BACKGROUNDDepressive symptoms is more prevalent in middle-aged and older populations, and the development of Internet technology has brought new challenges and opportunities for the mental health of this group. This study aims to explore the bidirectional relationship between Internet device diversity and levels of depressive symptoms through longitudinal data analysis, as well as investigate the mediating role of social participation.METHODSData for the analysis was retrieved from the China Health and Retirement Longitudinal Study (CHARLS) in 2015, 2018, and 2020, which included 9259 middle-aged and older adults. The sum of the number of devices mentioned by each participant was calculated to assess the diversity of internet device use, a weighted total score of how frequently and to what extent individuals had participated in various social activities in the past month was calculated to measure the social engagement, and the Center for Epidemiological Studies Depression Scale was used to calculate depressive symptoms. A cross-lagged mediation model and bivariate correlation analysis were employed to examine the longitudinal relationships among depressive symptoms, internet device diversity, and social participation. Age, gender, highest education level, marital status, drinking status, smoking history, self-rated health, sleep duration, and chronic health conditions were adjusted as covariates.RESULTSThe results from the three waves of longitudinal data revealed mutual predictive relationships among internet device diversity, social participation, and depressive symptoms. The mediating effect was significant (β = -0.612, 95 % CI -0.796 to -0.429), with significant gender differences observed. Furthermore, our findings indicated that social participation functioned as a protective factor against depressive symptoms, and the association between social participation and depressive symptoms was not a straightforward linear relationship. This suggests that participation in social activities (either high or low) is associated with a lower risk of depressive symptoms compared to no social participation, with no gender differences.CONCLUSIONSThis study suggests that there is a relationship between internet device diversity and lower levels of depressive symptoms among middle-aged and older adults in China. Social participation mediates the relationship between internet device diversity and depressive symptoms, and different levels of social engagement have different effects on depressive symptoms. It is recommended to increase support for social participation among middle-aged and older adults, strengthen structural social capital, and better leverage the use of the internet to reduce depressive symptoms levels and promote the physical and mental well-being of this population.