Should Patients with Schizophrenia be Treated with Antipsychotics Forever?

A paper that has gained a lot of attention over the last couple of months was just published in print in the current issue (August 2018) of the American Journal of Psychiatry. A Finnish registry study examined the risk of rehospitalization or death in 8719 patients with the first episode of schizophrenia over the observation period 1996 to 2014 (Tiihonen et al., 2018).


The risk of rehospitalization or death was lowest in patients who received continuous treatment and increased with the duration of treatment when the medication was discontinued. If medication was discontinued immediately after discharge, the risk of rehospitalization or death increased by 1.63-fold (95% CI = 1.52 – 1.75). It further increased more than threefold, when the medication was discontinued after 2-5 years of treatment ( HR = 3.26, 95% CI = 2.07-5.13), and more than sevenfold (HR = 7.28, 95% CI = 2.78- 19.05) after discontinuation of the medication after more than five years. Thus, the authors disprove the clinically widespread opinion that after years of clinical stability the risk of relapse is lower than shortly after a psychotic episode. They conclude that prolonged therapy is in principle beneficial for all patients with schizophrenia.

However, the finding can also be interpreted in the sense of the supersensitivity hypothesis: According to this view, patients would become more and more sensitive to discontinuation of medication as their duration of treatment increases, which is why the risk of recurrence increases with the duration of treatment. The authors, however, counter that only a small proportion of relapses occurred within the first six months after stopping the medication, which would be expected if they were due to supersensitivity of dopaminergic systems. Nevertheless, the graphs published in their paper clearly illustrate that in all treatment groups the relative relapse risk is highest shortly after stopping the medication, and that it is continuously decreasing over the years. Animal experimental data suggest that functional effects of supersensitivity may persist for one year and longer after discontinuation of antipsychotic medication (Yin et al., 2017).

There is no doubt that many patients with schizophrenia benefit from long-term antipsychotic treatment. Bot some probably don’t. Important questions remain: If supersensitivity is not an issue, as the authors and René Kahn in an accompanying editorial claim, why is relapse risk increasing over time? Why is it not a more or less stable factor? Both Tiihonen and colleagues and the editorial neglect that the brain certainly adapts to continuous treatment. We have to understand the mechanisms, and we have to identify the patients that do not benefit from the currently available antipsychotics and develop “treatment resistance”, which certainly has some iatrogenic aspects. Such studies will begin at the Central Institute of Mental Health next year.

Please see also the related post “Antipsychotics for long-term treatment – when to stop? Or better not?”, which I published here on April 2nd, 2018.