Effect of psychotic symptoms on cognitive impairment in patients with bipolar disorder.

BACKGROUND: A considerable proportion of patients with bipolar disorder (BD) have psychotic symptoms during the illness. This subset of BD due to its genetic susceptibility and family segregation has considerable overlap with schizophrenia. However, the extent to which BD patients with psychotic symptoms and without psychotic symptoms differ on neurocognitive functions is still not completely clear. AIM: The aim of this study was to examine the neurocognitive functions in BD patients with psychotic symptoms and BD without psychotic symptoms in comparison with healthy volunteers (HVs).
MATERIALS AND METHODS: Thirty patients with Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition BD (16 with psychotic symptoms) and thirty age- and sex-matched HVs were recruited in the study. Clinical severity was assessed using structured rating scales. The presence of psychotic symptoms was assessed using the Lifetime Dimensions of Psychosis Scale (LDPS). All patients underwent tests, namely continuous performance test, Stroop Word-Color interference test, and Wisconsin Card Sorting Test, to measure executive functions. Differences between groups were examined using analysis of covariance with age and sex as covariates.
RESULTS: There was a significant difference between groups on the performance of all the three tests, with patients performing poorer than HVs. While the HVs differed from both BD with psychotic symptoms and without psychotic symptoms, there was no difference between BD patients with and without psychotic symptoms. There was no significant correlation between LDPS score and scores on neurocognitive tests.
CONCLUSION: The study findings, at least with respect to cognitive function, suggest that BD with psychotic symptoms may not be a categorically distinct subtype of BD. Cognitive functions need to be assessed in all patients with BD, regardless of psychotic symptoms, and remediation interventions need to be provided. Copyright:

Bipolar disorder (BD) is a chronic episodic illness, with lifetime estimates of 1.0% for BD-I and 1.1% for BD-II.[1] Several studies have reported neurocognitive impairments in patients with BD in the areas of executive function, verbal memory, psychomotor speed, and sustained attention.[234] These neurocognitive impairments are seen not only during acute mood episodes,[5] but are also present in varying intensity during euthymic periods.[6] A recent meta-analysis reported executive function, verbal memory, psychomotor speed, and sustained attention to be the most affected cognitive domains.[3]

A subset of patients with BD has psychotic symptoms in addition to mood symptoms during the course of illness.[7] This subset of BD due to its genetic susceptibility and family segregation has considerable overlap with schizophrenia.[8] Neuroimaging studies have delineated differences between BD patients with psychotic symptoms and without psychotic symptoms.[9] However, the extent to which BD patients with psychotic symptoms and without psychotic symptoms differ on neuropsychological functions is still controversial as studies have reported conflicting results; while some authors have reported poorer cognitive performance in BD patients with psychotic symptoms than in those without psychotic symptoms,[5710] others have reported absence of such difference[26111213] or a modest difference.[2] It is important to note that BD patients with psychotic symptoms at any time during illness are reported to have effect on the neurocognitive functioning in BD-I; that is, regardless of the cross-sectional status of the psychotic symptoms, the history of psychotic symptoms at any moment during the illness (present or past episodes) is considered in previous studies[71213] and is shown to have differences in neurocognitive functions compared to those without psychotic symptoms at any time during the illness.[712]

However, till date, no study has examined differences in neurocognitive functions between BD patients with and without psychotic symptoms from India. Considering the regional differences in BD prevalence and course, it is important to examine the nature of cognitive deficits from diverse cultural backgrounds.[141516] Moreover, as the presence of psychotic symptoms can influence the outcome of BD,[17] it is critical to examine the differences between BD patients with psychotic symptoms and without psychotic symptoms. Hence, in this study, we aimed to examine the effect of the presence of psychotic symptoms on cognitive deficits in BD.

MATERIALS AND METHODS

Thirty patients with BD were recruited from the psychiatry department of a hospital. Thirty age- and sex-matched healthy volunteers (HVs) were recruited from the same area by word of mouth and flyers. Written informed consent was obtained from the participants before they were enrolled in the study. All patients were examined using the Mini-International Neuropsychiatric Interview (MINI),[18] and those satisfying the Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition Text Revision (DSM-IV TR)[19] criteria for BD were included in the study. We used the DSM-IV TR as the study was conceptualized and started in the year 2013. Patients with concurrent comorbid DSM-IV TR axis-I disorder, head injury, and neurological disorders were excluded from the study. All HVs were also examined using the MINI-Plus, and those with current or past DSM-IV TR axis-I psychiatric disorders were excluded from the study. Those HVs with a history of axis-I disorder in first-degree relatives were also excluded from the study. None of the participants had substance dependence or harmful use. Handedness was assessed using the Edinburgh handedness inventory,[20] and we recruited only right-handed individuals. The study was approved by the institute's ethics committee. The patients continued to take their prescribed medications while participating in the study.

Clinical assessments

Relevant clinical information pertinent to illness, including age at illness onset, clinical details of previous bipolar episodes, family history, and treatment details, was collected from patient, informant, and case records by an experienced psychiatrist. The clinical severity of current mood symptoms was assessed using the Hamilton Depression Rating Scale (HDRS)[21] and Young Mania Rating Scale (YMRS).[22] We assessed the presence of psychotic symptoms, namely hallucinations and/or delusions within a mood episode, using the Lifetime Dimensions of Psychosis Scale (LDPS).[23] The LDPS allows to characterize the positive, negative, bizarre, and disorganized symptom factors in the lifetime of an individual in addition to mood-related symptoms. This instrument was used earlier while examining differences between psychotic and nonpsychotic BD.[24] Based on the LDPS, 16 patients were classified as having BD with psychotic symptoms and 14 as BD without psychotic symptoms. Out of the 30 patients, 17 had a prior history of depressive or manic episodes and 13 patients had one episode of mania. Twenty-two of them were treated with lithium, twenty with valproate, two with carbamazepine, and five with lamotrigine, and all participants had received atypical antipsychotics during the course of the illness.

Neuropsychological assessments

The following neuropsychological tests, measuring different domains of executive functions, were administered: (a) Continuous Performance Test-Identical pairs (CPT) to measure sustained attention,[25] (b) Stroop test to measure response inhibition,[26] and (c) Wisconsin card sorting test (WCST) to measure set shifting.[27] An experienced clinical psychologist (GA) administered the neuropsychological tests with a uniform environment and standard procedure to all participants. These tests are validated in the Indian population and have been used for both clinical and research purposes. For the Stroop effect score, the reading time and naming time were converted into seconds. The reading time was subtracted from the naming time to get the Stroop effect score.

Statistical analysis

Statistical analysis was done using the Statistical Package for Social Sciences software version 24 (IBM Corp., Armonk, NY, USA). Group differences between BD patients and HVs in age and sex were compared using independent t-test and Chi-square test, respectively. To examine the difference between BD patients and HVs in cognitive functions, Mann–Whitney U-test was conducted. To examine the effect of psychotic symptoms, analysis of covariance was conducted with age and sex as covariates between the following three groups: BD patients with psychotic symptoms, BD patients without psychotic symptoms, and HVs. Further, post hoc Bonferroni tests were conducted to examine the differences between pairs. To examine the relation between clinical variables and cognitive functions, Pearson's correlation test was conducted between scores on the LDPS, YMRS, HDRS, Simpson Angus scale, and age at onset of illness. Considering the multiple correlation analysis conducted, a conservative P < 0.01 was considered statistically significant.

RESULTS

There was no statistically significant difference between BD patients (M: F – 18:12) and HVs (M: F – 18:12) in sex ratio (P = 0.604) or age (BD patients – 24.73 ± 6.76 years; HVs – 24.53 ± 7.00 years; t = 0.113; P = 0.911). Among patients with BD, the average duration of illness was 3.43 ± 3.67 years, the average age at the onset of illness was 21.3 ± 5.05 years, and the average number of episodes was 2.4 ± 1.71. BD patients with psychotic symptoms, BD patients without psychotic symptoms, and HVs did not differ on age and sex distribution. There was no significant difference between BD patients with psychotic symptoms and BD patients without psychotic symptoms in age at onset, average number of episodes, YMRS score, or HDRS score. However, there was a significant difference between the two in duration of illness; BD patients with psychotic symptoms had longer illness duration than those without psychotic symptoms [Table 1].

Table 1

Demographic and clinical characteristics of bipolar disorder patients with psychotic symptoms, bipolar disorder patients without psychotic symptoms, and healthy volunteers

Characteristics	BDP+ (n=16)	BDP− (n=14)	HV (n=30)	F/t/χ2	P
Sex ratio (male:female)	11:5	7:7	18:12	1.09	0.58
Age (mean±SD)	26.00±5.76	23.29±7.70	24.53±7.00	0.59	0.56
Years of education	13.56±2.85	13.28±2.23	13.77±2.31	0.18	0.83
Duration of illness (years)	4.68±4.23	2.00±2.18	-	2.13	0.04
Age at onset (years)	21.31±3.62	21.28±6.47	-	0.01	0.98
Average number of episodes	2.87±1.70	1.85±1.61	-	1.67	0.10
MMSE score	29.37±1.25	29.57±1.08	30.00±0.00	3.31	0.04
YMRS	6.56±12.23	5.07±8.62	-	0.38	0.70
HDRS	2.93±4.04	1.21±3.09	-	1.29	0.20
LDPS	25.06±4.99	1.35±5.07		12.88	<0.001

BDP+ – Bipolar disorder with psychotic symptoms; BDP− – Bipolar disorder without psychotic symptoms; HV – Healthy volunteers; F – ANOVA; t – Independent t-test; SD – Standard deviation; YMRS – Young’s Mania Rating Scale; HDRS – Hamilton Depression Rating Scale; LDPS – Lifetime Dimensions of Psychosis Scale; MMSE – Mini–Mental State Examination; ANOVA – Analysis of variance

Difference between bipolar disorder patients and healthy volunteers on cognitive functions

Patients with BD had significant cognitive impairment in comparison to HV: (a) in the CPT, BD patients had significantly higher number of wrong responses, missed responses, and lower number of correct responses; (b) in the Stroop test, BD patients had significantly higher Stroop effect time; (c) in the WCST, BD patients had higher number of errors, perseverative responses, perseverative errors, nonperseverative errors, and lower number of categories completed. However, there was no difference in CPT – response time, WCST – trials to complete first category, and WCST – failure to maintain set [Table 2].

Table 2

Group differences between bipolar disorder patients and healthy volunteers on cognitive functions

Cognitive test measure	BD patients (n=30)	HVs (n=30)	Mann-Whitney U-test	P
CPT
Correct responses	25.97±10.11	35.83±5.4	174	<0.001
Wrong responses	40.16±55.08	7±4.2	142	0.002
Missed responses	18.13±10.24	8.1±5.6	176	<0.001
Response time	0.59±0.48	0.49±0.44	343	0.43
Stroop
Effect	215.33±149.73	119.7±40.7	207	0.001
Percentile	33.70±29.70	66.3±31.4	192	<0.001
WCST
Total number of correct responses	69.36±13.83	70.8±10.8	445	0.64
Total number of errors	49.20±22.02	25.2±20.4	179	<0.001
Perseverative response	29.06±23.06	11.8±9	166	<0.001
Perseverative errors	24.66±17.10	10.7±7.9	161	<0.001
Nonperseverative errors	24.56±16.14	15.1±13.9	254	0.019
Categories completed	3.86±1.88	5.4±1.33	228	<0.001
Trials to complete first category	22.96±15.49	20±11.7	438	0.40
Failure to maintain set	1.06±1.41	0.76±1.16	397	0.37

CPT – Continuous performance test; WCST – Wisconsin card sorting test; BD – Bipolar disorder; HV – Healthy volunteers

Difference among psychotic patients versus nonpsychotic patients versus healthy volunteers on cognitive function

There was a significant difference between the three groups on cognitive tests; on CPT, a significant group effect was seen in the number of correct responses, wrong responses, and missed responses but not in response time. Similarly, significant difference was seen in Stroop effect time as well as Stroop-performance percentile. In the WCST, significant difference was seen in total errors, perseverative responses, perseverative errors, and categories completed. There was no significant difference between the three groups in nonperseverative errors, trials to complete the first category, and failure to maintain set [Table 3]. On post hoc analysis, there were significant differences between HVs and both BD patients with psychotic symptoms and BD patients without psychotic symptoms. However, there was no significant difference between BD patients with and without psychotic symptoms [Table 4].

Table 3

Group difference between bipolar disorder patients with psychotic symptoms, bipolar disorder patients without psychotic symptoms, and healthy volunteers on cognitive functions

Cognitive test measure	BDP− (n=14)	BDP+ (n=16)	HVs (n=30)	F	P
CPT
Correct responses	28.00±7.76	24.18±11.75	35.83±5.4	11.99	<0.001
Wrong responses	50.78±71.09	30.87±35.89	7±4.2	6.24	0.004
Missed responses	16.21±8.21	19.81±11.74	8.1±5.6	11.75	<0.001
Response time	0.69±0.65	0.51±0.24	0.49±0.44	0.77	0.46
Stroop – Effect	219.00±186.45	211.75±114.85	119.7±40.7	5.84	0.005
WCST
Total number of correct responses	68.64±11.85	70.00±15.73	70.8±10.8	0.13	0.88
Total number of errors	45.07±23.30	52.81±20.91	25.2±20.4	10.65	<0.001
Perseverative response	24.35±16.32	33.18±27.54	11.8±9	9.37	0.001
Perseverative errors	21.14±12.83	27.75±20.03	10.7±7.9	10.38	<0.001
Nonperseverative errors	24.00±16.23	25.06±16.56	15.1±13.9	2.96	0.060
Categories completed	4.35±1.90	3.43±1.82	5.4±1.33	8.67	0.001
Trials to complete first category	19.57±11.09	25.93±18.36	20±11.7	1.25	0.29
Failure to maintain set	0.57±1.01	1.50±1.59	0.76±1.16	1.86	0.165

BD – Bipolar disorder; BDP+ – BD with psychotic symptoms; BDP− – BD without psychotic symptoms; HVs – Healthy volunteers; F – Analysis of covariance with age and sex as covariates; CPT – Continuous performance test; WCST – Wisconsin Card Sorting test

Table 4

Mean differences between the three groups, namely bipolar disorder patients with psychotic symptoms, bipolar disorder patients without psychotic symptoms, and healthy volunteers on cognitive measures

Group 1	Group 2	Mean difference	SE	P
CPT correct responses
BDP−	BDP+	3.8125	2.95907	0.608
HV	BDP−	7.8333	2.61710	0.012
HV	BDP+	11.6458	2.50310	<0.001
CPT wrong responses
BDP−	BDP+	19.9107	14.17832	0.497
HV	BDP−	−43.7857	12.53978	0.003
HV	BDP+	−23.8750	11.99351	0.154
CPT missed responses
BDP−	BDP+	−3.5982	3.01635	0.714
HV	BDP−	−8.0810	2.66776	0.011
HV	BDP+	−11.6792	2.55154	<0.001
CPT response time
BDP−	BDP+	0.1823	0.17013	0.865
HV	BDP−	−0.1931	0.15047	0.613
HV	BDP+	−0.0108	0.14391	1.000
Stroop effect
BDP−	BDP+	7.2500	40.49399	1.000
HV	BDP−	−99.3000	35.81423	0.022
HV	BDP+	−92.0500	34.25408	0.028
WCST total number of correct responses
BDP−	BDP+	−1.3571	4.58183	1.000
HV	BDP−	2.2238	4.05232	1.000
HV	BDP−	0.8667	3.87579	1.000
WCST total number of errors
BDP−	BDP+	−7.7411	7.77596	0.971
HV	BDP−	−19.8714	6.87732	0.016
HV	BDP+	−27.6125	6.57773	<0.001
WCST perseverative response
BDP−	BDP+	−8.8304	6.35651	0.511
HV	BDP−	−12.5571	5.62191	0.088
HV	BDP+	−21.3875	5.37700	0.001
WCST perseverative errors
BDP−	BDP+	−6.6071	4.84123	0.533
HV	BDP−	−10.4429	4.28175	0.054
HV	BDP+	−17.0500	4.09522	<0.001
WCST nonperseverative errors
BDP−	BDP+	−1.0625	5.55837	1.000
HV	BDP−	−8.8333	4.91601	0.233
HV	BDP+	−9.8958	4.70185	0.119
WCST categories completed
BDP−	BDP+	0.9196	0.59093	0.376
HV	BDP−	1.1095	0.52264	0.114
HV	BDP+	2.0292	0.49987	<0.001
WCST trials to complete first category
BDP−	BDP+	−6.3661	5.00337	0.625
HV	BDP−	0.4286	4.42515	1.000
HV	BDP+	−5.9375	4.23238	0.498
WCST failure to maintain set
BDP−	BDP+	−0.9286	0.46188	0.147
HV	BDP−	0.1952	0.40850	1.000
HV	BDP+	−0.7333	0.39071	0.197

BDP+ – BD with psychotic symptoms; BDP− – BD without psychotic symptoms; HVs – Healthy volunteers; P – Post hoc Bonferroni analysis following ANOVA; ANOVA – Analysis of variance

Relation between clinical variables and cognitive functions

There was a significant positive correlation between YMRS score and WCST – nonperseverative errors and trials to complete the first category, indicating that those with higher severity of mania had higher errors on the WCST. There was no significant relation between performance on cognitive functions and other clinical variables, namely LDPS score, HDRS score, Simpson Angus scale score, and age at onset of illness [Table 5].

Table 5

Correlation between clinical variables and performance on cognitive functions

Cognitive test measure	Significance	LDPS score	YMRS score	HDRS score	Simpson Angus scale core	Age at onset of illness
CPT correct responses	r	−0.207	−0.291	−0.038	−0.021	−0.025
	P	0.273	0.119	0.842	0.914	0.894
CPT wrong responses	r	−0.177	−0.002	−0.231	0.077	−0.110
	P	0.351	0.993	0.219	0.687	0.562
CPT missed responses	r	0.192	0.270	0.039	0.030	0.016
	P	0.309	0.149	0.838	0.877	0.934
CPT response time	r	0.086	0.005	0.064	0.331	−0.036
	P	0.651	0.978	0.737	0.074	0.849
Stroop effect	r	0.006	0.012	−0.130	−0.238	−0.095
	P	0.974	0.950	0.493	0.206	0.619
WCST total number of correct responses	r	−0.048	−0.340	0.150	0.131	00.400
	P	0.800	0.066	0.428	0.489	00.028
WCST total number of errors	r	0.247	0.438	−0.097	0.139	−0.390
	P	0.188	0.015	0.611	0.464	0.033
WCST perseverative response	r	0.134	0.013	−0.057	0.150	−0.278
	P	0.479	0.945	0.764	0.428	0.137
WCST perseverative errors	r	0.130	0.007	−0.011	0.173	−0.302
	P	0.494	0.969	0.955	0.360	0.104
WCST nonperseverative errors	r	0.197	0.589	−0.116	0.005	−0.211
	P	0.296	0.001	0.540	0.980	0.263
WCST categories completed	r	−0.309	−0.387	0.097	−0.174	0.246
	P	0.097	0.035	0.610	0.358	0.190
WCST trials to complete first category	r	0.165	0.519	−0.085	0.297	0.103
	P	0.383	0.003	0.654	0.112	0.586
WCST failure to maintain set	r	0.280	0.021	0.297	0.206	0.455
	P	0.134	0.910	0.111	0.274	0.011

CPT – Continuous performance test; WCST – Wisconsin card sorting test; YMRS – Young’s Mania Rating Scale; HDRS – Hamilton Depression Rating Scale; LDPS – Lifetime Dimensions of Psychosis Scale

DISCUSSION

The study findings report that BD-I patients have significant cognitive impairment in executive functions in comparison to age- and sex-matched healthy controls. This impairment in executive function is consistent with prior reports[5] and suggests neural disruption in prefrontal network. The study findings also suggest that there was no significant effect of psychotic symptoms on the cognitive functions as BD patients with psychotic symptoms and without psychotic symptoms were not significantly different. Our findings suggest that BD patients with psychotic symptoms may not be a categorically distinct subtype compared to BD patients without psychotic symptoms. Several studies have reported significant overlap between BD patients with psychotic symptoms and schizophrenia using genetic, neuroimaging, and electrophysiological assessment.[8] However, whether the neurobiology of BD patients with psychotic symptoms and without psychotic symptoms is different is not conclusive and yet to be explored in detail.

Our findings of absence of difference between patients with and without psychotic symptoms are similar to those of earlier studies,[28293031] but differ from those of others.[25732] Several differences may explain this discrepancy in findings. First, a few previous studies included both Type I and Type II BD,[323334] but we included only BD Type I to decrease the heterogeneity. Second, all participants in our study were on treatment with antipsychotics in comparison with earlier studies, considering that atypical antipsychotics are now a first-line agent for the treatment of both phases of BD, mania, and depression.[35] As antipsychotics and anticholinergics are known to cause cognitive deficits,[36] one cannot rule out the possibility of the confounding effect of antipsychotics. Third, the participants in this study were enrolled from a peripheral center, with a considerable proportion of patients being the first contact. As majority of the previous studies recruited patients from university hospitals, it is possible that the BD patients with psychosis recruited in earlier studies were more severely ill than those of the present study. Fourth, a considerable number of patients were in the first episode of mania in the present study. One cannot rule out the possibility that the cognitive deficits may worsen with longer duration of illness as the number of manic episodes had significant influence on the cognitive deficits in a previous meta-analysis.[2] Fifth, the two subgroups were matched on several illness variables such as age at onset of illness and number of episodes, which may influence the cognitive performance. Finally, we examined selected cognitive domains, and it is possible that the BD patients with psychosis may differ in other cognitive domains such as verbal memory.[2]

The study findings need to be considered in the background of the following limitations. First, the sample size was small. While the sample size was adequate to compare BD versus HV, the sample size was small to compare the subgroups. Hence, one cannot rule out type II statistical error for the absence of difference between BD patients with and without psychotic symptoms and our findings need to be considered preliminary, awaiting replication in a larger sample. Second, all patients were on treatment with antipsychotics at some point of time during illness. Although the effect of antipsychotics on cognitive performance in BD is controversial,[7] it is important to examine those patients who have not received treatment with antipsychotics. While such a study is ideal and required for definitive answers, inclusion of patients on antipsychotics increased the generalizability of our findings as atypical antipsychotics are considered first-line treatment for both depressive and manic phases.[36] Third, in view of the smaller number of participants, we were not able to do further analysis between those with mood-incongruent and mood-congruent psychotic symptoms. Future studies with a larger sample will be able to address this question. While these limitations need to be considered before the interpretation of results, it is important to also note the strengths of the study. BD patients and HVs were individually matched on age, sex, and education, controlling for potential confounding effects. The study participants were recruited from a hospital which caters to both first-contact and referred patients. Recruitment of patients from a nontertiary care hospital increased the generalizability significantly compared to earlier studies.

CONCLUSION

To summarize the study findings suggest that (a) BD patients have significant cognitive deficits compared to HVs, (b) there was no significant difference between BD patients with psychotic symptoms and BD patients without psychotic symptoms, and (c) both subgroups have significant deficits compared to HVs. The study findings provide further support to the existing literature that BD patients with psychotic symptoms may not be a distinct subtype. Regardless of the presence or absence of psychotic symptoms, cognitive functions need to be assessed in all patients with BD and intensive remediation interventions need to be provided. There is an acute need for novel pharmacological interventions for the treatment of cognitive deficits in patients with BD. The study findings also warrant future studies to include patients from the community and not from tertiary centers alone to increase generalizability.

Financial support and sponsorship

This study was supported by grant from the Medical Research Council of Maharashtra under “STAR Research Project” by the Government of Maharashtra, India.

Conflicts of interest

There are no conflicts of interest.