BACKGROUND The clinical relevance of raised levels of circulating cytokines in bipolar disorder is still unclear. Cytokines influence neurotransmitters, neuroplasticity, and white matter integrity. An inconsistent literature suggests that higher cytokine levels could hamper antidepressant response. Total sleep deprivation (TSD) and light therapy (LT) prompt a rapid antidepressant response and can provide a model treatment to study predictors of response. METHODS We studied at baseline 15 immune-regulating compounds in 37 consecutively admitted inpatients with a major depressive episode in the course of bipolar disorder (DSM-5 criteria) and in 24 controls. Thirty-one patients (84%) had a lifetime history of drug resistance. Patients were administered 3 TSD + LT cycles in 1 week (study period: 2010-2012). Data were analyzed with age- and false-discovery-rate-corrected analysis of variance and were tested as predictors in a regressive model. RESULTS Twenty-three patients (62%) responded to treatment (Inventory of Depressive Symptomatology IDS-C score < 12). Five highly intercorrelated compounds (IL-8, MCP-1, IFN-γ, IL-6, TNF-α) showed higher levels in nonresponder patients as compared to responders, corrected for multiple comparisons (respectively F = 6.138, PFDR = .0134; F = 6.197, PFDR = .0134; F = 4.785, PFDR = .0255; F = 3.782, PFDR = .0441; F = 3.764, PFDR = .0441). A principal component analysis identified a single component that explained 84% of variance of these cytokines (Q² = 0.15), and a high factor score significantly predicted worse response (b = -0.692; W = 4.34, P = .037). A higher body mass index correlated with higher cytokines (r = 0.430, P = .010), indirectly hampering response (b = -0.0192, P = .013). CONCLUSIONS Proinflammatory compounds reflecting an M1-like proinflammatory state of monocytes/macrophages are associated with a poor response to antidepressant TSD + LT treatment in bipolar depression.