html xmlns=”http://www.w3.org/1999/xhtml” xmlns:mml=”http://www.w3.org/1998/Math/MathML” xmlns:epub=”http://www.idpf.org/2007/ops”>
Major depressive disorder (MDD) primarily involves mood disturbances, but its association with cognitive deficits is now well established (Austin, Mitchell, & Goodwin, 2001; Lee, Hermens, Porter, & Redoblado-Hodge, 2012). Furthermore, preliminary evidence suggests that non-specific cognitive impairment, such as disruption of arousal-activation, contributes to difficulties encountered in performing various effortful tasks (Weingartner, Cohen, Murphy, Martello, & Gerdt, 1981) observed in patients with MDD.
Historically, cognitive deficits were considered a residual effect resulting from the acute phase of a major depressive episode (MDE). Accumulating evidence suggests that cognitive deficits are trait-related rather than state-related (Boeker et al., 2012).
The implication of the foregoing observation is that there is an underlying neurobiological vulnerability to MDD. More specifically, the association between cognitive deficits and its independent expression beyond an acute MDE provides a basis for examining this relationship using Gottesman and Gould’s criteria for an “endophenotype” (Gottesman & Gould, 2003). Conventionally, an endophenotype constitutes a simpler phenotype used to aid in characterizing the disorder and its underlying substrates. Therefore, as measures of cognitive function improve in specificity and sensitivity (Austin et al., 2001; Beats, Sahakian, & Levy, 1996; Den Hartog, Derix, Van Bemmel, Kremer, & Jolles, 1999; Marazziti, Consoli, Picchetti, Carlini, & Faravelli, 2010; Purcell, Maruff, Kyrios, & Pantelis, 1997), our present understanding of their neuroanatomical correlates and risk factors will be augmented with the identification of endophenotypes associated with specific cognitive functions (Clark, Chamberlain, & Sahakian, 2009).
The aim of the present review is to analyze findings from recent studies on cognitive deficits in MDD and their characteristics and evolution at different stages of the disease. First, cognitive deficits, which are well described at the acute phase of an MDE, might persist beyond clinical recovery, have been related to neuroimaging abnormalities, and could have trait-like characteristics. This hypothesis raises the question of the presence of cognitive deficits before clinical onset, consisting in an underlying vulnerability to MDD. Moreover, longer-term studies of MDD patients provide information about progressive, cumulative, even neurotoxic impact of MDD on cognitive function, potentially leading to a higher risk of poorer functional outcome, depressive relapses, or even dementia.
Cognitive deficits persist beyond clinical recovery
Although several studies support the hypothesis that cognitive impairment is state-dependent (Austin et al., 2001; Beats et al., 1996; Biringer et al., 2005), recent literature with longitudinal assessments of neurocognitive function demonstrates that cognitive functions, such as memory (Gallagher, Robinson, Gray, Porter, & Young, 2007), attention (Weiland-Fiedler et al., 2004), and executive function (Grant, Thase, & Sweeney, 2001; Paelecke-Habermann, Pohl, & Leplow, 2005), are impaired at the acute stage of an MDE and this persists during remission. Moreover, the persistence of some cognitive deficits (e.g. episodic memory) has been reported to remain following longer-term resolution (e.g. three years) of symptomatic and social function (Airaksinen, Wahlin, Larsson, & Forsell, 2006). The persistence of neurocognitive deficits is clinically meaningful, as it has a significant impact on the patient’s ability to reach functional recovery. Indeed, persisting cognitive deficits are predictive of the level of functional recovery six months after hospitalization for MDD, independent of mood-state and medication (Jaeger, Berns, Uzelac, & Davis-Conway, 2006).
Neuroimaging studies also help to disentangle trait- versus state-dependent cognitive deficits because many cognitive functions rely on well-characterized neural circuits. Accordingly, it is interesting to note that some mood-state-dependent neurophysiological abnormalities are located in regions where metabolic activity increases during normal or pathological emotional states (i.e. ventrolateral prefrontal cortex, orbitofrontal prefrontal cortex, amygdala) or decreases in depressed patients (i.e. hippocampus, anterior cingulate cortex) (Bremner, Vythilingam, Vermetten, Vaccarino, & Charney, 2004). Likewise, other abnormalities, such as amygdala hyperactivity, have been reported to persist after clinical remission (Drevets, 2000; Paelecke-Habermann et al., 2005). Persistence of hypo- or hyper-activities may also disturb signals in cortico-subcortical networks, chronically influencing cognitive function (Paelecke-Habermann et al., 2005). Post-mortem studies of patients with recurrent MDEs report reductions in cortex volume and histopathological changes in brain regions pertinent to cognitive function (i.e. medial and orbital prefrontal cortex) (Davidson, Pizzagalli, Nitschke, & Putnam, 2002; Drevets, 2000).
A separate, but related, line of questioning is the distinction of early versus late stages of MDD. Assessing the presence of cognitive abnormalities after an acute MDE does not account for the fact that individuals who remit following a single MDE differ significantly from those who experience recurrent MDEs, wherein the probability of remission is usually highly correlated to the number of past depressive episodes (Kessing, Andersen, Mortensen, & Bolwig, 1998). Indeed, there is little evidence that cognitive deficits are present at the early stage of the illness. In a recent meta-analysis of 15 independent studies in adults with a first episode of MDD, Lee et al. (2012) demonstrated that during the MDE psychomotor speed, attention, visual learning and memory, and executive functioning were significantly and consistently reduced in first-episode patients compared with healthy controls. Psychomotor speed and memory performance improved and were associated with clinical factors, whereas deficits in attention and executive function were not, suggesting that these latter two cognitive functions may be more clearly trait markers (Lee et al. 2012; Paelecke-Habermann et al., 2005). These results are consistent with the findings of a systematic review of studies in which neuropsychological assessments were conducted at baseline and after follow-up in MDD patients: verbal learning and memory were more related to clinical state, whereas attention and executive function represented more trait-like markers of MDD (Douglas & Porter, 2009).
After clinical remission, Reppermund and colleagues (Reppermund, Ising, Lucae, & Zihl, 2009) found that performance in cognitive tests improved but was still impaired in 57 percent of 53 remitted patients, in a large number of cognitive domains. No significant differences in cognitive performance between remitted and non-remitted patients were observed; likewise, no correlation between cognitive measures and depressive symptoms were found. Cognitive impairment at baseline did not differ between patients with first versus recurrent MDEs, suggesting that cognitive deficits in MDD have trait-like features (Reppermund et al., 2009). The most recent meta-analysis of cognitive deficits in euthymic patients who had an MDE gathered 27 empirical studies (nMDD = 895; nControl = 997) and demonstrated that all studied cognitive functions (i.e. processing speed, visual memory, verbal memory, executive functions, working memory, attention) were impaired in patients when compared with the healthy controls; however, it should be noted that the magnitude of the deficits were modest when recurrent episodes were excluded (Bora, Harrison, Yucel, & Pantelis, 2013).
Considering that cognitive deficits in MDD are present at an early stage of the disorder, have trait-like features, persist after remission, and may even progress, the possibility that cognitive deficits precede the onset of MDD has to be raised.
Cognitive deficits before clinical onset
Studying asymptomatic relatives of patients with MDD, or at-risk subjects with mild to moderate symptoms, may help to elucidate whether cognitive deficits exist before the onset of an MDE, providing a basis for hypothesizing that cognitive dysfunction constitutes risk for MDD.
Christensen and colleagues led a case-control study with 203 healthy monozygotic and dizygotic twins with (high-risk) or without (low-risk, control group) a co-twin with affective disorder, and compared their cognitive performance. Healthy high-risk twins have discrete but persistent cognitive deficits including language processing, declarative memory, and executive function when compared with low-risk twins. The foregoing observations support the hypothesis that cognitive abnormalities are (1) shared by unaffected relatives; (2) involve genetic transmission; and (3) may be present before the onset of an affective disorder (Christensen, Kyvik, & Kessing, 2006).
A separate study of young women with no personal history of MDD but increased genetic risk of MDD demonstrated more cognitive deficits than control subjects in declarative memory, which was associated with increased cortisol secretion. No significant effects of the 5-HTT allelic variations were observed; however, the aforementioned findings suggest that cognitive deficit may precede the onset of clinical depression in at-risk subjects (Mannie, Barnes, Bristow, Harmer, & Cowen, 2009).
Supporting the previous article favoring the fact that at-risk subjects might also have cognitive abnormalities, Frodl and colleagues assessed the connectivity of brain areas potentially involved in cognitive skills (i.e. right frontal and orbitofrontal lobe, corpus callosum), comparing subjects at risk for MDD versus controls. At-risk subjects had changes in white matter fiber connections compared with healthy controls, mostly in case of early-life adversity (Frodl et al., 2012). This observation suggests that differences in neural fiber connections, in pathways involved in cognitive processes, may occur before the onset of the disease in at-risk subjects for MDD. Further investigations would be useful to disentangle the interactions between neuroplasticity and the effect of stress, and their protecting or enhancing role in the occurrence of depressive episodes.
If cognitive deficits are detected following an MDE, during periods of remission, as well as among individuals at increased risk of developing MDD, it is plausible that cognitive deficits represent a valid endophenotype. However, the challenge then resides in the clinical utility afforded by measures of cognitive deficits. Indeed, MDD is associated with non-recovery, recurrence, and chronicity, with accumulating evidence for further impairments to cognition with each subsequent episode. Taken together, the prognosis of treatment for MDD should take into consideration the number of past episodes, treatment response (Majer et al., 2004), cognitive performance, quality of life, and social functioning (McIntyre et al., 2013).
Burden of depressive history: are cognitive deficits worsening with time and recurrence of episodes (the “neurotoxicity” hypothesis)?
If cognitive deficits are being observed during, before, but also after an MDE, the fact that cognitive deficits may worsen with the recurrence of the depressive episodes can be proposed. Nevertheless, before any definite conclusion, it is important to assess which aspect of depressive disorder is concerned. Indeed, age at onset, age at interview, and number of past depressive episodes are overlapping concepts (e.g. older patients with young age at onset are more exposed to past depressive episodes). Proposing that past episodes are being neurotoxic implies the need to distinguish the roles of these factors.
Role of age of illness onset
Bora et al. (2013) have emphasized that not all studies report cognitive impairment in euthymic MDD patients (Biringer et al., 2005). In a meta-analysis of 27 existing studies that had examined cognition in euthymic patients with MDD, they found that cognitive impairments were obviated in euthymic patients with MDD across all examined domains, particularly for psychomotor speed and verbal memory. More severe impairments were associated with later onset (i.e. after 60), suggesting a role for vascular or neurodegenerative factors in some patients (Bora et al., 2013). Rapp et al. (2005) conducted a neuropsychological assessment of geriatric patients with MDD and reported marked deficits in episodic memory among those with recurrent MDD when compared with age-matched subjects without a past history of an MDE as well compared with patients with late-onset MDD. Deficits in attention and executive function were predominant in those with late-onset MDD in comparison with other patient groups; moreover, these deficits were associated with comorbid cardiovascular illness (Rapp et al., 2005). These results are consistent with a separate study by Grant et al. (2001) wherein a cohort of more than 100 unmedicated adult (average age 39 years) outpatients with MDD were assessed using a battery of standardized neuropsychological tests. Results indicated that patients with MDD did not exhibit significant impairments in memory, attention, or executive function. Higher symptom severity was only mildly associated with deficits in attentional shifting and psychomotor speed; likewise, longer depressive index episodes were associated with greater impairments on some aspects of executive function compared with shorter episodes. There was no relationship between the number of episodes and neuropsychological performance. However, age-corrected correlations between age at onset of first episode and performance in most of the neuropsychological tests were reported as significant. For example, a later age of onset was associated with poorer cognitive performance (Grant et al., 2001).
Role of lifetime illness severity: number of recurrent episodes of illness, total duration of illness, and number of hospitalizations
In subsets of patients with MDD, cognitive deficits may progress as a function of the number of depressive episodes, suggesting a toxic impact of depressive illness on brain functions with a cumulative effect over the lifetime (McIntyre, 2013). Some studies clearly supported the idea that with more past depressive episodes, different cognitive functions worsen, including response latencies, executive function, attention process, and working memory (Table 9.1). Beats and colleagues found a correlation between longer response latencies and number of prior depressive episodes (Beats et al., 1996). Neuropsychological tests were used to compare performance of 20 patients in recovery from recurrent unipolar depression, with 20 healthy patients: Bhardwaj and colleagues found that deficits in executive function, even after controlling for subclinical depressive symptoms, were correlated with the number of previous MDEs (Bhardwaj, Wilkinson, Srivastava, & Sharma, 2010). In a study examining 40 remitted patients and healthy controls, Paelecke-Haberman et al. (2005) found that not only remitted patients displayed impaired performance in attention and executive function in comparison with healthy controls, but also patients with a severe course of their illness (more than three depressive episodes) had lower outcome in executive function than patients with a mild course of their illness. Weiland-Fiedler et al. (2004) compared unmedicated outpatients with MDD with healthy controls and demonstrated that neurocognitive deficits (e.g. attention and working memory) remained significant, even after correction for residual depressive symptoms, after three months or longer of medication-free status following remission. Preiss et al. (2009) compared the neuropsychological performance of healthy subjects to that of remitted MDD outpatients; on average, patients displayed residual cognitive deficits in performing tasks of executive function, memory, and attention when compared with controls. No significant correlations between performance on any cognitive test and number of MDEs were reported. The number of past hospitalizations was related to Trail Making Test performance, which suggests that patients with more hospitalizations in their medical history perform worse in tasks involving attention and executive function (Preiss et al., 2009). Furthermore, Hasselbalch and colleagues reported that the cumulative duration of MDEs was associated with a decreased global cognitive function in remitted MDD patients several years (i.e. 12 years) following remission (Hasselbalch, Knorr, Hasselbalch, Gade, & Kessing, 2013).
&: for word fluency; °: during acute depression; #: at recovery; ^: for executive function only; €: cumulative duration of depressive episodes; $: for executive functions in severe MDD ( ≥ 3 episodes); £: for learning and memory; §: correlation with number of hospitalizations.
HAM-D: Hamilton Depression Rating Scale; MADRS: Montgomery–Åsberg Depression Rating Scale; HADS: Hospital and Anxiety Depression Scale; BADS: Behavioral Assessment of Dysexecutive Syndrome; CAMCOG: Cambridge Cognitive Examination; CANTAB: Cambridge Neuropsychological Test Automated Battery; FTT: Finger Tapping Test; (R)AVLT: (Rey) Auditory Verbal Learning Test; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; RCFT: Rey–Osterrieth Complex Figure Test; TMT: Trail Making Test; WCST: Wisconsin Card Sorting Test; WMS: Wechsler Memory Scale.
Related posts:
HPA axis and cognitive dysfunction in mood disorders
Hot and cold cognition in major depressive disorder
Implications of cognitive impairments on functional outcomes in major depressive disorder
Treatment of cognitive dysfunction in adults with major depressive disorder
Measuring the mind: detecting cognitive deficits and measuring cognitive change in patients with depression
A novel treatment targeting cognitive dysfunction in mood disorders
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
