate of plasma NfL change over time. Rates of change were extracted from LMEMs, using time (in years) as a fixed effect. Random slopes and intercepts of time per participant were included to produce individual regression coefficients per participant as previously described.²¹ Statistically significant values are in bold. Est.FE = estimated fixed effects; SE = standard error.

Survived Bonferroni correction for multiple testing.