Development of a Novel Staging Model for Affective Disorders Using Partial Least Squares Bootstrapping: Effects of Lipid-Associated Antioxidant Defenses and Neuro-Oxidative Stress

Although, staging models gained momentum to stage define affective disorders, no attempts were made to construct mathematical staging models using clinical and biomarker data in patients with major depression and bipolar disorder. The aims of this study were to use clinical and biomarker data to construct statistically derived staging models, which are associated with early lifetime traumata (ELTs), affective phenomenology, and biomarkers. In the current study, 172 subjects participated, 105 with affective disorders (both bipolar and unipolar) and 67 controls. Staging scores were computed by extracting latent vectors (LVs) from clinical data including ELTs, recurring flare ups and suicidal behaviors, outcome data such as disabilities and health-related quality of life (HR-QoL), and paraoxonase (PON)1 actvities and nitro-oxidative stress biomarkers. Recurrence of episodes and suicidal behaviors could reliably be combined into a LV with adequate composite reliability (the "recurrence LV"), which was associated with female sex, the combined effects of multiple ELTs, disabilities, HR-QoL, and impairments in cognitive tests. All those factors could be combined into a reliable "ELT-staging LV" which was significantly associated with nitro-oxidative stress biomarkers. A reliable LV could be extracted from serum PON1 activities, recurrent flare ups, disabilities, and HR-QoL. Our ELT-staging index scores the severity of a relevant affective dimension, shared by both major depression and bipolar disorder, namely the trajectory from ELTs, a relapsing course, and suicidal behaviors to progressive disabilities. Patients were classified into three stages, namely an early stage, a relapse-regression stage, and a suicidal-regression stage. Lowered lipid-associated antioxidant defenses may be a drug target to prevent the transition from the early to the later regression stages.