Objective: To explore the relationship between family intergenerational contact, depressive symptoms and self-rated health status in the elderly, so as to provide insights into promoting health of the elderly and actively coping with population aging. Methods: Demographic information, intergenerational contact frequency, self-rated health status and depressive symptoms in the elderly at ages of 65 years and older were collected through the Chinese Longitudinal Healthy Longevity Survey (CLHLS). Structural equation model was established to analyze the effects of intergenerational contact on self-rated health and the mediating role of depressive symptoms between family intergenerational contact and self-rated health in the elderly. Results: Totally 15 497 elderly participants were enrolled, with 6 727 males (43.41%) and 8 770 females (56.59%), and there 12 583 individuals lived with family members (81.19%). The median age was 85.00 (interquartile range, 16.00) years. The scores of self-rated health and depressive symptoms were (3.39±0.87) and (2.22±0.56), respectively. The results of structural equation model analysis showed that after adjusting for age, gender and residence, frequent intergenerational contact had a direct positive effect on self-rated health (effect size=0.038, P<0.001), and could improve the depressive symptoms (effect size=0.089, P<0.001) in the elderly. In addition, depressive symptoms had a negative effect on self-rated health in the elderly (effect size=-0.422, P<0.001). Conclusions Frequent intergenerational contact can improve the self-rated health status in the elderly, and indirectly improve the self-rated health level by alleviating depressive symptoms.

Objective:To analyze the factors influencing the theoretical score of graduation examination of standardized residency training in a class-A tertiary general hospital in Shanghai from 2018 to 2021, and establish prediction models, and to provide a reference for further improving the quality of residency training.Methods:We collected the data of 556 residents who participated in the residency training graduation examination at the hospital from 2018 to 2021. Binary regression and logistic regression analyses were used to determine the association of residents' basic information, their routine assessment results at the bases, and the base performance assessment score (indicating the bases' training quality) with theoretical graduation examination results; and prediction models for the theoretical graduation examination score were established, using binary regression and logistic regression methods to construct model datasets. SPSS 26.0 was used for statistical analysis.Results:From 2018 to 2021, a total of 556 people participated in the residency training graduation examination at the hospital. The first-time pass rate of the first station theoretical examination station was 97.48% (542/556). Trainees graduating from double first-class schools had the highest first-time pass rate of 98.09% (359/366); doctoral students had the highest first-time pass rate of 98.59% (140/142); and residents of the two-year training program had the highest first-time pass rate of 98.34% (297/302). According to the binary regression analysis of categorical variables such as residents' basic information, the base level was associated with whether residents passed the theoretical examination ( P<0.05). According to the logistic regression analysis of residents' routine assessment scores and the base performance assessment score, the annual theoretical assessment score, the first-time theoretical score of the national medical licensing examination, the annual professional performance test score, and the base performance assessment score were key factors affecting the theoretical graduation examination score ( P<0.05). In the regression analysis associated with the theoretical graduation examination score, the annual professional performance test score showed the highest regression coefficient and the highest variance inflation factor, indicating strong multicollinearity. Therefore, the annual professional performance test score was important in predicting the theoretical score of the graduation examination. Conclusions:This study used binary regression and logistic regression models to analyze the possible factors affecting the theoretical graduation examination score and the degree of association, and also established two prediction models of the theoretical graduation examination score to predict trainees at risk. Through implementing the concept of hierarchical progression, adding personalized tutoring, realizing information management for residency training, and providing intelligent early intervention, the pass rate of residency training graduation examination is expected to be further improved, ultimately achieving the purpose of improving the quality of residency training.