Emotional blunting in anxiety and depression: neurobiology and psychopathology

Guy M. GOODWIN, MD, PhD, FMedSci
University of Oxford – Department of Psychiatry
Warneford Hospital
Oxford, UK

Emotional blunting in anxiety
and depression: neurobiology
and psychopathology

by G. M. Goodwin, United Kingdom

Approaches to understanding anxiety and depression have been polarized between the neurobiological and the psychopathological. These alternative narratives to explain the same phenomena of anxiety and depression have given rise to competing approaches to treatment based on medicines and psychotherapy, respectively. They have long demanded unification and in recent years significant convergence between the two has been made possible by the development of cognitive neuroscience. This has been based on the analysis of cognition and its neural underpinning as revealed by human functional imaging. Both depression and anxiety are characterized by emotional biases. In depression, there is often an apparent blunting of positive emotional experience; in anxiety, fear and threat experiences may be magnified. It is now possible to analyze antidepressant drug effects in terms of the cognitive biases associated with the processing of emotionally relevant information in perception or memory. And the mechanisms underlying retraining or exposure in cognitive behavior therapy can be analyzed with reference to the underlying neural changes. One consequence of this, as will be shown here, is the new understanding that, just as antidepressants have the surprising effect of treating depression, a disorder of the emotions, it is possible that the serotonergic treatments may also give rise to emotional blunting. This was originally based on clinical observation, but with appropriate scales and double- blind designs, antidepressants may be differentially implicated in the effect. The emphasis on mechanisms of emotional processing, rather than simple phenomenology, offers novel ways of approaching how emotions are blunted in depression, and to what extent this is a fundamental feature of depressed mood and its remission.

Medicographia. 2012;34:295-299 (see French abstract on page 299)

For many years, approaches to understanding anxiety and depression have polarized somewhat between the neurobiological and what might imprecisely be called the psychopathological. These alternatives are effectively narratives to explain the same phenomena of anxiety and depression. But they have given rise to competing treatment traditions unduly focused on medicines on the one hand, and psychotherapy on the other. They have long demanded unification, but have lacked common ground on which to effect a rapprochement. Very recently, significant convergence between the two has been made possible by advances in cognitive neuroscience. One consequence of this, as will be shown here, is the new understanding that, just as antidepressants have the surprising effect of treating depression, a disorder of the emotions, it is possible that drugs may also give rise to emotional blunting. Furthermore, the traditional pragmatic emphasis on symptoms has shifted towards the effects of mood disorder and anxiety on emotional processing.

The neurobiological model

Neurobiological approaches to depression and anxiety take their origins from the clinical research that proved the effectiveness of treatments like the tricyclic antidepressants and the benzodiazepine anxiolytics. Because these medicines were effective—and in the case of the antidepressants perhaps most effective—in patients with the most severe illness, they provided an obvious starting place for biological accounts of the underlying neural basis for mood and anxiety disorder. The ideas that developed through the 1960s and subsequently were based on the seminal finding that the tricyclic antidepressants were reuptake inhibitors for the neurotransmitters serotonin and noradrenaline (and to a lesser extent dopamine).1 These transmitter systems seem to possess many of the properties that one might expect in a system regulating just those global changes in behavior and experience that characterize depression or normal mood. Thus, they were demonstrated to fulfill regulatory roles in sleep, sexual function, appetite, and attention. Disorders of function in all these domains are obvious in patients with depressive episodes; therefore, the idea that these symptoms are the consequence of changes in neurotransmitter levels or overall system function was for many years attractive and heuristically useful in generating innovative drug development. Independent clinical research showed that a different class of drugs, inhibitors of monoamine oxidase (MAO), also possessed antidepressant and anxiolytic properties. The convergence of therapeutic effect between the reuptake inhibitors on the one hand and the MAO inhibitors on the other seemed to confirm that the monoamine systems would be the locus for the neurobiology of depression and potentially also anxiety. Drug innovation focused on trying to produce more acceptable versions of the original classes of antidepressant drugs and so we now have drugs selective for reuptake systems of serotonin and noradrenaline, and reversible enzyme inhibitors for the MAO enzyme subtypes, all less toxic in overdose.

Had depression or anxiety been a monoamine disorder in the sense that Parkinson’s disease is a dopamine disorder, then both the anatomy of any pathology and the biochemistry of the relevant neuronal systems should have been obviously disturbed. In fact, it turned out that this was not correct. It was difficult to demonstrate large and invariant changes in the metabolism of the relevant neurotransmitters in the functional neuropathology of depression.2 In so far as there is a consensus about what the postmortem data shows, it is really focused on changes in synaptic proteins in the cortex, and these are most evident in patients with bipolar disorder.3 A purely neurobiological formulation of depression etiology, inferred from treatment effects, therefore ran into a disappointing cul-de-sac. Indeed, this failure to demonstrate any convincing fit between the monoamine transmitters and the pathology of the condition exhausted its heuristic potential and the consequence has been wholesale disinvestment by Big Pharma from psychiatry, largely because the science of new discovery has proved so difficult beyond the monoamines.

The psychopathological or cognitive model

Throughout this era, the alternative narrative associated with clinical psychologists and their approach to the treatment of depression and anxiety with cognitive behavior therapy (CBT) has gained substantial support. The cognitive model of mood disorder stemmed from the seminal ideas of Aaron Beck.4 Having been brought up in the prevailing tradition of psychoanalysis in the United States, which stressed the unconscious causes of psychopathology, Beck’s insight was to ask very simply whether conscious cognitive distortions might underlie and maintain the states that we recognize clinically as depression and anxiety. In the case of depression, he proposed a triad consisting of negative views of the self, the world, and most particularly, the future. Beck took this to characterize all depressions regardless of the apparent clinical subtype, and moreover he proposed that the degree of negative thinking related directly to the non-cognitive symptoms of depression such as sleep disturbance, etc. In its origins, this was not a theory of causation of depression, but a theory of how depression might come to be maintained and how therefore it might be reversed with a therapy that emphasized a correction of these cognitive biases.

CBT is intended to change how you think and what you do. It was strongly developed on the basis of formal cognitive theories of emotion and its regulation, and its evaluation was both more rational and more scientific than psychodynamic therapies had been. However, it never concerned itself very much with whether it had an underlying biological basis. Instead, it provided a pragmatic rationale to the psychotherapy which allowed treatment innovation, but within a closed high-level theory separate from brain function. This separation was reinforced by widely held views in psychology that the mind was really best addressed without reference to the brain. Until recently, this had largely remained unchallenged and CBT has, like antidepressant medication, run into some of the same limits to innovation, at least it has seemed so to a noninvolved observer such as myself.

The application of neuroscience

The study of the emotions using the methods of neuroscience has generally lagged behind the investigation of other brain functions. The idea that the brain represents emotional aspects of external stimuli in particular brain areas is nevertheless an old one and has long emphasized the role of limbic cerebral cortex and associated subcortical structures such as the striatum, amygdala, and thalamus in providing the neural basis for the emotional world of animals and man. For example, our understanding of fear circuits in animals has been well advanced for many years, but has been slow in finding translation to understanding clinical anxiety in neurobiological terms.5,6 Given the emphasis on the amygdala as a key structure, it was not surprising that imaging studies in patients with anxiety showed enhanced processing of fear stimuli (eg, fearful faces),7 even with subliminal stimuli.8 Findings in depression have generally been less consistent in demonstrating simple biases in emotional processing, despite the overwhelming clinical evidence for negative biases at a conscious level. In part, this will reflect at a behavioral level the impact of global impairments of memory and executive function that make depressed subjects difficult to test.9

Over 10 years ago, Catherine Harmer and I started to look at what effects antidepressants have on emotional processing, first in healthy volunteers and subsequently in patients.10,11 The findings have been surprisingly consistent in demonstrating that selective serotonin reuptake inhibitors (SSRIs), for example, reduce the capacity of individuals to detect negative emotion in the faces of others, increase the access to positive as compared with negative self-referential words, and reduce fear-potentiated startle responses. These findings allow a reformulation of how antidepressants may actually work, which can explain both their potential limitations and also why the biological changes that are seen in depressed states are not simply expressions of abnormal monoamine biochemistry.

The critical hypothesis is that antidepressants can change emotional bias, which is determined by relatively subtle alterations in function in the monoamine systems. On this view of drug treatment, what is most critical about a patient when depressed could be the extent to which they show automatic biases in their processing of emotional stimuli and the extent to which, when once these are reversed, recovery can then occur spontaneously.12 The critical feature, of course, is the link between a drug effect and a cognitive/emotional effect which is essentially automatic. However, emotional bias appears to be the target both for drugs, via automatic processing effects, and CBT, via its explicit focus on conscious emotional bias.

While study of the mechanisms and actions of antidepressants has embraced cognitive theory, cognitive theory has increasingly embraced neuroscience. The increased availability of brain imaging modalities has allowed cognitive science to be based simultaneously on observing behavior and on observing brain function; this has transformed attitudes in psychology. This fertile environment may foster new ideas that bridge the biological and cognitive narratives of depression. Imaging patients who have received cognitive therapy or volunteers who are studied while making emotional judgments has suggested a critical link between the automatic processing of emotional signals in the amygdala and the regulation of this activity from the frontal cortex.13 These top-down mechanisms are increasingly seen as the likely target for psychologically orientated therapies. On this view, therapy can strengthen this control by retraining habits and biases.14

This unified view of treatment efficacy, implicating neurobiological changes due to both drug action (on automatic processing mechanisms) and cognitive therapy (on the more voluntary aspects of cognitive control) is likely to be increasingly fruitful in the future. It invites a much closer examination of the extent to which drug treatments can facilitate psychological interventions and vice versa.

While this argument has been developed in relation to depression, it is just as likely to apply to anxiety as well, where psychological and drug interventions are both well developed. In relation to anxiety, the model of fear-potentiated startle can be used translationally in both animals and healthy volunteers.15 Thus, the potentiation of startle responses to distressing emotionally charged images is reduced by citalopram, but not by reboxetine, suggesting a greater potential efficacy for SSRIs in anxiety indications.12 Just as for major depression, the independent impacts of psychological training and drug treatment have the potential to inform treatment mechanisms in man.

Antidepressants and emotional blunting

This fusion of two previously disparate fields has thrown up a further question of remarkable interest, which is whether antidepressants themselves may actually lead to what is essentially emotional blunting. This was first proposed for patients who complained of reduced sexual interest while receiving treatment with SSRIs for depression. The data came from the Laukes Emotional Intensity Scale (LEIS),16 a self-report instrument comprising 18 questions to rate as compared with one’s “usual” state along a 5-point scale (a lot less/somewhat less/same as usual/somewhat more/a lot more). Significant reductions occurred in 12 of the 18 items, including ability to cry, irritation, care about others’ feelings, sadness, erotic dreaming, creativity, surprise, anger, expression of their feelings, worry, sexual pleasure, and interest in sex.16 It was suggested that antidepressants blunt not only sexual experience and interest, but also other aspects of emotional experience and behavior. While this is a very interesting hypothesis, these data did little to confirm it, because the discriminating items overlapped very obviously with what might have been residual symptoms of depression.

While an impact on sexual function could be formulated in terms of reduced function of relevant innervation in the sex organs and brainstem, any effect on the emotions might be expected to be more central within the brain. In any case, although the phenomenon of emotional blunting by SSRIs seems to have accorded very well with the experience of clinicians, it has been little investigated in the subsequent decade. When we reopened the question three years ago, there was no adequate scale with which to capture the experience which had been imputed. We therefore undertook a qualitative analysis of what patients actually complained of, a detailed inventory of what complaints were most frequent,17 the design of a scale to capture this experience, and its testing under conditions where patients would be blind to their treatment. This program has been, in part, motivated by the hypothesis that the new antidepressant agomelatine would lack the impact on emotional experience of the SSRIs. It is ongoing, but the results so far are thought provoking. They pose the central question of whether emotional blunting is simply a drug side effect, or an untreated aspect of depression, or some interaction between the two.

Our qualitative investigation of patient experience generated four dimensions. The four dimensions were “not caring,” “emotional detachment,” “reduction in positive emotion,” and “general reduction in emotions.” Several of the dimensions are highly correlated with depressive symptoms and so one would predict would be sensitive to depression. Indeed, they provide an alternative emotionally relevant depression scale, which we may call the Oxford Depression Questionnaire or ODQ. The development of the scale is still in progress. However, the performance of some of the items has been examined under double-blind conditions during the treatment of patients in a major depressive episode with either agomelatine or escitalopram. The findings are preliminary because the patient numbers were relatively small (the trial had already started when the study was added to it and the scale was only available for English-speaking patients). Figure 118 shows the findings for one of the most interesting of the items and poses key questions about the relationship between depression itself and the scale items and the response after treatment with different antidepressants.

Our provisional conclusions are that emotional blunting is a real effect and that it is more likely to occur when patients are taking a selective serotonergic drug than when they are taking agomelatine. Agomelatine does not increase the availability of serotonin in the brain, and the implication is that emotional blunting is a serotonergic side effect. What is not decided is whether such effects are nevertheless related to depression, which is a possibility as emotional blunting is a marked feature of being acutely depressed. This is illustrated by the baseline responses to the statement “My emotions lack intensity” in Figure 1: the percentage of patients endorsing this item implies a very common experience in major depressive episodes.18 With treatment with either an SSRI, escitalopram (10- 20 mg), or agomelatine (25-50 mg), this rate falls as would be expected for a simple symptom of depression. At week 24 however, about twice as many patients continued to endorse this item in the escitalopram group as in the agomelatine group, despite equivalent remission on the Hamilton ratings (see legend, Figure 1). This is the first, albeit preliminary, double- blind evidence for a blunting effect of an SSRI, but it is also ambiguous because of the high baseline levels. The raised levels at week 24 could be a delayed effect of the SSRI per se or an aspect of the depressed state poorly treated by an SSRI. This raises the question addressed in the next paragraph.

Figure 1
Figure 1. Patients endorsing the statement, “My emotions lack

Percentage of patients (n=66) endorsing the statement at baseline (W0), and
after 2, 12, and 24 weeks (W2, W12, W24) of treatment with either agomelatine
(blue) or escitalopram (light red). Mean Hamilton scale scores at W24 were
6.1±5.6 (median 4) for agomelatine- and 6.1±4.0 (median 4) for escitalopramtreated
Abbreviation: W, week.
Based on reference 18: Corruble et al. Eur Psychiatry. 2011;26(suppl 1):619.
Abstract P02-24.

Does emotional blunting occur in healthy volunteers who take SSRIs?

The first studies of antidepressant effects in healthy volunteers were short-term and subjects showed no changes in mood and complained of no marked side effects. However, it is possible that with longer treatment periods, some of the effects on emotional processing would translate into emotional blunting. The basis for saying this is that the primary effect of SSRI treatment in several tests was to reduce the processing of negative stimuli, rather than to increase the salience of positive stimuli. This was seen for the perception of negative emotion in faces, both behaviorally12 and when imaging neural activity in the region of the amygdala.19 In behavioral experiments, agomelatine had no impact on the perception of negative facial expressions, except sadness.20 This approach has been supplemented in related studies, where, again in healthy volunteers, the SSRI citalopram reduced activation to positive appetitive stimuli in the ventral striatum and the ventral medial/ orbitofrontal cortex, whereas the noradrenergic reuptake inhibitor reboxetine did not.21 Citalopram decreased neural responses to aversive stimuli more than did reboxetine. Treatment of depression may be conceptualized as decreasing punishment signals and increasing reward. SSRIs appear to achieve the former more effectively than the latter and therein may lie their potential for blunting emotional experience. These experiments were short-term and there were no subjective changes in the rating of pleasant and unpleasant stimuli by healthy volunteers. However, it is an open question whether, under these changed neural processing biases, an experience of reduced emotion might become more marked. The prediction would be that this effect would only occur with serotonergic medicines and not with drugs acting like agomelatine and reboxetine through non serotonergic mechanisms.

Acknowledgment. Dr Goodwin holds grants from Bailly Thomas, the Medical Research Council (MRC), the National Institute for Health Research (NIHR), Servier, and the Wellcome Trust, and has advised or received honoraria from AstraZeneca, Bristol Myers Squibb (BMS), Boehringer Ingelheim, Cephalon/Teva, Janssen-Cilag, Eli Lilly, Lundbeck, Otsuka, P1Vital, Pfizer, Roche, Servier, Schering Plough, and Takeda.

1. Cooper J, Bloom F, Roth R. The biochemical basis of neuropharmacology. 8th ed. New York, NY: Oxford University Press; 2003.
2. Harrison PJ. The neuropathology of primary mood disorder. Brain. 2002;125: 1428-1449.
3. Eastwood SL, Harrison PJ. Synaptic pathology in the anterior cingulate cortex in schizophrenia and mood disorders. A review and a Western blot study of synaptophysin, GAP-43 and the complexins. Brain Res Bull. 2001;55(5):569-578.
4. Beck AT. Cognitive therapy: past, present, and future. J Consult Clin Psychol. 1993;61(2):194-198.
5. Ledoux J. Fear and the brain: where have we been, and where are we going? Biol Psychiatry. 1998;44(12):1229-1238.
6. Ledoux JE. Emotion, memory and the brain. Sci Am. 1994;270(6):50-57.
7. Bishop SJ. Neurocognitive mechanisms of anxiety: an integrative account. Trends Cogn Sci. 2007;11(7):307-316.
8. Monk CS, Telzer EH, Mogg K, et al. Amygdala and ventrolateral prefrontal cortex activation to masked angry faces in children and adolescents with generalized anxiety disorder. Arch Gen Psychiatry. 2008;65(5):568-576.
9. Austin MP, Mitchell P, Goodwin GM. Cognitive deficits in depression: possible implications for functional neuropathology. Br J Psychiatry. 2001;178:200-206.
10. Harmer CJ, Hill SA, TaylorMJ, Cowen PJ, Goodwin GM. Toward a neuropsychological theory of antidepressant drug action: Increase in positive emotional bias afterpotentiationofnorepinephrine activity. Am J Psychiatry. 2003;160(5):990-992.
11. Harmer CJ, O’Sullivan U, Favaron E, et al. Effect of acute antidepressant administration on negative affective bias in depressed patients. Am J Psychiatry. 2009; 166(10):1178-1184.
12. Harmer CJ, Goodwin GM, Cowen PJ. Why do antidepressants take so long to work? A cognitive neuropsychological model of antidepressant drug action. Br J Psychiatry. 2009;195(2):102-108.
13. Goldapple K, Segal Z, Garson C, et al. Modulation of cortical-limbic pathways in major depression: treatment-specific effects of cognitive behavior therapy. Arch Gen Psychiatry. 2004;61(1):34-41.
14. Browning M, Holmes EA, Murphy SE, Goodwin GM, Harmer CJ. Lateral prefrontal cortex mediates the cognitive modification of attentional bias. Biol Psychiatry. 2010;67(10):919-925.
15. Vaidyanathan U, Patrick CJ, Cuthbert BN. Linking dimensional models of internalizing psychopathology to neurobiological systems: affect-modulated startle as an indicator of fear and distress disorders and affiliated traits. Psychol Bull. 2009;135(6):909-942.
16. Opbroek A, Delgado PL, Laukes C, et al. Emotional blunting associated with SSRI-induced sexual dysfunction. Do SSRIs inhibit emotional responses? Int J Neuropsychopharmacol. 2002;5(2):147-151.
17. Price J, Cole V, Goodwin GM. Emotional side-effects of selective serotonin reuptake inhibitors: qualitative study. Br J Psychiatry. 2009;195(3):211-217.
18. Corruble E, Bélaidi C, Goodwin GM. Agomelatine versus escitalopram in major depressive disorders: a randomized double-blind, long term study focusing on sleep satisfaction and emotional blunting. Eur Psychiatry. 2011;26(suppl 1): 619. Abstract P02-24.
19. Harmer CJ, Mackay CE, Reid CB, Cowen PJ, Goodwin GM. Antidepressant drug treatment modifies the neural processing of nonconscious threat cues. Biol Psychiatry. 2006;59(9):816-820.
20. Harmer CJ, de Bodinat C, Dawson GR, et al. Agomelatine facilitates positive versus negative affective processing in healthy volunteer models. J Psychopharmacol. 2011;25(9):1159-1167.
21. McCabe C, Mishor Z, Cowen PJ, Harmer CJ. Diminished neural processing of aversive and rewarding stimuli during selective serotonin reuptake inhibitor treatment. Biol Psychiatry. 2010;67(5):439-445.

Keywords: antidepressant; emotion; emotional blunting; major depression; side effect