Network perspectives in population mental health : understanding suicidality and help-seeking as complex phenomena

Abstract

Recent years have seen incessant rises in the incidence and prevalence of mental health problems among the younger generation (collectively referring to 10-35 year-olds) globally. Emerging findings further suggest that the burden of suicide, already a leading cause of the younger generation's deaths, could worsen in the long term due to the COVID-19 pandemic's spillover effects. Urgent concrete action to buck this precarious trend is therefore imperative to avert a looming societal crisis on multiple fronts. However, even though scholars widely agree that suicide is a complex problem, scientific endeavors hitherto have not addressed its full complexity.

Through the lens of complexity science, this thesis conceptualizes population mental health issues such as incidence of suicidality and low help-seeking tendency among the younger generation as emergent phenomena resulting from complex, nonlinear interplays between components of a dynamic social system: Various micro- to macro-level determinants of mental health would concurrently act on each person -- who is in turn embedded on social networks and interconnected with one another through network ties -- to shape the younger generation’s mental health within a given population setting.

This thesis broadly espouses a multidisciplinary perspective in addressing such complex networks, drawing upon complexity science toolkits and network science methodologies and integrating them with social theories and domain knowledge from social sciences. Successive studies in this thesis examine networked systems of increasing scope, using data on Hong Kong's younger generation as showcases to inform policy directions for early intervention of the younger generation’s suicidality.

Study I conceptualizes comorbidity between distress and suicidality as a network of symptoms with overlapping connectivity and identifies bridge symptoms linking the two syndromes. It then examines potential efficacies of individual distress symptoms' treatment in reducing suicidality prevalence through in silico interventions. Findings here serve to inform which specific symptoms should be prioritized for early intervention of suicidality.

The thesis then investigates nuances around help-seeking, a social behavior with relatively untapped potential in altering the younger generation’s trajectory from distress to suicidality. Study II employs latent class analysis to characterize underlying typologies of help-seeking behaviors among the younger generation and examine their respective suicide risk profiles. Shifts in these patterns before and during the COVID-19 pandemic are further analyzed to better safeguard against mental health implications of similar future scenarios.

Finally, Study III proposes a methodological innovation to overcome empirical limitations in measuring complex networks. Combining Approximate Bayesian Computation, agent-based modelling, and theoretically -informed network model formulation, it infers relative effects of unobserved micro-macro social forces in shaping the younger generation's population-level help-seeking tendency. Resulting models are also explored as policy sandboxes in guiding policy decisions.

This thesis strives to demystify nuanced interrelations between the many determinants of mental health, and subsequently pinpoint precise symptoms, constructs, and policy foci to guide strategies for early intervention of the younger generation's suicidality. Its end-goals are twofold: To spur a paradigm shift in suicide prevention research and to better position policymakers in steering population health measures, ultimately effecting real and timely positive change in society.

(499 words)