Early effects of antidepressants – a clue to mechanism of action

The delay in onset of effect of antidepressant treatment can be frustrating both for the patient and the clinician. It is also a puzzle in terms of the mechanism of action of these drugs. However, research bridging the psychological and pharmacological treatments is shedding light on this issue.

Early changes before clinical effects

Patients with depression often have a negative affective bias – they are likely to interpret ambiguous situations in a more negative than positive way, and pay more attention to negative cues in the environment than to positive cues.1 This is a key target of treatment with cognitive behavior therapy (CBT).2 Research indicates, however, that antidepressant drugs also improve this negative affect bias very early on in treatment, before clinical effects on mood are apparent. Further, research has shown that even a single dose of an antidepressant can have positive effects on emotional processing (i.e. the ability to discern positive facial expressions on a computerized task).3,4 This early change in negative affect processing has also been shown to have predictive value for patients who are likely to show clinical effects from antidepressant treatment.5,6

These positive effects of antidepressants on negative affective bias were also modulated by the patient’s environment. That is, the therapeutic effects predicted by initial effects on negative bias were not seen in patients who were isolated or who lived in non-supportive, difficult environments. This underlines the need to take into account the patient’s general well-being and social environment when considering the possible effects of antidepressant treatment.


Link between negative bias and brain circuitry

The early effects of antidepressants on negative bias have also been linked to changes in frontolimbic circuitry that are implicated in depression. Patients with depression show an increased responsiveness in the amygdala to negative stimuli, which can be normalized by treatment with antidepressants. Again, this change in reactivity precedes the appearance of clinical responses to the treatment.7 Moreover, a change in emotional reactivity in frontolimbic structures was found to differentiate between responders and non-responders to antidepressant treatment before the clinical effects became apparent.8

Reduction in reactivity of the amygdala to negative stimuli precedes the onset of clinically apparent symptom relief


Cognitive treatments targeting negative bias

These early effects of antidepressant drugs provide a bridge between pharmacological and psychological therapies for depression. Negative bias is a key target for CBT, and changes in this affective bias can be detected after one session of psychological treatment. Like the antidepressant effects, this change in negative bias precedes the onset of clinically apparent symptom relief. This has directed attention towards interventions that target cognitive bias directly, to help managing or preventing recurrences of depression. Specific training of patients to attend to positive cues in their environment was associated with a reduction in residual depression and anxiety.9 This emphasizes the advantages of synergistic use of both pharmacological and psychological therapies in the long-term management of depression.

Specific training of patients to attend to positive cues in their environment was associated with a reduction in residual depression and anxiety


Future uses of this technology

Models of affective bias may also be able to predict patient responses to antidepressant treatments – this hypothesis is currently being tested in the PReDicT trial. Early prediction of treatment response would allow clinicians to make timely changes to treatment for patients unlikely to respond. In addition, tests of negative bias could in the future be used as a potential biological marker of antidepressant efficacy, in order to help identify new therapies

Our correspondent’s highlights from the symposium are meant as a fair representation of the scientific content presented. The views and opinions expressed on this page do not necessarily reflect those of Lundbeck.


1. Urban EJ, et al. PLoS One 2018;13(9):e0203574.

2. Bruijniks SJE, et al. Clin Psychol Sci 2019;7(4):668-92.

3. Harmer CJ, et al. Br J Psychiat 2009;195(2):102-8.

4. Harmer CJ, et al. Am J Psychiat 2009;166(10):1178-84.

5. Tranter R, et al. J Affect Disord 2009;118(1-3):87-93.

6. Browning M, et al. Eur Neuropsychopharmacol 2019;29(1):66-75.

7. Godlewska BR, et al. Psychol Med 2012;42(12):2609-17.

8. Godlewska BR, et al. Transl Psychiatry 2016;6(11):e957.

9. Browning M, et al. Biol Psychiatry 2012;72(7):572-9.