Sleep regularity index as a novel indicator of sleep disturbance in stroke survivors: a secondary data analysis

Abstract

Sleep disturbance is common but often overlooked after stroke. Regular sleep is increasingly recognised as important for overall health, yet little is known about how sleep regularity changes after stroke. This study examined differences in the Sleep Regularity Index (SRI) between stroke survivors and healthy controls using actigraphy data from an existing dataset (~ 1 week per participant). Data were analysed for 162 stroke survivors (mean age 61 ± 14 years, 5 ± 5 years post-stroke, 89 males) and 60 controls (mean age 57 ± 17 years, 32 males). Stroke survivors had significantly lower SRI scores than controls (p = 0.001), indicating less regular sleep. In the stroke group, higher SRI correlated with longer total sleep time (p = 0.003) and better self-reported sleep quality (p = 0.001) but not with other sleep metrics. Lower SRI was associated with worse depressive symptoms (p = 0.006) and lower quality of life (p = 0.001) but not with disability (p = 0.886) or time since stroke (p = 0.646). These findings suggest that sleep regularity is disrupted post-stroke and may influence well-being. Future research should explore interventions to improve sleep regularity and related health outcomes in stroke survivors.

Keywords: Actigraphy; Circadian rhythm; Post-stroke recovery; Quality of life; Sleep-wake patterns.

Fig. 1

(A) Relationship between sleep regularity (SRI) and total sleep time (TST) for the stroke cohort (dashed line) and healthy controls (solid line). Plots that show the sleep/wake cycle according to the SRI calculation for four participants, two classified as irregular sleepers (B = stroke, C = control), with SRI values in the first quintile, and two regular sleepers (D = stroke, E = control), with SRI values in the fifth quintile. The colours reflect ‘irregular’ (pink), ‘normal’ (orange) and ‘regular’ (teal) sleepers. The figure displays results from the current analysis, though the design is based on similar figures created by other studies^,.

Fig. 2

Violin plot showing the distribution of SRI scores in the stroke survivors and healthy controls. Higher SRI is indicative of more regular sleep. There was a significant effect of group, adjusting for age and sex (p = 0.001).

Fig. 3

A scatterplot of the correlation between SCI and SRI for the whole dataset (including both stroke survivors and healthy controls), with better self-reported sleep associated with more regular sleep: r = 0.25, p < 0.001.

Fig. 4

Plotted residuals accounted for age and sex to demonstrate partial correlations for the stroke cohort between the sleep regularity index (SRI) and (A) total sleep time (TST), (B) wake after sleep onset (WASO), (C) fragmentation index, (D) sleep efficiency, (E) sleep condition indicator (SCI), α = 0.01 (Bonferroni corrected).

Fig. 5

Plotted residuals accounted for age and sex to demonstrate partial correlations for the healthy control cohort between the sleep regularity index (SRI) and (A) and total sleep time (TST), (B) wake after sleep onset (WASO), (C) fragmentation index, (D) sleep efficiency, (E) sleep condition indicator (SCI), α = 0.01 (Bonferroni correction for multiple comparisons).

Fig. 6

Plotted residuals controlled for age and sex to demonstrate partial correlations between (A) SRI and the Patient Health Questionnaire depression scale (PHQ-8) (n = 113 stroke survivors). (B) SRI and the Stroke Impact Scale (n = 54 stroke survivors).