AIMS: Early detection and management of sleep disturbances can improve postoperative outcomes given the high prevalence of sleep disturbances and unrecognized cognitive impairment in older surgical patients. There is an association between sleep disturbances and cognitive impairment in the general population. However, the relationship in older surgical patients has not been systematically investigated. The objective of this study was to assess the prevalence and trajectory of preoperative and postoperative sleep disturbances in older surgical participants with and without suspected cognitive impairment (sCI). METHODS: Two hundred and fifty-two participants aged ≥ 65 years undergoing non-cardiac surgery were recruited. The primary outcome was the prevalence and trajectory of sleep disturbances measured by the Pittsburgh Sleep Quality Index (PSQI) in participants with and without sCI preoperatively, 30, 90, and 180 days postoperatively. The main exposure, preoperative sCI, was operationalized as screening positive on one or more of the following cognitive screening tools: Centers for Disease Control and Prevention cognitive question (answered "yes"), Ascertain Dementia Eight-item Questionnaire (≥2), Telephone Montreal Cognitive Assessment (≤18), and Modified Telephone Interview for Cognitive Status (≤31). Sleep disturbances were defined as a PSQI score >
5. Mixed effects logistic regression models with random intercepts were used for the dichotomous outcome of sleep disturbances. RESULTS: One hundred and eight participants (43%) screened positive for preoperative sCI. The prevalence of preoperative sleep disturbances was higher in participants with sCI versus without (63% vs 47%, P = 0.02). Postoperatively in both groups, the prevalence of sleep disturbances was lower at 30, 90, and 180 days, compared to the preoperative assessment and overall trajectories did not differ significantly. Female sex and depression were associated with poorer postoperative sleep, regardless of cognitive status. CONCLUSION: Sleep disturbances and suspected cognitive impairment are highly prevalent in surgical cohorts. Targeting conditions such as depression which affect sleep, may improve postoperative outcomes.