Research paperThe lack of meaningful association between depression severity measures and neurocognitive performance☆
Introduction
Mild to moderate deficits in neurocognitive functioning are common in major depression (for review see Austin et al., 2001, Rock et al., 2014). However, there has been a vexing lack of correlation between measures of depression severity and these neurocognitive deficits that have led to a range of conclusions including that (a) there may not be a consistent association (McClintock et al., 2010), that (b) the association is only evident when considering large numbers of patients across a wide range of depression severity (McDermott and Ebmeier, 2009), or that (c) neurocognitive deficits only occur in a subsample of patients (e.g., see Iverson et al., 2011).
Studies of the association between neurocognitive functioning and depression severity as assessed via standard depression rating scales have been markedly inconsistent: some report an association (Cataldo et al., 2005, Naismith et al., 2003) while others find none (Abas et al., 1990, Brown et al., 1994). A meta-analysis cumulating across multiple studies found a moderate correlation between depression severity and performance on tasks of episodic memory, processing speed, and executive function (McDermott and Ebmeier, 2009), but included studies of both unmedicated and medicated patients, which may distort these associations. Although some studies have excluded medicated patients (Baudic et al., 2004, Martin et al., 1991), most include them (Airaksinen et al., 2004, Biringer et al., 2005, Harvey et al., 2004, Kessing, 1998, Kiosses et al., 2001, Naismith et al., 2003, Neu et al., 2005, Paradiso et al., 1997, Stordal et al., 2004, Taylor et al., 2002, Wang et al., 2006), and medication status was not reported in others (see McDermott and Ebmeier, 2009, Rohling et al., 2002). Depressed patients with psychosis have been included in some studies as well (Burt et al., 2000, Naismith et al., 2003, Majer et al., 2004).
Different depression rating scales have been used in these studies, including the 17- and 24-item Hamilton Depression Rating Scale (HDRS-17; HDRS-24; Grant et al., 2001, Burt et al., 2000), the Beck Depression Inventory (BDI; Wang et al., 2006), the Clinical Interview for Depression (CID; Murphy et al., 2003), the Montgomery-Asberg Depression Scale (MADRS; Beblo et al., 1999) or a combination of such measures (e.g., Kessing, 1998, Martin et al., 1991, Porter et al., 2003). However, even among studies using the same scales, there is little consistent association between symptom severity and neurocognitive difficulty (McClintock et al., 2010).
The purpose of this study was to directly test the contention that there is little to no association between neurocognition and depression severity, using a relatively large sample of unmedicated non-psychotic patients (n = 262), by carefully examining associations between neuropsychological test performance and a number of aspects of symptom severity. Initially, we examined relationships between test performance and summary scores of two standard depression rating scales – the Hamilton Depression Rating Scale and the Beck Depression Inventory.
We then decomposed these scales into factors to account for these scales’ heterogeneity. Previous studies from our lab have found that both the HDRS-24 and the BDI are not unitary scales (Grunebaum et al., 2005), and that factor-analytically derived “symptom clusters” from these scales may be more informative about underlying biological factors (such as regional cerebral glucose metabolism; Milak et al., 2005, Milak et al., 2010) and clinical characteristics (risk for suicide attempt and suicidal ideation; Grunebaum et al., 2005, Keilp et al., 2012, Keilp et al., 2018) than total scores. We speculated that these more specific symptom factor scores may better identify sets of symptoms associated with depression-related neurocognitive difficulty.
Finally, we added a specific measure of subjective cognitive complaint to our symptom assessment. The lack of association between measures of depression severity and actual neurocognitive performance may simply reflect the absence of specific questions about neurocognitive performance in standard depression rating scales. Both the HDRS-24 and the original BDI only ask broadly about “work difficulties” in a single item. The revised BDI-II has modified this single item to ask more specifically about concentration difficulties, but this is the only item in this scale that asks directly about neurocognitive performance. In this study, we obtained more specific information about subjective neurocognitive complaints using a separate scale, the Cognitive Failures Questionnaire (CFQ-25; Broadbent et al., 1982), a 25-item questionnaire including questions about cognitive lapses, to determine if these specific questions about cognition were more closely related to actual neurocognitive performance. This scale can also be decomposed into factors assessing memory complaints, distractibility, lapses in concentration (“blunders”), and naming failures, which provide an even more detailed characterization of patients’ complaints (Wallace et al., 2002, Pollina et al., 1992). If the depression severity scales themselves failed to correlate with neurocognitive performance, the CFQ-25 might indicate the types of questions to be added to symptom assessments to improve the characterization of these deficits.
Section snippets
Participants
Participants were 262 unmedicated patients with unipolar major depression enrolled in various clinical and biological studies (predictive, genetic, neurochemical, and/or imaging) at an academic medical center. A minimum score of ≥16 on the first seventeen items of the HDRS was required at the time of admission to the study (although at time of assessment, some patients had experienced some remission of symptoms, as noted below). Patients with psychosis were excluded. All patients were medically
Demographic and clinical characteristics
Depressed patients were older than healthy volunteers and had a higher proportion of females (both of which were accounted for in demographic adjustments of test scores), but with equivalent levels of educational attainment and comparable estimated intelligence (Table 1.). Samples were relatively well-educated with above-average estimated intelligence.
Depressed patients ranged in age from 18 to 72 and patients’ average age of onset was in their early 20′s, with multiple prior episodes, and a
Discussion
Depression severity measures, as well as ratings of subjective cognitive complaint, were only very weakly related to actual neurocognitive performance in this sample of unmedicated MDD patients, to a degree that did not meet our minimal significance standards. In all cases, across the entire sample, less than 5% of shared variance was explained by symptom severity, whether these included standard depression severity total scores, derived symptom factor scores, or specific questions regarding
Acknowledgements
The authors acknowledge the contributions of the staff of the Clinical Evaluation Core of the Conte Center for the Study of Suicidal Behavior. We also acknowledge the generosity of the patients who volunteered for this study in a time of significant personal distress.
References (77)
- et al.
Executive dysfunction and the course of geriatric depression
Biol. Psychiatry
(2005) - et al.
Psychometric properties of the beck depression inventory: twenty-five years of evaluation
Clin. Psychol. Rev.
(1988) - et al.
Neurocognitive predictors of antidepressant clinical response
J. Affect. Disord.
(2014) - et al.
Impulsivity in depressed children and adolescents: A comparison between behavioral and neuropsychological data
Psychiatry Res.
(2005) - et al.
The Continuous Performance Test, identical pairs version (CPT-IP): I. New findings about sustained attention in normal families
Psychiatry Res.
(1988) - et al.
Cognitive disturbance in outpatient depressed young adults: evidence of modest impairment
Biol. Psychiatry
(2001) - et al.
Symptom components of standard depression scales and past suicidal behavior
J. Affect. Disord.
(2005) - et al.
Executive functions and updating of the contents of working memory in unipolar depression
J. Psychiatr. Res.
(2004) - et al.
Identifying a cognitive impairment subgroup in adults with mood disorders
J. Affect. Disord.
(2011) - et al.
Suicidal ideation declines with improvement in the subjective symptoms of major depression
J. Affect. Disord.
(2018)