Defining remission in depression: is MADRS enough? Was the question posed at a CINP satellite symposium funded by Lundbeck . The answers reflect the complexity of the condition, and reveal an important disparity between the outcomes assessed by clinicians and those most valued by patients, which include cognition.
Diverse faces of depression
Depression is a complex phenomenon. Currently, diagnosis requires the presence of any five of the nine features listed in DSM-5, which means that two people can be diagnosed with depression while having only one feature in common.
So it perhaps not surprising that a recent field trial in the United States and Canada suggested that the reliability of the DSM-5 definition – assessed as the degree to which two clinicians agree independently on the presence of depression in a particular patient -- is open to question.
It is not clear that scales in everyday clinical use – such as the HAM-D or MADRS – reflect the diversity in depressive symptoms. Nor is it clear that they evaluate aspects of the illness that are of greatest concern to patients, Koen Demyttenaere, of the University of Leuven, Belgium, told the Seoul meeting. MADRS has a high sensitivity to change, simply because that was the way its items were chosen. This is a helpful psychometric property, but the scale may not capture the experience that is most important to those who suffer from depression.
It is not clear that clinical rating scales cover aspects of depression of greatest concern to patients
In fact, according to a recent study by Koen Demyttenaere and colleagues from other Belgian centres, there is considerable discordance between what physicians have in mind when they talk of being cured of depression and what patients mean by it. The study included 426 outpatients with a diagnosis of MDD.
Patients focus on positive affect
Physicians ranked alleviation of negative feelings (such as despair and hopelessness) as of greatest importance while what patients wanted most was to have a meaningful life, to enjoy it, to feel satisfied with themselves, and to be able to concentrate. This focus on outcomes involving positive affect was greater at three-month follow-up than at the start of antidepressant treatment, and it was higher in patients with recurrent depression than in those experiencing their first episode.
MDD in reality is a set of overlapping features including somatic and affective symptoms, functioning, and frequently also anxiety. One point of agreement in the Belgian study mentioned above is that both physicians and depressed patients ranked somatic symptoms as relatively unimportant.
Rudolf Uher and colleagues from London’s Institute of Psychiatry reviewed data from the STAR*D and GENDEP studies and included information from the observer-rated MADRS and HAM-D17 scales and the self-reported Beck Depression Inventory. Factor analysis suggested three dimensions – observed mood, cognition, and neurovegetative symptoms (which covered sleep and appetite) – and six factors.
Affect and cognition
One symptom cluster centred on what the authors termed “interests and activities”. This factor included interest, enjoyment, concentration, decisiveness, ability to feel, activity and energy, and sex.
Cognition and positive mood are very closely associated
This cluster is essentially cognition and positive affect, Professor Demyttenaere argued. And the fact that it predicts poor outcomes independently of depression severity at baseline suggests it is one on which we should focus. The prominence of such symptoms should have implications for treatment strategies and the evaluation of response, Uher and colleagues had suggested.
In contrast, of the 17 items in HAM-D17, only one relates to a patient’s work and interests. And of the ten items in the MADRS, only one (difficulty in concentrating) relates to cognition.
Changed emotional processing precedes clinical benefit
Enhancing patients’ capacity to experience the world in a positive way seems to be an integral part of successful therapy for depression. Objective assessment of emotion-related cognition can help us understand how antidepressants work and provides a measure of early changes that patients aren’t necessarily aware of, according to data presented by Catherine Harmer, Professor of Cognitive Neurosciences at the University of Oxford, UK.
Everyday cognitive operations are influenced by emotion
In a study of healthy controls and patients with depression, short-term administration of antidepressant brought patients’ capacity to recognize happy faces close to that of controls. This effect was not seen with placebo. A study involving memory for words with positive emotional connotation showed that antidepressant treatment achieved a similar normalisation of performance.
Patients whose processing of happy facial expressions had improved two weeks after starting antidepressants were more likely than those showing little change to experience clinical benefit at six weeks. This is a phenomenon that has now been independently replicated. So it seems that pharmacological therapy can set the scene for patients to start perceiving the world more positively. This early, objectively measured change paves the way to therapeutic response -- even though it is only later that patients start to feel better and rating scales pick up measurable improvements in symptoms.
Correlates in brain function
The corticolimbic network, and particularly the amygdala, is thought to be important in determining emotion. Functional MRI (fMRI) studies have found a relationship between vulnerability to depression and high responsiveness to negative stimuli in the amygdala. This suggests that the brain prone to depression is treating negative experiences as more deserving of attentional resources than positive ones.
By countering negative bias in emotional processing, therapy sets the scene for patients to start perceiving the world more positively
But it seems that bias towards negative facial expressions seen on fMRI can be reduced by a week of antidepressant therapy -- even though depressive symptoms at this stage have not been lessened.
Such early drug-induced reduction in negative bias may be a helpful step towards therapeutic response. But it is not the only factor. A change in emotional processing with antidepressant seems more effective when combined with aspects of the environment such as good interpersonal support, Professor Harmer has suggested.
Pathways to improvement
Cognitive dysfunction has been recognized as part of MDD since the condition was first described. Yet it has not been prioritized for treatment, nor in the evaluation of response.
Judith Jaeger, (Albert Einstein College of Medicine, New York, USA) listed a number of reasons for now giving cognition the attention it deserves:
- Around a half of all patients with MDD say they are troubled by cognitive problems.
- Between 30% and 50% of patients with MDD have objectively established cognitive dysfunction. The overall extent of dysfunction is roughly 0.5 of a standard deviation -- about equivalent to a blood alcohol level of 0.05%. Our consensus as a society is that this is a meaningful difference; and it is also meaningful clinically.
- Patients who report subjective impairment are not necessarily the same as those with objective evidence of cognitive dysfunction – which means it is important to measure it.
- Functional recovery from depression tracks improved cognitive function, and this relationship is independent of the persistence or resolution of mood symptoms.
- This suggests that cognitive impairment is not attributable to depressed mood per se.
- Residual functional disability is strongly associated with residual impairment of cognition.
- Though the data are not longitudinal, the work of Gorwood and others suggests that cognitive function declines as the number of episodes of depression increases.
It is therefore important to assess cognitive function, and also to assess whether there has been a change in cognitive function with onset of a depressive episode: many patients experience a reduction in cognitive capacities even though they remain within the normal range.
In assessing a patient’s cognition and the effects of treatment, we no longer need a full neuropsychological test battery. Computer-based tools that quickly and reliably measure changes in cognitive capacity are becoming available and have potential for routine application in the clinic, Dr Jaeger told the meeting.
The symposium was chaired by Professor Bernhard Baune (University of Adelaide, Australia) who, at the outset, drew attention to the number of citations attracted by the FOCUS study as an indication of growing interest in the cognitive dimension of depression.