AbstractBackgroundSince sleep is important for physiological memory function, sleep disturbance may contribute to subjective memory complaints and objective memory impairments independent of brain pathology. We investigated the diagnostic utility of the Pittsburgh Sleep Quality Index (PSQI) as an independent test to differentiate patients with and without dementia at the initial clinical diagnostic interview in dedicated memory clinics.MethodsPSQI was administered to consecutive new patient referrals to two memory clinics over a 2‐year period, independent of other tests (interview, neuropsychology, imaging) which were used to establish diagnoses according to standard diagnostic criteria (DSM‐IV), and results were compared.ResultsIn a cohort of 310 patients, 50% fulfilling clinical diagnostic criteria for dementia, PSQI scores and subscores showed suboptimal sensitivity, specificity, positive and negative predictive values (PPV, NPV) and area under the Receiver Operating Characteristic (ROC) curve for the diagnosis of dementia, and likelihood ratios were unimportant. However, the proportion of poor sleepers was significantly larger in the non‐demented as compared to the demented patient group.ConclusionsPSQI scores were suboptimal for dementia diagnosis using traditional diagnostic parameters, but were pragmatically useful in identifying non‐demented ‘memory complainers’ with poor sleep quality. This observation may have therapeutic implications for these patients. Copyright © 2009 John Wiley & Sons, Ltd.