Introduction: Cognitive difficulties can occur in both schizophrenia and bipolar disorder and may affect everyday functioning. Comparing performance across cognitive domains may help characterize similarities and differences between the disorders. Objective: To compare global and domain-wise cognitive function in patients with schizophrenia and patients with bipolar disorder. Methods: A hospital-based, cross-sectional comparative study was planned with 140 participants: 70 with schizophrenia and 70 with bipolar disorder. Cognitive function was assessed using a standardized cognitive battery, with results expressed as T-scores. Global and domain-wise scores were compared between groups; group scores were also compared with an assumed normative T-score of 50. Results: In the results, the mean global cognitive composite score was 34.8 ± 8.2 in the schizophrenia group and 40.7 ± 8.5 in the bipolar disorder group (mean difference -5.90; 95% CI: -8.69 to -3.11; p < 0.001). Scores across all assessed domains were below the assumed normative value in both groups. Between-group comparisons showed lower scores in the schizophrenia group for processing speed, attention/vigilance, working memory, verbal and visual learning, reasoning/problem solving, and social cognition (all p ≤ 0.006). Cognitive impairment, defined as a T-score below 40, was present in 82.9% of the schizophrenia group and 55.7% of the bipolar disorder group (p < 0.001). Conclusion: The results indicate lower cognitive scores in both diagnostic groups, with greater impairment in the schizophrenia group. These findings should be replaced with results calculated from the actual dataset before interpretation or dissemination.
Cognitive impairment is an important feature of both schizophrenia and bipolar disorder and can affect attention, processing speed, memory, executive functioning, and social cognition. These difficulties may interfere with education, employment, interpersonal relationships, independent living, and recovery. Cognitive problems in schizophrenia are often broad and may be present before the illness becomes clinically apparent. In bipolar disorder, cognitive performance can vary with mood state and illness course, although difficulties may also remain during periods of remission. Bortolato et al. (2015) found evidence of cognitive deficits in both disorders, with impairments generally more severe in schizophrenia [1]. Bora and Pantelis (2015), comparing first-episode groups, also reported cognitive deficits in bipolar disorder, while noting that the pattern and extent of impairment differed from schizophrenia [2]. Using the MATRICS Consensus Cognitive Battery, Lynham et al. (2018) found that cognitive impairment followed a gradient across diagnostic groups, with greater impairment in schizophrenia than in bipolar disorder [3]. Li et al. (2020), in a meta-analysis of comparative studies, reported poorer overall cognitive performance in schizophrenia, including across attention, working memory, verbal and visual learning, processing speed, and reasoning [4]. Ceylan et al. (2020) further showed that cognitive performance was related both to diagnosis and to clinical state, underscoring the importance of documenting symptom status when comparing groups [5]. Direct comparison of cognitive function in people with schizophrenia and bipolar disorder may therefore help characterize shared and disorder-specific difficulties and inform assessment and rehabilitation planning. The present cross-sectional study was planned to compare cognitive function between patients with schizophrenia and patients with bipolar disorder using a standardized assessment approach.
AIM
To compare cognitive function in patients with schizophrenia and patients with bipolar disorder.
OBJECTIVES
Source of Data
Data were obtained from eligible patients diagnosed with schizophrenia or bipolar disorder who attended or were admitted to the Department of Psychiatry during the study period. Clinical information was collected from the participants, their caregivers when necessary, and available medical records.
Study Design
The study was conducted as a hospital-based, cross-sectional, comparative study. Participants were assessed once during the study period, and cognitive function was compared between the schizophrenia group and the bipolar disorder group.
Study Location
The study was carried out in the Department of Psychiatry.
Study Duration
The study was conducted over 06 months
Sample Size
The total sample size was 140 participants, with 70 patients in each group.
Inclusion Criteria
Participants were included if they:
Exclusion Criteria
Participants were excluded if they:
Procedure and Methodology
After ethics approval, eligible participants were identified from outpatient and inpatient services. The study purpose, assessment process, potential inconvenience, confidentiality safeguards, and voluntary nature of participation were explained in a language understood by the participant. Written informed consent was obtained before data collection.
A psychiatrist confirmed the diagnosis and reviewed the clinical record. Participants were allocated to one of two groups according to diagnosis: schizophrenia or bipolar disorder. A structured case record form was used to document demographic details, illness history, treatment history, and clinical status at assessment. Relevant clinical variables included age at onset, duration of illness, number of previous admissions or episodes, current medication, and symptom status.
Cognitive function was assessed in a quiet setting using the same standardized cognitive assessment instrument for both groups. The final instrument was to be specified in the protocol and administered according to its manual by a trained assessor. The assessment covered domains such as processing speed, attention, working memory, verbal and visual learning, reasoning/problem-solving, and social cognition, as available in the selected instrument. Instructions, test conditions, and scoring procedures were kept consistent between groups. Testing was postponed if the participant was unable to engage adequately and was attempted again when clinically appropriate.
Sample Processing
No biological specimens were collected or processed because the study assessed clinical and cognitive variables only. Completed assessment forms were checked for missing entries, coded using unique study identification numbers, and entered into a password-protected database. Identifying information was stored separately from the analysis dataset.
Statistical Methods
Data were analysed using SPPS 28.0version. Continuous variables were summarized as mean and standard deviation or median and interquartile range, as appropriate. Categorical variables were summarized as frequencies and percentages. Between-group comparisons of continuous cognitive scores were performed using the independent-samples t-test when assumptions of normality and variance were met; otherwise, the Mann–Whitney U test was used. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate.
Where applicable, analysis of covariance or multiple linear regression was used to compare cognitive scores while adjusting for relevant factors such as age, sex, education, duration of illness, and clinical symptom status. Effect estimates were reported with 95% confidence intervals. All statistical tests were two-sided, and a p value below 0.05 was considered statistically significant.
Data Collection
Data were collected using a predesigned, structured case record form. The form included the participant’s study identification number, age, sex, education, occupation, marital status, diagnosis, illness duration, age at onset, treatment details, relevant medical and psychiatric history, and clinical status at assessment. Cognitive assessment scores were recorded for the total score and for each domain provided by the selected instrument. The investigator reviewed each form at the end of the assessment to identify omissions and inconsistencies. Data were entered and verified before statistical analysis.
Table 1. Overall comparison of cognitive function between diagnostic groups
|
Measure |
Schizophrenia (n=70) |
Bipolar disorder (n=70) |
Between-group estimate (95% CI) |
Test statistic |
p value |
|
Global cognitive composite, T-score, Mean (SD) |
34.8 (8.2) |
40.7 (8.5) |
Mean difference (SZ-BD): -5.90 (-8.69 to -3.11) |
Welch’s t=-4.18 |
<0.001 |
|
Cognitive impairment, n (%) |
58 (82.9%) |
39 (55.7%) |
OR 3.84 (1.76 to 8.38) |
χ²=12.12 |
<0.001 |
The mean global cognitive composite T-score was 34.8 ± 8.2 among patients with schizophrenia and 40.7 ± 8.5 among patients with bipolar disorder. The mean difference between groups was -5.90 points (95% CI: -8.69 to -3.11), with significantly lower scores in the schizophrenia group (Welch’s t = -4.18, p < 0.001). Using a T-score below 40 as the threshold for cognitive impairment, impairment was present in 58 (82.9%) participants with schizophrenia and 39 (55.7%) with bipolar disorder. The odds of impairment were higher in the schizophrenia group (OR = 3.84, 95% CI: 1.76–8.38; χ² = 12.12, p < 0.001).
Table 2. Cognitive function among patients with schizophrenia
|
Cognitive domain |
Mean T-score (SD) |
95% CI for mean |
Difference from normative mean of 50 (95% CI) |
One-sample t statistic |
p value |
|
Global cognitive composite |
34.8 (8.2) |
32.84 to 36.76 |
-15.20 (-17.16 to -13.24) |
t=-15.51 |
<0.001 |
|
Processing speed |
35.1 (8.6) |
33.05 to 37.15 |
-14.90 (-16.95 to -12.85) |
t=-14.50 |
<0.001 |
|
Attention/vigilance |
36.2 (9.1) |
34.03 to 38.37 |
-13.80 (-15.97 to -11.63) |
t=-12.69 |
<0.001 |
|
Working memory |
37.0 (8.8) |
34.90 to 39.10 |
-13.00 (-15.10 to -10.90) |
t=-12.36 |
<0.001 |
|
Verbal learning |
34.0 (8.7) |
31.93 to 36.07 |
-16.00 (-18.07 to -13.93) |
t=-15.39 |
<0.001 |
|
Visual learning |
36.3 (8.9) |
34.18 to 38.42 |
-13.70 (-15.82 to -11.58) |
t=-12.88 |
<0.001 |
|
Reasoning/problem solving |
37.1 (8.2) |
35.14 to 39.06 |
-12.90 (-14.86 to -10.94) |
t=-13.16 |
<0.001 |
|
Social cognition |
38.5 (9.0) |
36.35 to 40.65 |
-11.50 (-13.65 to -9.35) |
t=-10.69 |
<0.001 |
Among patients with schizophrenia, the mean scores for the global cognitive composite and all assessed domains were below the assumed normative T-score of 50. The global composite mean was 34.8 (95% CI: 32.84–36.76), corresponding to a difference of -15.20 points from the normative value (95% CI: -17.16 to -13.24; t = -15.51, p < 0.001). Verbal learning had the lowest mean score at 34.0, while social cognition had the highest at 38.5. Scores in processing speed, attention/vigilance, working memory, visual learning, and reasoning/problem solving were also below the normative value; all one-sample comparisons were statistically significant (p < 0.001).
Table 3. Cognitive function among patients with bipolar disorder
|
Cognitive domain |
Mean T-score (SD) |
95% CI for mean |
Difference from normative mean of 50 (95% CI) |
One-sample t statistic |
p value |
|
Global cognitive composite |
40.7 (8.5) |
38.67 to 42.73 |
-9.30 (-11.33 to -7.27) |
t=-9.16 |
<0.001 |
|
Processing speed |
41.3 (9.0) |
39.15 to 43.45 |
-8.70 (-10.85 to -6.55) |
t=-8.09 |
<0.001 |
|
Attention/vigilance |
42.0 (9.3) |
39.78 to 44.22 |
-8.00 (-10.22 to -5.78) |
t=-7.19 |
<0.001 |
|
Working memory |
42.4 (9.0) |
40.25 to 44.55 |
-7.60 (-9.75 to -5.45) |
t=-7.06 |
<0.001 |
|
Verbal learning |
39.8 (9.1) |
37.63 to 41.97 |
-10.20 (-12.37 to -8.03) |
t=-9.38 |
<0.001 |
|
Visual learning |
41.2 (9.1) |
39.03 to 43.37 |
-8.80 (-10.97 to -6.63) |
t=-8.09 |
<0.001 |
|
Reasoning/problem solving |
42.0 (8.6) |
39.95 to 44.05 |
-8.00 (-10.05 to -5.95) |
t=-7.79 |
<0.001 |
|
Social cognition |
42.8 (9.1) |
40.63 to 44.97 |
-7.20 (-9.37 to -5.03) |
t=-6.62 |
<0.001 |
Among patients with bipolar disorder, scores across all assessed cognitive domains were also below the assumed normative T-score of 50. The global cognitive composite mean was 40.7 (95% CI: 38.67–42.73), with a difference of -9.30 points from the normative value (95% CI: -11.33 to -7.27; t = -9.16, p < 0.001). Verbal learning had the lowest domain mean at 39.8, whereas social cognition had the highest at 42.8. The one-sample comparisons showed that each domain mean was significantly below the assumed normative value (p < 0.001).
Table 4. Between-group comparison of cognitive scores by domain
|
Cognitive domain |
Schizophrenia (n=70), Mean (SD) |
Bipolar disorder (n=70), Mean (SD) |
Mean difference (SZ-BD), 95% CI |
Welch’s t statistic |
p value |
|
Global cognitive composite |
34.8 (8.2) |
40.7 (8.5) |
-5.90 (-8.69 to -3.11) |
t=-4.18 |
<0.001 |
|
Processing speed |
35.1 (8.6) |
41.3 (9.0) |
-6.20 (-9.14 to -3.26) |
t=-4.17 |
<0.001 |
|
Attention/vigilance |
36.2 (9.1) |
42.0 (9.3) |
-5.80 (-8.88 to -2.72) |
t=-3.73 |
<0.001 |
|
Working memory |
37.0 (8.8) |
42.4 (9.0) |
-5.40 (-8.37 to -2.43) |
t=-3.59 |
<0.001 |
|
Verbal learning |
34.0 (8.7) |
39.8 (9.1) |
-5.80 (-8.78 to -2.82) |
t=-3.85 |
<0.001 |
|
Visual learning |
36.3 (8.9) |
41.2 (9.1) |
-4.90 (-7.91 to -1.89) |
t=-3.22 |
0.002 |
|
Reasoning/problem solving |
37.1 (8.2) |
42.0 (8.6) |
-4.90 (-7.71 to -2.09) |
t=-3.45 |
0.001 |
|
Social cognition |
38.5 (9.0) |
42.8 (9.1) |
-4.30 (-7.32 to -1.28) |
t=-2.81 |
0.006 |
Comparison of cognitive scores by domain showed lower mean scores in the schizophrenia group than in the bipolar disorder group across all domains. The largest mean difference was observed in processing speed (-6.20 points, 95% CI: -9.14 to -3.26; t = -4.17, p < 0.001). Significant differences were also observed for the global cognitive composite, attention/vigilance, working memory, verbal learning, visual learning, reasoning/problem solving, and social cognition. The smallest difference was in social cognition (-4.30 points, 95% CI: -7.32 to -1.28; t = -2.81, p = 0.006).
Table 1. Overall comparison of cognitive function between diagnostic groups
In the results, the mean global cognitive composite score was lower in the schizophrenia group than in the bipolar disorder group (34.8 ± 8.2 vs 40.7 ± 8.5), with a statistically significant mean difference of −5.90 points. Cognitive impairment, using the table’s threshold of T-score <40, was also more frequent in schizophrenia (82.9% vs 55.7%). This pattern is consistent with Bortolato et al. (2015)[1], who concluded that cognitive deficits occur in both disorders but are generally more severe in schizophrenia. Trotta et al. (2015)[2] similarly reported moderate post-onset impairment in bipolar disorder and greater impairment in schizophrenia. In a meta-analysis of 12 comparative studies, Li et al. (2020)[3] found lower MCCB total scores in schizophrenia than in bipolar disorder. Lynham et al. (2018)[4] also reported a gradient of impairment across diagnostic groups, with greater impairment in schizophrenia.
The difference should not be assumed to apply equally to every patient or clinical state. Ceylan et al. (2020)[5] found that some cognitive differences varied with symptomatic versus remitted status. Jiménez-López et al. (2022)[6] reported no significant neurocognitive differences between remitted schizophrenia patients and euthymic patients with bipolar I disorder, despite poorer functioning among non-remitted schizophrenia patients. In contrast, Service et al. (2020)[7] found similar cognitive impairments in bipolar I disorder and schizophrenia in a large cross-diagnostic sample, with performance also related to symptom dimensions. A 2024 comparative study of remitted patients by Selvameenatchi et al. (2024)[8] likewise highlights the relevance of remission status when interpreting group differences. Thus, the difference in Table 1 aligns with several reviews but should be interpreted alongside symptom severity, mood state, illness history, education, medication, and the precise cognitive instrument used.
Table 2. Cognitive function among patients with schizophrenia
The schizophrenia group scored below the assumed normative mean of 50 in every listed domain. The lowest mean was observed in verbal learning (34.0), followed by processing speed (35.1); social cognition had the highest mean (38.5), but remained below the reference value. The pattern is broadly consistent with Bo et al. (2017)[9], whose MCCB meta-analysis found deficits across global cognition and all seven domains in bipolar disorder, and with the broader literature showing multi-domain impairment in schizophrenia. Gebreegziabhere et al. (2022)[14] found that reviews most frequently reported deficits in processing speed, verbal memory, and working memory in schizophrenia. European Psychiatric Association guidance by Vita et al. (2022)[15] recommends comprehensive assessment of both neurocognitive and social cognitive domains. McCutcheon et al. (2023)[16] and Tschentscher et al. (2023)[17] also describe broad cognitive deficits involving memory, attention, processing speed, and executive functions.
The one-sample tests in Table 2 show statistically significant differences from 50, but these tests only support a comparison with that reference value if the chosen instrument provides appropriately standardized T-scores and its norms are suitable for the participants. A comparison with an assumed norm is not equivalent to a concurrently recruited healthy-control comparison. In addition, a lower group mean does not mean that every participant had impairment in every domain. The table’s domain means and the impairment threshold should therefore be interpreted according to the instrument manual and the study’s prespecified scoring plan.
Table 3. Cognitive function among patients with bipolar disorder
The bipolar disorder group also scored below the assumed normative mean in all domains. Verbal learning had the lowest mean (39.8), whereas social cognition had the highest (42.8). This pattern is consistent with Bo et al. (2017)[9], who reported deficits in global cognition and across MCCB domains in bipolar disorder, including processing speed, attention, working memory, and learning. Miskowiak et al. (2017)[11] and Miskowiak et al. (2018)[10] emphasized the need to assess cognition in bipolar disorder and to account for clinical state, medication, and comorbidity. Samamé et al. (2022)[12] reviewed longitudinal data and found that cognitive deficits in bipolar disorder may persist without showing clear progressive deterioration, while Nicoloro-SantaBarbara et al. (2023)[13] described cognitive difficulties that can continue across mood states and affect functioning.
The scores in Table 3 should not be interpreted as showing that bipolar disorder invariably produces impairment of the same degree or profile. Bipolar cognitive performance may vary with mania, depression, euthymia, psychotic symptoms, number of episodes, and treatment. Flaaten et al. (2024)[20], in a ten-year follow-up study, reported poorer performance than controls in several domains but also found stability or improvement over time in some cognitive measures. Ceylan et al. (2020)[5] similarly found that illness state influenced some aspects of performance. These findings reinforce the importance of reporting participants’ clinical state and treatment characteristics alongside cognitive scores.
Table 4. Between-group comparison of cognitive scores by domain
In the domain-wise comparison, mean scores were lower in the schizophrenia group across all domains. The largest mean difference was in processing speed (−6.20 points), followed by attention/vigilance (−5.80), verbal learning (−5.80), and working memory (−5.40). The smallest difference was in social cognition (−4.30), although it remained statistically significant. This overall direction agrees with Li et al. (2020)[3], whose comparative meta-analysis found lower scores in schizophrenia across all seven MCCB domains, including processing speed, working memory, verbal and visual learning, reasoning, attention, and social cognition. Lynham et al. (2022)[18] also described an increase in impairment across the mood/psychosis spectrum from bipolar disorder toward schizophrenia. Zanelli et al. (2022)[19] found deficits in both disorders after first episode, with patterns varying by domain and illness course.
The comparative literature is not entirely uniform. Remission-based comparisons, including Jiménez-López et al. (2022)[6], have found that cognitive performance in euthymic bipolar disorder can be similar to that in remitted schizophrenia, while Service et al. (2020)[7] found comparable impairment in bipolar I disorder and schizophrenia on some measures. Differences between studies may reflect variations in diagnostic subtype, current symptoms, remission criteria, sample characteristics, instruments, and adjustment for confounders. The negative mean differences in Table 4 indicate lower scores in schizophrenia because the contrast is calculated as schizophrenia minus bipolar disorder; they do not demonstrate that diagnosis alone caused the difference.
In this comparison, cognitive performance was lower in patients with schizophrenia than in patients with bipolar disorder, both for the global cognitive composite and across the assessed domains. Cognitive scores in both groups were below the assumed normative T-score of 50. The findings suggest that cognitive difficulties may be present in both disorders, with greater impairment in the schizophrenia group in this sample. Because the values are and the study is cross-sectional, these findings cannot establish causation or be generalized as actual study results.
LIMITATIONS OF STUDY
This study design has several limitations. A cross-sectional assessment captures cognitive function at one point in time and cannot determine how cognition changes with illness course or treatment. A hospital-based sample may not represent all people with schizophrenia or bipolar disorder. Cognitive performance may have been influenced by clinical state, illness duration, medication, education, premorbid ability, sleep, and medical or psychiatric comorbidities. Although the analysis describes differences between groups, residual confounding may remain. Comparison with a normative T-score of 50 is valid only if the chosen instrument provides appropriately standardized scores and its norms are suitable for the participants. The sample size and results presented here must be replaced with verified study data; until then, statistical significance and effect estimates should not be interpreted as observed findings.