Association of serum biomarker neurofilament light concentration with post-stroke depression: A preliminary study.

OBJECTIVE

To investigate serum neurofilament light (sNfL) levels in acute ischemic stroke and to assess whether sNfL are related to the severity of disease and a potential prognostic marker of post-stroke depression (PSD) during a 3-month follow-up period.

METHODS

This was a single-center prospective cohort study. The sNfL concentration was measured in baseline samples using the Simoa platform- Single Molecule Array technology. A psychiatrist administered the Structural Clinical Interview for Diagnostic and Statistical Manual IV to all patients and made a diagnosis of PSD 3 months after stroke. The logistic regression was used to examine the association between sNfL and PSD.

RESULTS

In total, 236 ischemic stroke cases were included and finished the follow-up. In the follow-up, 55 patients were defined as PSD, thus the incidence rate was 23.3% (95% confidence intervals [CI]: 17.9%-28.7%). Significant differences were observed between the sNfL levels in patients with PSD (124.8 pg/ml [interquartile range {IQR}: 59.6-159.2]) and in patients without PSD (35.9 pg/ml [IQR: 18.2-60.4]) levels (P < 0.001). After adjusting for age, family history of depression, marital status, National Institutes of Health and Stroke Scale score, C-reactive protein and homocysteine levels, sNfL levels independently predicted the development of post-stroke depression. The crude and adjusted odds ratios [OR] (and 95%CI) of PSD associated with an IQR increase for sNfL were 3.38(2.29, 4.98) and 2.65(1.59, 4.04), respectively. According to receiver operating characteristic curves (ROC) curves, the cut-off value of sNfL to predict PSD was 111.4 pg/ml with an area under the curve (AUC) of 0.84(95% CI, 0.78-0.90) and with the highest sensitivity (61.8%) and specificity (95.4%).

CONCLUSIONS

In this study, elevated level of sNfL is associated with higher risk of 3-month depression in patients with ischemic stroke and makes early diagnoses of depression. The study needs replication to ensure the validity of our preliminary results.