Background and hypothesisUse of clozapine in treatment-resistant schizophrenia is often limited due to risk of adverse effects. Cross-sectional associations between clozapine treatment and low immunoglobulin levels have been reported, however prospective studies are required to establish temporal relationships. We tested the hypothesis that reductions in immunoglobulin levels would occur over the first 6 months following initiation of clozapine treatment. Relationships between immunoglobulin levels and symptom severity over the course of clozapine treatment were also explored. DesignThis prospective observational study measured immunoglobulin (Ig) levels (A, M and G) in 56 patients with treatment-resistant schizophrenia at 6-, 12- and 24-weeks following initiation with clozapine. Clinical symptoms were also measured at 12 weeks using the positive and negative syndrome scale (PANSS). ResultsIgA, IgG and IgM all decreased during clozapine treatment. For IgA and IgG the reduction was significant at 24 weeks (IgA: β = –32.66, 95% CI = -62.38, −2.93, p = 0.03; IgG: β = -63.96, 95% CI = -118.00, −9.31, p = 0.02). For IgM the reduction was significant at 12 and 24 weeks (12 weeks: β = –23.48, 95% CI = -39.56, −7.42, p = 0.004; 24 weeks: β = –33.12, 95 %CI = -50.30, −15.94, p = <0.001). Reductions in IgA and IgG during clozapine treatment were correlated with reductions in PANSS-total over 12 weeks (n = 32, IgA r = 0.59, p = 0.005; IgG r = 0.48, p = 0.03). ConclusionsThe observed reductions in immunoglobulin levels over six months of clozapine treatment add further evidence linking clozapine to secondary antibody deficiency. Associations between Ig reduction and symptom improvement may however indicate that immune mechanisms contribute to both desirable and undesirable effects of clozapine.