Cognitive deficits, depressive symptoms, insight, and medication adherence in remitted patients with schizophrenia.

OBJECTIVE: To study the relationship between cognitive deficits, depressive symptoms, insight, and medication adherence in remitted patients with schizophrenia.
METHODS: Fifty-four patients aged 18-60 years, with schizophrenia in remission, were evaluated for adherence using the Medication Adherence Rating Scale (MARS), cognitive deficits using the Schizophrenia Cognition Rating Scale (SCoRS), depressive symptoms using the Calgary Depression Scale for Schizophrenia (CDSS), and insight using the Beck Cognitive Insight Scale.
RESULTS: Twenty-one (38.9%) patients were found to be nonadherent to their medication. A significant negative correlation was found between MARS total score with SCoRS attention/vigilance (r = -0.28), attitude toward negative side effects of the psychotropic medication with SCoRS total score (r = -0.36), SCoRS attention/vigilance (r = -0.27), verbal learning and memory (r = -0.32), reasoning and problem-solving (r = -0.30), and social cognition (r = -0.28). A significant negative correlation was found between CDSS total score with MARS total score (r = -0.50), medication adherence behavior (r = -0.44), and attitude toward negative side effects of psychotropic medication (r = -0.60). MARS total score significantly positively correlated with years in remission (r = 0.29).
CONCLUSIONS: Poor medication adherence was seen in more than one-third of remitted patients with schizophrenia and was associated with global cognitive deficits, depressive symptoms, and the number of years in remission.

INTRODUCTION

Schizophrenia is a severe and chronic mental illness affecting 1% of the world's population.[1] The presence of cognitive deficits in schizophrenia was recognized early almost a century ago, but an understanding of their importance and influence on the prognosis has increased dramatically over time. Cognitive impairment which often predates the illness is now considered as one of the cardinal features of schizophrenia, with affected patients typically performing 0.8–1.5 standard deviations (SDs) below the level of healthy controls across many cognitive domains.[2] Several meta-analyses have shown prominent cognitive deficits in schizophrenia in areas such as digit symbol coding, verbal learning, and general IQ, with up to 80% of patients being cognitively impaired.[234] More importantly, cognitive deficits are clinically relevant because they are considered to contribute to social impairments and treatment outcome in schizophrenia.[5]

Patients who do not follow the treatment schedule and drug regimens prescribed by physicians can be described as noncompliant or nonadherent.[6] Psychotic illness impairs insight and judgment and places the patients with schizophrenia at increased risk for medication nonadherence.[7] Nonadherence may occur in up to 50% of the patients with schizophrenia receiving antipsychotics.[8] These patients are more vulnerable to relapse following medication nonadherence, requiring frequent admissions to the hospital as well as longer stay, thereby consuming more resources.[9] Improving adherence in persons with schizophrenia has the potential to reduce the morbidity and suffering of patients and families in addition to decreasing the cost of rehospitalization.[10]

There is a dearth of studies analyzing the association between the cognitive deficits and medication nonadherence in patients with schizophrenia in remission, specifically from the Indian subcontinent. The objectives of the current study were to examine the cognitive deficits, depressive symptoms, insight, and medication adherence and their relationship in remitted patients with schizophrenia.

METHODS

Participants

This was a hospital-based cross-sectional study conducted over a 2-year period from September 2012 to September 2014 at Kasturba Medical College, Manipal, a tertiary care center in southern India. The institutional ethics committee approved the study. Participants included outpatients aged between 18 and 60 years, of either sex, diagnosed with schizophrenia according to the ICD-10 DCR,[11] and fulfilling Andreasen et al.'s[12] criteria for remission. Written informed consent was taken from the patients and informants in their native language. Those with any other comorbid major psychiatric disorder, substance dependence (except nicotine), or major medical disorder were excluded from the study.

Tools

The sociodemographic and clinical details were collected using a semistructured pro forma designed for the study. The 30-item clinician-rated Positive and Negative Syndrome Scale (PANSS)[13] was used to assess psychopathology. Andreasen et al.'s[12] criteria for symptomatic remission include mild (score 2 or less) over a 6-month period of simultaneous rating in the following items of PANSS: delusion (P1), unusual thought content (G9), hallucinatory behavior (P3), conceptual disorganization (P2), mannerisms/posturing (G5), blunted affect (N1), social withdrawal (N4), and lack of spontaneity (N6). The Schizophrenia Cognition Rating Scale (SCoRS)[14] was used to assess cognitive impairment in schizophrenia and the degree to which it affects day-to-day functioning. It assesses problems in attention, memory, motor skills, speech, and problem-solving within the past 2 weeks. Higher scores reflect a greater degree of impairment. To assess adherence, the Medication Adherence Rating Scale (MARS)[15] was used. It is a 10-item yes/no self-reporting multidimensional instrument describing three dimensions: “medication adherence behavior,” “attitude toward taking medication,” and “negative side effects and attitudes toward psychotropic medication.” The 9-item Calgary Depression Scale for Schizophrenia (CDSS)[16] was used for assessing the depressive symptoms in patients with schizophrenia. Each item is rated from 0 to 3 in a time period of the last 2 weeks. The 15-item Beck Cognitive Insight Scale (BCIS)[17] was used to evaluate patients' self-reflectiveness and their overconfidence in their interpretations of their experiences. A list of 20 items comprising reasons for nonadherence prepared for this study was also used. Each item was rated as yes/no. All the patients meeting the inclusion and exclusion criteria were evaluated on MARS, SCoRS, CDSS, BCIS, and scale for nonadherence by the first author.

Statistical analysis

Data collected for 54 patients were analyzed using the Statistical Package for the Social Sciences-version 16.0 for Windows® (SPSS Inc., Chicago, IL, USA). Group differences in socio-demographic and clinical variables between adherent and non-adherent patients were described using Mann-Whitney U test (Z) and Pearson's chi-square test (X2)/Fischer exact test for continuous and categorical variables, respectively. Pearson's correlation coefficient (r) was used to study the association between continuous variables. Binary logistic regression analysis was done with adherence as the outcome variable and years in remission, total depression score, and SCoRS total score as predictor variables. The variance explained was reported using pseudo-R squares (Cox and Snell and Nagelkerke). The level of significance was set at P < 0.05 (2-tailed).

RESULTS

Sociodemographic and clinical characteristics

The sample characteristics are summarized in Table 1. The mean age of the sample was 33.80 (SD 8.96) years, and the mean duration of the illness was 11.80 (SD 7.08) years. More than half (55%) of the samples were male and most belonged to middle socioeconomic status. The frequencies of positive responses on MARS items are depicted in Figure 1. Among them, 21 (38.9%) were found to be nonadherent on their medication based on the MARS score. Table 2 shows the reasons for medication nonadherence in the patients.

Table 1

Sociodemographic and clinical characteristics (n=54)

Variables	Mean (SD)
Age	33.8 (8.9)
Education in years	10.7 (3.9)
Total duration of illness	11.8 (7.1)
Years in remission	3.3 (2.3)
MARS
Total score	16.2 (2.8)
Medication adherence behavior	6.6 (1.3)
Attitude toward the medication	6.7 (1.2)
Attitude toward negative side effects of psychotropic medication	2.9 (0.9)
SCoRS
Total score	37.3 (11.7)
Working memory	8.0 (3.4)
Attention/vigilance	7.9 (2.6)
Verbal learning and memory	9.9 (3.7)
Speed of processing	5.7 (1.9)
Reasoning and problem-solving	1.8 (1.0)
Social cognition	3.8 (1.6)
CDSS total score	7.00 (5.24)
BCIS
Self-reactivity	21.70 (4.76)
Self-certainty	11.93 (3.87)
Composite index score	9.20 (6.38)
Variables	n (%)
Gender
Male	30 (55.6)
Female	24 (44.4)
Occupation
Unemployed	15 (27.8)
Employed	27 (50.0)
Homemakers/students	12 (22.2)
Socioeconomic status
Lower	11 (20.4)
Middle	37 (68.5)
High	6 (11.1)
Marital status
Married	29 (53.7)
Single	25 (46.3)
Family history of mental illness
Present	13 (24.1)
Absent	41 (75.9)

MARS – Medication Adherence Rating Scale; SCoRS – Schizophrenia Cognition Rating Scale; CDSS – Calgary Depression Rating Scale for Schizophrenia; BCIS – Beck Cognitive Insight Scale; SD – Standard deviation

Figure 1

Medication Adherence Rating Scale items in patients (n = 54)

Table 2

Reasons for nonadherence among patients

Variables	n (%)
I am all right	24 (44.4)
Stigma	22 (40.7)
Adverse effects	21 (38.9)
Unaware of long-term medication	20 (37.0)
Need to take medication lifelong	19 (35.2)
Fed up on taking medications for many years	19 (35.2)
Felt that I would become dependent on medication	18 (33.3)
Just counseling is enough	15 (27.8)
Taking medicines for many years so stopped	15 (27.8)
Told by someone not to take medication	13 (24.1)
Confused with the medication regimen	11 (20.4)
Every time I am seen by a new doctor	10 (18.5)
Need to wait hours for the prescription	9 (16.7)
Can’t afford medication	9 (16.7)
Busy at work	8 (14.8)
Did not like the taste of the medication	6 (11.1)
Psychiatric medicines interfere with other medications	5 (9.3)
Native place is far off	4 (7.4)
Family members do not know the need for medication	3 (5.6)
No improvement with medication	3 (5.6)
Doctor did not tell me the use of medications	3 (5.6)
Other reasons	2 (3.7)
Nobody to supervise to medication	1 (1.9)
Doctor is unfriendly	0 (0)

Correlation between the Medication Adherence Rating Scale with Schizophrenia Cognition Rating Scale, Calgary Depression Rating Scale, and Beck Cognitive Insight Scale

A significant negative correlation was found between CDSS total score with MARS total score (r = −0.50, P = 0.001), medication adherence behavior (r = −0.44, P = 0.001), and attitude toward negative side effects of psychotropic medication (r = −0.60, P = 0.001). A significant negative correlation was found between MARS total score with SCoRS attention/vigilance (r = −0.28, P = 0.04), attitude toward negative side effects of the psychotropic medication with SCoRS total score (r = −0.36, P = 0.01), SCoRS attention/vigilance (r = −0.27, P = 0.05), verbal learning and memory (r = −0.32, P = 0.02), reasoning and problem-solving (r = −0.30, P = 0.03), and social cognition (r = −0.28, P = 0.05). No significant correlation was found between the MARS total score and its domains with BCIS. MARS total score significantly positively correlated with years in remission (r = 0.29, P = 0.04).

Group differences between adherent and non-adherent patients

Table 3 shows the group differences in socio-demographic and clinical variables among the adherent (N = 33) and nonadherent patients (N = 21). The only significant difference between the adherent and non-adherent group was for SCoRS total score; it was lower in adherent patients (p= 0.04). No significant differences were found with any other socio-demographic or clinical variable.

Table 3

Group differences in socio-demographic and clinical variables between adherent and non-adherent patients (N=54)

Variables	Mean (SD)		Mann-Whitney U (Z)	P
	Adherent (n=33)	Non-adherent (n=21)
Age	35.1 (9.7)	31.8 (7.4)	280.5 (−1.17)	0.24
Education years	10.7 (4.1)	10.6 (3.7)	317.0 (−0.53)	0.60
Total illness duration in years	12.9 (7.8)	10.1 (5.6)	282.0 (−1.14)	0.25
Years in remission	3.9 (2.5)	2.5 (1.6)	246.5 (−1.81)	0.07
CDSS total score	5.9 (4.9)	8.6 (5.5)	246.0 (−1.79)	0.07
SCoRS
SCoRS total score	12.1 (4.3)	14.0 (3.1)	231.5 (−2.05)*	0.04
Working memory	1.6 (0.8)	2.0 (0.8)	259.0 (−1.67)	0.10
Attention/vigilance	2.2 (0.9)	2.6 (0.6)	270.0 (−1.46)	0.15
Verbal memory	2.3 (0.9)	2.6 (0.7)	277.0 (−1.32)	0.19
Speed of processing	2.2 (0.8)	2.5 (0.7)	259.5 (−1.66)	0.10
Reasoning and problem-solving	1.8 (1.1)	2.1 (0.9)	280.5 (−1.23)	0.22
Social cognition	1.9 (0.8)	2.2 (0.9)	286.0 (−1.14)	0.26
BCIS
Self-reflectiveness	21.4 (4.7)	20.6 (4.8)	310.5 (−0.64)	0.52
Self-certainty	12.0 (3.8)	11.7 (3.9)	334.0 (−0.22)	0.82
Composite Index score	9.4 (5.7)	8.8 (7.4)	345.5 (−0.02)	0.99
Variables	n (%)	n (%)	χ2	P
Gender
Male	18 (54.5)	12 (57.1)	0.04	0.85
Female	15 (45.5)	9 (42.1)
Occupation#
Unemployed	11 (33.3)	4 (19.0)	-	0.27
Employed	17 (51.5)	10 (47.6)
Homemakers/students	5 (15.2)	7 (33.3)
Socioeconomic status#
Lower	5 (15.2)	6 (28.6)	-	0.47
Middle	24 (72.7)	13 (61.9)
High	4 (12.1)	2 (9.5)
Marital status
Married	17 (51.5)	12 (57.1)	0.16	0.69
Unmarried	16 (48.5)	9 (42.9)
Family history of mental illness
Present	8 (24.2)	5 (23.8)	0	0.97
Absent	25 (75.8)	16 (76.2)

*P<0.05 (two-tailed), #Fisher’s exact test. SCoRS – Schizophrenia Cognition Rating Scale; BCIS – Beck Cognitive Insight Scale; SD – Standard deviation

Association with nonadherence

Table 4 describes binary logistic regression analysis using backward method, showing variables associated with nonadherence. The number of years in remission was found to be associated with nonadherence (B = 0.74, 95% confidence interval [CI]: 0.55 − 1). The number of years in remission explained 8.4%–11.4% of the total variance in nonadherence.

Table 4

Binary logistic regression analysis using backward method showing variables associated with nonadherence (n=54)

	Unstandardized coefficients		P	Exp B	95% CI for Exp B
	B	SE			Lower	Upper
Step 1
Years in remission	−0.26	0.16	0.10	0.77	0.57	1.05
Total depression score	0.07	0.06	0.22	1.08	0.96	1.21
SCoRS total score	0.08	0.08	0.30	1.09	0.93	1.27
Constant	−1.27	1.25	0.30	0.28
Step 2
Years in remission	−0.27	0.15	0.08	0.76	0.57	1.03
Total depression score	0.09	0.06	0.14	1.09	0.97	1.22
Constant	−0.23	0.71	0.74	0.79
Step 3
Years in remission	−0.30	0.15	0.05*	0.74	0.55	1.00
Constant	0.48	0.52	0.36	1.62

*P<0.05; Pseudo R2 for step 3: Cox and Snell=0.084, Nagelkerke=0.114; SCoRS – Schizophrenia Cognition Rating Scale; SE – Standard error; CI – Confidence interval

DISCUSSION

In the current study, 21 (38.9%) patients were found to be nonadherent on their medication. The reason for nonadherence varied from forgetting medication in 35%, carelessness about taking medication in 26%, whereas 41% felt that it was better to stop medication and 39% reported that they felt worse on medication. The attitude toward medication was also negative; 42% of patients felt that they need medication only when sick, one-third said that it is unnatural for their mind to be controlled on medication, a quarter said that their thoughts are unclear on medication, and another third said that they cannot prevent getting sick by staying on medication. For more than half of the patients, negative side effects were important, with 48% being said that they feel like a zombie on medications and 63% being said that they felt sluggish. In the study by Srinivasan and Thara,[18] which analyzed 254 outpatients for a period of about 2 years, it was found that among the 148 (58%) outpatients who were found to be nonadherent, 86 (58%) refused medication whatever amount of conviction was put in. About 74 (50%) nonadherent patients were treated as outpatients and given medication at least once without their knowledge by the family members under the psychiatrist's advice.

In the current study, among the factors for nonadherence, the five leading causes included patients perceiving that they are all right (44%), followed by stigma (41%), adverse effects (39%), unaware of taking long-term medication (37%), and being fed upon taking medication for many years (35%). In the study conducted by Cooper et al.[19] on 634 patients, 217 (34%) reported incomplete adherence to their psychiatric medication. Reasons given included forgetting, losing, and running out (37%); thinking medication unnecessary (24%); reluctance to take drugs (19%); and side effects (14%), which were similar to our study.

Cognitive deficits and medication adherence

Studies have shown that patients with schizophrenia generally perform poorly on cognitive tests and that they specifically have problems in areas of attention, memory, and problem-solving.[20] Cognitive deficits may decrease the patients' ability to adhere to their treatment regimens and may affect the patients' ability to determine the benefits of medication and the subsequent need for treatment. Deficits in information processing of complex medication regimens can also lead to problems adhering to medication instruction.[21] In our study, global cognitive deficits (SCoRS total score) had a significant negative correlation with the attitude toward negative side effects of the psychotropic medication, with a medium effect size (r = −0.36). Furthermore, a significant negative correlation was found between deficits in attention/vigilance with MARS total score and negative side effects of the psychotropic medication, with small to medium effect sizes. Also, deficits in verbal learning and memory, deficits in reasoning and problem solving and deficits in the social cognition correlated negatively with negative side effects of the psychotropic medication, with small to medium effect sizes.

The findings of the current study are similar to the findings of Jeste et al.,[21] which concludes that deficits in cognitive functions, especially conceptualization and memory, were strongly correlated to the patient's ability to manage his/her medication, and Cuffel et al.'s[22] study which showed that neurocognitive impairment is associated with lower medication adherence. On the other hand, Buchanan[23] showed that there were no significant differences in the cognitive functioning among the adherent and nonadherent patients. Similarly, Boyer et al.[24] found that there was no significant association between neurocognition and nonadherence.

In our study, global cognitive deficits (SCoRS total score) were lower in adherent patients as compared to the nonadherent group. It is possible that those with cognitive deficits would forget to take medications daily without supervision. Nevertheless, it still remains to be examined whether cognitive deficits might actually lead to medical nonadherence in patients with schizophrenia, in prospective trials.

Depression and medication adherence

A significant negative correlation was found between the depressive symptoms and MARS total score (r = −0.50), medication adherence behavior (r = −0.44), and attitude toward negative side effects of psychotropic medication (r = −0.60), with medium-to-strong effect sizes. Our findings are in contrast with Na et al.'s[25] study that found lower depressive symptoms to be associated with good medication adherence. It is possible that those with depressive symptoms have less of cognitive deficits, which explains better adherence to medications. Indeed, there was small-to-medium effect positive correlation between depression score and cognitive deficits score, although it was not statistically significant.

Cognitive insight and medication adherence

No significant correlation was found between the MARS total score and its domains with any of the individual domains of BCIS (cognitive insight) in our study, which is similar to the findings of Boyer et al.[24] that did not find any direct association between awareness of illness and nonadherence. In contrast, Roy et al.[26] found that 12% of the patients were nonadherent to medication because of lack of insight and 6% because of a lack of awareness about their mental illness. Similar findings were reported by Cuffel et al.[22] and Smith et al.,[27] which showed that treatment adherence is higher with greater levels of insight about the illness. Furthermore, a recent systematic review[28] found that positive attitude toward medication and illness insight had a consistent association with better adherence. Interestingly, Beck et al.[29] based on the mediational model using structural equation modelling, argue that it is the attitudes rather than global insight into the illness which is important for adherence, and interventions that address the attitudes may be more effective in improving adherence.

Sociodemographic variables and medication adherence

In the present study, a significant positive correlation was found between medication nonadherence and the number of years in remission, albeit with small effect size (r = 0.29). Logistic regression analysis also found only the number of years in remission to be associated with nonadherence (B = 0.74, 95% CI: 0.55 − 1). The number of years in remission explained 8.4%–11.4% of the total variance in nonadherence. This could be related to the patients perceiving that they are not in need of any medication due to a long symptom-free period. Indeed, more than 40% of those having nonadherence reported that they stopped medications as they felt “I am all right” and more than 35% reported being “fed upon taking medications since many years.” This finding is similar to that of Roy et al.,[26] which showed that almost 28% of the people who find improvement stop being adherent to the medication. Furthermore, it could be related to stigma and adverse effects of medications, as reported by 40% and 38% patients, respectively. Another possible reason is lack of knowledge about the need to take medication for the long term, in 33% of patients. This leads to the possibility that psychoeducation of patients and family may lead to a reduction in stigma, clarifying myths and the need for long-term medications, thus improving adherence and outcome.[30]

There was no relationship between medication adherence with any other sociodemographic or clinical variable. Similarly, studies like Buchanan[23] showed that sociodemographic characteristics were not significantly associated with medication adherence.

CONCLUSIONS

This study found medication adherence to be poor in more than a third of remitted patients with schizophrenia, and it was influenced by global cognitive deficits, depressive symptoms, and the number of years in remission.

The limitations of the study are small sample size, and as it was a hospital-based study, the findings cannot be generalized to the community. Furthermore, no objective method was used to nonadherence and cognitive deficits in the patients. Future studies could also include patients from the community, day care, and rehabilitation centers to obtain a more heterogeneous sample, which reflects the real trends of the medication adherence. Furthermore, a validated neuropsychological battery for cognitive assessment could be used to confirm the cognitive deficits objectively, and the evaluation of nonadherence can be enhanced using objective methods such as pill counters.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.