Prevalence, predictors, and patterns of patient reported non-motor outcomes six months after stroke: a prospective cohort study

Abstract

Background: Adverse non-motor outcomes have a major impact on patients and caregivers after stroke, but knowledge of their prevalence, predictors and patterns across multiple health domains remains limited; we therefore aimed to obtain these data in a large observational prospective cohort study.

Methods: We included data from the Stroke Investigation Group in North and Central London (SIGNAL) registry based at the University College London Hospitals (UCLH) Comprehensive Stroke Service which serves a multi-ethnic population of ∼1.6 million people. In adult patients diagnosed with acute stroke due to cerebral ischaemia or intracerebral haemorrhage (ICH) from January 2017 to January 2020 we evaluated non-motor outcomes (anxiety, depression, fatigue, sleep disturbance, social participation, pain, bowel dysfunction, bladder dysfunction, mood problems, communication problems, activities of daily living (ADL), memory and thinking problems) at 6-month follow-up. We evaluated baseline predictors in multivariable logistic regression, and correlations between domains using kappa statistics.

Findings: Follow-up was complete for 3080 of 3338 (92.3%) eligible surviving patients (2534 ischaemic stroke, 547 with ICH; mean age 71.2 years, 1379 (44.8%) female, 1774 (59.3%) white). The most prevalent adverse non-motor outcomes were fatigue 1756 (57%), reduced social participation 1694 (55%), sleep disturbance 1663 (54%), and constipation 1355 (44%). The rates of adverse non-motor outcomes in ⩾ 1, ⩾ 2, ⩾ 3, ⩾ 4, and ⩾ 5 domains were 2310 (75%), 1571 (51%), 1519 (49%), 1232 (40%), and 801 (26%), respectively. Factors associated with adverse non-motor outcomes included stroke due to ICH, stroke severity, previous stroke, or history of cardiovascular disease. We identified moderate correlations between fatigue and sleep disturbance (kappa = 0.72); memory and thinking impairment and reduced ADL (kappa = 0.68); and communication problems and ADL (kappa = 0.70).

Interpretation: Adverse non-motor outcomes are highly prevalent and often multiple at 6-months after stroke: 75% have at least one affected domain; fatigue, sleep disturbance, and reduced social participation each affect over 50% of survivors, and 26% of patients report ≥5 adverse outcomes. Our findings suggest an urgent need to better detect and mitigate these outcomes to improve quality of life after stroke.

Funding: The National Institute for Health and Care Research (NIHR) UCLH Biomedical Research Centre.

Keywords: Autonomic; Language; Life satisfaction; Memory outcomes; Neuropsychological; Non-motor outcomes; Patient reported health; Quality of life after stroke; Sensory; Sleep.

Fig. 1

Prevalence of adverse non-motor outcomes. Prevalence estimates were calculated using unadjusted proportion analysis with their 95% Confidence Intervals.

Fig. 2

Co-occurrence of adverse non-motor outcomes. Bars show the proportions of patients with each adverse non-motor outcome who have additional adverse non-motor outcomes in ≥1, ≥2, ≥3 ≥ 4 and ≥5 other domains.

Fig. 3

Patient characteristics associated with adverse non-motor outcomes. Each figure shows the results of multivariable analysis for individual non-motor domains across multiple outcomes, detailing the associated sociodemographic and clinical factors. Adjusted odds ratios (aORs) with 95% confidence intervals (CIs) are presented, where ORs greater than 1 indicate a significant positive association, while ORs equal to or less than 1 indicate no significant association. Each outcome domain is analysed separately, highlighting the diverse influences of sociodemographic and clinical variables on health outcomes. Logistic regression model of non-motor outcome domains, ranked by prevalence: A) Fatigue, B) Reduced Social Participation, C) Sleep Disturbance, and D) Constipation. The model is adjusted for clinical and sociodemographic characteristics (i.e., age, sex, stroke type, stroke severity, and cardiovascular risk factors).

Fig. 4

Correlations between adverse non-motor outcomes. Heatmap illustrating the correlations between individual adverse non-motor outcome domains. The intensity of the colour reflects the strength of the correlation, with red indicating a strong positive correlation and blue indicating a weak or no correlation.