I've been trying to understand the position of Steve Hyman (a former NIMH director), who has been amply quoted in an recent article in Quartz, saying among ohter things that

There’s not “an iota of direct evidence” for the theory that a chemical imbalance causes depression.

I looked into Hyman's papers and I found a fairly cited (~500 in Google Scholar) paper of his some 20 years ago in AJP, co-authored with Eric J. Nestler, in which they propose that

drug-induced neural plasticity [...] underlies the long-term actions of psychotropic drugs in the brain [...] Biological investigation in psychiatry has often focused too narrowly on synaptic pharmacology, especially on neurotransmitter turnover and neurotransmitter receptors. A paradigm is presented to help understand the long-term effects of psychotropic drugs, including the latency in onset of their therapeutic actions.

The paper further details that

Acute dosing of the antidepressants produces little in the way of subjective or behavioral effects (other than side effects). Initial doses of the antipsychotic drugs produce subtle motor effects or some degree of sedation, and initial doses of the stimulants produce euphoria and locomotor activation. However, substantial improvement in depressive or psychotic symptoms, or the development of addiction to psychostimulant drugs, occurs only if the drugs are taken at adequate dosage and with adequate frequency and chronicity. For example, depressive symptoms would not be expected to improve if an antidepressant were taken at too low a dose, only occasionally, or for too short a period of time.

We therefore conceptualize antidepressant-induced increases in synaptic serotonin or norepinephrine, antipsychotic- induced blockade of dopamine neurotransmission, and cocaine- and amphetamine-induced increases in synaptic dopamine as initiating events for longerterm changes in neural function. It is the adaptive response of the nervous system to adequate repeated perturbations mediated through these initial targets that produces the therapeutic responses to antidepressants or antipsychotic drugs or, in the case of the psychostimulants, addiction [...].

I'm somewhat familiar with the slow, cumulative effect of monoamine antidepressants and with the mechanism of addiction. Hyman and Nestler can be probably be forgiven for not anticipating the ketamine-as-fast-antidepressant fervor 20 years later. But what I want to ask here is:

What is the evidence that anti-psychotics also work in this delayed manner? Is it uncontroversially so?

  • 1
    $\begingroup$ typical or atypical antipsychotics? $\endgroup$
    – faustus
    Dec 30, 2017 at 18:35
  • $\begingroup$ @faustus: they don't particularize their statements to one sub-class of antipsychotics anywhere in the paper, as far as I can tell. $\endgroup$ Dec 30, 2017 at 18:36
  • $\begingroup$ are you interested in the first quote, or the answer.re: antipsychotics mechanism of action? the quote is a broader statement about biological psychiatry. it's a position that i share, though i accept that it's not a conventional one. $\endgroup$
    – faustus
    Dec 30, 2017 at 18:39
  • $\begingroup$ @faustus: not in this question. It's just for context how I got to read the Hyman and Nestler position paper. The "chemical imbalance" has been derided by other professionals in the field as a silly, dumbed down version of the various monoamine hypotheses proposed by researchers. I was trying to understand in exactly what way Hyman rejects that "chemical imbalance". Once a straw man is set up, there are many ways to burn it down... $\endgroup$ Dec 30, 2017 at 18:55

1 Answer 1


There seem to be multiple studies generally pointing to a delay of clinical antipsychotic action of these drugs, typically between a week and two weeks (e.g.., Zedkova et al (2011) and Mousavi et al. (2013)). However, there is evidence that antipsychotics may start working after a day, or even within a few hours after starting the meds. See for an interesting review Agid et al. (2006).

From reading these papers my understanding is that a lot of it depends on where the researcher puts the threshold of clinical effectiveness and which outcome measure is used to define that effectiveness.

- Agid et al., J Psychiatry Neurosci (2006); 31(2): 93–100
- Mousavi et al., J Res Pharm Pract (2013); 2(4): 138–44
- Zedkova et al., Neuro Endocrinol Lett (2011); 32(5): 667-70


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.