Objective: To evaluate clinical phenotypes and nocturnal sleep structure characteristics with mild cognitive impairment due to Alzheimer's disease (AD). Methods: Based on cerebrospinal fluid or positron emission tomography imaging to detect the concentration or deposition status of β-amyloid protein (Aβ) to distinguish between MCI due to AD and control group. All patients underwent polysomnography monitoring (PSG), APOE gene testing, Montreal Cognitive Assessment (MoCA), Hamilton Anxiety Scale (HAMA), and Hamilton Depression Scale (HAMD) assessments. Results: ① Significant differences were found in clinical characteristics such as sleep disturbance, body mass index (BMI), education years, occupations, combined with risk factors, and APOEε4 gene mutation (P<0.05); ② There were statistically significant differences in MoCA and HAMA scores between these two groups (P<0.05); ③Significant differences in N1/total sleep time (TST)%, N2/TST%, N3/TST% and rapid eye movement sleep time (REM)/TST%, average blood oxygen concentration during night-sleep time, sleep-related apnea-hypopnea index (AHI), obstructive sleep apnea (OSA) index between these two groups (P<0.05); ④Clinical phenotypes including sleep disorders, APOEε4 mutation, multiple risk factors, cognitively demanding occupations and sleep microstructural parameters were significantly correlated with MCI due to AD (P<0.05); ⑤Gender, sleep disturbance, APOEε4 gene mutation, BMI and polysomnographic indices (N1/TST%, N2/TST%, REM/TST% and OSA index) were key factors in MCI due to AD (P<0.05); ⑥N1/TST%, N2/TST%, REM/TST% and OSA index demonstrated significant linear correlations with MoCA scores (P<0.01). Conclusion: Patients with MCI due to AD exhibited distinct clinical features, including a higher prevalence of sleep disorders, elevated APOEε4 allele carrier rates, lower prevalence of multiple risk factors, a higher proportion of cognitively demanding occupations and lower BMI. Sleep parameters, including N1/TST%, N2/TST%, REM/TST% and OSA index, demonstrated significant correlations with cognitive impairment. Characteristic changes in the proportion of N1, N2 and REM sleep among MCI patients may serve as potential biomarkers for early identification of MCI due to AD.