Abstract
Depression is one of the most prevalent and disabling mental health conditions worldwide, with an estimated 5.7% of adults currently experiencing depression, around 332 million people globally. Rumination, a repetitive pattern of thinking that involves comparing one’s circumstances with how one would ideally like them to be without engaging in problem-solving, is the cognitive vulnerability most strongly associated with depression. Despite the evidence, there are no studies that have employed causal discovery and inference to investigate whether rumination has a causal effect on depressive mood using N-of-1 psychological time series data. To address this gap, this study conducted a secondary analysis of publicly available data on an American Latino Female diagnosed with Major Depressive Disorder (MDD) who reported on her feelings 4 times per day over 39 days. To conduct causal discovery, the Vector Autoregressive Linear Non-Gaussian Acyclic Model (VAR-LiNGAM) algorithm was used. To conduct causal inference, the g-formula, implemented via a Monte Carlo simulation, was used to estimate the Contemporaneous Effect, the 1-lag Structural Direct Effect and the 1-lag Total Effect of rumination on depressed mood. The findings indicated that high rumination was associated with a 15.6-point increase in depressed mood contemporaneously and a 1.7-point increase in depressed mood approximately 4 hours later. This evidence further substantiates the importance of interventions designed to reduce ruminative thinking and, consequently, improve depressive symptoms.