Correlates and management of comorbid anxiety disorders in schizophrenia.

BACKGROUND: Only a few studies have examined the treatability of anxiety disorders in schizophrenia, even though it is generally accepted that in the absence of schizophrenia, the anxiety disorders are safely and effectively treatable. AIM: The aim of this study was to study the relation of anxiety disorders with the positive and negative symptoms of schizophrenia and the effect of treatment of different anxiety disorders in schizophrenia patients.
MATERIALS AND METHODS: The study was carried out on inpatients of a tertiary care psychiatric hospital using a purposive sampling technique. The schizophrenia patients were evaluated for psychopathology and the presence of anxiety disorder at baseline. After being prescribed with antipsychotic medication in a suitable dose for 8 weeks, they were followed up at monthly intervals for the course of both schizophrenia and anxiety disorders. Thereafter, an selective serotonin reuptake inhibitor (SSRI) was also prescribed to the schizophrenia patients with comorbid anxiety disorder, and the patients were again followed up for a period of 8 weeks to assess the progress of schizophrenia and anxiety disorder.
RESULTS: The prevalence of anxiety disorder in 93 schizophrenia patients included in the present study was 45.16%. The most common comorbid anxiety disorders in schizophrenia patients were panic disorder (18.27%), social anxiety disorder (9.68%), obsessive-compulsive disorder (8.60%), and agoraphobia (6.45%). Schizophrenia patients with anxiety disorder had a significantly higher positive score of the Positive and Negative Symptom Scale for Schizophrenia (PANSS) and a significantly lower score on the negative scale and the general psychopathology scale of the PANSS, as compared to the scores of the schizophrenia group without anxiety disorders. Schizophrenia patients with anxiety disorders responded well to the combination of SSRIs and antipsychotics but not antipsychotics alone.
CONCLUSIONS: Comorbid anxiety disorders are common in schizophrenia. Schizophrenia patients with anxiety disorders differ significantly from those without anxiety disorders in their basic psychopathology. These anxiety disorders are quite responsive to the SSRIs but not to antipsychotics alone. Further, there is a shorter duration of illness in schizophrenia patients with anxiety disorders as compared to schizophrenia patients without anxiety disorders assigning a prognostic significance to the presence of comorbid anxiety disorders in schizophrenia.

In psychiatry, the conventional systems of diagnosis with their hierarchical assumptions have played a role in diverting attention from assigning dual diagnosis for co-occurring syndromes. The basic assumption shared by hierarchical system of classification is that there is a hierarchy of diagnoses and that diagnoses that are higher on the hierarchy subsume diagnosis lower on the hierarchy. Therefore, it was assumed that the diagnosis of schizophrenia could somehow explain or account for the presence of lower diagnosis such as anxiety disorders in patients diagnosed with schizophrenia. This assumption is connected to diagnostic reductionism, that is, the tendency to reduce all of the signs and symptoms shown by persons with schizophrenia to schizophrenia alone. Such reductionism undoubtedly contributes to the widespread tendency to treat schizophrenia as if it were a single unitary disorder and thus prevent clinicians and researchers from paying adequate attention to the accessory symptomatology shown by these patients outside of the core symptoms. The newer editions of classificatory systems have made some progress in recognizing these comorbid syndromes, but remnants of the old thinking persist. These remnants may continue to interfere with the recognition of comorbid anxiety disorders in schizophrenia and their therapeutic implications.[123]

As a result of these advances in recent years, there have been consistent reports of increased prevalence of comorbid anxiety disorders in schizophrenia.[45678] Despite this, little work has been done to establish the clinical validity and practical utility of this phenomenon. Of special concern is that only a few studies have examined the treatability of anxiety disorders in schizophrenia, even though it is generally accepted that in the absence of schizophrenia, the anxiety disorders are safely and effectively treatable. Findings from small preliminary case studies suggest that panic attacks, obsessive-compulsive disorder (OCD), and posttraumatic stress disorder (PTSD) in the presence of schizophrenia may be treatable as well. In view of the paucity of studies in this area, the present study was undertaken to assess the relation between anxiety disorders and the symptomatology of schizophrenia, that is, relation with the positive and negative symptoms. In addition, the response of anxiety disorders to medication (antipsychotics and selective serotonin reuptake inhibitors [SSRIs]) was also evaluated.

MATERIALS AND METHODS

This prospective, hospital-based study conducted at Ranchi Institute of Neuropsychiatry and Allied Sciences (RINPAS), Kanke, Ranchi, a 500-bedded postgraduate teaching psychiatric hospital and tertiary referral center for the patients with psychiatric disorders from the states of Jharkhand and Bihar. The study sample was selected using purposive sampling technique and comprised 93 consecutively admitted inpatients fulfilling ICD-10 DCR criteria for schizophrenia.[3] The study was conducted with the approval of the Ethics Committee of the Institute. The inclusion and exclusion criteria for the study group are given below.

Inclusion criteria for the study group

Inpatients of RINPAS Meeting ICD-10 DCR criteria for schizophrenia

Patients in the age range of 18–55 years

Patients who are drug naïve or off drugs for a period of 1 month or off depot preparation for at least 3 months

Cooperative for the interview.

Exclusion criteria for the study group

Patients not willing to give informed consent

History of mental retardation

History of substance abuse/neurological disorder/head injury/major physical illness.

After explaining the purpose of the study and obtaining informed consent from the schizophrenia patients, their sociodemographic characteristics were recorded on a specially prepared pro forma. The participants were then assessed using the Mini-International Neuropsychiatry Interview,[9] Brief Psychiatric Rating Scale (BPRS),[10] and Positive and Negative Symptom Scale for Schizophrenia (PANSS)[11] for a baseline assessment of psychopathology. Thereafter, the schizophrenia patients were divided into two groups: schizophrenia with anxiety disorder and schizophrenia without anxiety disorder. The schizophrenia patients with various anxiety disorders were assessed using the appropriate anxiety scales: Acute Panic Inventory,[12] Mobile Inventory for Agoraphobia,[13] Yale–Brown Obsessive-Compulsive Scale (YBOCS),[14] Liebowitz Social Anxiety Disorder Scale (LSADS),[15] Generalized Anxiety Disorder-7 Scale (GAD-7),[16] and Impact of Event Scale.[17]

All the schizophrenia patients were prescribed with appropriate antipsychotic medications (haloperidol, trifluoperazine, olanzapine, risperidone, and clozapine) in an adequate dose. They were then evaluated monthly using the above-mentioned scales for psychopathology and appropriate anxiety disorder scale as well as the objective and subjective observations of the patients for the course of both schizophrenia as well as the anxiety disorders symptoms. At the end of 8 weeks, an SSRI (fluoxetine) was prescribed in addition only to schizophrenia patients with anxiety disorders, while the schizophrenia patients without anxiety disorders were continued on their antipsychotic medications. Again, the course of the anxiety disorder and the effect of the same on the schizophrenia symptoms were evaluated for 8 weeks.

The obtained results were then tabulated for evaluation using the aid of computer program SPSS 17 (IBM, ISA). Analysis of general sociodemographic variables, clinical characteristics, and scores on psychopathology scales (BPRS scale and PANSS scale) were done using the t-test, Chi-square test, and Mann–Whitney U-test to identify significant differences between schizophrenia patients with and without anxiety disorders. The association between the different anxiety disorder scores at baseline and psychopathology scales (BPRS scale and PANSS scale) was obtained using Spearman's bivariate correlation. Comparison of psychopathology scores before and after treatment was assessed using the Wilcoxon test. In this study, a level of significance of <0.05 (two-tailed) was taken to consider a result statistically significant.

RESULTS

The study group of 93 schizophrenia patients consisted of 79 male patients and 14 female patients with a mean age of 31.30 (±8.27) years. There was no significant difference between schizophrenia patients with anxiety disorders and schizophrenia patients without anxiety disorders with respect to the sociodemographic variables [Table 1]. Comparison of the clinical characteristics of schizophrenia patients with and without anxiety disorders showed that the two groups differed significantly on duration of illness, the type of schizophrenia, and the drug status [Table 2]. There were no significant differences between the two groups on age of onset, mode of onset, precipitating factors, past history, and family history of psychiatric illness [Table 2]. The prevalence of anxiety disorder was 45.16% (n = 42) in the schizophrenia patients. Further analysis revealed that the prevalences of various types of anxiety disorders in schizophrenia patients were as follows: panic disorder 18.28% (n = 17), social anxiety disorder 9.68% (n = 9), OCD 8.60% (n = 8), agoraphobia 6.50% (n = 6), PTSD 1.08% (n = 1), and GAD 1.08% (n = 1).

Table 1

Comparison of sociodemographic characteristics of schizophrenia patients with and without anxiety disorders

Variables	Category	Schizophrenia with anxiety disorder (n=42), n (%)	Schizophrenia without anxiety disorders (n=51), n (%)	t/χ2
Age (years)		32.5±9.8	30.0±6.7	3.507
Gender	Male	33 (78.6)	39 (76.5)	3.911
	Female	9 (21.4)	12 (23.5)
Education	Matric and below	35 (83.3)	43 (84.3)	0.650
	Above matric	7 (16.7)	8 (15.7)
Occupation	Farmer	18 (42.9)	19 (37.2)	1.263
	Employed	6 (14.3)	9 (17.7)
	Homemaker	7 (16.7)	6 (11.8)
	Unemployed	11 (26.1)	17 (33.3)
Marital status	Unmarried	14 (33.3)	16 (31.4)	0.841
	Married	22 (52.4)	28 (54.9)
	Others	6 (14.3)	7 (13.7)
Religion	Hindu	33 (78.6)	38 (74.5)	0.850
	Others	9 (21.4)	13 (25.5)
Mother tongue	Hindi	29 (69.4)	36 (70.6)	1.738
	Others	13 (20.6)	15 (29.4)
Income	<3000	26 (61.9)	34 (66.7)	5.811
	3001-7000	7 (16.7)	13 (25.5)
	>7000	9 (21.4)	4 (7.8)
Residence	Rural	31 (73.8)	40 (78.4)	2.994
	Urban	11 (26.2)	11 (21.6)
Family type	Nuclear	13 (31.0)	18 (35.3)	0.700
	Joint	29 (69.0)	33 (64.7)

Table 2

Clinical characteristics of patients of schizophrenia with anxiety disorders and without anxiety disorders

Variables	Schizophrenia with anxiety disorders (n=42)	Schizophrenia without anxiety disorders (n=51)	t/χ2	P
Illness duration (years), mean±SD	5.14±5.25	9.22±5.38	t=26.903	0.01**
Age at onset (years), mean±SD	27.28±7.97	25.16±8.36	t=4.216	0.98
BPRS scores, mean±SD	45.25±3.69	46.57±3.59	t=1.723	0.10
Positive subscale of PANSS, mean±SD	24.86±3.54	21.20±5.24	t=4.001	0.01**
Negative subscale of PANSS, mean±SD	21.69±4.01	26.03±4.39	t=4.976	0.01**
General psychopathology scale of PANSS, mean±SD	43.06±3.75	45.38±3.59	t=3.027	0.01**
Diagnosis, n (%)
Paranoid	36 (85.7)	24 (47.1)	χ2=24.502	0.01**
Others	6 (14.3)	27 (52.9)
Mode of onset, n (%)
Abrupt	1 (2.4)	1 (2.0)	χ2=3.012	0.99
Acute	4 (9.5)	3 (5.9)
Insidious	37 (88.1)	47 (92.1)
Precipitating factor, n (%)
Present	7 (16.7)	7 (13.7)	χ2=7.393	0.764
Absent	35 (83.3)	44 (86.3)
Drug status, n (%)
Drug naïve	19 (45.2)	14 (27.5)	χ2=7.839	0.005**
Drug free	22 (54.8)	37 (72.5)
Past history of psychiatric illness, n (%)
Present	3 (7.1)	5 (9.8)	χ2=4.377	0.931
Absent	39 (92.6)	46 (90.2)
Family history of psychiatric illness, n (%)
Present	12 (28.6)	15 (23.5)	χ2=11.442	0.30
Absent	30 (71.4)	38 (74.5)
Comorbid anxiety disorder, n (%)
Panic disorder	17 (40.47)	-	-	-
Social anxiety disorder	9 (21.42)	-	-	-
OCD	8 (19.04)	-	-	-
Agoraphobia	6 (14.28)	-	-	-
PTSD	1 (2.38)	-	-	-
GAD	1 (2.38)	-	-	-

**Statistically significant at 0.01 level. SD - Standard deviation; BPRS - Brief Psychiatric Rating Scale; PANSS - Positive and Negative Symptom Scale for Schizophrenia; OCD - Obsessive-compulsive disorder; PTSD - Posttraumatic stress disorder; GAD - Generalized anxiety disorder

Overall, the schizophrenia group with anxiety disorder had a significantly higher positive score of the PANSS, a significantly lower negative score on PANSS, and a significantly lower score on the general psychopathology scale of the PANSS, as compared to the scores of the schizophrenia group without anxiety disorders. In contrast, the values obtained for the BPRS for the two groups were not significantly different [Table 2]. Distribution of psychopathology scores among various anxiety disorders is summarized in Table 3. There was a significant correlation between positive scores on PANSS and Liebowitz Social Anxiety Scale (LSAS) scores in social anxiety disorder. In agoraphobia patients, a significant correlation was noted between scores on mobility inventory and positive, negative, and general psychopathology of PANSS and the BPRS.

Table 3

Correlation of psychopathology scores and various anxiety disorder scores in schizophrenia patients with comorbid anxiety disorders

Type of anxiety disorder	PANSS			BPRS	Score on anxiety scale (name in parentheses)	Spearman correlation of anxiety scores with
	Positive score	Negative score	General psychopathology scale			Positive score	Negative score	General psychopathology scale	BPRS
Panic disorder (n=17)	26.12±2.55	21.0±3.0	42.53±3.94	44.65±3.79	40.82±4.14 (Acute Panic Inventory)	0.238	0.289	0.041	0.281
						P=0.357 (NS)	P=0.260 (NS)	P=0.876 (NS)	P=0.275 (NS)
Social anxiety disorder (n=9)	24.56±2.74	22±2.78	43.11±3.86	44.89±3.59	67.44±5.73 (Liebowitz Social Anxiety score)	0.807**	0.435	−0.009	0.160
						P=0.009 (S)	P=0.242 (NS)	P=0.983 (NS)	P=0.682 (NS)
OCD (n=8)	23.63±4.57	23.36±5.85	44.50±3.51	46.75±3.49	32.88±2.53 (YBOCS)	0.780	0.570	0.551	0.405
						P=0.022 (NS)	P=0.141 (NS)	P=0.157 (NS)	P=0.319 (NS)
Agoraphobia (n=6)	26.44±2.31	20±2.47	43.54±3.76	44.74±3.21	4.33±1.03 (Mobility Inventory)	0.896*	0.955**	0.893*	0.941**
						P=0.016 (S)	0.003 (S)	0.016 (S)	P=0.005 (S)

*Significant at 0.05; **Statistically significant at 0.01. NS - Not significant; S - Significant; BPRS - Brief Psychiatric Rating Scale; PANSS - Positive and Negative Symptom Scale for Schizophrenia; OCD - Obsessive-compulsive disorder; YBOCS - Yale-Brown Obsessive-Compulsive Scale

After the baseline evaluation, the schizophrenia patients were given antipsychotics both from the first generation as well as the second generation (haloperidol, trifluoperazine, olanzapine, risperidone, and clozapine) in an adequate dose and their effect on the schizophrenia as well as the anxiety disorder was assessed. There was a significant decrease in the severity of the panic disorder after the administration of antipsychotic drugs but not in the severity of OCD, social anxiety disorder, or agoraphobia. After a period of 8 weeks, the patients were prescribed fluoxetine in addition and the progress of anxiety disorder assessed. It was seen that in all the cases, the severity of the anxiety disorder had decreased in a significant manner [Table 4].

Table 4

Comparison of scores on anxiety scale at baseline, after antipsychotics, and after antipsychotics and fluoxetine treatment in schizophrenia patients with comorbid anxiety disorders

Type of anxiety disorder	Score on anxiety scale (name in parentheses)			Wilcoxon test comparing 1 and 2	Wilcoxon test comparing 2 and 3
	1. At baseline	2. After antipsychotics	3. After antipsychotics and fluoxetine
Panic disorder (n=17)	40.82±4.14 (Acute Panic Inventory)	39.18±3.25 (Acute Panic Inventory)	29.82±3.56 (Acute Panic Inventory)	−2.95	−3.63
				P=0.003 (S)	P=0.000 (S)
Social anxiety disorder (n=9)	67.44±5.73 (Liebowitz Social Anxiety Score)	66.89±5.25 (Liebowitz Social Anxiety Score)	57.44±5.88 (Liebowitz Social Anxiety Score)	−1.89	−2.68
				P=0.059 (NS)	P=0.007 (S)
OCD (n=8)	32.88±2.53 (YBOCS)	32.63±2.20 (YBOCS)	23.13±2.17 (YBOCS)	−1.414	−2.54
				P=0.157 (NS)	P=0.011 (S)
Agoraphobia (n=6)	4.33±1.03 (mobility inventory)	4.17±0.41 (Mobility Inventory)	3.67±0.52 (Mobility Inventory)	−0.577	−1.732
				P=0.564 (NS)	P=0.083 (NS)

OCD - Obsessive-compulsive disorder; YBOCS - Yale-Brown Obsessive-Compulsive Scale; NS - Not significant; S - Significant

DISCUSSION

The current study has attempted to improve over the methodological limitations of the previous studies. These studies had included a small number of patients in the clinical samples and concentrating on a one or few of the anxiety disorders, namely, panic disorder and OCD. Hence, increasing the sample size as well as including all the anxiety disorders with specific treatments for anxiety disorders is a methodological improvement over earlier studies.

Characteristics of schizophrenia patients with and without anxiety disorder

In the present study, no significant differences were found in the sociodemographic variables of schizophrenia patients without anxiety disorders and schizophrenia patients with anxiety disorders [Table 1]. This is similar to the findings of many other studies.[51819202122]

There was a significant difference between the duration of schizophrenic illness in schizophrenics with anxiety disorder and schizophrenics without anxiety disorder [Table 2], as the duration of illness was longer in cases of schizophrenia patients without anxiety disorders as compared to schizophrenia patients with anxiety disorders, which is a novel finding and had not been reported by the previous studies.

A greater association of anxiety disorder was found with the paranoid subtype of schizophrenia (P ≤ 0.01), which had been shown by the previous studies as well.[232425] There was a higher rate of treatment received by patients having schizophrenia only as compared to the other group reported by others also.[26] No significant difference was found in terms of the other clinical variables between schizophrenia patients without anxiety disorders and schizophrenia patients with anxiety disorders, which is similar to the other studies done in this regard.[25]

Comparison between the scores obtained on various psychopathology scales of schizophrenia with and without anxiety disorders

There was no significant difference between the two groups on BPRS for schizophrenia with anxiety disorders and schizophrenia without anxiety disorders, which were similar to the other studies.[20] However, a significant difference was found between the two groups on positive, negative, and general psychopathology scale of PANSS. These findings were similar to the previous studies with different studies showing different results. Emsley et al.[26] found a significant difference for the positive subscale of PANSS between the two groups which was also replicated later,[23] which is in agreement with our findings. In contrast to the above findings, Tibbo et al.[18] found a significant difference on the general psychopathology scores of the two groups reflecting the link between the general psychopathology scale and the anxiety disorders as the general psychopathology scale includes anxiety or anxiety-related symptoms (e.g., anxiety, tension, and active social avoidance).

Psychopathology scores of schizophrenia and anxiety disorder scales

The mean values obtained on negative, positive, and general psychopathology scale of PANSS as well as the BPRS scores were much higher than the corresponding studies done in the West,[519] which is probably due to the long duration of untreated illness in Indian schizophrenic population, especially belonging to the rural population. The values for the psychopathology scales for panic disorder are in close agreement as those found by the other researchers.[1923] In addition, the values obtained on positive, negative, and general psychopathology scale of PANSS and BPRS and on LSAS are in accordance with the Western population,[2728] whereas that for OCD is lower than that reported earlier, which might be because of the inclusion of only acute schizophrenics in the earlier study.[2728] The scores on the positive, negative, and general psychopathology scale of PANSS, BPRS scores, and GAD-7 scores are similar to that reported by other researchers[518] and also for PTSD.[2930]

Correlation between anxiety disorder scores and psychopathology scores

There was no significant correlation between acute panic inventory scores and PANSS scores, that is, positive scores, negative scores, and general psychopathology scores. This is in agreement with an earlier study[20] but is in contrast with a few related studies which have found a significant correlation between the positive scale of PANSS and panic disorder severity scales.[232627] The acute panic inventory score was not correlated significantly with the BPRS score either, which again is in agreement with other studies.[19]

Some researchers have found a higher level of psychopathology in schizophrenia patients with a comorbid diagnosis of panic disorder,[232831323334] but this was not observed in the present study.

The current study showed no significant correlation between the LSADS scores and PANSS positive subscale score, which is in close agreement with a few studies,[1927] whereas a few reported a positive correlation between social anxiety score and PANSS negative score,[35] while another study found no correlation between LSADS scores with BPRS and PANSS scores.[6]

There was no significant correlation between obsessive-compulsive symptoms and schizophrenia symptoms as reflected by the YBOCS scores and the BPRS and PANSS scores. Similar to this, a few studies have also found no significant correlation between the YBOCS scores and psychopathology scales.[213637] A few have found a positive correlation between YBOCS scores and positive scores of PANSS[273839] or negative score of PANSS.[2140] In contrast, some have found a negative correlation between YBOCS scores and negative scores of PANSS.[41] As there was a single schizophrenia patient with PTSD as well as GAD, statistical analysis could not be done. In earlier studies, a significant correlation between PTSD scores and psychopathology scores for schizophrenia was found reflecting a higher severity of psychopathology in schizophrenic patients with comorbid PTSD.[3042434445] Similarly, a significant correlation was found between the GAD-7 scores and the general psychopathology scale of PANSS.[5] As in the current study, majority of the study sample belonged to the group of chronic schizophrenics, which is different from the other studies with patients mostly belonging to acute schizophrenia and may explain the differences in the results with the other studies.

Effect of treatment on schizophrenia and the comorbid anxiety disorders

There was a significant difference in the treatment outcome after the administration of olanzapine and fluoxetine, which has also been reported by other researchers.[1719464748] Although reports of improvement of panic disorder and agoraphobia with cognitive behavior therapy have been reported by a few of the researchers,[4950] this could not be corroborated because of the less number of patients as well as their limited hospital stay. There was a significant difference between the pre- and post-treatment YBOCS scores after the administration of fluoxetine (P ≤ 0.01) but not after the administration of olanzapine. These findings are similar to the findings of the previous studies, with no improvement in the OCD after the administration of antipsychotics[51] but a good improvement after the administration of anti-obsessional agents.[484952] For social anxiety disorder as well, there was a significant difference between the pre- and post-treatment values after the administration of fluoxetine (P ≤ 0.01), but not after the administration of olanzapine which is similar to the findings of a previous study.[5354] Given that anxiety disorders are relatively responsive to treatment, greater awareness of their comorbidity with schizophrenia should yield worthwhile clinical benefits to the patient.

Limitations

Limitations of the study include possible selection bias being a tertiary care hospital-based study where patients with more severe illness are more likely to be included. No distinction was made between the clinical correlates of patients with acute schizophrenia and chronic schizophrenia. Moreover, the sample consisted mostly of men (81%), thus the findings of the study cannot be generalized. In the treatment aspect also, the effect of only one SSRI was studied.

CONCLUSIONS

Comorbid anxiety disorders are quite common in schizophrenia. Schizophrenia patients with anxiety disorders differ significantly from those without anxiety disorders in their basic psychopathology, that is, positive, negative, and cognitive domains. Further, there is a shorter duration of illness in schizophrenia patients with anxiety disorders as compared to schizophrenia patients without anxiety disorders assigning a prognostic significance to the presence of anxiety disorders. Moreover, these anxiety disorders are quite responsive to the SSRIs but not responsive to the antipsychotics alone. Hence, SSRIs may be tried in such cases for the effective treatment of the anxiety disorders with schizophrenia, leading to the effective relief of distress of such patients.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.