The Montreal Cognitive Assessment (MoCA) was developed as a cognitive screening tool to help practitioners identify individuals with mild cognitive impairment.Reference Nasreddine, Phillips and Bedirian 1 It has since shown sensitivity in detecting cognitive impairment in a number of other clinical populations (e.g. cerebrovascular disease, Parkinson’s disease, Huntington’s disease, brain tumors, epilepsy).Reference Julayanont, Phillips, Chertkow and Nasreddine 2 Common to these various populations is a high prevalence of mood dysfunction.Reference Benedetti, Bernasconi and Pontiggia 3 - Reference Rickards 5 However, the influence of mood symptomatology, particularly depressive symptoms, on the MoCA remains unclear and represents an important area of investigation, especially because individuals presenting with depressive symptoms are common among memory clinic referrals.Reference Knapskog, Barca and Engedal 6
Both recent and a lifetime history of depressive symptomatology represent an increased risk factor for cognitive impairment and dementia.Reference Gao, Huang, Zhao and Ma 7 - Reference Wilson, Barnes and Mendes de Leon 11 In human and animal studies, depressive symptomatology has been linked to abnormal function, reduced volumes, and decreased connectivity in various brain regions (e.g. prefrontal cortex, amygdala).Reference Drevets 12 - Reference Sheline 14 In line with this, depressive symptomatology has been linked to deficits in various cognitive domains,Reference Steffens, Otey and Alexopoulos 10 , Reference Ganguli, Du, Dodge, Ratcliff and Chang 15 particularly executive functions, and dysfunction in associated brain regions.Reference Dotson, Davatzikos, Kraut and Resnick 16 - Reference Royall, Palmer, Chiodo and Polk 18
Consequently, the impact of mood symptoms on a highly sensitive global cognitive tool, such as the MoCA, may be important to clarify differential diagnostic issues, especially in family practice and memory clinic settings where full neuropsychological batteries are not readily available. However, at this point our understanding of the impact of mood symptoms on the MoCA is limited with few exceptions, primarily involving community as opposed to clinical samples.Reference Baer, Tabri, Blair, Bye, Li and Pushkar 19 - Reference Freitas, Simões, Alves and Santana 21 Therefore, in the current study, we examined whether depressive symptoms impacted MoCA performance by employing a memory clinic sample of individuals with depressive symptoms, with the prediction that they would demonstrate impaired scores on the overall MoCA as well as MoCA subtasks, particularly those reflecting executive/attentional domains.
Methods
Participants
Hospital records were reviewed for patients cognitive test scores on the MoCAReference Nasreddine, Phillips and Bedirian 1 at the time of their initial consultation to the Cognitive Neurology and Alzheimer’s Research Centre at Parkwood Institute. Charts obtained ranged from 2008 (when the MoCA was added to our cognitive screening battery) to April 2014. We included individuals with depressive symptoms as well as individuals diagnosed with Alzheimer’s disease or frontotemporal degeneration (FTD) and healthy volunteers.
The group with depressive symptoms (n=102) was composed of individuals who were referred to the clinic because of cognitive complaints (largely memory and concentration difficulties). They endorsed depressive symptoms in the mild depression range or higher on the Beck Depression Inventory (BDI; score >9)Reference Beck, Ward, Mendelson, Mock and Erbaugh 22 and did not have a prescription for a cholinesterase inhibitor or memantine (Namenda). The patients were primarily taking medications for management of cholesterol and blood pressure. Exclusion criteria for this group with depressive symptoms included a diagnosis at presentation or follow-up of a major neurocognitive disorder, 23 a neurological disorder (e.g. epilepsy, Korsakoff syndrome, obstructive hydrocephalus, cerebral palsy), or mild cognitive impairment,Reference Winblad, Palmer and Kivipelto 24 except when the mild cognitive impairment criteria was fulfilled but depressive symptomatology was deemed the primary explanatory factor by the clinicians (SP and EF). Patients were also excluded when basic laboratory tests results indicated treatable disorders underlying their cognitive complaints (e.g. low vitamin B12, hypothyroidism). Chart review of the clinical consultation revealed that a history of emotional dysfunction was reported by 31.5% of these patients, largely depression and/or anxiety symptoms (26.5%), and to a lesser extent patients reported panic disorder (2%), bipolar disorder (1%), and posttraumatic stress disorder (2%).
A majority of the group with depressive symptoms attended the clinic once (72.5%). A total of 43.1% of these patients were scheduled for a follow-up visit to track any cognitive changes, but only 39.5% returned. Of those returning (28 patients), 39.2% were seen once, with an average follow-up time between first and last clinic visits of 6 months; 17.9% were seen twice, with an average follow-up time of 18 months; and 42.9% were seen three or more times, with an average follow-up time of 29 months. Of the 28 patients followed, two were eventually diagnosed with Alzheimer’s disease after approximately 3 years of follow-up. After the initial consultation, 56.9% of patients with depressive symptoms were not scheduled for follow-up because depressive symptoms were considered essential/primary to their cognitive status; thus, they were recommended to follow-up with their primary care physicians for treatment (either directly or through psychological/psychiatric referral), and to be re-referred should cognitive complaints persist.
Alzheimer’s disease was diagnosed according to the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s disease and Related Disorders Association (n=137).Reference McKhann, Drachman and Folstein 25 FTD patients met Neary et alReference Neary, Snowden and Gustafson 26 criteria and included those presenting with behavior (n=46) or language symptoms (progressive nonfluent aphasia, n=12; semantic dementia, n=9). Study exclusion for the dementia groups included any history of traumatic brain injury, stroke, tumor or other neurologic or psychiatric disorder that could account for their symptoms.
Healthy volunteers (n=41) included individuals recruited from the community through posters and newspaper advertisements, caregiver support groups, or approached while accompanying friends or family to clinic visits. They were included if they reported general good physical and cognitive health, and excluded if they reported medical, psychological, or motor conditions that impacted their activities of daily living or their performance on cognitive tests. They reported no cognitive complaints or changes more than what is typical for normal aging relative to their peers and did not have a prescription for a cholinesterase inhibitor or memantine (Namenda). In addition, they were not recruited on the basis of normative cutoffs on a cognitive or mood measure.
Procedures and Materials
All participants completed the MoCA (with education correction appliedReference Nasreddine, Phillips and Bedirian 1 ) and the BDI.Reference Beck, Ward, Mendelson, Mock and Erbaugh 22 Additional cognitive screening measures administered to all participants as part of a standard brief cognitive battery at our clinic were: Mini-Mental State Examination,Reference Folstein, Folstein and McHugh 27 Rivermead Behavioral Memory Test Prose Recall Passage,Reference Wilson, Cockburn and Baddeley 28 Trail Making Test,Reference Strauss, Sherman and Spreen 29 verbal fluency (FAS and animals),Reference Strauss, Sherman and Spreen 29 the Western Aphasia Battery Naming subtest,Reference Kertesz 30 and the Clock Drawing Test.Reference Rouleau, Salmon and Butters 31
Results
MoCA scores for the group with depressive symptoms and patients with
Alzheimer’s disease and FTD are shown in Table 1 and Figure 1. As
indicated in Table 1, half of the
group with depressive symptoms scored below the suggested normal MoCA cutoff of
26. Across the groups, mean MoCA scores were significantly different using a
one-way analysis of variance (ANOVA), F(3, 343)=124.82,
p<.05,
$$\eta _{p}^{2} $$
=.52 (Figure 1).
Post-hoc analyses revealed that as a group, those with depressive symptoms
(mean [M]=25.05, standard deviation [SD]=3.29) scored significantly lower than
healthy controls (M=27.59, SD=1.69), p<.05, but
significantly higher than the Alzheimer’s (M=14.47, SD=5.68) and FTD groups
(M=17.33, SD=6.96), p<.05 for all (Bonferroni correction applied; Table 1). Because education is known to
influence MoCA scores,Reference Nasreddine, Phillips and Bedirian
1
we examined the impact of education on group comparisons by removing
all participants with lower education (<10 years of education) and
re-running the analyses (20% of participants removed). The ANOVA results
obtained were unchanged and Bonferroni corrected post hoc contrasts showed a
marginal difference between those with depressive symptoms and healthy
controls, p=0.1.
Figure 1 Mean and individual data points on the Montreal Cognitive Assessment (MoCA) for each group (AD = Alzheimer’s disease; FTD = frontotemporal dementia; Depressive = Group with depressive symptoms; Controls = healthy controls).
Table 1 Demographics, BDI ratings, and MoCA scores by group
AD=Alzheimer’s disease; BDI=Beck Depression Inventory; Dep=group with depressive symptoms; F=female; FTD=frontotemporal dementia; HC=healthy control; M=male; MoCA=Montreal Cognitive Assessment; SD=standard deviation.
* Dep, HC<FTD<AD; p<0.05 for all using one-way analysis of variance.
† Dep, HC, FTD>AD; p<0.05 for all using one-way analysis of variance.
‡ Dep>AD, FTD, HC; FTD>HC; p<0.05 for all using one-way analysis of variance.
§ HC>Dep>FTD>AD, p<0.05 for all using one-way analysis of variance.
To examine cognitive domains affected in those with depressive symptoms, we compared their MoCA subtask scores to healthy controls using independent samples t-tests. These analyses demonstrated that participants with depressive symptoms scored significantly lower than healthy controls on visuospatial/executive, repetition, and fluency components of the MoCA, p<0.005 for all (Bonferroni correction applied; Figure 2). These results remained unchanged when participants with lower education were removed (<10 years of education; 9% of participants removed), p<0.005 for all (Bonferroni correction applied). Analyses on supplementary cognitive screens revealed that healthy controls scored significantly better than patients with depressive symptoms on delayed prose recall, Clock Drawing Test, verbal fluency (FAS and animals), and the Trail Making Test, p<0.005 for all (Table 2; Bonferroni correction applied).
Figure 2 MoCA sub-task performance (mean percentage and standard error bars) of the group with depressive symptoms and healthy controls (Visuo/EF = visuospatial/executive function; * significant group differences).
Table 2 Cognitive test scores showing means and standard deviations
Highest possible scores: MMSE=30; prose recall - immediate=21; prose recall - delay=21; clock - command=10; clock - copy=10; WAB naming=20; TMT A and B discontinued after 300 seconds.
Dep=group with depressive symptoms; HC=healthy control; MMSE=Mini-Mental State Examination; TMT=Trail Making Test; WAB=Western Aphasia Battery.
*Significant results using independent samples t-tests, p<.005 (Bonferroni corrected).
Discussion
In this study, we examined the influence of mood symptoms on MoCA performance in a memory clinic sample of patients endorsing depressive symptoms. Half of the group with depressive symptoms scored below the MoCA-suggested cutoff for cognitive impairment. As a group, these individuals scored below healthy controls, but above those with Alzheimer’s disease and frontotemporal dementia. In addition, MoCA subtask analysis revealed a pattern of executive/attentional dysfunction in the group endorsing depressive symptoms.
The findings of this study demonstrate the susceptibility of individuals with depressive symptoms to obtain impaired scores on the MoCA, not unlike other cognitive screens (prose recall, clock drawing, verbal fluency, and trail making) (see Musso et al for similar MoCA findings in severe mental illness impacting moodReference Musso, Cohen and Auster 32 ). Thus, these results provide an important caveat/potential explanatory factor for clinicians when using the MoCA to interpret a patient’s global cognition. In a memory clinic setting, with cognitive impairment indicated by the MoCA in the context of depressive symptoms, differential diagnosis may be aided by detailed assessment of functional capacities, as well as cognitive investigations with other domain-specific screens or in-depth neuropsychological evaluation. Such additional testing often reveals more severe cognitive deficits across various domains in those with dementia relative to individuals with depressive symptoms.Reference Christensen, Griffiths, MacKinnon and Jacomb 33 , Reference Wright and Persad 34 Additionally, objective cognitive impairment in those presenting with depressive symptoms necessitates pharmacological or psychotherapeutic intervention; however, cognitive benefits may potentially be moderated by underlying neurodegenerative pathology.
Together with global cognitive decline, deficits in a range of cognitive domains are associated with mood dysfunction, especially in those with late-life onset of mood symptoms (age 50-65 years).Reference Herrmann, Goodwin and Ebmeier 35 In the present study, individuals with depressive symptoms showed an executive/attentional pattern of weakness relative to healthy controls on the MoCA, consistent with a recent meta-analysis showing the impact of mood dysfunction on multiple executive/attentional domains.Reference Snyder 36 Moreover, these observed executive/attention deficits are in line with theories of mood dysfunction that posit dysregulation in prefrontal regions (asymmetric ventromedial vs dorsolateral function)Reference Koenigs and Grafman 37 and cortico-striatal-pallidal-thalamic circuits (dysfunction in prefrontal subregions and anterior cingulate cortex connections with basal ganglia and thalamic structures).Reference Bora, Harrison, Davey, Yücel and Pantelis 38 , Reference Price and Drevets 39
A strength of this work is the observed influence of mood symptoms on the MoCA in a large cohort of patients with depressive symptoms in a clinical setting, unlike research thus far that primarily involves community samples. Nevertheless, this work is limited by the inclusion of patients with depressive symptoms who presented with subjective cognitive complaints, the majority of whom were without a psychiatric diagnosis at the time of consultation. Thus, research is needed to clarify whether our findings extend to individuals with a diagnosed mood disorder. Of the 40% of these individuals with depressive symptoms who returned for follow-up, only two eventually met criteria for Alzheimer’s disease, suggesting that the majority of these patients did not progress to a neurodegenerative disease. A longer longitudinal follow-up may be necessary to clarify whether the group with depressive symptoms will eventually develop neurodegenerative disease.
Moreover, although we used of a healthy control comparison group, it might be argued than a better control group would include patients presenting to the clinic with similar characteristics as our group with depressive symptoms (cognitive complaints, same inclusion and exclusion criteria), but who do not meet criteria for depressive symptomatology (BDI<10). Notably, these patients (n=40; retrieved from our clinic database) had similar mean MoCA scores as our group with depressive symptoms (p>0.05) and 43% fell below MoCA cutoff (lower than the 50% in the group with depressive symptoms). Although these patients did not meet BDI criteria for depressive symptoms, they represent a questionable control group in a memory clinic setting because depressive symptomatology cannot be ruled out in these patients; 23% reported either a history of depression or stated at least two symptoms that comprise a major depressive episode according to Diagnostic and Statistical Manual of Mental Disorders IV criteria. 40 Furthermore, as the BDI threshold for depressive symptoms is likely variable depending on the medical population studied,Reference Wang and Gorenstein 41 future work with patients diagnosed by Diagnostic and Statistical Manual of Mental Disorders criteria will help to further clarify the sensitivity of the MoCA to depressive symptomatology, an approach not typically afforded to memory clinics.
In conclusion, individuals endorsing depressive symptoms often demonstrate scores on the MoCA that fall below the normal cutoff. Consequently, mood status, documented through routine mood screening, should be considered an important caveat when interpreting MoCA scores. Longitudinal research, with the involvement of early neuropathological biomarkers, is necessary to better inform this mood-cognition relationship and the MoCA’s prognostic value to differential diagnosis in patients presenting depressive symptoms.
Disclosures
The authors have nothing to disclose.
