Abstract

The Montreal Cognitive Assessment (MoCA) is a relatively newly designed test that was developed as a tool to screen patients with mild cognitive problems that are not typically detected by the Mini-Mental State Exam (MMSE). While early research suggests that the MoCA is more sensitive to subtle cognitive impairment than the MMSE, there is concern about potential decreased specificity when using the MoCA. The aim of the present study was to examine the comparative utility of using the MoCA and the MMSE to detect subtle cognitive impairment among a group of 82 middle-aged U.S. military veterans referred for outpatient neuropsychological testing. Using receiver operating characteristic analyses, the MoCA was shown to be a better predictor of subtle cognitive impairment on neuropsychological testing than the MMSE. When using an adjusted cutoff, the MoCA was shown to be more sensitive (i.e., 0.72 vs. 0.52) and nearly as specific as the MMSE (0.75 vs. 0.77).

Introduction

Screening tests are of great use to clinicians in identifying suspected cognitive impairment. The Montreal Cognitive Assessment (MoCA: Nasreddine et al., 2005) is a relatively new screening test that was “developed as a tool to screen patients who present with mild cognitive complaints and usually perform in the normal range on the MMSE” (p. 696). Initial studies have shown that the MoCA may, indeed, be more sensitive to cognitive impairment than the Mini-Mental State Exam (MMSE) across a variety of populations, including mild cognitive impairment (MCI) and mild Alzheimer's disease (Nasreddine et al., 2005), transient ischemic attack and stroke (Pendlebury, Cuthbertson, Welch, Mehta, & Rothwell, 2010) and Parkinson's disease (Nazem et al., 2009). However, there is growing concern that its increased sensitivity comes with an accompanying decrease in specificity.

In the initial validation study of 94 individuals with clinically diagnosed MCI, the authors of the MoCA showed that, using a cut-off score of ≤25 on both measures, the MMSE had a sensitivity of 18% for detecting MCI, while the sensitivity of the MoCA was higher, falling at 90% (Nasreddine et al., 2005). In the same study, the MMSE had a sensitivity of 78% for detecting mild Alzheimer's disease among a separate group of patients clinically diagnosed with that disorder (N = 93), while the MoCA detected 100%. Specificity was calculated separately among a group of 90 healthy elderly controls and was found to be 87% for the MoCA and 100% for the MMSE.

Other studies using clinical samples, rather than healthy controls, have not been quite so optimistic about the specificity of the MoCA. For example, in a study of 110 cardiovascular outpatients, using a cut-off of ≤23, the MoCA's sensitivity for detecting amnestic MCI was 100% and for multiple-domain MCI was 83%, while specificity rates were only 50% and 52%, respectively (McLennan, Mathias, Brennan, & Stewart, 2011). Using the publisher-recommended cutoff of ≤25 on the MoCA, as used in the initial validation study, furthered lowered specificity. Specifically, the MoCA's sensitivity was 100% for detecting amnestic MCI and 83% for detecting multiple-domain MCI. However, specificity rates decreased to 29% and 30%, respectively.

Similarly, in a separate study comparing the MoCA with the MMSE among 95 patients with recent cerebral infarct or hemorrhage, specificity of the MoCA for detecting cognitive impairment was less than ideal (Godefroy et al., 2011). Using raw scores and “published cutoffs,” which were undeclared, the authors reported that the MoCA was indeed more sensitive than the MMSE, with sensitivity levels falling at 94% and 66%, respectively. However, the specificity of the MoCA was far worse than the MMSE, falling at 42% and 97% for each measure, respectively. Notably, due to concern about the effects of age and education on MoCA and MMSE scores, the authors computed new cutoffs for the MoCA (<20) and the MMSE (<24). Using these new cutoffs, it was shown that the MoCA was no longer more sensitive than the MMSE, with sensitivity levels falling at 67% and 70% for each measure, respectively. However, while still less than the specificity of the MMSE, which was 97%, the specificity of the MoCA, which was 90%, was much better than when using published cutoffs. Based on their findings, the authors concluded that the good sensitivity of the MoCA documented in most published studies is primarily attributable to the choice of cut-off scores, which favored sensitivity at the cost of specificity. They further recommended that the refinement of MoCA norms was mandatory for determining optimal cut-off scores.

Given the aforementioned concerns about the sensitivity and specificity of the MoCA, the aim of the present study was to examine the comparative utility of using the MoCA and the MMSE to detect subtle cognitive impairment. To provide optimal objectivity, subtle cognitive impairment was measured with a comprehensive neuropsychological battery. The sample employed was a group middle-aged U.S. military veterans clinically referred for outpatient cognitive testing, which was an ideal group for addressing the published proposal that the MoCA may be useful as a brief screening tool for cognitive impairment in neurologic diseases in younger populations (Ratchford, Ochoa, & Finney, 2008). Consistent with previously published findings of greater sensitivity of the MoCA, it was hypothesized that, using published cutoffs, impairment on the MoCA would be seen with greater frequency than impairment on the MMSE. It was also hypothesized that the MoCA would be a better predictor of subtle cognitive impairment than would the MMSE. However, consistent with previously published findings, it was predicted that the optimal cutoff (i.e., determined using sensitivity/specificity values) for predicting cognitive impairment using the MoCA would be lower than the published suggested cut-off value of ≤25.

Method

Participants

Data were collected from the files of 82 outpatients referred to the lead author (K.A.W.) for neuropsychological testing at a Department of Veterans Affairs (VA) Medical Center. Referral sources included the Psychiatry Ambulatory Care Clinic, Polytrauma Clinic, Operation Iraqi Freedom/Operation Enduring Freedom Clinic, Primary Medical Clinics, and the Neurology Clinic. No patients were diagnosed with mental retardation. Consecutive referrals were reviewed for cases that passed the Test of Memory Malingering (Tombaugh, 1996) and were administered all neuropsychological measures used in the present study. As a rule, all patients referred to the lead author were administered the core neuropsychological battery (outlined in the Measures and Procedures section) used in the present study. The only exceptions were those patients referred for a retest, in which case they were administered a different stand-alone symptom validity test, and those patients who were referred for testing related to attention deficit hyperactivity disorder (ADHD), in which case they were generally given a different battery. Thus, patients referred for a retest or testing for ADHD were generally excluded from this study. It should be noted that, in the present study, all patients were primarily referred for neuropsychological evaluation to assess the potential presence of cognitive dysfunction, not primarily to assess for the presence of psychiatric disorder. See Table 1 for diagnoses mentioned by the referral source or in the patient's history at the time the patient was referred for cognitive testing.

Table 1.

Diagnoses mentioned by referral source or in patient history at time of referral

Diagnosis Number of total sample (N = 82) 
Traumatic Brain Injurya 14 (17.1%) 
 Mild Traumatic Brain Injury 78.6% 
 Moderate-to-Severe 21.4% 
Other Major Neurological Diagnosis 23 (28.0%) 
 Stroke 30.4% 
 Seizures 17.4% 
 Dementia/Capacity Issues 13.0% 
 Subarachnoid Hemorrhage 4.3% 
 Parkinson's Disease 4.3% 
 Multiple System Atrophy 4.3% 
 Unexplained Coma 4.3% 
 Global Atrophy 4.3% 
 Heparin-Induced Thrombosis 4.3% 
 Ischemia and Hypoxia During Hip Surgery 4.3% 
 Resected Frontal Lobe Brain Tumor 4.3% 
 “Chemobrain” 4.3% 
No Obvious Major Neurologic Dysfunction 45 (54.9%) 
 Psychiatric Diagnosis Only 22 (26.8%) 
  Anxiety/Depression 54.5% 
  Depression and PTSD 18.2% 
  ADHD/Learning Disability 9.1% 
  Bipolar Disorder 4.5% 
  Bipolar Disorder/PTSD 4.5% 
  Transient Global Amnesia 4.5% 
  Depression/Anxiety/Tic Disorder 4.5% 
 Neurologic Diagnosis Only 2 (2.4%) 
  TIA/Minor Stroke/Small Vessel Ischemic Disease 100% 
 Psychiatric and Neurologic Diagnoses 13 (15.9%) 
  Neurological Conditionb  
  Transient Ischemic Attack/SVID/Lacunar Infarcts 61.5% 
  Obstructive Sleep Apnea/Sleep Related Hypoventilation 23.1% 
  Pseudoseizures 7.7% 
  Pituitary Adenoma 7.7% 
  Toxic Exposure 7.7% 
 Psychiatric Conditionb  
  Depression and/or Anxiety 76.9% 
  Post-Traumatic Stress Disorder 30.8% 
Diagnosis Number of total sample (N = 82) 
Traumatic Brain Injurya 14 (17.1%) 
 Mild Traumatic Brain Injury 78.6% 
 Moderate-to-Severe 21.4% 
Other Major Neurological Diagnosis 23 (28.0%) 
 Stroke 30.4% 
 Seizures 17.4% 
 Dementia/Capacity Issues 13.0% 
 Subarachnoid Hemorrhage 4.3% 
 Parkinson's Disease 4.3% 
 Multiple System Atrophy 4.3% 
 Unexplained Coma 4.3% 
 Global Atrophy 4.3% 
 Heparin-Induced Thrombosis 4.3% 
 Ischemia and Hypoxia During Hip Surgery 4.3% 
 Resected Frontal Lobe Brain Tumor 4.3% 
 “Chemobrain” 4.3% 
No Obvious Major Neurologic Dysfunction 45 (54.9%) 
 Psychiatric Diagnosis Only 22 (26.8%) 
  Anxiety/Depression 54.5% 
  Depression and PTSD 18.2% 
  ADHD/Learning Disability 9.1% 
  Bipolar Disorder 4.5% 
  Bipolar Disorder/PTSD 4.5% 
  Transient Global Amnesia 4.5% 
  Depression/Anxiety/Tic Disorder 4.5% 
 Neurologic Diagnosis Only 2 (2.4%) 
  TIA/Minor Stroke/Small Vessel Ischemic Disease 100% 
 Psychiatric and Neurologic Diagnoses 13 (15.9%) 
  Neurological Conditionb  
  Transient Ischemic Attack/SVID/Lacunar Infarcts 61.5% 
  Obstructive Sleep Apnea/Sleep Related Hypoventilation 23.1% 
  Pseudoseizures 7.7% 
  Pituitary Adenoma 7.7% 
  Toxic Exposure 7.7% 
 Psychiatric Conditionb  
  Depression and/or Anxiety 76.9% 
  Post-Traumatic Stress Disorder 30.8% 

aTraumatic brain injury categorization was based on self-report and is of questionable validity.

bSubcategories listed below this group are not mutually exclusive.

In terms of patient demographics, age of participants ranged from 29 to 62, with a mean age of 56.1 years (SD = 9.6). Highest year of education completed by participants ranged from 7 to 20, with a mean level of 13.1 years (SD = 2.7). In terms of gender, 79 (96.3%) participants were men and 3 (3.7%) were women. Sixty-nine of 82 participants (84.1%) were Caucasian and 13 (15.9%) were African American.

Measures and Procedures

At the time of the study, participants' medical records were retrospectively reviewed and cases were categorized into groups according to whether they demonstrated cognitive impairment (n= 29) or not (n= 53). Cognitive impairment was defined as a score of ≤1.5 SD below the mean on two domains of neuropsychological testing. Neuropsychological tests were administered within the following domains: memory, sustained attention, language, executive function, and visual-spatial construction. For puposes of the present study, scores included in the memory domain were age-scaled scores from Logical Memory II and Visual Reproduction II of the Wechsler Memory Scale-IV (Wechsler, 2009). Scores included in the sustained attention domain were normatively based Z-scores (Saykin et al., 1995) for total comissions and total omissions on the Adult Vigilance Task of the Gordon Diagnostic System (Gordon, McClure, & Aylward, 1996). Scores in the language domain included normatively based T-scores on the Boston Naming Test (Kaplan, Goodglass, & Weintraub, 1983) and the Category Fluency Test (Gladsjo, Miller, & Heaton, 1999). Scores within the executive function domain included Total Errors T-score on the Wisconsin Card Sorting Test-64 Card Version (Kongs, Thompson, Iverson, & Heaton, 2000). Scores within the visual-spatial construction domain included age-scaled scores from the Block Design from the Wechsler Adult Intelligence Scale-IV (Wechsler, 2008) and Visual Reproduction Copy from the Wechsler Memory Scale IV (Wechsler, 2009). In addition, scores from the anxiety (PT: Psychasthenia) and Depression (D) clinical scales of the Minnesota Multiphasic Personality Inventory-2 (MMPI-2: Butcher et al., 2001) were collected from patient files, as were scores from the MMSE (Folstein, Folstein, & McHugh, 1975) and the MoCA (Nasreddine et al., 2005). A cutoff of ≤26 was chosen to indicate impairment on the MMSE (Crum, Anthony, Bassett, & Folstein, 1993). A cutoff of ≤25 was used to indicate impairment on the MocA, as suggested in the test manual. Administration of all measures took place as part of routine clinical care and followed the standard procedures as defined in the respective test manuals.

Statistical Analyses

Except where indicated, statistical analyses were calculated using SPSS Version 19 (SPSS, 2010). Initial analyses consisted of computing the frequency with which patients scored below cutoffs on the MoCA and the MMSE. Next, independent t-tests were conducted to compare group differences in MoCA and MMSE scores between the group demonstrating cognitive impairment and the group not demonstrating cognitive impairment. The α was set at 0.05 for this and all other analyses. Cohen's d effect sizes, along with 95% confidence intervals (CIs) for these group differences were calculated using a computerized program provided by Devilly (2004).

Linear regression analyses were conducted to evaluate the incremental validity of the MoCA compared with the MMSE in discriminating persons who demonstrated cognitive impairment versus those who did not. For the first analysis, MMSE score was entered in the first block and the MoCA score in the second block. The F (change) statistic was used to evaluate whether or not the MoCA contributed incrementally to the prediction of group membership. In the second analysis, the order of entry was reversed, with the MoCA score entered in the first block and MMSE entered in the second block. In the latter analyses, the usefulness of the MMSE in incrementally contributing to the MoCA in predicting cognitive impairment was evaluated by looking at the significance of the F (change) statistic.

A receiver operating characteristic (ROC) curve analysis was used to evaluate the usefulness of varying MoCA and MMSE cutoffs in predicting cognitive impairment. As part of the ROC analysis, the sensitivity and specificity of the MoCA and MMSE at various cutoffs was examined. After examining the sensitivity and specificity of the MoCA and MMSE at the cutoffs that were chosen a priori for this study, the Youdon Index was used to choose the optimal cutoff for each measure and then sensitivity and specificity were reconsidered. The Youdon Index is calculated by sensitivity + specificity − 1 (Youdon, 1950).

Following the ROC analysis, positive and negative predictive values (PPV and NPV) were calculated. As explained by O'Bryant and Lucas (2006), PPV refers to the likelihood that an individual has condition X (i.e., cognitive impairment as evidenced by scoring ≤1.5 SDs from the mean on two cognitive domains on neuropsychological testing) given positive findings on test Y (i.e., falls below cut-off score on the MoCA; Glaros & Kline, 1988; McCaffrey, Palav, O'Bryant, & Labarge, 2003). NPV is defined as the likelihood that the individual does not have condition X (i.e., is not demonstrating cognitive impairment as evidenced not scoring <1.5 SDs from the mean on two cognitive domains on neuropsychological testing) given a negative finding on test Y (i.e., scores above cutoff on the MoCA; Glaros & Kline, 1988; McCaffrey, Palav, O'Bryant, & Labarge, 2003). Both PPV and NPV were calculated using the formulae presented in O'Bryant and Lucas (2006). An estimated base rate of the condition in question (in this case, cognitive impairment as evidenced by neuropsychological testing) is needed to calculate PPV and NPV. In the present study, a base rate of 35% was employed, as that was the percentage of patients who demonstrated cognitive impairment.

Results

The MoCA score was ≤25 in 48 patients, whereas the MMSE score was ≤26 in only 27 patients. With regard to those scoring in the impaired range on the MoCA, neuropsychological impairments on comprehensive testing were most frequently seen within the memory domain (54%, n= 26/48), followed by attention (44%; n= 21/48), language (40%, n= 19/48), and executive function (38%; n= 18/48). With regard to those scoring below the cutoff on the MMSE, neuropsychological impairments on comprehensive testing were also most frequently seen within the memory domain (63%, n= 17/27), followed by attention (48%, n= 13/27), executive function (37%, n= 10/27), and language (33%, n= 9/27). No one with an impaired MoCA score or MMSE score demonstrated impairment within the visual-spatial construction domain. Post hoc t-tests showed that, for patients who completed the MMPI-2 (n= 73/82), there were no significant differences on the Depression and Psychasthenia scales between patients scoring above and below the cutoff on the MoCA. There were also no significant differences in Depression and Psychasthenia scores for those scoring above and below the cutoff on the MMSE.

Results of the t-tests showed that there were significant differences in MoCA and MMSE scores between the groups with and without cognitive impairment on neuropsychological testing (Table 2). The effect sizes for these group differences are considered large (Lipsey, 1990) and were 1.10 and 0.77 for the MoCA and MMSE, respectively. In the regression analyses (Table 3), the MoCA added significantly to the MMSE in predicting cognitive impairment (R2 change = .11, p < .01). However, the MMSE did not add significantly to the MoCA in predicting cognitive impairment (R2 change = .01, p = .298). Review of β-weights for the analyses suggested that the MoCA provided the greatest relative contribution in predicting group membership.

Table 2.

Group differences in MoCA and MMSE scores among groups with and without cognitive impairment

  No Cognitive Impairment (n = 53; M [SD]) or n [% of group]) Cognitive Impairment (n= 29; M [SD]) or n [% of group]) t(80)
 
Cohen's d
 
t p-value d 95% CI 
MMSE (/30) 27.8 (1.6) 26.1 (2.7) 3.45 0.005 0.77 0.30–1.23 
MoCA (/30) 25.6 (2.8) 22.1 (3.5) 4.83 0.001 1.10 0.62–1.59 
  No Cognitive Impairment (n = 53; M [SD]) or n [% of group]) Cognitive Impairment (n= 29; M [SD]) or n [% of group]) t(80)
 
Cohen's d
 
t p-value d 95% CI 
MMSE (/30) 27.8 (1.6) 26.1 (2.7) 3.45 0.005 0.77 0.30–1.23 
MoCA (/30) 25.6 (2.8) 22.1 (3.5) 4.83 0.001 1.10 0.62–1.59 
Table 3.

Hierarchical regression analyses: Predicting cognitive performance (impaired vs. not impaired)

Model Block Scale Final β R2 R2change Fchange p-value 
MMSE −0.12 .13    
MoCA −0.40 .24 .11 10.90 .001 
MoCA −0.40 .22    
MMSE −0.12 .23 .01 1.10 .298 
Model Block Scale Final β R2 R2change Fchange p-value 
MMSE −0.12 .13    
MoCA −0.40 .24 .11 10.90 .001 
MoCA −0.40 .22    
MMSE −0.12 .23 .01 1.10 .298 

As shown by the ROC analysis, for the MoCA, the area under the curve of 0.777 (95% CI: 0.670–0.883) suggests that the predictive information captured by the MoCA score was reasonably good. The same was true for the MMSE, where the area under the curve was smaller, but still acceptable at 0.679 (95% CI: 0.554–0.805) (Fig. 1). Using ≤26 as the cutoff, the MMSE was moderately sensitive (0.52) to cognitive impairment and had moderate specificity (0.77) (Table 4). In comparison, the MoCA (cutoff <25) had excellent sensitivity (0.86), but poor specificity (0.57). For the MoCA, using the optimal cutoff (i.e., ≤23) as determined by the Youden index, improved the specificity to 0.75, which was nearly equivalent to the MMSE, but decreased the sensitivity to 0.72, the latter of which was still far higher than the MMSE. Use of the Youdon index for the MMSE resulted in the same suggested cutoff as the one chosen apriori for this investigation (i.e., ≤26). As can be seen in Table 4, using a cutoff of ≤23 on the MoCA resulted in a PPV of 0.61 and a NPV of 0.83.

Table 4.

The classification accuracy of MoCA and MMSE cutoffs in predicting cognitive impairment

 Cut-off ≤ Sensitivity Specificity PPV NPV 
MMSE 20 0.03 1.0 1.0 0.66 
 21 0.07 1.0 1.0 0.67 
 22 0.14 1.0 1.0 0.68 
 23 0.21 1.0 1.0 0.70 
 24 0.24 0.98 0.87 0.71 
 25 0.38 0.85 0.58 0.72 
 26a 0.52 0.77 0.55 0.75 
 27 0.55 0.68 0.48 0.74 
 28 0.83 0.38 0.42 0.81 
 29 0.93 0.11 0.36 0.74 
MoCA 17 0.07 1.0 1.0 0.67 
 18 0.21 1.0 1.0 0.70 
 19 0.24 0.98 0.87 0.71 
 20 0.28 0.94 0.72 0.71 
 21 0.41 0.89 0.67 0.74 
 22 0.41 0.81 0.54 0.72 
 23a 0.72 0.75 0.61 0.83 
 24 0.76 0.68 0.56 0.84 
 25 0.86 0.57 0.52 0.88 
 26 0.93 0.47 0.49 0.93 
 27 0.93 0.44 0.47 0.92 
 28 0.97 0.09 0.36 0.85 
 29 0.97 0.04 0.35 0.71 
 Cut-off ≤ Sensitivity Specificity PPV NPV 
MMSE 20 0.03 1.0 1.0 0.66 
 21 0.07 1.0 1.0 0.67 
 22 0.14 1.0 1.0 0.68 
 23 0.21 1.0 1.0 0.70 
 24 0.24 0.98 0.87 0.71 
 25 0.38 0.85 0.58 0.72 
 26a 0.52 0.77 0.55 0.75 
 27 0.55 0.68 0.48 0.74 
 28 0.83 0.38 0.42 0.81 
 29 0.93 0.11 0.36 0.74 
MoCA 17 0.07 1.0 1.0 0.67 
 18 0.21 1.0 1.0 0.70 
 19 0.24 0.98 0.87 0.71 
 20 0.28 0.94 0.72 0.71 
 21 0.41 0.89 0.67 0.74 
 22 0.41 0.81 0.54 0.72 
 23a 0.72 0.75 0.61 0.83 
 24 0.76 0.68 0.56 0.84 
 25 0.86 0.57 0.52 0.88 
 26 0.93 0.47 0.49 0.93 
 27 0.93 0.44 0.47 0.92 
 28 0.97 0.09 0.36 0.85 
 29 0.97 0.04 0.35 0.71 

aOptimal cut-off score determined by the Youden Index.

Fig. 1.

ROC curve.

Fig. 1.

ROC curve.

Discussion

The primary aim of the present study was to examine the comparative utility of using the MoCA versus the MMSE to detect subtle cognitive impairment among middle-aged U.S. military veterans on a comprehensive battery of neuropsychological tests. The study was undertaken to both confirm early findings that the MoCA is more sensitive than the MMSE to subtle cognitive impairment and, also, to explore the possibility that the increased sensitivity of the MoCA is associated with an undesirable decrease in specificity. The findings of the present study were in line with the a priori hypotheses made.

As predicted, below cut-off performance was seen with far greater frequency on the MoCA than on the MMSE. Specifically, the MoCA score was impaired in 59% (n= 48/82) of the sample, while the MMSE was only impaired in 33% (n= 27/82). Also, as predicted, using area under the curve analyses and hierarchical regression analyses, the MoCA was shown to be a better predictor of subtle cognitive impairment on neuropsychological testing than the MMSE. Analysis of the data using published cutoffs for the MMSE and MoCA (≤25 and ≤26, respectively) did bear out the concerns voiced by other researchers. Specifically, while the MoCA was more sensitive to cognitive impairment than the MMSE (0.86 vs. 0.52), it did indeed show poorer specificity (0.57 vs. 0.77). However, it should be noted that the specificity of the MoCA was greatly improved by adjusting the cutoff to ≤23, as suggested by the Youdon Index. Using that cutoff, the MoCA was nearly as specific as the MMSE at its optimal cutoff (i.e., 0.75 vs. 0.77), while remaining considerably more sensitive (i.e., .72 vs. 0.52).

Thus, given the small loss in specificity (0.02) with a substantial increase in sensitivity (0.20), results of this study support use of the MoCA (cutoff of <23), as opposed to the MMSE, in identifying subtle cognitive impairment among a mixed group of middle-aged U.S. veteran outpatients referred for neuropsychological testing. In the present sample, in terms of PPV, using a MoCA cut-off of ≤23, a clinician would have a 61% probability of being correct in suspecting a patient of experiencing subtle cognitive impairment given a below cut-off performance on the MoCA. In terms of NPV, the same clinician would have an 83% probability of being correct in not suspecting cognitive impairment in a patient with above cut-off performance on the MoCA. While perhaps not ideal, these figures exceed estimates for the MMSE, which were 55% PPV and 75% NPV.

The use of a well-validated screening test to evaluate mental status can be a valuable clinic strategy, as it maximizes resources by allowing clinicians to quickly sort-out which patients are in need of more comprehensive, but time-consuming, neuropsychological testing. In this sense, sensitivity may be of greater importance than specificity because the goal is to identify as many patients as possible who are likely to demonstrate impairment on more extensive neuropsychological testing and who may, consequently, require further diagnostic procedures. The latter being the case, clinicians may do well to tolerate the loss of specificity using higher cutoffs of the MoCA to reap the benefits of a substantial gain in sensitivity when compared with the MMSE.

Future research may address the applicability of these findings to patient populations at various risk for cognitive impairment. In the present study, participants were either suspected of demonstrating cognitive impairment and/or were in a diagnostic group that is estimated to have a high base rate of cognitive impairment. Thus, the results of this study may not be representative of those that may be found in the general population. An attempt was made to compensate for the bias in our sample by using a high base rate of cognitive impairment when calculating the PPVs and NPVs of the MoCA and MMSE. However, statistically speaking, it is always true that PPVs will increase and NPVs will decrease with an increase in the base rate. As such, it may be that, in the general population, where the base rate of cognitive impairment is likely much lower, the MoCA and the MMSE would demonstrate lower positive predictive power and higher negative predictive power.

Conflict of Interest

None declared.

References

Butcher
J. N.
Graham
J. R.
Ben-Porath
Y. S.
Tellegen
A.
Dahlstrom
W. G.
Kaemmer
B.
MMPI-2: Manual for administration and scoring (Rev. ed.)
 , 
2001
Minneapolis
University of Minnesota Press
Crum
R. M.
Anthony
J. C.
Bassett
S. S.
Folstein
M. F.
Population-based norms for the Mini-Mental State Examination by age and educational level
Journal of the American Medical Association
 , 
1993
, vol. 
269
 (pg. 
2386
-
2391
)
Devilly
G. J.
Effect Size Generator for Windows: Version 2.3.
 , 
2004
Australia
Centre for Neuropsychology, Swinburne University
Folstein
M. F.
Folstein
S. E.
McHugh
P. R.
Mini-mental state: A practical method for grading the cognitive state of patients for the clinician
Journal of Psychiatric Research
 , 
1975
, vol. 
12
 (pg. 
189
-
198
)
Gladsjo
J. A.
Miller
S. W.
Heaton
R. K.
Norms for letter and category fluency: Demographic corrections for age, education, and ethnicity
 , 
1999
Lutz, FL
Psychological Assessment Resources
Glaros
A. G.
Kline
R. B.
Understanding the accuracy of tests with cutting scores: The sensitivity, specificity, and predictive value model
Journal of Clinical Psychology
 , 
1988
, vol. 
44
 (pg. 
1013
-
1023
)
Godefroy
O.
Fickl
A.
Roussel
M.
Auribault
C.
Bugnicourt
J. M.
Lamy
C.
, et al.  . 
Is the Montreal Cognitive Assessment superior to the Mini-Mental State examination to detect poststroke cognitive impairment?
Stroke
 , 
2011
, vol. 
42
 (pg. 
1712
-
1716
)
Gordon
M.
McClure
F. D.
Aylward
G. P.
The Gordon Diagnostic System: Instruction manual and interpretive guide
 , 
1996
DeWitt, NY
Gordon Systems
Kaplan
E.
Goodglass
H.
Weintraub
S.
Boston Naming Test
 , 
1983
Philadelphia
Lea & Febiger
Kongs
S. K.
Thompson
L. L.
Iverson
G. L.
Heaton
R. K.
Wisconsin Card Sorting Test – 64 Card Version
 , 
2000
Lutz, FL
Psychological Assessment Resources
Lipsey
M. W.
Design Sensitivity
 , 
1990
Newberry Park, CA
Sage
McCaffrey
R. J.
Palav
A.
O'Bryant
S. E.
Labarge
A. S.
Practitioner's guide to symptom base rates in clinical neuropsychology
 , 
2003
New York
Plenum
McLennan
S. N.
Mathias
J. L.
Brennan
L. C.
Stewart
S.
Validity of the Montreal Cognitive Assessment (MoCA) as a screening test for mild cognitive impairment (MCI) in a cardiovascular population
Journal of Geriatric Psychiatry and Neurology
 , 
2011
, vol. 
24
 
1
(pg. 
33
-
38
)
Nasreddine
Z. S.
Phillips
N. A.
Bedirian
V.
Charbonneau
S.
Whitehead
V.
Collin
I.
, et al.  . 
The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment
Journal of the American Geriatrics Society
 , 
2005
, vol. 
53
 (pg. 
695
-
699
)
Nazem
S.
Siderowf
A. D.
Duda
J. E.
Have
T. T.
Colcher
A.
Horn
S. S.
, et al.  . 
Montreal Cognitive Assessment performance in patients with Parkinson's disease with “normal” global cognition according to mini-mental state examination score
American Geriatric Society
 , 
2009
, vol. 
57
 
2
(pg. 
304
-
308
)
O'Bryant
S. E.
Lucas
J. A.
Estimating the predictive value of the Test of Memory Malingering: An illustrative example for clinicians
The Clinical Neuropsychologist
 , 
2006
, vol. 
20
 (pg. 
533
-
540
)
Pendlebury
S. T.
Cuthbertson
F. C.
Welch
S. J.
Mehta
Z.
Rothwell
P. M.
Underestimate of cognitive impairment by Mini-Mental State Examination versus the Montreal Cognitive Assessment in patients with transient ischemic attack and stroke
Stroke
 , 
2010
, vol. 
41
 (pg. 
1290
-
1293
)
Ratchford
T. L.
Ochoa
M.
Finney
G.
Normative data for the Montreal Cognitive Assessment (MoCA) in young adults. P05.128. Presented at the American Academy of Neurology Meeting
Neurology
 , 
2008
, vol. 
70
 
Suppl. 1
pg. 
A283
 
Saykin
A. J.
Gur
R. C.
Gur
R. E.
Shtasel
D. L.
Flannery
K. A.
Mozley
L. H.
, et al.  . 
Normative neuropsychological test performance: Effects of age, education, gender, and ethnicity
Applied Neuropsychology
 , 
1995
, vol. 
2
 
2
(pg. 
79
-
88
)
2010
Chicago
SPSS
 
SPSS for Windows, Rel 19.0.0
Tombaugh
T. N.
Test of memory malingering (TOMM)
 , 
1996
North Tonowanda, NY
Multi-Health Systems
Wechsler
D.
The Wechsler Memory Scale – Fourth Edition
 , 
2009
San Antonio, TX
Pearson
Wechsler
D. A.
Wechsler Adult Intelligence Scale – IV
 , 
2008
San Antonio, TX
Psychological Corporation
Youdon
W.J.
Index for rating diagnostic tests
Cancer
 , 
1950
, vol. 
3
 (pg. 
32
-
35
)