BackgroundObstructive sleep apnea （OSA） represents a potential modifiable risk factor for cognitive impairment， while spatial navigation performance serves as a sensitive marker to detect such deficits at an early stage. Currently， polysomnography （PSG） is the gold standard for OSA diagnosis， with the apnea hypopnea index （AHI） used to stratify the disease severity. Nevertheless， AHI alone cannot fully explain the individual variability in cognitive impairment among OSA patients. Existing evidence indicates that sleep architecture disruption impairs spatial memory consolidation and subsequently alters spatial navigation through pathways independent of AHI， yet this hypothesis remains unvalidated in routine clinical OSA cohorts. ObjectiveTo compare the global cognitive function and spatial navigation performance across OSA patients of varying severities， and to analyze associated factors of spatial navigation performance， so as to provide clinical evidence for the early identification and intervention of cognitive impairment in OSA patients. MethodsThis cross-sectional study consecutively enrolled 135 outpatients from the Mental Health Center of Inner Mongolia Autonomous Region from March 8， 2024 to September 12， 2025. All participants completed PSG monitoring and fulfilled the OSA diagnostic criteria outlined in the Guidelines for the Diagnosis and Treatment of Adult Obstructive Sleep Apnea. Subjects were stratified into mild （5 ≤AHI < 15 events/h）， moderate （15 ≤ AHI < 30 events/h）， and severe （AHI ≥ 30 events/h） groups based on AHI values， and all PSG sleep architecture parameters were extracted for statistical analysis. Global cognitive function was evaluated using the Mini-Mental State Examination （MMSE）， the Trail Making Test （including TMT-A and TMT-B）， and two subtasks from the THINC Integrated Toolkit （THINC-it）： the Spotter test （choice reaction time， CRT） and the Codebreaker test （digit symbol substitution test， DSST）. Spatial navigation performance was assessed through the Santa Barbara Sense of Direction Scale （SBSOD） and the computerized AMUNET spatial navigation test. ResultsAmong the 135 OSA patients， 46 （34.07%） were classified into the mild group， 47 （34.82%） into the moderate group， and 42 （31.11%） into the severe group. In terms of cognitive function， there were no statistically significant intergroup differences across three groups in CRT， consecutive correct DSST responses， MMSE scores， and completion time of TMT-A and TMT-B （P>0.05）. Regarding spatial navigation performance， no statistically significant differences were observed among three groups in SBSOD scores and the outcomes of the four AMUNET subtasks （P>0.05）. Spearman correlation analyses revealed that the performance on the delayed allocentric AMUNET subtask was positively correlated with stage N2 duration， percentage of stage N2 sleep （N2%）， and rapid eye movement （REM） latency （r=0.277， 0.296， 0.239， P_FDR<0.05）. In contrast， this subtask exhibited negative correlations with REM duration， percentage of REM sleep （REM%）， stage N3 duration and percentage of stage N3 sleep （N3%） （r=-0.223–-0.175， P_FDR<0.05）. SBSOD score was positively correlated with the completion time of TMT-A and TMT-B （r=0.236， 0.278）， and negatively correlated with MMSE scores （r=-0.311）. For AMUNET indices， performance on the allocentric navigation subtask showed a positive correlation with CRT （r=0.013）. Moreover， performance of both egocentric and allocentric navigation subtasks yielded positive correlations with the completion time of TMT-A and TMT-B （r=0.191–0.265， P_FDR<0.05）. Performance on the egocentric， allocentric， and combined egocentric–allocentric navigation subtasks was negatively correlated with MMSE scores and the consecutive correct DSST responses （r=-0.318–-0.207， P_FDR<0.05）. Hierarchical multiple liner regression analysis demonstrated that after controlling for baseline characteristics and global cognitive function， N2% emerged as a positive predictor of the delayed allocentric navigation subtask performance （β=0.242， 95% CI： 0.104–1.463）. ConclusionPatients with OSA across all severity strata exhibit comparable global cognitive function and spatial navigation performance. Spatial memory is associated with sleep architecture parameters， among which N2% acts as an independent positive predictor of spatial memory performance. ［Funded by Natural Science Foundation Project of Inner Mongolia Autonomous Region （number， 2024QN08050）； Science and Technology Project for High-Level Clinical Specialty Development of Public Hospitals of the Capital Region under the Autonomous Region Health Commission of Inner Mongolia Autonomous Region in 2023 （number， 2023SGGZ047）］

Objective To explore and analyze the timing and postoperative effect of proximal femoral nail anti-rotation( PFNA) for the treatment of elderly femoral intertrochanteric fracture. Methods From September 2011 to March 2013, there were 36 elderly patients with femoral intertrochanteric fractures in our hospital, including 6 cases of type A1, 16 cases of type A2, and 14 cases of type A3. All the 36 pa-tients received PFNA and systematic rehabilitation exercise. Treat the underlying disease and strengthen the nursing to prevent thrombosis at the same time. Hip joint function was evaluated by the Harris score system. Results The average hospitalization of patients is 4 weeks. The wound healing was good in 30 cases and wound infection occured in 2 cases who recovered after subsequent treatment, and the other 4 pa-tients died due to complicated internal medicine diseases. 21 cases of patients were followed up,and their hip joint function recovered well according to the Harris score system, including 12 cases of excellent, 6 cases of good, and 3 cases of eligible, with an excellent and good rate of 85%. Conclusion The application of PFNA for the treatment of elderly femoral intertrochanteric fractures has the advantages of less damage to the local blood supply and better recovery of fracture healing and hip joint function, which is an ideal treatment for the elderly pa-tients with femoral intertrochanteric fractures.