Objectives: To further test the hypothesis that past illness episodes and delay of long-term treatment do not limit maintenance treatment response among patients with manic-depressive illnesses (MDI).
Methods: In a sample of 764 MDI patients in Cagliari and Berlin, 77% of whom had bipolar disorder (BPD), we: (i) correlated treatment latency or pretreatment episode counts versus hospitalized morbidity during treatment; (ii) correlated treatment duration versus pretreatment morbidity; (iii) correlated treatment latency versus pretreatment or treated morbidity; (iv) modeled factors associated with longer treatment latency; (v) compared treatment latencies at extremes of treatment outcomes, and (vi) compared pretreatment morbidity within 2 years of the longest versus shortest treatment latency quartiles.
Results: Pretreatment morbidity was strongly correlated with shorter treatment latency, but morbidity during treatment was unrelated to treatment latency, pretreatment episode counts, sex, diagnosis, treatment type or treatment duration. In multivariate modeling, treatment latency was longer among patients who had experienced an early onset of illness, mainly in depressive disorders (BPD II and major depression) and among women, but was unrelated to morbidity during treatment. Patients with no illness recurrences during treatment and those who were ill at least 50% of the time had similar treatment latencies. Pretreatment morbidity occurring just prior to the initiation of long-term treatment was very similar at the highest and lowest treatment latencies.
Conclusions: These findings support the therapeutically favorable conclusion that prior episode counts and treatment delay have little association with morbidity during prophylaxis with mood-stabilizing agents. Comparisons of morbidity during versus before treatment in episodic disorders are misleading because overall morbidity becomes diluted with longer time-at-risk, whereas therapeutic intervention is typically determined by immediately preceding illness.