There is scarce information on the complex and bidirectional relationship between sleep disorders and stroke in underserved populations. We assessed the prevalence of sleep-related symptoms in Atahualpa (a rural South American population) to determine the impact of stroke in the overall quality of sleep.
Operational definitions and ethics considerations of this study have been described elsewhere [1-3]. During a door-to-door survey, all Atahualpa residents aged ≥40 years with a stroke (case-patients) and a matched sample of healthy subjects (controls) were screened with validated field instruments to assess differences in the occurrence of sleep-related symptoms across both groups. Analysis of data was carried out using sas version 9·3 (SAS Institute Inc., Cary, NC, USA), and statistical significance was tested by the conditional logistic regression for matched pair data.
The study included 81 persons (27 stroke cases and 54 healthy subjects). Mean age was 71 ± 11 years, 52% were men, and 26% had psychological distress. As case-patients and controls were properly matched, there were no differences in age, gender, and psychological distress across both groups (Table 1). Sleep duration was shorter among case-patients than controls (P = 0·03). Otherwise, there were no significant differences in total sleep quality, daytime somnolence, and clinically significant insomnia across case-patients and controls.
|Stroke patients (n = 27)||Controls (n = 54)||Significance|
|Age (men ± SD)||70·6 ± 10·4||70·8 ± 10·6||Matched|
|Sleep hours (mean ± SD)||7·3 ± 1·3||8 ± 1·5||p = 0·03|
|Poor sleep quality (%)||37%||39%||n.s.*|
|Daytime somnolence (%)||56%||32%||n.s.*|
Our findings are in concordance with recent reports from industrialized nations showing an increased risk of cardiovascular diseases among persons with short sleep duration, probably related to endothelial dysfunction [4, 5]. In this cross-sectional survey, we could not assess causality. However, the fact that we included all persons with a stroke found in the community, together with the criteria used for selection of controls and the evaluation of sleep-related symptoms argue for the strength of our results. Further longitudinal studies in underserved populations are warranted to settle the cause-and-effect relationship of this association.