The direct action of antidepressants remediates negative biases in affective processing (AP) at a neuropsychological level. These actions occur early in treatment before an improved mood. Majority of research has been conducted using antidepressants mainly affecting serotonin or noradrenaline activity in major depressive disorder (MDD). 

Introduction

The acute neuropharmacological actions of most antidepressants are characterized by enhancing serotonin and/or noradrenaline neurotransmission, which is immediately detectable after administration. However, repeated administration of the drug over multiple weeks is required before a clinically important therapeutic effect is observed. It’s thought that delayed onset is due to numerous neuroadaptive changes that happen between first administration and the onset of the desired effect. Recently, focus has shifted to neurobiological models aiming to explain the delay by introducing the notion  of antidepressant-induced activation of second messengers and subsequent changes in gene expression. But these models are limited as they fail to explain how these actions relate to an improvement in mood and overall symptomatic relief. Furthermore, antidepressants don’t elevate mood in healthy individuals. 

These findings lead to our hypothesis: antidepressants don’t act as direct mood enhancers, but rather give way to the cognitive neuropsychological model (CNM) which aims to explain the gap between first administration of an antidepressant and the onset of its mood improving effect.

The CNM in antidepressant response

In the context of MDD, individuals tend to extract positive affective information leading to negative affective biases (NAB), being the basis and fuel of MDD at once. CNM suggests that, at a neuropsychological level, direct action of antidepressants leads to an extraction of negative affective information from a variety of social and affectively loaded stimuli, and that these actions occur early in the treatment prior to an improved mood. The suggested model allows for a time interval ranging from first drug intake to a clinically important elevation in mood in later treatment.

The role of NABs in depression

MDD patients and at-risk individuals display negative biases in emotional and reward processing across cognitive domains. Such negative biases link to abnormal activity in the limbic and striatal circuitry involved in the initial appraisal and memory of affective stimuli. MDD patients show abnormal responses toward reward, punishment and performance feedback. These deviant responses are associated with reduced function in frontostriatal systems in depressed, remitted depressed, and at-risk individuals.

NAB is linked to relapse and exhibited by people at risk of MDD by virtue of high neuroticism and family/personal history of MDD. Concludes that NAB appears to play a role in the vulnerability to/aetiology of MDD rather than being a secondary consequence of low mood.

The effects of antidepressants on AP

The repeated, subchronic, and acute effects of antidepressants on AP have been assessed in healthy volunteers and MDD.

Subchronic antidepressant (SSRI and SNRI) treatment increases processing of positive affective information, like positive self-descriptive word and propensity to perceive ambiguous facial expressions as happy, in healthy volunteers. A single administration of citalopram or reboxetine is sufficient to increase recognition of happy faces. Administration of citalopram or reboxetine results in associated neural alterations, including attenuated amygdala responses to aversive stimuli, increased fusiform responses to happy vs neutral facial expressions, and increased respectively decreased frontoparietal activity during classification or recognition of positive vs. negative self-descriptive words.

Important note is that all these effects of antidepressants on AP occur in the absence of a significant elevation in subjective mood.

Single administration of reboxetine restores normal balance of positive to negative processing in facial expression recognition and recall of self-descriptive words in MDD patients in the absence of any changes in subjective mood. An 8-week-treatment of an SSRI in MDD patients normalizes hyperactive amygdala, ventral striatal, and frontoparietal cortical responses to negative affective information. 8-week-treatment of an SNRI also normalizes hyperactive anterior cingulate responses to negative affective information.  Repeated SSRI treatment can also normalize hyperactive responses to positive affective information.

In a subchronic treatment, citalopram was found to reduce ventral striatal and medial orbitofrontal activity in response to reward, while reboxetine resulted in the opposite.  Two-week treatment of SNRI and SSRI treatment found SNRI to enhance reward-related ventral striatal activity in healthy volunteers, while SSRI diminishes neural processing of both aversive and rewarding stimuli resulting in general dampening of negative and positive affective experiences. The latter provides a plausible neural mechanism for the emotional blunting phenomenon observed in some MDD patients during treatment.

Can the model explain the delayed onset of clinical effect and symptomatic relief?

Acute antidepressant treatment can produce early unconscious changes requiring interaction with the social environment in AP prior to mood improvement - subjective delay in response to antidepressants.

The requirement for changes in NABs and interaction with social environment helps explain some variance in clinical response to antidepressant treatments. Treatment resistant MDD patients tend to have long-standing NABs or social environments that don’t allow an improvement in mood. Therefore, changes in AP could account for the global resolution of the multiple syndrome domains in MDD. A reduction in the preoccupation with negative information would improve mood, reduce social withdrawal, and increase resources for other cognitive tasks. A range of antidepressants decrease submissive and increase affiliative problem solving behaviours in healthy volunteers, demonstrating that reduced NABs are translated into improved social interactions/behaviour.

Clinical applications

CNM suggests the possibility to predict the effectiveness of a certain antidepressant for a certain individual in the long-term from the magnitude of the initial effects of the antidepressants on behavioural or neural measures of affective processes, like the magnitude of the increase in recognition of happy facial expressions produced by a single dose of citalopram or reboxetine.

Since AP appears to be altered by a range of antidepressants with different neurochemical actions, behavioural and neural measures of AP could be a useful predictor of the therapeutic potential of a new antidepressant. To be effective, the agent must be able to alter affective processes.

Processing measures can be used in the development and efficacy of screening novel antidepressants and the tailored treatment to individuals.