Background:
The CHA
2
DS
2
–VASc score was initially applied to stratify stroke risk in patients with atrial fibrillation (AF) and was found to be effective in predicting all-cause mortality outcomes. To date, it is still unclear whether circulating long non-coding RNAs (lncRNAs) as emerging biomarkers, can improve the predictive power of the CHA
2
DS
2
–VASc score in stroke and all-cause mortality.
Methods:
Candidate lncRNAs were screened by searching the literature and analyzing previous RNA sequencing results. After preliminary verification in 29 patients with AF, the final selected lncRNAs were evaluated by Cox proportional hazards regression in 192 patients to determine whether their relative expression levels were associated with stroke and all-cause mortality. The c-statistic, net reclassification improvement (NRI), and integrated discrimination improvement of the patients were calculated to evaluate the discrimination and reclassification power for stroke and all-cause mortality when adding lncRNA expression levels to the CHA
2
DS
2
–VASc score model.
Results:
Five plasma lncRNAs associated with stroke and all-cause mortality in AF patients were selected in our screening process. Patients with elevated H19 levels were found to have a higher risk of stroke (hazard ratio [HR] 3.264, 95% confidence interval [CI]: 1.364–7.813,
P
= 0.008). Adding the H19 expression level to the CHA
2
DS
2
–VASc score significantly improved the discrimination and reclassification power of the CHA
2
DS
2
–VASc score for stroke in AF patients. In addition, the H19 level showed a marginally significant association with all-cause mortality (HR 2.263, 95% CI: 0.889–5.760,
P
= 0.087), although it appeared to have no significant improvement for the CHA
2
DS
2
–VASc model for predicting all-cause mortality.
Conclusions:
Plasma expression of H19 was associated with stroke risk in AF patients and improved the discriminatory power of the CHA
2
DS
2
–VASc score. Therefore, lncRNA H19 served as an emerging non-invasive biomarker for stroke risk prediction in patients with AF.