. 2021 May 1;18:105. doi: 10.1186/s12974-021-02138-7

Table 2.

Association of baseline sGFAP and sNfL with clinical disability parameters in AQP4-IgG⁺ and MOG-IgG⁺ patients

	Patient group (n)	sGFAP				sNfL
	Patient group (n)	η_ρ² for interaction, p	Standardized effect size	β (95% CI)	p	η_ρ² for interaction, p	Standardized effect size	β (95% CI)	p
EDSS^a	AQP4-IgG⁺ (33)	0.10, p = 0.04	1.30	1.78 (0.52–3.04)	0.007	0.06, p = 0.11	1.09	1.58 (− 0.58–3.75)	0.15
EDSS^a	MOG-IgG⁺ (16)	0.10, p = 0.04	− 0.48	− 0.52 (− 2.26–1.30)	0.59	0.06, p = 0.11	− 0.29	− 0.43 (− 2.55–1.70)	0.69
MSFC^a	AQP4-IgG⁺ (25)	0.14, p = 0.03	− 1.28	− 0.73 (− 1.30 to − 0.16)	0.01	0.05, p = 0.20	− 1.75	− 1.05 (− 2.13−0.03)	0.06
MSFC^a	MOG-IgG⁺ (12)	0.14, p = 0.03	0.76	0.43 (− 0.46–1.32)	0.33	0.05, p = 0.20	− 0.37	− 0.22 (− 1.26–0.82)	0.67
9-HPT^a	AQP4-IgG⁺ (32)	0.11, p = 0.04	− 1.03	− 0.007 (− 0.013 to − 0.001)	0.03	< 0.01, p = 0.70	− 0.82	− 0.006 (− 0.017–0.005)	0.28
9-HPT^a	MOG-IgG⁺ (14)	0.11, p = 0.04	0.65	0.004 (− 0.004–0.013)	0.32	< 0.01, p = 0.70	− 0.47	− 0.003 (− 0.015–0.008)	0.56
PASAT^a	AQP4-IgG⁺ (27)	0.05, p = 0.19	− 1.00	− 12.7 (− 25.1 to − 0.3)	0.045	0.13, p = 0.03	− 1.86	− 23.0 (− 43.7 to − 2.4)	0.03
PASAT^a	MOG-IgG⁺ (13)	0.05, p = 0.19	0.21	2.7 (− 17.0–22.4)	0.78	0.13, p = 0.03	0.44	5.5 (− 15.9–26.9)	0.61
T25-FW^b	AQP4-IgG⁺ (30)	0.01, p = 0.61	0.19	0.027 (− 0.105–0.158)	0.69	0.01, p = 0.54	0.31	0.043 (− 0.179–0.265)	0.70
T25-FW^b	MOG-IgG⁺ (14)	0.01, p = 0.61	− 0.21	− 0.029 (− 0.212–0.154)	0.75	0.01, p = 0.54	0.89	0.122 (− 0.099–0.343)	0.27

^aLinear model using log-transformed sGFAP or sNfL values, including age as well as the log-transformed interval since the last attack as covariates. Furthermore, an interaction term of baseline sGFAP or sNFL (log-transformed) and group was included to assess the statistical significance of inter group differences

^bLinear model using log-transformed sGFAP or sNfL and log-transformed T25-FW values, including age as well as the log-transformed interval since the last attack as covariates. Furthermore, an interaction term of baseline sGFAP or sNFL (log-transformed) and group was included to assess the statistical significance of inter group differences

Note that a higher EDSS score indicates a worse functional status, whereas a higher MSFC score indicates a better functional status. The EDSS [23] is the most common score to rate global neurological dysfunction secondary to MS and NMOSD. The MSFC [24] is a more complex, multidimensional scoring system for neurological impairment in MS and NMOSD, which consists of three components. These components, which may each be used individually as well, are the 9-HPT, PASAT, and T25-FW. The 9-HPT assesses upper extremity function and dexterity. PASAT, in rating the processing speed of auditory input and calculation ability, quantifies cognitive impairment. T25-FW addresses lower extremity function based on walking speed

9-HPT 9-hole peg test, AQP4-IgG aquaporin-4 immunoglobulin G, β regression coefficient, CI confidence interval, η_ρ² partial eta-squared, EDSS expanded disability status scale, MOG-IgG myelin oligodendrocyte protein immunoglobulin G, MSFC multiple sclerosis functional composite, n number, NMOSD neuromyelitis optica spectrum disorder, PASAT paced auditory serial addition test, sGFAP serum glial fibrillary acidic protein, sNfL serum neurofilament light chain protein, T25-FW timed 25-foot walk