Medication adherence in bipolar disorder: Exploring the role of predominant polarity

Abstract

Objectives

Medication non-adherence is one important reason behind sub-optimal outcome from treatment of bipolar affective disorder (BPAD). Though various reasons behind medication non-adherence have been identified, little is known about the medication adherence patterns across various predominant polarities (PP) in BPAD.

Methods

100 euthymic patients of BPAD were purposively recruited and the PP were determined. Subsequently, Morisky Medication adherence scale (MMAS); Global Assessment of Functioning (GAF); Oslo Social Support Scale and World Health Organization Quality of Life scale- Brief version (WHOQOL-Bref) were administered. Analysis of covariance (ANCOVA) was done to estimate the difference of scores of MMAS after adjusting for any potential confounders.

Results

Overall, 44 patients with manic PP (MPP), 17 with depressive PP (DPP) and 39 with indeterminate PP (IPP) were recruited. It was found that patients who presented with DPP showed significantly higher medication adherence as compared to MPP.

Conclusion

Knowledge of PP of a patient of BPAD can be useful in anticipating medication adherence and treatment outcome. The major limitations included non-probability sampling, cross-sectional design and limited generalizability of the results.

Keywords

bipolar disorder medication adherence predominant polarity euthymia compliance

Introduction

Bipolar affective disorder (BPAD) is a chronic disorder usually with an episodic course, causing significant morbidity to the patients. This disorder is usually characterized by the occurrence of multiple episodes of either manic/hypomanic episodes (characterized by euphoric or irritable mood state, increase in psychomotor activities and other symptoms) or depressive episodes (characterized by low mood, anhedonia, anergia and other symptoms), often with good inter-episodic remission. The treatment of this disorder largely involves controlling the index mood episode and also prevention of relapse by use of prophylactic medicines, commonly referred to as mood stabilizers. Over the past years, the treatment of BPAD has been fortified with the advent of a number of evidence-based treatments. However, in-spite of that, the outcome of the treatment often leaves a lot to desire. One of the major issues pertaining to the sub-optimal response to pharmacotherapy is medication adherence.¹ Medication adherence is defined as ‘the degree to which a patient’s behavior matches agreed-upon recommendations from a prescribing clinician’² The prevalence of medication non-adherence has been found to be ranging between 20-60% in the past studies.^3–6 The variety in the range is largely attributed to difference of the study population, tools used in the measurement and recall methods. However, the significant burden arising out of medication non-adherence was ubiquitous across studies.

The reasons behind non-adherence to medication have received substantial attention from the researchers. A number of patient-level barriers like young age, being single, poor social support, more severe symptoms, high frequency of affective episodes, other medical or psychiatric comorbidity etc. have been identified.⁷ In addition, a number of external barriers were also identified in the studies like poor side-effect profile, complex treatment regime, choice of medication and stigma.⁸ Various factors arising out of the disorder has also been implicated for treatment non-adherence. Various studies in the past had found that bipolar I subtype, suicidality or presence of psychotic symptoms were more likely to be treatment non-adherent.^9,10

Till now, the most common approach to study this problem has been to compare medication adherence in BPAD with other severe mental illnesses, most commonly schizophrenia. But, BPAD in itself, is a complex disorder and further research needs to be invested in delineating the various sub-types in BPAD. Predominant polarity (PP) is one such sub-type that garnered our interest by allowing classification of patients of BPAD by predominance of one polarity of affective episodes over the longitudinal course of the illness. Though there are variety of definitions, the stricter criterion deducts that, if two-thirds of the episodes in a patient of BPAD are of a particular polarity, the PP of that patient is the corresponding PP.¹¹ Patients with two-third majority of manic episodes are classified as manic PP (MPP); whereas patients with two-third majority of depressive episodes are classified as depressive PP (DPP). But, if neither manic nor depressive episodes gain two-thirds predominance in the course the patients is classified as indeterminate PP (IPP).

Our current understanding tells us that patients of various PP can vary from each other in terms of their cognitive functioning,¹² temperament or internalized stigma¹³; basically allowing us to predict important clues in the further course of the patient. However, no study has been conducted about the patterns of treatment adherence in various PP in patients of BPAD. This information could also become important in the background of the fact that prescription patterns in various PP is unique.¹⁴ Studies till now have shown that the medication adherence can vary according to the types of medication, pill burden and perceived side-effects.¹⁵ Thus, it will be extremely beneficial for a clinician in decision-making to have the knowledge of when to expect medication non-adherence. With this background, the current study was conducted to compare medication adherence in euthymic patients of BPAD across the PP of their treatment regimen.

Methodology

The current study was conducted in a tertiary center from Eastern India. The study obtained ethical clearance from the institute ethics committee. 100 patients of BPAD in remission were purposively sampled if they adhered to the inclusion criteria. The inclusion criteria were i) aged above 18 years, ii) presenting with a reliable informant, iii) able to converse and comprehend in English, Hindi or Bengali to the level of primary education and iv) willing to provide written informed consent. Remission of BPAD was defined as absence of clinically significant mood symptoms, current Young Mania Rating Scale score of <5, Hamilton Rating Scale for Depression score of <8 and no change in maintenance medication by more than 50% or hospitalization in the past 2 months.¹⁶ Subjects were not recruited if clinical information is inadequate.

The study was conducted using a cross-sectional design with patients being recruited after fulfilling the inclusion criteria. They were then administered a semi-structured questionnaire to collect data regarding socio-demographic and clinical variables. Subsequently, all the tools mentioned as follows were administered: a) Morisky Medication adherence scale (MMAS)¹⁷- an 8-item self-reported measure of treatment adherence; b) Global Assessment of Functioning (GAF)¹⁸- a numeric scale used by mental health clinicians and physicians to rate subjectively the social, occupational, and psychological functioning of an individual; c) Oslo Social Support Scale¹⁹- a 3-item scale with scores ranging from 3–14 with higher scores indicating higher social support and d) World Health Organization Quality of Life scale- Brief version (WHOQOL-Bref)²⁰- an international cross-culturally comparable quality of life assessment instrument comprising of 26 items and 4 domains.

The statistical analysis was conducted using the SPSS version 24. Descriptive statistics were compiled. Test for Normality was done by Kolmogorov Smirnov test. Recruited subjects were divided into three PP based on two-third predominance as per the criteria given by Colom et al.¹¹ Comparisons of socio-demographic and clinical variable were done using chi-square-test (for categorical variables) and one-way analysis of variance for continuous variables (ANOVA) (Kruskal-wallis test for non-parametric data) between the three predominant polarities. Subsequently, analysis of covariance (ANCOVA) was done to estimate the difference of scores of MMAS after adjusting for any potential confounders. The level of significance was kept at 5%.

Results

Overall, in our study, 100 euthymic patients were recruited of which there were 44 patients with MPP, 17 with DPP and 39 with IPP. Table 1 shows the comparison of the various socio-demographic and clinical parameters in the three groups. It was found that the patients with MPP had a significantly lower age; higher percentage of patients being not married (being single or divorced) and lower total duration of illness and duration of seeking treatment.

Table 1.

Comparison of socio-demographic and clinical variables across various predominant polarity.

Variables	MPP (n=44)	DPP (n=17)	IPP (n=39)	Chi-square,(df); p-value	ANOVA/Kruskal-Wallis,(df); p-value	Post-hoc pairwise comparison (Tukey’s test)
Variables	Mean (S.D) or frequency (%)	Mean (S.D) or frequency (%)	Mean (S.D) or frequency (%)	Chi-square,(df); p-value	ANOVA/Kruskal-Wallis,(df); p-value	Post-hoc pairwise comparison (Tukey’s test)
Sociodemographic variables
Age	30.16 (10.4)	42.76 (11.6)	39.23 (13.2)		9.127; 2; 0.001*	(DPP, IPP > MPP)
Gender				0.17; 2; 0.918
Male	26 (59%)	11 (64.7%)	24 (61.5%)
Female	18 (41%)	6 (35.3%)	15 (38.5%)
Marital status
Single	19 (43.1%)	3 (17.6%)	10 (25.6%)	10.274; 4; 0.03*
Currently married	17 (38.6%)	13 (76.4%)	26 (66.6%)
Widowed/divorced	8 (18.3%)	1 (6%)	3 (7.7%)
Education				10.703; 10; 0.09
Clerk/shop-owner/farmer	5 (11.3%)	4 (23.5%)	2 (5.1%)
Skilled worker	14 (31.8%)	5 (29.4%)	18 (46.1%)
Semi-skilled/unskilled	16 (36.3%)	4 (23.5%)	4 (10.2%)
Retired/housewife	9 (20.6%)	4 (23.5%)	5 (38.6%)
Family type
Nuclear	26 (59%)	13 (76.4%)	25 (64.1%)	1.608; 2; 0.448
Joint	18 (41%)	4 (23.6%)	14 (35.9%)
Clinical parameters
Total duration of illness^@ (months)	90.73 (89.3)	202.59 (138.3)	153.64 (145.8)		5.903; 2; 0.004*	(DPP, IPP > MPP)
Number of previous episodes^@	4.27 (3.4)	5.53 (3.0)	5 (3.5)		0.975; 2; 0.381
Duration of treatment^@ (months)	63.68 (76.6)	149.76 (90.6)	107.44 (117.6)		5.337; 2; 0.006*	(DPP, IPP > PP)

^@-Kruskal Wallis Test, *p < 0.05; MPP-manic predominant polarity; DPP-depressive predominant polarity; IPP-indeterminate predominant polarity.

Table 2 which depicts the comparison of the clinical parameters in the three groups, showed that scores on MMAS and WHOQOL-Physical domain were significantly higher for DPP as compared to others. Subsequently, to look for difference in medication adherence in the three groups after adjusting for the effects of the potential confounders; analysis of covariance (ANCOVA) was conducted using age, duration of treatment and WHOQOL-Physical as covariates. Only one amongst total duration of treatment and total duration of illness was chosen as a covariate to avoid collinearity because both the parameters were significantly correlated (Pearson co-efficient = 0.875; p = 0.001). Table 3, which depicts the results of the ANCOVA, shows that there was significant difference in the estimated marginal means of MMAS scores between patients with DPP and MPP (but not with IPP).

Table 2.

Comparison of measures on clinical parameters in patients of BPAD with manic, depressive & indeterminate predominant polarity.

	MPP (n = 44)	DPP(n = 17)	IPP(n = 39)	ANOVA (F); dF; p- value	Post-hoc pairwise comparison (Tukey’s test)
MMAS	5.27 (2.22)	6.86 (1.51)	5.63 (2.11)	3.637; 2; 0.03*	DPP > MPP, IPP
GAF	3.82 (1.18)	3.59 (0.87)	3.85 (1.22)	0.318; 2; 0.729	–
Oslo Social Support Scale^@	0.91 (0.8)	1.24 (0.7)	1 (0.7)	1.043; 2; 0.356	–
WHOQOL Physical	22.55 (4.6)	24.41 (3.8)	21.28 (4.0)	3.188; 2; 0.046*	DPP > MPP, IPP
WHOQOL Psychological	18.64 (4.1)	19.76 (3.0)	18.03 (3.7)	1.22; 2; 0.3	–
WHOQOL Social	8.95 (2.5)	10.24 (2.1)	8.49 (2.9)	2.585; 2; 0.81	–
WHOQOL Environment	28.7 (5.6)	29.06 (4.8)	26.69 (5.0)	1.942; 2; 0.14	–

^@-Kruskal Wallis Test, *p < 0.05; MPP-manic predominant polarity; DPP-depressive predominant polarity; IPP-indeterminate predominant polarity.

Table 3.

Analysis of covariance of scores on the Morisky Medication Adherence Scale in patients of BPAD with manic, depressive & indeterminate predominant polarity (using age, duration of treatment and WHOQOL-Physical as covariates).

Discussion

Our study found that patients who presented with DPP showed significantly higher medication adherence as compared to MPP and IPP. This difference of medication adherence persisted between DPP and MPP patients when the effects of age, duration of treatment and WHOQOL-Physical were adjusted. Though there is no literature to directly compare our results, it can be said, that our result is supported by another study which showed that the medication adherence is poorer in patients with bipolar I.⁹ This is because conceptually, there is a high chance that patients with MPP had a higher prevalence of bipolar I disorder. In our sample of 100 patients, the proportion of MPP was higher than DPP and IPP, which was in line with various studies conducted on PP from Indian sub-continent.¹³ This finding becomes even more important considering that medication non-adherence was found to be higher in patients of MPP.

It has been found in the previous research that patients with poor medication adherence tended to have prominent cognitive impairments. The most common cognitive impairments associated with medication non-adherence were impairments in verbal memory and executive dysfunction. This is further supported by the results of the study by Belizario et al.¹² which showed that patients with MPP tended to present with more significant cognitive deficits as compared to other PP and healthy controls. In this context, it becomes imperative to discuss the psychological model of medication adherence which has been named as ‘Health Belief Model’.²¹ According to this theory the two major deciding factors for medication adherence are a) patient being able to comprehend his own condition and b) ability to perceive the benefits of being compliant on medication. It is postulated that both these factors are affected by the cognitive deficits and the effects are even more magnified in patients with MPP. In addition, the ever-growing volume of research on cognitive impairment in BPAD, has been able to identify that the impairment of neurocognitive profile in BPAD is very similar to what is seen in other severe mental illnesses like schizophrenia. Authors have commented that the differences in neurocognitive impairments amongst these disorders are rather quantitative than qualitative.²² By the use of cluster analysis approach in patients of BPAD, authors have been able to identify distinct groups in BPAD, some with normal profile, others with modest impairment and some with more profound impairments.²³ We believe that our study adds a new dimension to this research question, and further research regarding the neurocognitive profiling of patients of various PP can provide us with more answers.

Another important factor that plays a role in medication adherence is the treatment-emergent side-effects. The common side-effects reported to be troublesome by the patients include extrapyramidal symptoms or neuroleptic-induced dysphoria.²⁴ It has been proven that drugs that are commonly used to treat mania (first generation antipsychotics- including long-acting depot preparations) are more likely to cause these side-effects as compared to drugs used to treat depressive episodes (drugs like Quetiapine, Lurasidone or anti-depressants).²⁵ As a result, medication non-adherence could be higher in MPP (because of the higher proportion of manic episodes). However, with the advent of newer antipsychotic drugs like Cariprazine for the treatment of bipolar depression, the gap arising due to adverse effect profile between the two groups are now being narrowed down. Another important predictor of non-adherence to medication was found to be history of substance use in patients.²⁶ Various substance use disorders like alcohol use disorder and cannabis use disorders have been found to be poorly correlated to medication adherence. In the same vein, substance use disorders were also found to be higher in patients with MPP.^27,28

Finally, the relationship of the patient with the healthcare provider and social support is one of the most important predictor of medication adherence.²⁹ Current evidence shows that patients with MPP tended to have a poor social support,³⁰ which may play an adverse role in the management. In addition to that, higher internalized stigma can also play a deterrent role in medication adherence.⁸ Current evidence also shows that patients with MPP tended to report higher internalized stigma as compared to other PP.¹³

The strength of our study included use of validated tools and a stringent selection criteria. The major limitations of the study included non-probability method of sampling and use of cross-sectional design. In addition, since medication adherence was measured using a questionnaire, possibility of recall bias remains. Also, since the study was conducted in hospital settings, generalizability of the results to other sectors may be limited. Further studies in this field should be conducted using a larger sample size, prospective designs and more stringent methods of assessing medication adherence. 39% of our sample belonged to the IPP, which is considered to be an exclusive sub-group in the construct of PP. But, it should be remembered that, a longer follow-up for such patients could reclassify them to other PP, and therefore the results may vary. Also, it is imperative to replicate similar studies in other centers, more so in the light of the fact that the prevalence of different PP could vary across different countries.²⁷

To conclude, our study intended to find the difference of medication adherence in euthymic patients of BPAD across its various PP. It was found that patients who presented with DPP showed significantly higher medication adherence as compared to MPP, even after adjusting for possible confounding factors.

Study highlights

What is the current knowledge?

Medication adherence in bipolar disorder, which one of the prime reasons of sub-optimal therapeutic response, has a scope for improvement.

What is new here?

Medication adherence is worse in patients with manic predominant polarity as compared to patients with depressive predominant polarity.

Footnotes

Author Contribution

Conceived and designed the experiments: SG, NM, RA, GD, DKM & AP. Collection of data: SG, NM, RA Analyzed the data: AP. Wrote the first draft of the manuscript: AP. Contributed to the writing of the manuscript: SG, NM, RA, GD, DKM & AP. Agree with manuscript results and conclusions: SG, NM, RA, GD, DKM & AP. Jointly developed the structure and arguments for the paper: SG, NM, RA, GD, DKM & AP. Made critical revisions and approved final version: SG, NM, RA, GD, DKM & AP. All authors reviewed and approved of the final manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Arghya Pal

References

Berk

Hallam

Colom

, et al. Enhancing medication adherence in patients with bipolar disorder. Human Psychopharmacol 2010; 25: 1–16.

Haynes

Ackloo

Sahota

, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev, https://pubmed.ncbi.nlm.nih.gov/18425859/ (2008, accessed 30 March 2021)

Licht

Vestergaard

Rasmussen

, et al. A lithium clinic for bipolar patients: 2-year outcome of the first 148 patients. Acta Psychiatr Scand 2001; 104: 387–390.

Sajatovic

Valenstein

Blow

, et al. Treatment adherence with antipsychotic medications in bipolar disorder. Bipolar Disord 2006; 8: 232–241.

Colom

Vieta

Martinez-Aran

, et al. Clinical factors associated with treatment noncompliance in euthymic bipolar patients. J Clin Psychiatry 2000; 61: 549–555.

Lingam

Scott

Treatment non-adherence in affective disorders. Acta Psychiatr Scand 2002; 105: 164–172.

Leclerc

Mansur

Brietzke

Determinants of adherence to treatment in bipolar disorder: a comprehensive review.

J Affect Disord 2013; 149: 247–252.

Levin

Krivenko

Howland

, et al. Medication adherence in patients with bipolar disorder: a comprehensive review. CNS Drugs 2016; 30; 819–835.

Murru

Pacchiarotti

Amann

, et al. Treatment adherence in bipolar I and schizoaffective disorder, bipolar type. J Affect Disord 2013; 151: 1003–1008.

10.

Sajatovic

Biswas

Kilbourne

, et al. Factors associated with prospective long-term treatment adherence among individuals with bipolar disorder. Psychiatr Serv 2008; 59: 753–759.

11.

Colom

Vieta

Suppes

Predominant polarity in bipolar disorders: refining or redefining diagnosis?

Acta Psychiatr Scand 2015; 132: 324–326.

12.

Belizario

Gigante

de Almeida Rocca

, et al. Cognitive impairments and predominant polarity in bipolar disorder: a cross-sectional study. Int J Bipolar Disord 2017; 5: 15.

13.

Pal

A cross-sectional study of internalized stigma in euthymic patients of bipolar disorder across its predominant polarity. Indian J Soc Psychiatry 2020; 36: 73.

14.

Nivoli

AMA

Colom

Pacchiarotti

, et al. Treatment strategies according to clinical features in a naturalistic cohort study of bipolar patients: a principal component analysis of lifetime pharmacological and biophysic treatment options. Eur Neuropsychopharmacol 2013; 23: 263–275.

15.

Cochran

Gitlin

MJ.

Attitudinal correlates of lithium compliance in bipolar affective disorders.

J Nerv Ment Dis 1988; 176: 457–464.

16.

Pal

Sharan

Chadda

RK.

Internalized stigma and its impact in Indian outpatients with bipolar disorder.

Psychiatry Res 2017; 258: 158–165.

17.

De las Cuevas

Peñate

Psychometric properties of the eight-item Morisky Medication Adherence Scale (MMAS-8) in a psychiatric outpatient setting.

Int J Clin Heal Psychol 2015; 15: 121–129. Available from:/pmc/articles/PMC6224788/?report = abstract

18.

Aas

IM.

Global Assessment of Functioning (GAF): properties and frontier of current knowledge. Ann Gen Psychiatry 2010; 9: 20.

19.

Kocalevent

Berg

Beutel

, et al. Social support in the general population: standardization of the Oslo Social Support Scale (OSSS-3). BMC Psychol 2018; 6: 31.

20.

The World Health Organization. WHOQOL-Files, www.who.int/tools/whoqol/whoqol-bref (accessed 30 March 2021).

21.

García

Martínez-Cengotitabengoa

López-Zurbano

, et al. Adherence to antipsychotic medication in bipolar disorder and schizophrenic patients. J Clin Psychopharmacol 2016; 36: 355–371.

22.

Solé

Jiménez

Torrent

, et al. Cognitive impairment in bipolar disorder: treatment and prevention strategies. Int J Neuropsychopharmacol 2017; 20: 670–680.

23.

Burdick

Russo

Frangou

, et al. Empirical evidence for discrete neurocognitive subgroups in bipolar disorder: clinical implications. Psychol Med 2014; 44: 3083.

24.

Baldessarini

Perry

Pike

Factors associated with treatment nonadherence among US bipolar disorder patients.

Hum Psychopharmacol 2008; 23: 95–105.

25.

Colom

Vieta

Daban

, et al. Clinical and therapeutic implications of predominant polarity in bipolar disorder. J Affect Disord 2006; 93: 13–17.

26.

Gonzalez-Pinto

Mosquera

Alonso

, et al. Suicidal risk in bipolar I disorder patients and adherence to long-term lithium treatment. Bipolar Disord 2006; 8: 618–624.

27.

Pal

Predominant polarity in bipolar affective disorder: a scoping review of its relationship with clinical variables and its implications.

Indian J Psychol Med 2019; 41: 9–17.

28.

Baldessarini

Undurraga

Vázquez

, et al. Predominant recurrence polarity among 928 adult international bipolar I disorder patients. Acta Psychiatr Scand 2012; 125: 293–302.

29.

Zeber

Copeland

Good

, et al. Therapeutic alliance perceptions and medication adherence in patients with bipolar disorder. J Affect Disord 2008; 107: 53–62.

30.

Popovic

Torrent

Goikolea

, et al. Clinical implications of predominant polarity and the polarity index in bipolar disorder: a naturalistic study. Acta Psychiatr Scand 2014; 129: 366–374.