Abstract
Objectives.
To generate robust, demographically-adjusted regression-based norms for the Montreal Cognitive Assessment (MoCA) using a large sample of diverse older US adults.
Methods.
Baseline MoCA scores were examined for participants in the Systolic Blood Pressure Intervention Trial (SPRINT). A robust, cognitively-normal sample was drawn from individuals not subsequently adjudicated with cognitive impairment through 4 years of follow-up. Multivariable Beta-Binomial regression was used to model the association of demographic variables with MoCA performance and to create demographically-stratified normative tables.
Results.
Participants’ (N=5,338) mean age was 66.9 ± 8.8 years, with 35.7% female, 63.1% White, 27.4% Black, 9.5% Hispanic, and 44.5% with a college or graduate education. A large proportion scored below published MoCA cutoffs: 61.4% scored below 26 and 29.2% scored below 23. A disproportionate number falling below these cutoffs were Black, Hispanic, did not graduate from college, or were ≥75 years of age. Multivariable modeling identified education, race/ethnicity, age, and sex as significant predictors of MoCA scores (p<.001), with the best fitting model explaining 24.4% of the variance. Model-based predictions of median MoCA scores were generally 1 to 2 points lower for Black and Hispanic participants across combinations of age, sex, and education. Demographically-stratified norm-tables based on regression modeling are provided to facilitate clinical use, along with our raw data.
Conclusion.
By using regression-based strategies that more fully account for demographic variables, we provide robust, demographically-adjusted metrics to improve cognitive screening with the MoCA in diverse older adults.
Keywords: MoCA, norms, cognition, demographics
Introduction
Cognitive screening instruments are quick, easy to administer, and cost-efficient. Key to their utility, however, are relevant normative data to guide the interpretation of scores. Without them, screeners can misrepresent an individual’s performance and lead to incorrect diagnoses with unnecessary personal and societal detriments (Obermeyer et al., 2019). Unlike tests with population-based distributions that are unimodal and approximate Gaussian distributions (Mitrushina, Boone, & D’Elia, 1999), brief, cognitive screeners like the Montreal Cognitive Assessment (MoCA) (Nasreddine et al., 2005) often rely on norms that have unidirectional or highly skewed distributions (Busch, Chelune, & Suchy, 2006). Interpretation of scores relies on the base-rate of a given test score in the normative population compared to a population with cognitive impairment (Chelune, 2017; Ivnik et al., 2001), with cutoff scores chosen to optimally separate impaired from normal test performances. As such, the choice of appropriate cutoff scores is highly dependent on the composition of the normative group and its demographic characteristics.
This is particularly relevant, as screeners such as the MoCA are favored by a variety of medical and mental health professionals in clinical practice to help distinguish normal cognitive aging from mild cognitive impairment (MCI) and dementia in older adults (Gluhm et al., 2013; Nasreddine et al., 2005). The MoCA has been adopted as part of the test battery used by Alzheimer Disease Research Centers across the country (Besser et al., 2018). Despite its ubiquity, the original normative data for the instrument were developed on a small sample of only 90 individuals from Canada with scores >25 found to be diagnostically indicative of normal cognitive functioning (with 87% specificity). The racial and ethnic make-up of this initial cohort was predominantly White. However, because of the small sample size and lack of diversity, additional studies have attempted to develop more comprehensive normative data for the measure (Luis, Keegan, & Mullan, 2009; Rossetti et al., 2017).
While a growing number of studies have sought to establish normative data for important subpopulations like African Americans and Hispanics (Rossetti, Lacritz, Cullum, & Weiner, 2011), many have included small overall sample sizes (Goldstein et al., 2014; Sink et al., 2015), small proportions of African American and/or Hispanic subjects (Milani, Marsiske, & Striley, 2019), or have focused on individuals under the age of 75 (Rossetti et al., 2017). Nonetheless, a fairly consistent finding of these studies is that using the suggested threshold for impairment of <26 may “overpathologize neurologically intact individuals” (Waldron-Perrine & Axelrod, 2012), especially in racial/ethnic minorities. For example, data from the National Alzheimer’s Coordinating Center Uniform Data Set suggested optimal cutoffs for detecting mild cognitive impairment of less than 25, 24, and 23 respectively among non-Hispanic Whites, Hispanics, and non-Hispanic Blacks (Milani et al., 2018). Similarly, in the Dallas Heart Study, a large population-based diverse cohort, investigators found that approximately 66% of their sample scored <26 (Rossetti et al., 2011), even after accounting for education and other demographic factors. Other studies conducted with African American populations have also shown that lower cutoff scores yield better diagnostic accuracy (Goldstein et al., 2014; Rossetti et al., 2017; Sink et al., 2015). A recent meta-analysis of MoCA studies that included both White and African American participants suggested that a cut-off score of 23 both limited the likelihood of false-positive errors and provided optimal diagnostic accuracy (Carson, Leach, & Murphy, 2018).
While evidence is emerging that more demographically sensitive normative information for the MoCA is needed, there is also a growing recognition that existing normative data may be limited. Studies have typically been conducted in cross-sectional and community-based samples where there has been limited concurrent ascertainment of prevalent cognitive impairment or individuals with pre-clinical impairment (De Santi et al., 2008). For example, several longitudinal normative studies have documented that between 4–16% of individuals believed to have had normal cognitive functioning at baseline were cognitively impaired one year later (De Santi et al., 2008; Holtzer et al., 2008; Pedraza et al., 2010). The inclusion of cognitively impaired individuals tends to lower mean test scores and increase variability (Ritchie, Frerichs, & Tuokko, 2007), potentially leading to diagnostic thresholds that are too low and reducing predictive performance (Pedraza et al., 2010). There has been one normative study for the MoCA from Sweden that carefully curated a sample of healthy, cognitively normal older adults and found that age, sex and level of education were significant predictors of MoCA performance such that MoCA cut-offs for cognitive impairment ranged from <25 to <21 for the lowest educated and <26 to <24 for the highest educated (Borland et al., 2017); however, such a population sample obviously does not generalize to the diversity inherent to populations in the United States.
In the present study, we utilize baseline MoCA data from the Systolic Blood Pressure Intervention Trial (SPRINT MIND Investigators for the SPRINT Research Group et al., 2019), a randomized clinical trial of strategies for blood pressure control that recruited a large, diverse sample of older adults at increased risk for cardiovascular disease (N=9,361) between 2010 and 2013. We used repeated follow-up cognitive testing out to 4 years from baseline to select a robust normative sample of participants highly likely to be cognitively healthy and free of incident impairment. In addition, we employed regression-based strategies to jointly account for the influence of age, sex, education, and race/ethnicity on MoCA performance, and we present normative tables for easy utilization.
Methods
SPRINT Cohort Description
The design and methods for the SPRINT trial, including the ascertainment of cognitive status have been described in detail elsewhere (Ambrosius et al., 2014; SPRINT MIND Investigators for the SPRINT Research Group et al., 2019; SPRINT Research Group et al., 2015). Briefly, participants (N=9,361) were recruited from 102 sites in the US and Puerto Rico, were 50 years of age or older, and had a screening systolic blood pressure (SBP) between 130 and 180 mm Hg. Participants qualified for the trial if they had either previously diagnosed clinical or subclinical cardiovascular disease, chronic kidney disease (estimated glomerular filtration rate (eGFR) range 20 to <60 ml/min/1.73 m2) or a Framingham cardiovascular disease (CVD) risk score of 15% or greater, or if they were 75 years of age or older. Individuals residing in a nursing home, persons with a diagnosis of dementia (based on medical record review), or treated with medications primarily used for dementia therapy were excluded, as were persons with prevalent diabetes mellitus, history of stroke, symptomatic heart failure (or ejection fraction <35%), stage IV chronic kidney disease or proteinuria >1 gram / day.
Cognitive Assessments and Adjudication of Impairment
All participants were administered a cognitive screening battery at baseline and during follow-up (planned assessments every two years in follow-up) that included the MoCA, and tests of verbal learning and memory (Logical Memory I and II of Story C) and processing speed (Digit Symbol Coding) (Wechsler, 2008, 2009). As part of the ascertainment process for Mild Cognitive Impairment (MCI) and probable dementia during follow-up, participants scoring below pre-set race and education-specific thresholds on the MoCA [White participants scoring <19 (<12 year of education) or <21 (≥12 years of education) and non-white participants scoring <17 (<12 years of education) or <19 (≥12 years of education)] or participants with a drop of ≥5 points on the MoCA from a previous assessment, were administered the Functional Activities Questionnaire, a 10-item, proxy-reported measure of functional abilities (FAQ; range: 0 to 30; Pfeffer et al., 1982). Participants scoring >0 on the FAQ or ≤1 on the 5-point delayed recall subtest of the MoCA were also administered an extended cognitive battery that measured attention/concentration (Digit Span), verbal and nonverbal memory (Hopkins Verbal Learning Test-Revised, Rey-Osterreith Complex Figure-modified), language (Category Fluency-Animals), and executive function (Trail Making Test) (for more details see SPRINT MIND Investigators for the SPRINT Research Group et al., 2019). Participants not able to be tested in-person were administered a validated telephone battery (Rapp et al., 2012) with the FAQ administered to their proxy. Deceased participants’ pre-identified proxies were administered the Dementia Questionnaire (Kawas, et al., 1994). In addition, all participants were administered standardized measures of depressive symptom severity, health status and quality of life and reported current medications, medical problems, and health habits (smoking, alcohol use and physical activity). Hospitalizations and serious adverse events were also collected (SPRINT MIND Investigators for the SPRINT Research Group et al., 2019). Suspected cases of cognitive impairment were adjudicated by a panel of experts in late-life cognitive impairment composed of geriatricians, neurologists, neuropsychologists, and geropsychologists. Each case was assigned to two adjudicators, masked to treatment assignment, who independently reviewed all available data and used standardized criteria (McKhann et al, 2011; Albert et al, 2011) to classify the case as no cognitive impairment, MCI or probable dementia. Cases adjudicated as having MCI or probable dementia were then excluded. A ‘cannot classify’ assignment was made when cases were missing key information or otherwise unclassifiable. Agreements between adjudicators were recorded as final and disagreements were discussed by the full panel on regularly scheduled conference calls with the final classification decision determined by majority vote of panel members.
Creation of Robust, Cognitively-Normal Cohort at Baseline
In the trial, cognitive status was not adjudicated at baseline, and therefore an unknown number of participants might have had prevalent impairment at the time of randomization. Figure 1 summarizes a series of exclusion criteria we applied in order to create a cohort of participants highly likely to have been cognitively normal at baseline. First, 263 (2.8%) participants were excluded because they did not self-report their race/ethnicity or education level, had missing baseline data for the MoCA, or did not complete the 9-item Patient Health Questionnaire (PHQ-9) (Kroenke, Spitzer, & Williams, 2001), a measure of depressive symptom severity. An additional 717 (7.7%) participants were excluded because they had evidence of moderate to severe depressive symptoms (PHQ-9 score ≥10). Third, we restricted our cohort to participants who subsequently completed follow-up cognitive assessments at both years 2 and 4 (n=6,595), and also required that they not be adjudicated with MCI or probable dementia at either of these follow-up time points. This left a final robust normative cohort of 5,338 participants (57.0%).
Figure 1.

Flowchart of exclusions for building cognitively normal subgroup of SPRINT participants at baseline
MCI denotes Mild Cognitive Impairment; MoCA, Montreal Cognitive Assessment; PHQ-9, 9 item Patient Health Questionnaire.
Derivation of Regression-based Normative Values
Regression-based norms (Crawford & Howell, 1998; Crawford, Garthwaite, Denham, & Chelune, 2012) for the MoCA were derived using Beta-Binomial regression models (Arostegui, Núñez-Antón, & Quintana, 2012). Beta-Binomial models were chosen because they acknowledge that MoCA scores have bounded, discrete support (integer values from 0 to 30) and can accommodate distributional features such as skewness and ceiling effects that are common to cognitive screening test scores. Both the mean and variance parameters for the Beta-Binomial model were modeled as a function of age, education (<High school education, High school graduate, Associate’s degree/Trade school, or College graduate or higher), and race/ethnicity (White, Black, Hispanic). Model selection was performed on the basis of Akaike’s Information Criteria (AIC), likelihood ratio tests, and the Cox-Snell version of Nagelkerke’s Pseudo-R2, considering up to two-way interactions for the above predictors. We also investigated whether MoCA performance was associated with markers of cardiovascular risk, including systolic and diastolic blood pressure, eGFR, and HDL and total cholesterol. Age and the cardiovascular risk factors were modeled using P-splines (Eilers, Marx, & Durbán, 2015) to allow for potential non-linear effects. All modeling was performed using the gamlss package for the R Statistical Computing Environment (Stasinopoulos & Rigby, 2007).
Results
Demographic characteristics
Of the 9,361 participants initially randomized, 263 were excluded because of missing data, 717 were excluded due to having PHQ-9 scores that exceeded the study cut-off, and an additional 3,043 were excluded because they did not complete subsequent cognitive testing beyond baseline or were subsequently adjudicated with MCI or probable dementia. Demographic characteristics for our robust normative cohort (N=5,338) are presented by race/ethnicity in Table 1. Overall, participants had a mean age of 66.9 years (SD, 8.8 years), with 35.7% being female, 63.1% White, 27.4% Black, 9.5% Hispanic, and 44.5% had a college degree or higher. Across the race/ethnicity groups, Black and Hispanic participants tended to be younger than Whites (p < .001), were more likely to be female (p < .001), had lower educational attainment (p < .001), and were less likely to have a history of cardiovascular disease (p < .001). Black and Hispanic participants also had lower scores on average for the three cognitive tests administered at baseline, including the MoCA. For example, median scores on the MoCA were 25, 23, and 23 for White, Black, and Hispanic participants respectively (p<0.001).
Table 1.
Characteristics of Systolic Blood Pressure Intervention Trial participants included in analyses
| Characteristics | Overall N = 5338 |
White N = 3366 |
Black N = 1463 |
Hispanic N = 509 |
P Value |
|---|---|---|---|---|---|
| Age, years, mean (SD) | 66.9 (8.8) | 69.1 (8.5) | 63.1 (8.2) | 63.7 (8.2) | <0.001 |
|
| |||||
| Female sex, No. (%) | 1904 (35.7) | 1019 (30.3) | 664 (45.4) | 221 (43.4) | <0.001 |
|
| |||||
| Education, No. (%) | <0.001 | ||||
| <High school education | 290 (5.4) | 91 (2.7) | 139 (9.5) | 60 (11.8) | |
| High school graduate | 756 (14.2) | 404 (12.0) | 260 (17.8) | 92 (18.1) | |
| Associate’s degree | 1919 (35.9) | 1116 (33.2) | 668 (45.7) | 135 (26.5) | |
| College graduate | 1315 (24.6) | 933 (27.7) | 252 (17.2) | 130 (25.5) | |
| Graduate degree | 1058 (19.8) | 822 (24.4) | 144 (9.8) | 92 (18.1) | |
|
| |||||
| History of CVD, No. (%) | 974 (18.2) | 678 (20.1) | 212 (14.5) | 84 (16.5) | <0.001 |
|
| |||||
| Systolic blood pressure, mm Hg, mean (SD) | 139.3 (15.3) | 138.9 (15.1) | 139.7 (16.1) | 140.5 (13.9) | 0.04 |
|
| |||||
| Diastolic blood pressure, mm Hg, mean (SD) | 78.7 (11.7) | 77.3 (11.3) | 81.9 (12.1) | 78.7 (11.4) | <0.001 |
|
| |||||
| eGFR, ml/min/1.73 m2, mean (SD) | 73.7 (19.1) | 70.9 (17.4) | 77.7 (21.7) | 81.1 (18.3) | <0.001 |
|
| |||||
| Administered cognitive testing in Spanish, No. (%) | 354 (6.6) | 0 (0.0) | 0 (0.0) | 354 (69.5) | <0.001 |
|
| |||||
| Montreal Cognitive Assessment, median [IQR] | 24 [22 to 27] | 25 [23 to 27] | 23 [21 to 25] | 23 [20 to 26] | <0.001 |
|
| |||||
| Logical Memory II, median [IQR] | 9 [7 to 11] | 10 [8 to 12] | 8 [6 to 10] | 9 [6 to 10] | <0.001 |
|
| |||||
| Digit Symbol Coding Test, median [IQR] | 54 [45 to 63] | 57 [48 to 64] | 50 [41 to 59] | 51 [40 to 63] | <0.001 |
CVD denotes cardiovascular disease, eGFR estimated glomerular filtration rate, SD standard deviation, and IQR interquartile range.
Demographic Factors and Traditional Screening Thresholds for Impairment
Using the MoCA cutoff of < 23 suggested by a recent meta-analysis (Carson et al., 2018), 29.2% of participants in our cohort would have screened positive for cognitive impairment, whereas 61.4% of participants scored below the cutoff of < 26 recommended for detecting impairment by Nasreddine and colleagues (2005). Table A1 provides the frequency counts of participants by race/ethnicity, age, and education for each of these two threshold scores. Table 2 summarizes the univariate association based on logistic regression of age, education, race/ethnicity, sex, and cardiovascular risk factors as predictors of screening positive at each of these two thresholds, with counts and percentages representing the proportion within each subgroup scoring below thresholds of < 26 or < 23. Having less than a high school education was strongly associated with scoring below either threshold (odds ratio for MoCA score <23 compared to having a graduate degree = 17.87, 95% CI: 13.04 to 24.50). Similarly, participants 75 years or older as well as Black and Hispanic participants were more likely to have MoCA scores below 23 or 26. In contrast, sex was not associated with scoring below either threshold, and there were only weak associations for baseline blood pressure and eGFR.
Table 2.
Univariate associations between demographic variables and scoring below common thresholds on the Montreal Cognitive Assessment (MoCA)
| MoCA Score <26 | MoCA Score <23 | |||||
|---|---|---|---|---|---|---|
| Characteristic | N < 26 (%)c | Odds Ratio (95% CI) | P value | N < 23 (%)c | Odds Ratio (95% CI) | P value |
| Age (<75 years, N=4112) | 2430 (59.1) | 1147 (27.9) | ||||
| ≥75 years (N=1226) | 859 (70.1) | 1.66 (1.43, 1.92) | <0.001 | 411 (33.5) | 1.33 (1.15, 1.53) | <0.001 |
|
| ||||||
| Sex (Male, N=3434) | 2118 (61.7) | 999 (29.1) | ||||
| Female (N=1904) | 1171 (61.5) | 0.99 (0.88, 1.11) | 0.90 | 559 (29.4) | 1.01 (0.90, 1.15) | 0.84 |
|
| ||||||
| Education (Graduate degree, N=1058) | 474 (44.8) | 144 (13.6) | ||||
| <High school education (N=290) | 266 (91.7) | 13.66 (8.84, 21.09) | <0.001 | 214 (73.8) | 17.87 (13.04, 24.50) | <0.001 |
| High school graduate (N=756) | 597 (79.0) | 4.63 (3.74, 5.73) | <0.001 | 353 (46.7) | 5.56 (4.43, 6.97) | <0.001 |
| Associate’s degree (N=1919) | 1279 (66.6) | 2.46 (2.11, 2.87) | <0.001 | 613 (31.9) | 2.98 (2.44, 3.64) | <0.001 |
| College graduate (N=1315) | 673 (51.2) | 1.29 (1.10, 1.52) | 0.002 | 234 (17.8) | 1.37 (1.10, 1.72) | 0.006 |
|
| ||||||
| Race/Ethnicity (White, N=3366) | 1800 (53.5) | 695 (20.6) | ||||
| Black (N=1463) | 1108 (75.7) | 2.72 (2.37, 3.12) | <0.001 | 635 (43.4) | 2.95 (2.58, 3.37) | <0.001 |
| Hispanic (N=509) | 381 (74.9) | 2.59 (2.10, 3.2) | <0.001 | 228 (44.8) | 3.12 (2.57, 3.78) | <0.001 |
|
| ||||||
| Systolic BPa | - | 1.07 (1.01, 1.13) | 0.02 | - | 1.04 (0.98, 1.10) | 0.17 |
|
| ||||||
| Diastolic BPa | - | 0.93 (0.88, 0.99) | 0.02 | - | 0.93 (0.88, 0.99) | 0.02 |
|
| ||||||
| eGFRb | - | 1.07 (1.01, 1.13) | 0.02 | - | 1.00 (0.94, 1.06) | 0.92 |
|
| ||||||
| HDL cholesterola | - | 1.00 (0.95, 1.06) | 0.99 | - | 0.99 (0.93, 1.05) | 0.65 |
|
| ||||||
| Total cholesterola | - | 0.96 (0.91, 1.02) | 0.19 | - | 1.01 (0.95, 1.07) | 0.87 |
BP denotes blood pressure, CI confidence interval, eGFR estimated glomerular filtration rate, HDL high density lipoprotein, and SD standard deviation. Values in parentheses denote reference category.
Per 1 SD increase.
Per 1 SD decrease.
% denotes proportion within each subgroup scoring below MoCA threshold.
Multivariable Consideration of Demographic Influences on MoCA Performance.
Beta-Binomial regression modeling was used to build progressively more complex models as a function of age, education, race/ethnicity, language of administration, and sex. Table 3 displays pseudo-R2 values and p-values from likelihood ratio tests for each of the successive steps. In general, accounting for administration of the MoCA in Spanish to some Hispanic participants did not statistically improve model fit (p = .35), while age, education, race/ethnicity, and sex were all significant predictors. In subsequent analyses we therefore modeled Hispanic participants together regardless of language of administration. Education was clearly the strongest predictor, adding ~15% of explained variance to age, followed by race/ethnicity, with age and sex providing smaller improvements to explained variability by the model. None of the cardiovascular risk factors we examined significantly improved model fit. Our final model, additionally included race/ethnicity x sex and age x education interactions, and accounted for 24.4% of explained variance in MoCA performance.
Table 3.
Model Selection Criteria for Beta-Binomial Modeling of Montreal Cognitive Assessment Scores
| N=5338 | ||||
|---|---|---|---|---|
| Model No. | Model Predictors | Pseudo-R2 | Comparison | P value |
| 1 | Age | 0.014 | - | - |
| 2 | 1 + Education | 0.166 | Model 1 | <0.001 |
| 3 | 2 + Race/Ethnicity | 0.230 | Model 2 | <0.001 |
| 4 | 3 + Language of administration | 0.230 | Model 3 | 0.35 |
| 5 | 3 + Sex | 0.237 | Model 3 | <0.001 |
| 6 | 5 + Education x Race/Ethnicity | 0.240 | Model 5 | 0.31 |
| 7 | 5 + Education x Sex | 0.238 | Model 5 | 0.89 |
| 8 | 5 + Race/Ethnicity x Sex | 0.241 | Model 5 | <0.001 |
|
| ||||
| 9 | 8 + Age x Education | 0.244 | Model 8 | 0.002 |
|
| ||||
| 10 | 9 + Age x Race/Ethnicity | 0.245 | Model 9 | 0.47 |
| 11 | 9 + Age x Sex | 0.245 | Model 9 | 0.09 |
|
| ||||
| N=5327* | ||||
|
| ||||
| 9 | 8 + Age x Education | 0.245 | - | - |
|
| ||||
| 12 | 9 + Systolic Blood Pressure | 0.245 | Model 9 | 0.86 |
| 13 | 9 + Diastolic Blood Pressure | 0.245 | Model 9 | 0.21 |
| 14 | 9 + estimated Glomerular Filtration Rate | 0.245 | Model 9 | 0.69 |
| 15 | 9 + HDL cholesterol | 0.246 | Model 9 | 0.15 |
| 16 | 9 + Total cholesterol | 0.245 | Model 9 | 0.93 |
P value based on a likelihood ratio test comparing each model to the comparison model. Highlighted row denotes final selected model.
Removes 11 participants with missing estimated Glomerular Filtration Rate, HDL cholesterol, or total cholesterol
Demographically-Stratified Norms Tables.
Sample sizes by age, sex, race/ethnicity, and education are shown in Table A2. Figure 2 depicts estimated median and 5th percentile MoCA scores from the beta-binomial model as a function of age, stratified by sex, race/ethnicity, and education. More comprehensive estimates at every 5th year of age from age 55 to age 85, stratified according to sex, race/ethnicity, and education are presented in the Appendix as Tables A3–A8. In general, the model estimates indicate decreasing performance on the MoCA with increasing age and 1-to-2-point lower median scores for Black and Hispanic participants as compared to White participants, across all combinations of age, sex, and education. R Code for calculating regression-based percentile scores for a given MoCA score and specific combinations of demographic characteristics is available upon request from the last author, while the raw data underlying our analyses is available at Dryad (https://doi.org/10.5061/dryad.f7m0cfxwm).
Figure 2.

Predicted Montreal Cognitive Assessment Median and 5th percentiles from Beta-Binomial models as a Function of Age, Sex, Race/ethnicity, and Education
Numbers in colored blocks denote sample size by race/ethnicity within each sex and education category. Additional sample size breakdowns by age are shown in Table A2.
Discussion
The current study provides robust, regression-based normative data for the MoCA jointly accounting for the influence of age, sex, education and race/ethnicity derived from participants in a large, geographically diverse, clinical trial of hypertension control. Overall, regression modeling indicated that performance on the MoCA was negatively correlated with age and positively correlated with education, such that participants had increasingly better performance on the MoCA with higher levels of education. Similar to previous research (Milani et al., 2018; Rossetti et al., 2011), we observed a significant association between race/ethnicity and MoCA performance. Across all combinations of age, sex, and education, Black and Hispanic participants generally scored lower on the MoCA compared to White participants. In addition, White females performed marginally better than White males.
The median scores in our sample are clearly well below commonly cited thresholds for possible or probable cognitive impairment using the MoCA score (Carson et al., 2018; Nasreddine et al., 2005). This finding is consistent with a number of previous studies (Damian et al., 2011; Malek-Ahmadi et al., 2015; Rossetti et al., 2011) which have shown that traditional cut-points are prone to flagging a large number of older adults for suspected cognitive impairment based on the MoCA cutoffs employed. This highlights the point that while the use of static cut-points that do not acknowledge demographic differences may provide high sensitivity to MCI, it will inevitably produce poor specificity; particularly amongst older and less educated adults as well as members of racial/ethnic minority groups. The choice of appropriate cut-points for diagnostic screening tests ultimately depends on achieving a balance of sensitivity and specificity that meets the needs of the specific clinical or research setting (Chelune, 2017; Ivnik et al., 2001; Sackett, Straus, Richardson, & Haynes, 2000).
The normative data provided are “robust” in that detailed steps were taken to include participants who were free of incipient cognitive impairment. However, because comprehensive neuropsychological testing and formal adjudication were not performed at baseline, we cannot entirely rule out the possibility that some participants with impairment were included within our robust cohort. Nonetheless, given the requirement that participants completed cognitive testing out to four years of follow-up and were never subsequently adjudicated as having MCI or probable dementia, we believe the proportion of participants with prevalent MCI in our sample is likely to be exceedingly small (De Santi et al., 2008; Ritchie et al., 2007). Further, our sample is representative of an important and common population found in primary care and geriatric clinics – older men and women with documented histories of cardiovascular and/or kidney disease with mild to moderate hypertension. It is within these settings in which the MoCA is frequently administered and where these findings may have the greatest utility. Moreover, demographically-adjusted normative data may help to decrease potential systematic biases in health care decision making based on the MoCA, especially among racial/ethnic minority populations (Obermeyer, Powers, Vogeli, & Mullainathan, 2019).
We should note that our robust sample was essentially one of convenience and has potential limitations. Although our cohort of older adults was demographically and geographically diverse with a relatively large number of Black and Hispanic participants, the number of participants for certain combinations of age, sex, and lower levels of education was somewhat limited. Deriving normative scores through a regression-based strategy, as opposed to direct stratification, somewhat mitigates this concern by capitalizing on the benefits of continuous norming techniques (Voncken, Albers, & Timmerman, 2019). Second, participants in SPRINT were recruited specifically for having elevated cardiovascular risk, and may not generalize to the broader population of older adults, although more than 2/3 of older adults in this age group have hypertension. Third, individuals who volunteer for clinical trials may not be representative of all older adults. However, it was reassuring that baseline cardiovascular risk factors were generally not associated with MoCA performance after accounting for strong correlates such as age and education. Thus, our normative data should generalize to the broader population of community-dwelling older adults, especially given that hypertension is highly prevalent among older adults, affecting up to 75% of adults 65 years or older (Whelton et al., 2018).
In conclusion, this study offers clinicians and researchers a set of robust normative data for the widely used MoCA for a demographically diverse population of older adults. Our results underscore the significant variability in MoCA performance as a function of age, sex, educational attainment, and race/ethnicity, highlighting the need to apply demographically-adjusted norms as part of screening protocols for detecting cognitive impairment. It is important to note, however, that normative data are but one tool used to understand and classify cognitive performances; cognitive “impairment” can only be identified by interpreting test scores within the context of the patient’s history, presenting problem, and functional capacity (Guilmette et al., 2020). We provide our underlying data (and will provide code upon request) to apply our estimated regression models to future patients and cohorts, which should allow researchers and clinicians to more accurately interpret and classify performance on the MoCA.
Acknowledgements
The Systolic Blood Pressure Intervention Trial was funded by the National Institutes of Health (including the National Heart, Lung, and Blood Institute, the National Institute of Diabetes and Digestive and Kidney Diseases, the National Institute on Aging, and the National Institute of Neurological Disorders and Stroke) under contracts HHSN268200900040C, HHSN268200900046C, HHSN268200900047C, HHSN268200900048C, and HHSN268200900049C and interagency agreement A-HL-13-002-001. It was also supported in part with resources and use of facilities through the Department of Veterans Affairs. Azilsartan and chlorthalidone (combined with azilsartan) were provided by Takeda Pharmaceuticals International Inc.
Additional support was provided through the following National Center for Advancing Translational Sciences clinical and translational science awards: UL1TR000439 (awarded to Case Western Reserve University); UL1RR025755 (Ohio State University); UL1RR024134 and UL1TR000003 (University of Pennsylvania); UL1RR025771 (Boston University); UL1TR000093 (Stanford University); UL1RR025752, UL1TR000073, and UL1TR001064 (Tufts University); UL1TR000050 (University of Illinois); UL1TR000005 (University of Pittsburgh); 9U54TR000017-06 (University of Texas Southwestern Medical Center); UL1TR000105-05 (University of Utah); UL1 TR000445 (Vanderbilt University); UL1TR000075 (George Washington University); UL1 TR000002 (University of California, Davis); UL1 TR000064 (University of Florida); and UL1TR000433 (University of Michigan); and by National Institute of General Medical Sciences, Centers of Biomedical Research Excellence award NIGMS P30GM103337 (awarded to Tulane University). Additional support also provided by R01AG055606, the Wake Forest Claude Pepper Center (P30AG021332), and the Alzheimer’s Association. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH, the U.S. Department of Veterans Affairs, the United States Government, or the Alzheimer’s Association. This paper was not reviewed by the SPRINT Publications and Presentations Committee.
Appendix
TABLE A1.
Cohort counts by race/ethnicity, education, and age and frequency of participants meeting previously proposed thresholds for the Montreal Cognitive Assessment
| Age Group | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| <65 years | 65 to <75 years | 75 years or older | ||||||||
| Race/Ethnicity | Education | N | %<23 | %<26 | N | %<23 | %<26 | N | %<23 | %<26 |
| White | <High school education | 28 | 53.6% | 75.0% | 30 | 63.3% | 90.0% | 33 | 54.5% | 90.9% |
| High school graduate | 123 | 29.3% | 69.9% | 155 | 36.8% | 71.0% | 126 | 46.8% | 83.3% | |
| Associate’s degree | 416 | 18.0% | 47.8% | 377 | 27.9% | 61.8% | 323 | 33.7% | 73.4% | |
| College graduate | 322 | 7.8% | 37.0% | 346 | 13.9% | 46.0% | 265 | 19.6% | 57.4% | |
| Graduate degree | 217 | 5.5% | 29.0% | 342 | 7.9% | 35.1% | 263 | 14.4% | 52.9% | |
|
| ||||||||||
| Black | <High school education | 77 | 72.7% | 90.9% | 39 | 94.9% | 100.0% | 23 | 95.7% | 95.7% |
| High school graduate | 173 | 52.0% | 85.0% | 63 | 57.1% | 79.4% | 24 | 83.3% | 95.8% | |
| Associate’s degree | 448 | 34.8% | 71.9% | 164 | 47.6% | 80.5% | 56 | 57.1% | 91.1% | |
| College graduate | 160 | 22.5% | 58.1% | 65 | 33.8% | 69.2% | 27 | 40.7% | 85.2% | |
| Graduate degree | 70 | 22.9% | 60.0% | 46 | 26.1% | 54.3% | 28 | 39.3% | 85.7% | |
|
| ||||||||||
| Hispanic | <High school education | 25 | 76.0% | 92.0% | 22 | 77.3% | 100.0% | 13 | 84.6% | 92.3% |
| High school graduate | 51 | 62.7% | 84.3% | 30 | 53.3% | 73.3% | 11 | 63.6% | 100.0% | |
| Associate’s degree | 88 | 37.5% | 75.0% | 32 | 46.9% | 75.0% | 15 | 66.7% | 100.0% | |
| College graduate | 90 | 27.8% | 62.2% | 31 | 32.3% | 61.3% | 9 | 55.6% | 77.8% | |
| Graduate degree | 51 | 19.6% | 56.9% | 31 | 38.7% | 77.4% | 10 | 60.0% | 80.0% | |
Values denote sample size, % of participants with a MoCA score <23, and % of participants with MoCA score <26.
TABLE A2.
Cohort counts by race/ethnicity, education, and age
| Age (Years) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Race/Ethnicity | Sex | Education | <60 | 60 to <65 | 65 to <70 | 70 to <75 | 75 to <80 | 80 to <85 | ≥85 |
| White | Males | <High school education | 11 | 10 | 12 | 9 | 12 | 5 | 3 |
| High school graduate | 35 | 53 | 53 | 35 | 51 | 16 | 3 | ||
| Associate’s degree | 122 | 196 | 131 | 103 | 117 | 50 | 14 | ||
| College graduate | 117 | 147 | 160 | 88 | 105 | 47 | 17 | ||
| Graduate degree | 68 | 120 | 144 | 111 | 104 | 61 | 17 | ||
|
| |||||||||
| White | Females | <High school education | 3 | 4 | 6 | 3 | 9 | 3 | 1 |
| High school graduate | 13 | 22 | 36 | 31 | 40 | 7 | 9 | ||
| Associate’s degree | 41 | 57 | 64 | 79 | 75 | 47 | 20 | ||
| College graduate | 26 | 32 | 45 | 53 | 55 | 28 | 13 | ||
| Graduate degree | 8 | 21 | 41 | 46 | 51 | 22 | 8 | ||
|
| |||||||||
| Black | Males | <High school education | 31 | 14 | 6 | 14 | 6 | 5 | 2 |
| High school graduate | 74 | 40 | 22 | 10 | 10 | 2 | 0 | ||
| Associate’s degree | 150 | 100 | 45 | 35 | 14 | 5 | 4 | ||
| College graduate | 53 | 35 | 21 | 11 | 11 | 3 | 1 | ||
| Graduate degree | 18 | 16 | 15 | 12 | 11 | 3 | 0 | ||
|
| |||||||||
| Black | Females | <High school education | 14 | 18 | 9 | 10 | 6 | 2 | 2 |
| High school graduate | 39 | 20 | 21 | 10 | 6 | 3 | 3 | ||
| Associate’s degree | 113 | 85 | 50 | 34 | 27 | 5 | 1 | ||
| College graduate | 34 | 38 | 13 | 20 | 6 | 5 | 1 | ||
| Graduate degree | 16 | 20 | 12 | 7 | 10 | 4 | 0 | ||
|
| |||||||||
| Hispanic | Males | <High school education | 5 | 10 | 4 | 4 | 6 | 1 | 2 |
| High school graduate | 15 | 9 | 7 | 9 | 4 | 1 | 0 | ||
| Associate’s degree | 31 | 22 | 13 | 8 | 4 | 3 | 2 | ||
| College graduate | 32 | 19 | 7 | 7 | 3 | 2 | 1 | ||
| Graduate degree | 20 | 14 | 9 | 8 | 5 | 0 | 1 | ||
|
| |||||||||
| Hispanic | Females | <High school education | 6 | 4 | 5 | 9 | 2 | 1 | 1 |
| High school graduate | 10 | 17 | 7 | 7 | 3 | 3 | 0 | ||
| Associate’s degree | 22 | 13 | 8 | 3 | 5 | 1 | 0 | ||
| College graduate | 20 | 19 | 7 | 10 | 2 | 1 | 0 | ||
| Graduate degree | 12 | 5 | 7 | 7 | 3 | 1 | 0 | ||
Values denote sample size.
Table A3.
Predicted Montreal Cognitive Assessment scores estimated from a Beta-Binomial model by age and education for White males
| Age | ||||||||
|---|---|---|---|---|---|---|---|---|
| Education | Value | 55 | 60 | 65 | 70 | 75 | 80 | 85 |
| <High school education | P90 | 26 | 26 | 26 | 26 | 26 | 25 | 25 |
| P75 | 24 | 24 | 24 | 24 | 24 | 23 | 23 | |
| P50 | 22 | 22 | 22 | 22 | 22 | 21 | 21 | |
| P25 | 19 | 19 | 20 | 20 | 19 | 19 | 19 | |
| P10 | 17 | 17 | 17 | 17 | 17 | 17 | 17 | |
| P05 | 15 | 15 | 16 | 16 | 16 | 16 | 16 | |
| P01 | 12 | 13 | 13 | 13 | 14 | 14 | 14 | |
|
| ||||||||
| High school graduate | P90 | 28 | 27 | 27 | 27 | 26 | 26 | 25 |
| P75 | 26 | 26 | 26 | 25 | 25 | 24 | 24 | |
| P50 | 24 | 24 | 24 | 23 | 23 | 22 | 22 | |
| P25 | 22 | 22 | 22 | 21 | 21 | 20 | 20 | |
| P10 | 20 | 20 | 19 | 19 | 19 | 18 | 18 | |
| P05 | 18 | 18 | 18 | 18 | 17 | 17 | 17 | |
| P01 | 16 | 16 | 16 | 15 | 15 | 15 | 14 | |
|
| ||||||||
| Associate’s degree | P90 | 28 | 28 | 28 | 28 | 27 | 27 | 26 |
| P75 | 27 | 27 | 27 | 26 | 26 | 25 | 25 | |
| P50 | 25 | 25 | 25 | 25 | 24 | 23 | 23 | |
| P25 | 23 | 23 | 23 | 22 | 22 | 21 | 21 | |
| P10 | 21 | 21 | 21 | 20 | 20 | 19 | 19 | |
| P05 | 20 | 20 | 19 | 19 | 18 | 18 | 17 | |
| P01 | 17 | 17 | 17 | 16 | 16 | 15 | 15 | |
|
| ||||||||
| College graduate | P90 | 29 | 29 | 29 | 28 | 28 | 28 | 27 |
| P75 | 28 | 28 | 27 | 27 | 27 | 26 | 26 | |
| P50 | 26 | 26 | 26 | 26 | 25 | 25 | 24 | |
| P25 | 24 | 24 | 24 | 24 | 23 | 23 | 22 | |
| P10 | 23 | 23 | 22 | 22 | 21 | 21 | 20 | |
| P05 | 22 | 21 | 21 | 21 | 20 | 19 | 19 | |
| P01 | 19 | 19 | 19 | 18 | 18 | 17 | 16 | |
|
| ||||||||
| Graduate degree | P90 | 29 | 29 | 29 | 29 | 28 | 28 | 27 |
| P75 | 28 | 28 | 28 | 28 | 27 | 27 | 26 | |
| P50 | 27 | 27 | 26 | 26 | 26 | 25 | 25 | |
| P25 | 25 | 25 | 25 | 24 | 24 | 23 | 23 | |
| P10 | 23 | 23 | 23 | 22 | 22 | 22 | 21 | |
| P05 | 22 | 22 | 22 | 21 | 21 | 20 | 20 | |
| P01 | 20 | 20 | 19 | 19 | 19 | 18 | 18 | |
P90, P75, P50, P25, P10, P05, and P01 denote the 90th, 75th, 50th, 25th, 10th, 5th, and 1st percentiles respectively.
Table A4.
Predicted Montreal Cognitive Assessment scores estimated from a Beta-Binomial model by age and education for White females
| Age | ||||||||
|---|---|---|---|---|---|---|---|---|
| Education | Value | 55 | 60 | 65 | 70 | 75 | 80 | 85 |
| <High school education | P90 | 27 | 27 | 27 | 27 | 27 | 26 | 26 |
| P75 | 26 | 26 | 25 | 25 | 25 | 25 | 24 | |
| P50 | 23 | 23 | 23 | 23 | 23 | 23 | 22 | |
| P25 | 20 | 21 | 21 | 21 | 21 | 21 | 20 | |
| P10 | 18 | 18 | 18 | 19 | 19 | 19 | 19 | |
| P05 | 16 | 17 | 17 | 17 | 17 | 17 | 17 | |
| P01 | 13 | 13 | 14 | 14 | 15 | 15 | 15 | |
|
| ||||||||
| High school graduate | P90 | 28 | 28 | 28 | 28 | 27 | 27 | 26 |
| P75 | 27 | 27 | 27 | 26 | 26 | 25 | 25 | |
| P50 | 25 | 25 | 25 | 24 | 24 | 23 | 23 | |
| P25 | 23 | 23 | 23 | 22 | 22 | 21 | 21 | |
| P10 | 21 | 21 | 21 | 20 | 20 | 19 | 19 | |
| P05 | 19 | 19 | 19 | 19 | 19 | 18 | 18 | |
| P01 | 17 | 17 | 16 | 16 | 16 | 16 | 16 | |
|
| ||||||||
| Associate’s degree | P90 | 29 | 29 | 29 | 28 | 28 | 28 | 27 |
| P75 | 28 | 28 | 28 | 27 | 27 | 26 | 26 | |
| P50 | 26 | 26 | 26 | 25 | 25 | 24 | 24 | |
| P25 | 24 | 24 | 24 | 23 | 23 | 22 | 22 | |
| P10 | 22 | 22 | 22 | 21 | 21 | 20 | 20 | |
| P05 | 21 | 21 | 20 | 20 | 19 | 19 | 18 | |
| P01 | 18 | 18 | 18 | 17 | 17 | 16 | 16 | |
|
| ||||||||
| College graduate | P90 | 29 | 29 | 29 | 29 | 29 | 28 | 28 |
| P75 | 28 | 28 | 28 | 28 | 28 | 27 | 27 | |
| P50 | 27 | 27 | 27 | 26 | 26 | 26 | 25 | |
| P25 | 25 | 25 | 25 | 25 | 24 | 24 | 23 | |
| P10 | 24 | 23 | 23 | 23 | 22 | 22 | 21 | |
| P05 | 22 | 22 | 22 | 21 | 21 | 20 | 20 | |
| P01 | 20 | 20 | 19 | 19 | 18 | 18 | 17 | |
|
| ||||||||
| Graduate degree | P90 | 30 | 29 | 29 | 29 | 29 | 28 | 28 |
| P75 | 29 | 29 | 28 | 28 | 28 | 27 | 27 | |
| P50 | 27 | 27 | 27 | 27 | 26 | 26 | 25 | |
| P25 | 26 | 26 | 25 | 25 | 25 | 24 | 24 | |
| P10 | 24 | 24 | 24 | 23 | 23 | 23 | 22 | |
| P05 | 23 | 23 | 22 | 22 | 22 | 21 | 21 | |
| P01 | 20 | 20 | 20 | 20 | 20 | 19 | 19 | |
P90, P75, P50, P25, P10, P05, and P01 denote the 90th, 75th, 50th, 25th, 10th, 5th, and 1st percentiles respectively.
Table A5.
Predicted Montreal Cognitive Assessment scores estimated from a Beta-Binomial model by age and education for Black males
| Age | ||||||||
|---|---|---|---|---|---|---|---|---|
| Education | Value | 55 | 60 | 65 | 70 | 75 | 80 | 85 |
| <High school education | P90 | 24 | 24 | 24 | 24 | 23 | 23 | 23 |
| P75 | 22 | 22 | 22 | 22 | 22 | 21 | 21 | |
| P50 | 20 | 20 | 20 | 20 | 20 | 19 | 19 | |
| P25 | 17 | 17 | 17 | 17 | 17 | 17 | 17 | |
| P10 | 14 | 15 | 15 | 15 | 15 | 15 | 15 | |
| P05 | 13 | 13 | 14 | 14 | 14 | 14 | 14 | |
| P01 | 10 | 11 | 11 | 11 | 12 | 12 | 12 | |
|
| ||||||||
| High school graduate | P90 | 26 | 26 | 26 | 25 | 25 | 24 | 23 |
| P75 | 24 | 24 | 24 | 23 | 23 | 22 | 22 | |
| P50 | 22 | 22 | 22 | 21 | 21 | 20 | 20 | |
| P25 | 20 | 20 | 20 | 19 | 19 | 18 | 17 | |
| P10 | 18 | 18 | 18 | 17 | 17 | 16 | 16 | |
| P05 | 17 | 16 | 16 | 16 | 15 | 15 | 14 | |
| P01 | 14 | 14 | 14 | 13 | 13 | 13 | 12 | |
|
| ||||||||
| Associate’s degree | P90 | 27 | 27 | 27 | 26 | 26 | 25 | 24 |
| P75 | 26 | 25 | 25 | 25 | 24 | 23 | 23 | |
| P50 | 24 | 24 | 23 | 23 | 22 | 21 | 21 | |
| P25 | 22 | 21 | 21 | 20 | 20 | 19 | 18 | |
| P10 | 20 | 19 | 19 | 18 | 18 | 17 | 16 | |
| P05 | 18 | 18 | 18 | 17 | 16 | 16 | 15 | |
| P01 | 16 | 16 | 15 | 15 | 14 | 13 | 13 | |
|
| ||||||||
| College graduate | P90 | 28 | 28 | 28 | 27 | 27 | 26 | 26 |
| P75 | 26 | 26 | 26 | 26 | 25 | 25 | 24 | |
| P50 | 25 | 25 | 24 | 24 | 24 | 23 | 22 | |
| P25 | 23 | 23 | 23 | 22 | 21 | 21 | 20 | |
| P10 | 21 | 21 | 21 | 20 | 20 | 19 | 18 | |
| P05 | 20 | 20 | 19 | 19 | 18 | 18 | 17 | |
| P01 | 18 | 18 | 17 | 17 | 16 | 15 | 15 | |
|
| ||||||||
| Graduate degree | P90 | 28 | 28 | 28 | 28 | 27 | 27 | 26 |
| P75 | 27 | 27 | 27 | 26 | 26 | 25 | 25 | |
| P50 | 26 | 25 | 25 | 25 | 24 | 23 | 23 | |
| P25 | 24 | 23 | 23 | 23 | 22 | 22 | 21 | |
| P10 | 22 | 22 | 21 | 21 | 20 | 20 | 19 | |
| P05 | 21 | 21 | 20 | 20 | 19 | 19 | 18 | |
| P01 | 18 | 18 | 18 | 18 | 17 | 17 | 16 | |
P90, P75, P50, P25, P10, P05, and P01 denote the 90th, 75th, 50th, 25th, 10th, 5th, and 1st percentiles respectively.
Table A6.
Predicted Montreal Cognitive Assessment scores estimated from a Beta-Binomial model by age and education for Black females
| Age | ||||||||
|---|---|---|---|---|---|---|---|---|
| Education | Value | 55 | 60 | 65 | 70 | 75 | 80 | 85 |
| <High school education | P90 | 25 | 25 | 25 | 24 | 24 | 23 | 23 |
| P75 | 22 | 23 | 23 | 22 | 22 | 22 | 21 | |
| P50 | 20 | 20 | 20 | 20 | 20 | 19 | 19 | |
| P25 | 17 | 17 | 18 | 18 | 17 | 17 | 17 | |
| P10 | 14 | 15 | 15 | 15 | 15 | 15 | 15 | |
| P05 | 13 | 13 | 14 | 14 | 14 | 14 | 14 | |
| P01 | 10 | 11 | 11 | 11 | 11 | 12 | 12 | |
|
| ||||||||
| High school graduate | P90 | 26 | 26 | 26 | 25 | 25 | 24 | 24 |
| P75 | 25 | 25 | 24 | 24 | 23 | 22 | 22 | |
| P50 | 23 | 22 | 22 | 22 | 21 | 20 | 20 | |
| P25 | 20 | 20 | 20 | 19 | 19 | 18 | 18 | |
| P10 | 18 | 18 | 18 | 17 | 17 | 16 | 16 | |
| P05 | 16 | 16 | 16 | 16 | 15 | 15 | 14 | |
| P01 | 14 | 14 | 14 | 13 | 13 | 13 | 12 | |
|
| ||||||||
| Associate’s degree | P90 | 27 | 27 | 27 | 27 | 26 | 25 | 25 |
| P75 | 26 | 26 | 25 | 25 | 24 | 24 | 23 | |
| P50 | 24 | 24 | 23 | 23 | 22 | 22 | 21 | |
| P25 | 22 | 22 | 21 | 21 | 20 | 19 | 19 | |
| P10 | 20 | 19 | 19 | 18 | 18 | 17 | 16 | |
| P05 | 18 | 18 | 18 | 17 | 16 | 16 | 15 | |
| P01 | 16 | 15 | 15 | 14 | 14 | 13 | 13 | |
|
| ||||||||
| College graduate | P90 | 28 | 28 | 28 | 27 | 27 | 27 | 26 |
| P75 | 27 | 27 | 26 | 26 | 26 | 25 | 25 | |
| P50 | 25 | 25 | 25 | 24 | 24 | 23 | 23 | |
| P25 | 23 | 23 | 23 | 22 | 22 | 21 | 20 | |
| P10 | 21 | 21 | 21 | 20 | 20 | 19 | 18 | |
| P05 | 20 | 20 | 19 | 19 | 18 | 18 | 17 | |
| P01 | 18 | 17 | 17 | 16 | 16 | 15 | 15 | |
|
| ||||||||
| Graduate degree | P90 | 28 | 28 | 28 | 28 | 27 | 27 | 26 |
| P75 | 27 | 27 | 27 | 26 | 26 | 25 | 25 | |
| P50 | 26 | 25 | 25 | 25 | 24 | 24 | 23 | |
| P25 | 24 | 24 | 23 | 23 | 22 | 22 | 21 | |
| P10 | 22 | 22 | 21 | 21 | 20 | 20 | 19 | |
| P05 | 21 | 20 | 20 | 20 | 19 | 19 | 18 | |
| P01 | 18 | 18 | 18 | 17 | 17 | 17 | 16 | |
P90, P75, P50, P25, P10, P05, and P01 denote the 90th, 75th, 50th, 25th, 10th, 5th, and 1st percentiles respectively.
Table A7.
Predicted Montreal Cognitive Assessment scores estimated from a Beta-Binomial model by age and education for Hispanic males
| Age | ||||||||
|---|---|---|---|---|---|---|---|---|
| Education | Value | 55 | 60 | 65 | 70 | 75 | 80 | 85 |
| <High school education | P90 | 24 | 24 | 24 | 24 | 23 | 23 | 23 |
| P75 | 22 | 22 | 22 | 22 | 22 | 21 | 21 | |
| P50 | 19 | 20 | 20 | 19 | 19 | 19 | 19 | |
| P25 | 16 | 17 | 17 | 17 | 17 | 17 | 16 | |
| P10 | 14 | 14 | 14 | 15 | 15 | 15 | 14 | |
| P05 | 12 | 13 | 13 | 13 | 13 | 13 | 13 | |
| P01 | 9 | 10 | 10 | 10 | 11 | 11 | 11 | |
|
| ||||||||
| High school graduate | P90 | 26 | 26 | 26 | 25 | 25 | 24 | 23 |
| P75 | 24 | 24 | 24 | 23 | 23 | 22 | 21 | |
| P50 | 22 | 22 | 22 | 21 | 21 | 20 | 19 | |
| P25 | 20 | 19 | 19 | 19 | 18 | 18 | 17 | |
| P10 | 17 | 17 | 17 | 16 | 16 | 16 | 15 | |
| P05 | 16 | 16 | 15 | 15 | 15 | 14 | 14 | |
| P01 | 13 | 13 | 13 | 13 | 12 | 12 | 11 | |
|
| ||||||||
| Associate’s degree | P90 | 27 | 27 | 27 | 26 | 26 | 25 | 24 |
| P75 | 26 | 25 | 25 | 25 | 24 | 23 | 23 | |
| P50 | 24 | 23 | 23 | 22 | 22 | 21 | 20 | |
| P25 | 21 | 21 | 21 | 20 | 19 | 19 | 18 | |
| P10 | 19 | 19 | 18 | 18 | 17 | 17 | 16 | |
| P05 | 18 | 17 | 17 | 16 | 16 | 15 | 15 | |
| P01 | 15 | 15 | 14 | 14 | 13 | 13 | 12 | |
|
| ||||||||
| College graduate | P90 | 28 | 28 | 28 | 27 | 27 | 26 | 26 |
| P75 | 26 | 26 | 26 | 26 | 25 | 25 | 24 | |
| P50 | 25 | 25 | 24 | 24 | 23 | 23 | 22 | |
| P25 | 23 | 23 | 22 | 22 | 21 | 21 | 20 | |
| P10 | 21 | 21 | 20 | 20 | 19 | 18 | 18 | |
| P05 | 20 | 19 | 19 | 18 | 18 | 17 | 16 | |
| P01 | 17 | 17 | 16 | 16 | 15 | 14 | 14 | |
|
| ||||||||
| Graduate degree | P90 | 28 | 28 | 28 | 28 | 27 | 27 | 26 |
| P75 | 27 | 27 | 27 | 26 | 26 | 25 | 24 | |
| P50 | 25 | 25 | 25 | 24 | 24 | 23 | 23 | |
| P25 | 23 | 23 | 23 | 22 | 22 | 21 | 21 | |
| P10 | 21 | 21 | 21 | 20 | 20 | 19 | 19 | |
| P05 | 20 | 20 | 20 | 19 | 19 | 18 | 18 | |
| P01 | 17 | 17 | 17 | 17 | 16 | 16 | 15 | |
P90, P75, P50, P25, P10, P05, and P01 denote the 90th, 75th, 50th, 25th, 10th, 5th, and 1st percentiles respectively.
Table A8.
Predicted Montreal Cognitive Assessment scores estimated from a Beta-Binomial model by age and education for Hispanic females
| Age | ||||||||
|---|---|---|---|---|---|---|---|---|
| Education | Value | 55 | 60 | 65 | 70 | 75 | 80 | 85 |
| <High school education | P90 | 25 | 25 | 25 | 24 | 24 | 23 | 23 |
| P75 | 22 | 22 | 22 | 22 | 22 | 21 | 21 | |
| P50 | 19 | 19 | 19 | 19 | 19 | 19 | 18 | |
| P25 | 16 | 16 | 16 | 16 | 16 | 16 | 16 | |
| P10 | 13 | 13 | 14 | 14 | 14 | 14 | 14 | |
| P05 | 11 | 11 | 12 | 12 | 12 | 13 | 13 | |
| P01 | 8 | 8 | 9 | 9 | 10 | 10 | 10 | |
|
| ||||||||
| High school graduate | P90 | 26 | 26 | 26 | 25 | 25 | 24 | 23 |
| P75 | 25 | 24 | 24 | 23 | 23 | 22 | 21 | |
| P50 | 22 | 22 | 22 | 21 | 20 | 20 | 19 | |
| P25 | 19 | 19 | 19 | 18 | 18 | 17 | 17 | |
| P10 | 16 | 16 | 16 | 16 | 15 | 15 | 14 | |
| P05 | 15 | 15 | 15 | 14 | 14 | 14 | 13 | |
| P01 | 12 | 12 | 12 | 11 | 11 | 11 | 11 | |
|
| ||||||||
| Associate’s degree | P90 | 27 | 27 | 27 | 27 | 26 | 25 | 25 |
| P75 | 26 | 26 | 25 | 25 | 24 | 23 | 23 | |
| P50 | 24 | 23 | 23 | 22 | 22 | 21 | 20 | |
| P25 | 21 | 21 | 20 | 20 | 19 | 18 | 18 | |
| P10 | 18 | 18 | 18 | 17 | 16 | 16 | 15 | |
| P05 | 17 | 16 | 16 | 15 | 15 | 14 | 14 | |
| P01 | 14 | 13 | 13 | 12 | 12 | 12 | 11 | |
|
| ||||||||
| College graduate | P90 | 28 | 28 | 28 | 27 | 27 | 27 | 26 |
| P75 | 26 | 26 | 26 | 26 | 25 | 25 | 24 | |
| P50 | 25 | 24 | 24 | 24 | 23 | 23 | 22 | |
| P25 | 22 | 22 | 22 | 21 | 21 | 20 | 20 | |
| P10 | 20 | 20 | 20 | 19 | 18 | 18 | 17 | |
| P05 | 19 | 19 | 18 | 17 | 17 | 16 | 16 | |
| P01 | 16 | 16 | 15 | 15 | 14 | 13 | 13 | |
|
| ||||||||
| Graduate degree | P90 | 29 | 28 | 28 | 28 | 27 | 27 | 26 |
| P75 | 27 | 27 | 27 | 26 | 26 | 25 | 24 | |
| P50 | 25 | 25 | 25 | 24 | 24 | 23 | 22 | |
| P25 | 23 | 23 | 23 | 22 | 22 | 21 | 20 | |
| P10 | 21 | 21 | 20 | 20 | 19 | 19 | 18 | |
| P05 | 19 | 19 | 19 | 19 | 18 | 18 | 17 | |
| P01 | 16 | 16 | 16 | 16 | 15 | 15 | 15 | |
P90, P75, P50, P25, P10, P05, and P01 denote the 90th, 75th, 50th, 25th, 10th, 5th, and 1st percentiles respectively.
Footnotes
Conflicts of Interest: KMS is a full-time employee of Genentech and stockholder in F Hoffman La Roche. Genentech had no role in the funding, design, or analysis of SPRINT or of this manuscript. JDW receives funding from Biogen.
References
- Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC, Gamst A, Holtzman DM, Jagust WJ, Petersen RC, Snyder PJ, Carrillo MC, Thies B, & Phelps CH (2011). The diagnosis of mild cognitive impairment due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & dementia : the journal of the Alzheimer’s Association, 7(3), 270–279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ambrosius WT, Sink KM, Foy CG, Berlowitz DR, Cheung AK, Cushman WC, … SPRINT Study Research Group. (2014). The design and rationale of a multicenter clinical trial comparing two strategies for control of systolic blood pressure: the Systolic Blood Pressure Intervention Trial (SPRINT). Clinical Trials (London, England), 11(5), 532–546. 10.1177/1740774514537404 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Arostegui I, Núñez-Antón V, & Quintana JM (2012). Statistical approaches to analyse patient-reported outcomes as response variables: an application to health-related quality of life. Statistical Methods in Medical Research, 21(2), 189–214. 10.1177/0962280210379079 [DOI] [PubMed] [Google Scholar]
- Besser L, Kukull W, Knopman DS, Chui H, Galasko D, Weintraub S, … Morris JC (2018). Version 3 of the National Alzheimer’s Coordinating Center’s Uniform Data Set. Alzheimer Disease & Associated Disorders, 32(4), 1. 10.1097/WAD.0000000000000279 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Borland E, Nägga K, Nilsson PM, Minthon L, Nilsson ED, & Palmqvist S (2017). The Montreal Cognitive Assessment: Normative Data from a Large Swedish Population-Based Cohort. Journal of Alzheimer’s Disease, 59(3), 893–901. 10.3233/JAD-170203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Busch RM, Chelune GJ, & Suchy Y (2006). Using norms in neuropsychological assessment of the elderly. In Geriatric neuropsychology: Assessment and intervention (pp. 141–165). New York: Gilford. [Google Scholar]
- Carson N, Leach L, & Murphy KJ (2018). A re-examination of Montreal Cognitive Assessment (MoCA) cutoff scores. International Journal of Geriatric Psychiatry, 33(2), 379–388. 10.1002/gps.4756 [DOI] [PubMed] [Google Scholar]
- Chelune GJ (2017). Evidence-based practices in neuropsychology. In Bowden S (Ed.), Neuropsychological assessment in the age of evidence-based practice: Diagnostic and treatment evaluations (pp. 153–181). New York: Oxford University Press. [Google Scholar]
- Crawford JR, & Howell DC (1998). Regression equations in clinical neuropsychology: an evaluation of statistical methods for comparing predicted and obtained scores. Journal of Clinical and Experimental Neuropsychology, 20(5), 755–762. 10.1076/jcen.20.5.755.1132 [DOI] [PubMed] [Google Scholar]
- Crawford JR, Garthwaite PH, Denham AK, & Chelune GJ (2012). Using regression equations built from summary data in the psychological assessment of the individual case: Extension to multiple regression. Psychological Assessment, 24(4), 801–814. 10.1037/a0027699 [DOI] [PubMed] [Google Scholar]
- Damian AM, Jacobson SA, Hentz JG, Belden CM, Shill HA, Sabbagh MN, … Adler CH (2011). The Montreal Cognitive Assessment and the mini-mental state examination as screening instruments for cognitive impairment: item analyses and threshold scores. Dementia and Geriatric Cognitive Disorders, 31(2), 126–131. 10.1159/000323867 [DOI] [PubMed] [Google Scholar]
- De Santi S, Pirraglia E, Barr W, Babb J, Williams S, Rogers K, … de Leon MJ (2008). Robust and conventional neuropsychological norms: diagnosis and prediction of age-related cognitive decline. Neuropsychology, 22(4), 469–484. 10.1037/0894-4105.22.4.469 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Eilers PHC, Marx BD, & Durbán M (2015). Twenty years of P-splines. SORT-Statistics and Operations Research Transactions, 39(2), 149–186. Retrieved from https://www.raco.cat/index.php/SORT/article/view/302258 [Google Scholar]
- Gluhm S, Goldstein J, Loc K, Colt A, Liew C. Van, & Corey-Bloom J (2013). Cognitive performance on the mini-mental state examination and the montreal cognitive assessment across the healthy adult lifespan. Cognitive and Behavioral Neurology : Official Journal of the Society for Behavioral and Cognitive Neurology, 26(1), 1–5. 10.1097/WNN.0b013e31828b7d26 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Goldstein FC, Ashley AV, Miller E, Alexeeva O, Zanders L, & King V (2014). Validity of the montreal cognitive assessment as a screen for mild cognitive impairment and dementia in African Americans. Journal of Geriatric Psychiatry and Neurology, 27(3), 199–203. 10.1177/0891988714524630 [DOI] [PubMed] [Google Scholar]
- Guilmette TJ, Sweet JJ, Hebben N, Koltai D, Mahone EM, Spiegler BJ, Stucky K, Westerveld M, & Conference Participants (2020). American Academy of Clinical Neuropsychology consensus conference statement on uniform labeling of performance test scores. The Clinical neuropsychologist, 34(3), 437–453. [DOI] [PubMed] [Google Scholar]
- Holtzer R, Goldin Y, Zimmerman M, Katz M, Buschke H, & Lipton RB (2008). Robust norms for selected neuropsychological tests in older adults. Archives of Clinical Neuropsychology : The Official Journal of the National Academy of Neuropsychologists, 23(5), 531–541. 10.1016/j.acn.2008.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ivnik RJ, Smith GE, Cerhan JH, Boeve BF, Tangalos EG, & Petersen RC (2001). Understanding the diagnostic capabilities of cognitive tests. The Clinical Neuropsychologist, 15(1), 114–124. 10.1076/clin.15.1.114.1904 [DOI] [PubMed] [Google Scholar]
- Kawas C, Segal J, Stewart WF, Corrada M, & Thal LJ (1994). A validation study of the Dementia Questionnaire. Archives of neurology, 51(9), 901–906. [DOI] [PubMed] [Google Scholar]
- Kroenke K, Spitzer RL, & Williams JBW (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. 10.1046/j.1525-1497.2001.016009606.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Luis CA, Keegan AP, & Mullan M (2009). Cross validation of the Montreal Cognitive Assessment in community dwelling older adults residing in the Southeastern US. International Journal of Geriatric Psychiatry, 24(2), 197–201. 10.1002/gps.2101 [DOI] [PubMed] [Google Scholar]
- Malek-Ahmadi M, Powell JJ, Belden CM, O’Connor K, Evans L, Coon DW, & Nieri W (2015). Age- and education-adjusted normative data for the Montreal Cognitive Assessment (MoCA) in older adults age 70–99. Neuropsychology, Development, and Cognition. Section B, Aging, Neuropsychology and Cognition, 22(6), 755–761. 10.1080/13825585.2015.1041449 [DOI] [PubMed] [Google Scholar]
- McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR Jr, Kawas CH, Klunk WE, Koroshetz WJ, Manly JJ, Mayeux R, Mohs RC, Morris JC, Rossor MN, Scheltens P, Carrillo MC, Thies B, Weintraub S, & Phelps CH (2011). The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & dementia : the journal of the Alzheimer’s Association, 7(3), 263–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Milani SA, Marsiske M, Cottler LB, Chen X, & Striley CW (2018). Optimal cutoffs for the Montreal Cognitive Assessment vary by race and ethnicity. Alzheimer’s & dementia (Amsterdam, Netherlands), 10, 773–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mitrushina M, Boone KB, & D’Elia LF (1999). Handbook of normative data for neuropsychological assessment. New York: Oxford University Press. [Google Scholar]
- Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I, … Chertkow H (2005). The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. Journal of the American Geriatrics Society, 53(4), 695–699. 10.1111/j.1532-5415.2005.53221.x [DOI] [PubMed] [Google Scholar]
- Obermeyer Z, Powers B, Vogeli C, & Mullainathan S (2019). Dissecting racial bias in an algorithm used to manage the health of populations. Science (New York, N.Y.), 366(6464), 447–453. 10.1126/science.aax2342 [DOI] [PubMed] [Google Scholar]
- Pedraza O, Lucas JA, Smith GE, Petersen RC, Graff-Radford NR, & Ivnik RJ (2010). Robust and expanded norms for the Dementia Rating Scale. Archives of Clinical Neuropsychology : The Official Journal of the National Academy of Neuropsychologists, 25(5), 347–358. 10.1093/arclin/acq030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pfeffer RI, Kurosaki TT, Harrah CH, Chance JM, & Filos S (1982). Measurement of functional activities in older adults in the community. Journal of Gerontology. 37(3), 323–329. [DOI] [PubMed] [Google Scholar]
- Rapp SR, Legault C, Espeland MA, Resnick SM, Hogan PE, Coker LH, Dailey M, Shumaker SA, & CAT Study Group (2012). Validation of a cognitive assessment battery administered over the telephone. Journal of the American Geriatrics Society, 60(9), 1616–1623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ritchie LJ, Frerichs RJ, & Tuokko H (2007). Effective normative samples for the detection of cognitive impairment in older adults. The Clinical Neuropsychologist, 21(6), 863–874. 10.1080/13854040701557239 [DOI] [PubMed] [Google Scholar]
- Rossetti HC, Lacritz LH, Cullum CM, & Weiner MF (2011). Normative data for the Montreal Cognitive Assessment (MoCA) in a population-based sample. Neurology, 77(13), 1272–1275. 10.1212/WNL.0b013e318230208a [DOI] [PubMed] [Google Scholar]
- Rossetti HC, Lacritz LH, Hynan LS, Cullum CM, Van Wright A, & Weiner MF (2017). Montreal Cognitive Assessment Performance among Community-Dwelling African Americans. Archives of Clinical Neuropsychology : The Official Journal of the National Academy of Neuropsychologists, 32(2), 238–244. 10.1093/arclin/acw095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sackett DL, Straus SE, Richardson W, & Haynes RB (2000). Evidence-based medicine: How to practice and teach EBM (Churchill; 2nd, ed.). New York. [Google Scholar]
- Sink KM, Craft S, Smith SC, Maldjian JA, Bowden DW, Xu J, … Divers J (2015). Montreal Cognitive Assessment and Modified Mini Mental State Examination in African Americans. Journal of Aging Research, 2015, 872018. 10.1155/2015/872018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- SPRINT MIND Investigators for the SPRINT Research Group, Williamson JD, Pajewski NM, Auchus AP, Bryan RN, Chelune G, … Wright CB (2019). Effect of Intensive vs Standard Blood Pressure Control on Probable Dementia: A Randomized Clinical Trial. JAMA, 321(6), 553–561. 10.1001/jama.2018.21442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- SPRINT Research Group, Wright JT, Williamson JD, Whelton PK, Snyder JK, Sink KM, … Ambrosius WT (2015). A Randomized Trial of Intensive versus Standard Blood-Pressure Control. The New England Journal of Medicine, 373(22), 2103–2116. 10.1056/NEJMoa1511939 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stasinopoulos DM, & Rigby RA (2007). Generalized Additive Models for Location, Scale, and Shape (GAMLSS) in R. Journal Of Statistical Software, 23(7). [Google Scholar]
- Voncken L, Albers CJ, & Timmerman ME (2019). Model Selection in Continuous Test Norming With GAMLSS. Assessment, 26(7), 1329–1346. 10.1177/1073191117715113 [DOI] [PubMed] [Google Scholar]
- Waldron-Perrine B, & Axelrod BN (2012). Determining an appropriate cutting score for indication of impairment on the Montreal Cognitive Assessment. International Journal of Geriatric Psychiatry, 27(11), 1189–1194. 10.1002/gps.3768 [DOI] [PubMed] [Google Scholar]
- Wechsler D (2008). Wechsler Adult Intelligence Scale-Fourth Edition. San Antonio, TX: Pearson. [Google Scholar]
- Wechsler D (2009). Wechsler Memory Scale-Fourth Edition. San Antonio, TX: Pearson. [Google Scholar]
- Whelton PK, Carey RM, Aronow WS, Casey DE, Collins KJ, Dennison Himmelfarb C, … Wright JT (2018). 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task F. Hypertension (Dallas, Tex. : 1979), 71(6), 1269–1324. 10.1161/HYP.0000000000000066 [DOI] [PubMed] [Google Scholar]
