Background   Computerized cognitive training has gradually become an important method to prevent and alleviate the cognition decline of patients with mild cognitive impairment (MCI) due to its unique characteristics of precision, diversification and individualization.

  Objective   To explore the effect of computerized cognitive training with different training duration and frequency on the improvement of the cognitive function and daily life ability of the elderly with MCI.

  Methods   Military retirees (≥ 60 years) who admitted to the Memory Clinic of the Eighth Medical Center of Chinese PLA General Hospital from January to December in 2020 were continuously enrolled. Patients were divided into three groups by random number method, and received different computer-based cognitive training programs. The training time of G1 group was 30 min per time, 3 times a week; G2 group 18 min per time, 5 times a week; The training time in G3 group was unlimited, but the total training duration was 90 min. The program lasted for 14 weeks with the total training duration of 21 hours. Before and after 14 weeks of training, patients in the three groups were evaluated using MoCA, MMSE, and ADL scales, and the changes in cognitive function and ADL of the three groups were compared.

  Results   A total of 90 MCI patients were enrolled, with 30 cases in each group. During the study, 4 cases in G2 group and 4 cases in G3 group quitted, and symptom worsen in one case and he was unable to continue training. The dropped-out patients accounted for 10%, and finally 81 patients completed cognitive training. There were 14 males and 16 females in the G1 group, with an average age of (71.13 ± 8.37) years; 11 males and 14 females in the G2 group, with an average age of (73.40 ± 9.39) years; 10 males and 16 females in the G3 group, with an average age of (68.89 ± 5.71) years. No statistically significant difference in baseline data between the three groups was found, including MoCA, MMSE, and ADL. After 14 weeks of training, the MoCA, MMSE, ADL scores in the three groups increased significantly compared to those before intervention (all P＜0.05). After training, the MMSE score in the G1 group was significantly higher than that of the G2 group (P＜0.05), and there was no statistically significant difference in other indicators between the three groups (P＞0.05).

  Conclusion   Computer-based cognitive training can effectively improve the overall cognitive function and ADL in elderly patients with MCI. The amount of training and certain frequency are the key to ensure the effect of computerized cognitive intervention.