A longitudinal study of polysomnographic variables in patients with mild cognitive impairment converting to Alzheimer's disease.

The main condition at increased risk of dementia is considered to be mild cognitive impairment. Mild cognitive impairment has been defined as a transitional state between normal aging and dementia, of which it may represent a prodrome. The aim of our study was to evaluate whether sleep variables (both conventional and microstructural ones) in subjects with mild cognitive impairment correlate with conversion to dementia. Nineteen subjects with amnestic mild cognitive impairment (mean age 68.5 ± 7.0 years) and 11 cognitively intact healthy elderly individuals (mean age 69.2 ± 12.6 years) underwent ambulatory polysomnography for the evaluation of nocturnal sleep architecture and cyclic alternating pattern parameters. Amnestic mild cognitive impairment subjects were clinically and cognitively re-evaluated after 2 years, during routine follow-up, and further classified as amnestic mild cognitive impairment converters (that is, patients developing Alzheimer's disease, N = 11) and amnestic mild cognitive impairment non-converters. Compared with healthy elderly individuals, amnestic mild cognitive impairment showed disrupted sleep with decreased rapid eye movement sleep, cyclic alternating pattern rate and cyclic alternating pattern slow-wave-related phases (A1 index). Standard sleep architecture analysis did not show significant differences between the two subgroups of amnestic mild cognitive impairment, whereas cyclic alternating pattern analysis showed that cyclic alternating pattern rate, A1 index and A3 index are significantly reduced in converters compared with non-converters. Our data confirm that in amnestic mild cognitive impairment subjects there is a sleep impairment, particularly when considering more refined sleep parameters and that sleep variables at baseline are different among converters versus non-converters at the 2-year follow-up. Specific sleep alterations might represent potential further biomarkers for the diagnosis and prognosis of early-phase cognitive impairment.