Table 4.

Model 2: The association of mean non-inflammatory sNfL level with MRI atrophy and clinical measures at the 10-year follow-up, corrected for age, sex, DMT use, eTIV, BL T2LC and BL Gd+ LC*

MRI/clinical measure		Mean non-inflammatory sNfL level				Full model
	N	B	Std. B	95% CI	P value†	R2 adj.	P value
Total GM volume	36				NS
Total WM volume	36				NS
Total deep GM volume	36				NS
Thalamus volume	36				NS
Mean Cth	36				NS
logLesion volume‡	36				NS
Lesion count	36				NS
EDSS≥4§	40				NS
logT25FW‡	38				NS
logChange in T25FW‡	38				NS
logD9-HPT‡	38				NS
logChange in D9-HPT‡	38				NS
logND9-HPT‡	37				NS
logChange in ND9-HPT‡	37				NS
PASAT	40				NS
Change in PASAT	40				NS
Oral SDMT	35	0.548	0.473	0.196 to 0.900	0.003	0.380	<0.001

*Non-significant covariates removed from final model by backward elimination.

†Adjusted p values after controlling the false discovery rate (FDR) for multiple hypothesis testing.

‡Dependent variable log transformed due to non-normality (log-linear transformation).

§Analysed by logistic regression.

adj, adjusted; B, beta; BL, baseline; Cth, cortical thickness; D9-HPT, dominant hand Nine-Hole Peg Test; DMT, disease modifying therapy; EDSS, Expanded Disability Status Scale; eTIV, estimated total intracranial volume; Gd+, gadolinium-enhancing; GM, grey matter; LC, lesion count; ND9-HPT, non-dominant hand Nine-Hole Peg Test; PASAT, Paced Auditory Serial Addition Test; SDMT, Symbol Digit Modalities Test; sNfL, serum neurofilament light chain; Std, standardised; T25FW, timed 25-foot walk; WM, white matter.