Extended Data Fig. 2: Association of cumulative interpersonal sociopolitical conflicts and probable depression by sociodemographic groups, 2014–2019.
From: Interpersonal conflicts, social media use and depression associated with protests
The forest plot shows estimated odds ratios of probable depression in 2019 (wave 9) with 95% confidence intervals. The odds ratios of probable depression were obtained through marginal structural models adjusting for baseline sociodemographics, past history of exposure (that is, sociopolitical conflicts with family members or friends and colleagues), probable depression, family support, time spent on social media and online for sociopolitical news during protests, protest participation, and prior levels of other types of sociopolitical conflicts (that is, adjusted for prior sociopolitical conflicts with friends and colleagues in the model examining association between intrafamilial sociopolitical conflicts and probable depression). Weights were truncated at 1st and 99th percentiles. Cumulative interpersonal sociopolitical conflicts were defined as the mean score of the conflicts from 2014 to 2019 (waves 3 to 8).
