MATRICS Consensus Cognitive Battery

Schizophrenia: (calculated from statistics in Keefe et al., 2011; n = 323; mean age = 43.1 (10.4) years; PANSS total score = 67.5 (11.7))

MCCB subscales

 SEM

Composite T-score

4.295

Speed of processing

5.545

Attention/vigilance

5.728

Working memory

5.132

Verbal learning

4.861

Visual learning

6.922

Reason/problem solving

4.460

Social cognition

6.300

Schizophrenia: (calculated from statistics in Georgiades et al., 2017; n = 2621; mean age = 42.7 (10.52) years)

MCCB subscales

 SEM

Neurocognitive composite

4.112

Cognitive composite

4.299

Speed of processing

5.318

Attention/vigilance

5.718

Working memory

5.147

Verbal learning

4.940

Visual learning

6.676

Reason/problem solving

4.561

Social cognition

6.418

Schizophrenia: (calculated from statistics in Keefe et al., 2011)

MCCB subscales

 MDC

Composite T-score

11.91

Speed of processing

15.37

Attention/vigilance

15.88

Working memory

14.23

Verbal learning

13.47

Visual learning

19.19

Reason/problem solving

12.36

Social cognition

17.46

Schizophrenia: (calculated from statistics in Georgiades et al., 2017)

MCCB subscales

 MDC

Neurocognitive composite

11.40

Cognitive composite

11.92

Speed of processing

14.74

Attention/vigilance

15.85

Working memory

14.27

Verbal learning

13.69

Visual learning

18.50

Reason/problem solving

12.64

Social cognition

17.79

Schizophrenia: (Kern et al., 2011; Schizophrenia or Schizoaffective Disorder: n = 176, mean age = 44.0 (11.2), male = 76%; Brief Psychiatric Rating Scale (BPRS) total score = 47.3 (13.6); Community Residents: n = 300, mean age = 42.6 (11.6), male = 47%)

• A T-score of ≤ 43.9 on Speed of Processing distinguishes persons with schizophrenia from community residents (sensitivity 66.4%; specificity 87.6%). Discrimination was further improved with the additional consideration of Social Cognition performance.
• Among those with schizophrenia, a T-score of ≤ 30.5 on Speed of Processing distinguishes unemployed persons from employed persons (sensitivity 89.4%; specificity 34.5%). Discrimination was further improved with the additional consideration of Visual Learning and Attention/Vigilance.

Major Depressive Disorder: (Liang et al., 2020; n = 48; mean age = 39.7 (13.4); HDRS-17 total scores ≥ 18; Chinese sample and translation of MCCB)

Cut-off value, Sensitivity, and Sensitivity by domain for patients with major depressive disorder

Domain

Cut-off

Sensitivity

Specificity

Speed of processing

40

0.778

0.771

Attention/vigilance

37

0.667

0.743

Working memory

39

0.444

0.914

Verbal learning

40

0.622

0.800

Visual learning

45

0.644

0.857

Reason/problem solving

42

0.800

0.886

Social cognition

34

0.667

0.657

Major Depressive Disorder: (Liang et al., 2020)

Mean (SD) T-scores by MCCB domain and patient group

Domain

BD

(n = 43)

HC

(n = 35)

MDD

(n = 48)

SCH

(n = 32)

Speed of processing1

32.9 (12.8)

46.1 (12.3)

27 (14.2)

20 (14.6)

Attention/vigilance1

36 13.9)

43.9 (10.1)

31.7 (11.4)

28.3 (10.7)

Working memory1

41.4 (10)

47.6 (7.8)

39.7 (11.2)

36.3 (11)

Verbal learning1

39.8 (8)

49.4 (11.7)

37.2 (9.4)

32.6 (6.5)

Visual learning1

43 (10.2)

52.6 (8.6)

39.1 (12.2)

35.7 (12.6)

Reasoning/problem solving1

37.5 (7.5)

50 (7.9)

36.2 (6)

33.6 (4.9)

Social cognition2

37.5 (9.5)

37.6 (10.4)

30.9 (10.5)

29.5 (10.2)

Note—BD: bipolar disorder in the euthymic phase; HC: healthy control; MDD: major depressive disorders; SCH: schizophrenia

1p < 0.001

2p < 0.005

Major Depressive Disorder in Young Patients: (Liang et al., 2021; Y-MDD: n = 50, mean age = 17.9 (2.7) years, female = 34 (66%); Y-BD: n = 38, mean age = 18.9 (3.0) years, female = 20 (52.6%); HC: n = 51, mean age = 17.7 (3.2) years; age range for all subjects = 13-24; Chinese sample and translation of MCCB)

Mean (SD) T-scores by MCCB domain and young patient group

Domain

Y-MDD

Y-BD

HC

Speed of processing1

35.3(12.6)

28.3 (13.7)

43.8 (10.1)

Attention/vigilance2

37.2 (13.0)

28.4 (14.2)

38.6 (12.7)

Working memory2

44.2 (12.8)

36.2 (8.4)

45.4 (10.0)

Verbal learning1

44.6 (9.2)

39.4 (9.3)

52.2 (11.5)

Visual learning1

42.0 (13.4)

40.1 (10.7)

47.4 (10.7)

Reasoning/problem solving1

43.2 (10.9)

35.6 (7.6)

49.3 (8.9)

Social cognition

46.7 (15.3)

41.3 (15.2)

44.3 (10.2)

Note—Y-MDD: young patients with major depressive disorders; Y-BD: young patients with bipolar disorder in the euthymic phase; HC: healthy controls.

1p < 0.001

2p < 0.005

Schizophrenia: 

• Excellent test-retest reliability for overall composite score (ICC = 0.90) (Nuechterlein et al., 2008; n = 167; mean age = 44.0 (11.2); mean educational level = 12.4 (2.4) years; male = 76%; subjects diagnosed with schizophrenia or schizoaffective disorder, depressed type; tests at 4-week interval)
• Acceptable test-retest reliability for composite T-score (ICC = 0.88) (Keefe et al., 2011)
• Acceptable test-retest reliability for cognitive composite T-score (ICC = 0.88) (Georgiades et al., 2017)
  • Acceptable test-retest reliability for neurocognitive composite T-score (all domains except for Social Cognition) (ICC = 0.88)

Major Depressive Disorder: (Liang et al., 2020)

• Poor to Acceptable test-retest reliability (r = 0.59 to 0.81)

Schizophrenia: (Burton et al., 2013; n = 183, mean age = 44.45 (11.47), male = 127 (69.4%), community-dwelling outpatients diagnosed with schizophrenia or schizoaffective disorder) 

• Adequate internal consistency for one-factor model (Cronbach's alpha = 0.76)
• Poor to Adequate internal consistency for 3-factor model (Cronbach's alpha = 0.52 to 0.70)

Major Depressive Disorder: (Liang et al., 2020)

• Adequate to excellent internal consistency of Total MCCB for MDD patients at baseline (α = 0.83) and one month later (α = 0.74)
• Poor to excellent internal consistency for MDD patients at baseline for the four cognitive domains: Speed of processing (α = 0.69), Verbal learning (α = 0.88), Visual learning (α = 0.90), and Attention/vigilance (α = 0.90) (Working memory, Reason/problem solving, and Social cognition not tested)
• Poor to excellent internal consistency for MDD patients for the four cognitive domains at one month: Speed of processing (α = 0.58), Verbal learning (α = 0.87), Visual learning (α = 0.89), and Attention/vigilance (α = 0.84) (Working memory, Reason/problem solving, and Social cognition not tested)

Major Depressive Disorder in Young Patients: (Liang et al., 2021)

• Adequate to excellent internal consistency of Total MCCB in Y-MDD patients at baseline (α = 0.79) and two weeks later (α = 0.83)
• Poor to excellent internal consistency in Y-MDD patients at baseline for the four cognitive domains: Speed of processing (α = 0.61), Verbal learning (α = 0.83), Visual learning (α = 0.86), and Attention/vigilance (α = 0.87) (Working memory, Reason/problem solving, and Social cognition not tested)  
• Poor to excellent internal consistency in Y-MDD patients for the four cognitive domains two weeks later: Speed of processing (α = 0.58), Verbal learning (α = 0.87), Visual learning (α = 0.84), and Attention/vigilance (α = 0.90) (Working memory, Reason/problem solving, and Social cognition not tested)

Concurrent Validity:

Schizophrenia: 

• Adequate correlations between the MCCB composite score with the UCSD Performance-Based Skills Assessment-2 (UPSA-2) total score (r = 0.42) (Silverstein et al., 2010; n = 155; male = 102 (65%); outpatients or partial hospital patients with schizophrenia)
• Excellent correlations with the UCSD Performance-Based Skills Assessment-Brief (UPSA-B) (r = 0.60) (Keefe et al., 2011)
• Adequate correlations of all three MCCB factors with the UCSD Performance-Based Skills Assessment-Brief (UPSA-B) total score (Burton et al., 2013):
  • Speed of processing (r = 0.461)
  • Attention and working memory (r  = 0.468)
  • Learning (r = 0.465) 

Bipolar Disorder with Psychosis: (Sperry et al., 2015; Bipolar Disorder: n = 56, mean age = 30.5 (8.16), female = 57%; Schizophrenia: n = 37, mean age = 33.9 (11.12), female = 13.5%; Healthy Controls: n = 57, mean age = 25.7 (6,42), female = 59.6%)

• Adequate correlation between the Positive and Negative Syndrome Scale (PANSS) and the Processing Speed domain (r = -0.38, p < 0.01)
• Poor correlations between the PANSS and MCCB domains Working Memory (r = -0.29, p < 0.01), Verbal Memory (r = -0.24, p < 0.05), and the Overall Composite score (r = -0.28, p < 0.01)

Major Depressive Disorder: (Liang et al., 2020)

• Adequate to excellent correlations between MCCB domains and total scores on the Montreal Cognitive Assessment (MoCA) at baseline (r = 0.46 to 0.74)
• Adequate correlations between Speed of processing and verbal learning domains of the MCCB and total scores on the MoCA one month later (r = 0.45 and 0.47, respectively)

Major Depressive Disorder in Young Patients: (Liang et al., 2021)

• Adequate correlations between the recall domain of the MoCA and the T-scores of MCCB domains:
  • Speed of processing (r  = 0.47, p < 0.05)
  • Reasoning and problem solving (r = 0.53, p < 0.01)
  • Social cognition (r  = 0.41, p < 0.05)

Convergent Validity:

Schizophrenia: (Kaneda et al., 2013; n = 37; mean age = 38.4 (11.2); BPRS score = 41.0 (7.1); Japanese sample and translation: MCCB-J)

• Poor to excellent correlations between the MCCB-J and the Brief Assessment of Cognition in Schizophrenia (BACS-J) (r = 0.14 to 0.87)

Discriminant Validity:

Bipolar Disorder: (Van Rheenen et al., 2014; Bipolar Disorder: n = 50, mean age = 38, male = 16; Healthy Controls: n = 52, mean age = 34, male = 20; Australia sample)

• There is a significant group effect for cognitive domain performance overall among MCCB domain scores (F (7, 90) = 2.47, p < 0.05, Wilk’s λ = 0.84)
• Bipolar Disorder group showed significantly reduced performance on Speed of Processing (F (1, 96) = 7.33, p < 0.01), Working Memory (F (1, 96) = 9.78, p < 0.01), and Visual Learning (F (1, 96) = 8.19, p < 0.01)

Bipolar Disorder: (Russo et al., 2014; n = 173 (64 patients with BD: mean age = 41.2 (10.5), male = 50% and 109 healthy controls: mean age = 37.9 (11.6), male = 53.2%)

• Significantly worse performance for patients than healthy controls for all MCCB cognitive domains except for visual learning (p = 0.178) and reasoning/problem solving (p = 0.505)
• Significant negative correlations for healthy controls between the TEMPS-A irritability temperament and the MCCB domains attention (r = -0.315, p = 0.035) and social cognition (r = -0.386, p = 0.009)
• Significant positive correlations for BD patients between ratings on the cyclothymic temperament scale and MCCB scales processing speed (r = 0.432, p = 0.002), working memory (r = 0.379, p = 0.009), reasoning/problem solving (r = 0.438, p = 0.002), and global cognition as measured by an overall composite score (r = 0.467, p = 0.001)     

Major Depressive Disorder: (Liang et al., 2020; Major Depressive Disorder (MDD) n = 48; mean age = 39.7 (13.4); HDRS-17 total scores ≥ 18; Bipolar Disorder (BD) n = 43, mean age = 36.3 (11.9); Schizophrenia (SCH) n = 32, mean age = 42.0 (13.1); Healthy-control n = 35, mean age = 39.2 (14.2); Chinese sample and translation of MCCB)

• Significant differences found in every domain among the four patient groups, including Speed of Processing (F = 10.93, p < 0.001), Attention/Vigilance (F = 10.31, p < 0.001), Working Memory (F = 5.78, p < 0.001), Verbal Learning (F = 12.27, p < 0.001), Visual Learning (F = 6.61, p < 0.001), Reasoning and Problem Solving (F = 16.67, p < 0.001), and Social Cognition (F = 3.25, p = 0.002).
• The scores for Attention/Vigilance (F = 4.16, p = 0.002), Verbal Learning (F = 3.37, p = 0.007), and Visual Learning (F = 2.33, p = 0.046) were lower in the MDD group than in the BD group, but higher than in the SCH group.

• Subtests were selected based on discussion at the MATRICS consensus conference with 130 schizophrenia experts and professionals (Nuechterline et al., 2008, p. 203).
• The resulting items were then tested through the MATRICS psychometric and standardization study using a stratified sample of 300 individuals from the general community at five sites in differing geographic regions (Kern et al., 2008, p. 214).

The MCCB is considered the FDA gold standard outcome measure for assessing cognitive treatment effects in schizophrenia clinical trials (Georgiades et al., 2017, p. 172).

Schizophrenia: (Nuechterline et al., 2008)

• Excellent: There were no noticeable ceiling effects or constrictions of variance at the second testing.

Schizophrenia: (Keefe et al., 2011)

• Small change for composite T-score (Cohen’s d = 0.18a) (Keefe et al., 2011)
• Small changes for domains range from d = 0.20a for Speed of Processing to d = 0.03a for Social Cognition (Keefe et al., 2011)
• Small changes for overall cognitive composite (d = 0.15b) and neurocognitive composite (d = 0.17b) (Georgiades, et al., 2017)
• Small changes for domains range from d = 0.18b for Speed of Processing to d = 0.02b for Social Cognition (Georgiades, et al., 2017)

aEffect size for paired sample = (Baseline−Screening)/SD Screening.

bEffect size for paired sample = (Baseline−Screening)/pooled SD from Screening and Baseline.