Clinical correlates of medication adherence in individuals with bipolar disorder and comorbid hypertension

Abstract

Objective

Individuals with bipolar disorder (BD) have high rates of suboptimal medication adherence, medical illness, and premature mortality, largely from cardiovascular causes. This study examined the association between adherence to antihypertensive, adherence to BD medications, and clinical symptoms in patients with BD and comorbid hypertension (HTN). Participants were involved in an ongoing clinical trial.

Method

Inclusion criteria were a BD diagnosis, treatment with antihypertensives, adherence challenges, and poorly controlled HTN. Adherence was measured via self-report using the Tablets Routine Questionnaire and using eCAP, an electronic pill bottle that captures openings. Average systolic blood pressure (SBP) was calculated from 12 readings over one week. The Montgomery-Asberg Depression Rating Scale (MADRS) and the Brief Psychiatric Rating Scale (BPRS) assessed BD symptoms.

Results

A total of 83 participants with BD and HTN were included. Adherence to BD medications and antihypertensive medications were positively correlated. eCAP openings showed more missed doses than participants self-reported for antihypertensive adherence. BD medication adherence was positively correlated with BPRS at baseline, whereas antihypertensive adherence was negatively correlated with SBP at screening. Antihypertensive adherence improved and SBP decreased between screening and baseline.

Conclusions

Adherence levels fluctuated over time and differed based on measurement method in participants with comorbid BD and HTN. Self-reported BD adherence was positively related to global psychiatric symptoms and antihypertensive adherence was related to better SBP control. Monitoring both medication and blood pressure led to a change in self-reported adherence. BD symptom severity may interfere with medication adherence in patients with BD and should be considered in treatment planning.

Keywords

comorbidity mental illness medical illness adherence measurement psychiatric symptoms

Introduction

Individuals with bipolar disorder (BD) have high rates of suboptimal medication adherence,¹ cardiovascular disease,² and premature mortality, largely due to cardiovascular factors.³ In the general population, treatment with antihypertensives has led to dramatic reductions in national rates of cardiac events including heart attack and stroke.^4,5 However, those with BD often have multiple adherence barriers¹ which prevent them from getting the maximum benefit from effective antihypertensives. Many trials measure and intervene on medication adherence to either psychiatric conditions,^6-8 or chronic health conditions,^9–11 but there is very limited research addressing adherence with BD and hypertension (HTN) medication simultaneously. By better understanding medication-taking behavior for multiple conditions, we have a better chance at successfully intervening and improving patient care.

Intentional (eg, choosing not to take medication) and non-intentional (eg, difficulty with taking medication regularly) medication adherence is a complex multifactorial behavior whose impact on symptoms must be evaluated in the context of co-occurring health conditions.^1,12,13 Intentional medication adherence factors include medication attitudes and illness knowledge, side effects, as well as the overall number of medications (eg, since I must take these medications, perhaps I will forgo others), which may be applied differentially to various mental and chronic health conditions.^14-16 Similarly, unintentional adherence factors include forgetting, impaired planning and organization, and difficulties with cognitive flexibility, often present in BD and compounded by the complexity of one’s overall medication regimen for both psychiatric and other health conditions.^17,18 Finally, mood instability as well as health stability may play a central role in a patient’s ability to stay on track with their overall medication regimen,^19,20 supporting the importance of measuring and intervening on both conditions simultaneously.²¹

In people with multiple chronic conditions, such as BD and HTN, it is not clear that adherence and symptoms are uniformly aligned for each condition. Presuming that the same factors impact medication taking across conditions, one might expect that people with BD and HTN with adherence problems might experience the same barriers for all of their medications. Alternatively, there may be differential attitudes and different medication-taking patterns for psychiatric and somatic conditions in the same individual. Clarifying the relationship between adherence to medicines for multiple conditions, measured via self-report and objective methods, and examining the association between clinical symptoms and adherence patterns has the potential to better inform approaches to care.

This analysis is derived from screening and baseline data collected from the first 83 enrolled participants in an ongoing National Heart, Lung and Blood Institute (NHLBI)-funded randomized controlled trial (RCT) (R01HL149409) testing a novel behavioral mobile health (mhealth) intervention to promote medication adherence to both BD medications and antihypertensives in adults with BD and HTN (see published study protocol for details).²² The current analysis evaluated the relationship between adherence to BD medications and antihypertensives, measured via self-report and validated with objective pill bottle openings. We also looked at the relationship between adherence behavior and clinical symptoms (ie, BD symptoms, systolic blood pressure) and demographic attributes (ie, age, sex, and race) at two time points, screening and baseline. The mean number of days between screen and baseline was 49.69 ± 16.10 (N = 68) and there was no active treatment intervention during this time period, aside from medication and blood pressure monitoring. We hypothesized that there would be a significant correlation between adherence for BD and HTN medications. We also hypothesized that younger age, BD Type I, more severe symptoms, and earlier age of onset would be associated with worse adherence. Finally, we hypothesized that adherence to BD medications would negatively correlate with MADRS and BPRS scores while antihypertensive adherence would negatively correlate with SBP.

Method

Study overview

This ongoing 5-year 2-stage randomized controlled trial (RCT) is evaluating the efficacy of a brief, practical adherence intervention delivered via interactive text messaging (iTAB-CV) along with self-monitoring of medication taking, mood, and home blood pressure compared to self-monitoring alone. Those randomized to iTAB-CV receive daily text messages with predetermined content to address 11 salient domains as well as targeted customized messages for 2 months. This group is then randomized a second time to receive either a high (gradual taper from daily to weekly texts) or low booster (weekly texts) phase for an additional 2 months. The detailed RCT protocol has been published elsewhere.²²

Inclusion criteria

Study inclusion criteria were purposely broad in order to generalize to real-world patients with BD and HTN and required participants to be: (1) between the ages of 21 and 80 years old; (2) have a DSM-5 diagnosis of BD type I or II as confirmed by a structured psychiatric interview, the Mini International Neuropsychiatric Interview (MINI)²³; (3) be taking antihypertensives for clinically diagnosed HTN for at least 6 months; (4) have either current or past challenges with medication adherence to antihypertensives, and (5) have poorly controlled BP as measured by an average systolic BP ≥ 130 at two different measurement sittings. All participants provided written informed consent. The study was approved by the local Institutional Review Board (IRB) and all participants completed informed consent.

Exclusion criteria

In the interest of patient safety, individuals who were at high immediate risk for suicide were excluded from study participation. Secondly, given that the adherence intervention of the RCT trial was delivered via text messages in English, individuals who were monolingual, non-English speaking were excluded. Finally, individuals with an upper arm circumference >50 cm were excluded given the lack of validity of home blood pressure monitoring for such individuals.

Measures

Medication adherence

The Tablets Routine Questionnaire (TRQ) is a self-report measure which identifies the proportion of days with missed medication in the past 7 and past 30 days.^24,25 Lower scores represent better adherence, while higher scores represent worse adherence. In this study, two different TRQ measurements were obtained, TRQ-Bipolar Disorder (TRQ-BD) and TRQ-Hypertension (TRQ-HTN). We collected both past 7 and past 30 days adherence but used past 7 day TRQ in the analyses given that results were nearly identical and participants are more likely to have accurate recall of their behavior in the past week. TRQ-BD was assessed for each regularly scheduled medication prescribed for the treatment of BD (lithium, anticonvulsant, antipsychotic, antidepressant) at both screening and baseline. For individuals on more than 1 medication, an average TRQ-BD was calculated. TRQ-HTN was assessed for each regularly scheduled medication prescribed for the treatment of hypertension at both screening and baseline. For individuals on more than 1 medication, an average TRQ-HTN was calculated.

Electronic monitoring using an eCAP©²⁶ was used to validate the self-report adherence findings for antihypertensives only. The eCAP data was collected for past week and past month at baseline for the antihypertensive dosed most frequently (index drug). If there was more than 1 agent dosed at that frequency, the medicine that was most recently added to the treatment regimen was the index drug. Given that the eCAP was only distributed after the participant passed study screen, eCAP measurement was first collected at the baseline visit, which occurred between 30-60 days following screen.

As-needed or “prn” drugs were not monitored. In previous work by this study team, the correlation between a single “index” drug and all drugs was 95%, providing support for measuring 1 medication as a proxy for medication adherence.²⁷ To ensure accurate data collection of pill-taking behavior, patients were instructed to: (1) take doses of the index medication from the eCAP bottle only, (2) remove only 1 dose at a time, and (3) ingest the index medication immediately after removing it from the eCAP bottle. Adherence was defined as the ratio/percent of days with missed doses vs actual prescribed doses such that higher percentages indicated worse adherence. For example, missing 2 days out of a prescribed 7 days yielded an adherence percentage of 28.6%.

Clinical symptoms

The Montgomery-Asberg Depression Rating Scale (MADRS),²⁸ a clinician rated measure of depression, was administered at both screen and baseline. The MADRS is one of the most widely used measures of depression in clinical research with a large body of evidence of strong psychometric properties.^29,30 Total scores on MADRS range from 0-60, with higher scores indicating more severe depression.

The Brief Psychiatric Rating Scale (BPRS)³¹ was used to measure global psychopathology at baseline. The BPRS is a clinician rated measure that evaluates the spectrum of symptoms seen in individuals with BD (including mania, psychosis, depression and disorganization). There is extensive research on the BPRS with evidence for good internal consistency, test-retest reliability, and construct validity.^31-33 Total scores range from 18 to 126 with higher scores indicative of more severe symptoms.

Average systolic blood pressure (SBP) and diastolic blood pressure (DBP) were calculated from 12 readings over the course of a week, half of which were taken in the morning and half in the evening based on a BP protocol of the Hypertension Society and the American Heart Association [24, 25]. Participants were sent an Omron Series 3® upper arm blood pressure monitor at screening and underwent standardized training on its use by study staff. Participants measured their BP while a member of the study staff observed them over a video platform. In cases where real-time video observation on BP measurement was not possible or practical, other methods were permitted such as sending in a picture of the BP monitor via text or calling in the readings. If there was more than a 10 mmHg difference between the two readings for either SBP or DBP at a sitting, participants were asked to take another 2 readings and the 4 readings were averaged. Once 12 readings were obtained (6 in the AM and 6 in the PM), BP readings were averaged. These procedures were repeated for the week prior to the baseline assessment to ascertain average BP readings.

Screening and Baseline Procedures

While all data were remotely collected using RedCap³⁴ at one academic medical center in Northeast Ohio, participants were recruited using multiple methods including flyers, clinical referrals, ResearchMatch, social media, electronic medical records, and referrals from other medical centers in the United States. Following the screening period in which all participants were sent an eCAP and trained on its use, there was a waiting period of approximately 4 weeks during which adherence was monitored via eCAP and TRQ. The current analysis is limited to screening and baseline data, prior to randomization, for the first 83 participants who passed screen and attended their baseline visit.

Statistical analyses

Non-parametric Spearman’s rank correlations and point biserial correlations were used to examine the correlations between TRQ-BD and TRQ-HTN in relation to continuous and binary demographic (age, sex, race) and clinical variables (BD type, years with BD, years with hypertension, number of medications), respectively. We examined correlations between TRQ-BD and MADRS at screening and baseline, between TRQ-BD and BPRS (not administered at screen) at baseline and between TRQ-HTN and SBP at screening and baseline. Furthermore, the relationship between TRQ-HTN (self-report) and eCAP (objective pill openings; not measured at screen) was assessed at baseline. Finally, change in self-reported adherence as well as BP from screen to baseline were assessed using the non-parametric Wilcoxon signed rank test. A two-sided alpha of 5% was considered statistically significant.

Results

Demographics and clinical characteristics of this interim sample (N = 83) are presented in Tables 1 and 2, respectively. The mean age of the sample was 54.0 years ± 10.70 with a majority women (53, 63.9%). The sample was racially diverse with 37 (44.6%) Black participants. Mean years of education was 14.11 ± 2.74 and the majority of the sample had BD Type I (71, 85.5%).

Table 1.

Demographics of sample of 83 participants with bipolar disorder and hypertension.

	N, Mean (SD)
Variables	N	M (SD)	N (%)
Age	83	54.00 (10.70)
Education (in years)	83	14.11 (2.74)
Biological sex	83
Female			53 (63.90)
Male			30 (36.10)
Race	83
White or Caucasian			42 (50.60)
Black or African American			37 (44.60)
Other			4 (4.80)
Ethnicity	83
Hispanic/Latino			3 (3.60)
Not hispanic			79 (95.20)
Don’t know			1 (1.20)
Marital status	83
Single, never married			30 (36.10)
Married or Cohabitating			24 (28.90)
Divorced			22 (26.50)
Other (Separated/Widowed)			7 (8.50)
Living Situation	83
Alone			35 (42.20)
With others			48 (57.80)
Occupation	83
Full time			22 (26.50)
Part time			7 (8.40)
Unemployed (Expected to work)			3 (3.60)
Unemployed (Disabled)			44 (53.00)

Table 2.

Clinical characteristics of 83 participants with bipolar disorder and hypertension.

Variables	N, Mean (SD)		N (%)
Variables	N	M (SD)	N (%)
Bipolar diagnosis	83
Type 1			71 (85.50)
Type 2			12 (14.50)
Age of bipolar disorder onset	80	24.08 (11.87)
Psychiatry family history	82
Yes			60 (73.20)
No			17 (20.70)
Don’t know			5 (6.10)
Number of previous psychiatric hospitalizations	83	4.27 (7.78)
Age of HTN diagnosis	83	38.30 (11.28)
Age of BD diagnosis	83	34.58 (12.05)
Systolic blood pressure screen	83	142.13 (13.04)
Systolic blood pressure baseline	67	138.23 (17.17)
Diastolic blood pressure screen	83	89.00 (11.59)
Diastolic blood pressure baseline	67	87.50 (10.33)
Montgomery Asberg Depression Rating Scale (MADRS)
Screen	83	18.95 (11.01)
Baseline	68	16.90 (9.42)
Brief Psychiatric Rating Scale (BPRS) baseline	68	29.43 (6.55)
BMI	83	34.16 (8.32)
Smoking
No			57 (68.70)
Yes			26 (31.30)
High Cholesterol
No			20 (36.4)
Yes			32 (58.2)
Do not know			3 (5.5)

Adherence characteristics are presented in Table 3. Mean past week TRQ-BD was 30.38% ± 29.86 (N = 79) at screen and 20.76% ± 29.46 (N = 64) at baseline. Mean TRQ-HTN was 31.21% ± 22.87 (N = 83) at screen and 13.76% ± 18.77 (N = 68) at baseline. At baseline, eCAP (N = 50) was 35.43% ± 33.69. Correlations between adherence and clinical characteristics are presented in Table 4. With regard to the relationship between demographic characteristics and adherence, while there was no significant correlation between TRQ-BD and age at screen (P > .05), there was a statistically significant inverse correlation between TRQ-HTN and age at screen (r_s = −.25, P < .05, N = 83), such that older age was associated with better antihypertensive adherence. At baseline, however, TRQ-HTN and age were no longer correlated. With regard to race, adherence was worse in Whites than other racial groups based on TRQ-BD and TRQ-HTN at screen (r_s = 0.22, P < .05, N = 83) and TRQ-BD at baseline (r_s = 0.29, P < .05, N = 68). Another significant correlation was TRQ-BD and sex such that females had higher (worse) TRQ-BD scores than men at screen (r_s = 0.25, P < .05, N = 83. Finally, BD duration and TRQ-BD were significantly correlated at baseline such that longer duration was associated with worse adherence (r_s = 0.27, P < .05, N = 68). No such significant correlations were found between adherence and BD type, HTN duration, or number of HTN medications.

Table 3.

Adherence characteristics of 83 participants with bipolar disorder (BD) and hypertension (HTN).

Variables	N, Mean (SD)
Variables	N	M (SD)
Mean number of HTN medications
Screen	83	1.94 (0.99)
Baseline	68	2.04 (1.03)
Mean number of BD medications
Screen	83	2.00 (1.15)
Baseline	68	1.96 (1.09)
TRQ-HTN^a at screen (%)
Past 7 days	83	31.21 (22.87)
Past 30 days	83	25.69 (18.50)
TRQ-HTN at baseline (%)
Past 7 days	68	13.76 (18.77)
Past 30 days	68	11.59 (12.58)
TRQ-BD^b at screen (%)
Past 7 days	79	30.38 (29.86)
Past 30 days	79	26.37 (28.65)
TRQ-BD at baseline (%)
Past 7 days	64	20.76 (29.46)
Past 30 days	64	16.61 (24.68)
eCAP^c at baseline (%)
Past 7 days	50	35.43 (33.69)
Past 30 days	50	31.87 (29.50)

^aTRQ-BD = Tablets Routine Questionnaire – Bipolar Disorder; higher scores represent worse adherence.

^bTRQ-HTN = Tablets Routine Questionnaire – Hypertension; higher scores represent worse adherence.

^ceCAP = electronic pill cap; higher scores represent worse adherence.

Table 4.

Correlations between adherence at screen and baseline and clinical characteristics in participants with bipolar disorder and hypertension.

Time point	Variable	Screen		Baseline
Time point	Variable	TRQ-BD past week	TRQ-HTN past week	TRQ-BD past week	TRQ-HTN past week	eCAP past week
Screen	TRQ-BD^a past week	1
	TRQ-HTN^b past week	0.44; P = 0.000041	1
	MADRS^c	0.07; P = 0.52	−0.02; P = 0.85
	Mean systolic bp	0.19; P = 0.09	0.31; P = 0.0042
Baseline	TRQ-BD past week			1
	TRQ-HTN past week			0.31; P = 0.013	1
	eCAP past week			0.21; P = 0.16	0.35; P = 0.013	1
	MADRS			0.16; P = 0.20	−0.13; P = 0.29	−0.20; P = 0.16
	BPRS^d			0.25; P = 0.047	−0.05; P = 0.69	−0.15; P = 0.29
	Mean systolic bp			−0.29; P = 0.023	−0.02; P = 0.87	0.22; P = 0.13
Age		−0.12, P = 0.31	−0.25; P = 0.022	−0.07; P = 0.59	−0.02; P = 0.84	−0.20; P = 0.16
Sex (female vs male)		0.25; P = 0.028	0.11; P = 0.33	0.21; P = 0.10	−0.10; P = 0.42	−0.10; P = 0.42
Race (Black/Other vs White)		−0.22; P = 0.049	−0.22; P = 0.046	−0.29; P = 0.018	−0.14; P = 0.25	−0.14; P = 0.25
BD type (I vs II)		0.06; P = 0.60	0.17; P = 0.14	0.11; P = 0.97	0.09; P = 0.47	0.09; P = 0.47
BD duration		0.17; P = 0.14	0.05; P = 0.65	0.27; P = 0.030	0.01; P = 0.93	−0.05; P = 0.70
HTN duration		−0.10; P = 0.37	0.03; P = 0.82	0.03; P = 0.82	−0.00; P = 0.99	0.12; P = 0.40

^aTRQ-BD = Tablets Routine Questionnaire – Bipolar Disorder.

^bTRQ-HTN = Tablets Routine Questionnaire – Hypertension.

^cMADRS = Montgomery Asberg Depression Rating Scale.

^dBPRS = Brief Psychiatric Rating Scale.

With regard to the relationship between adherence measures and clinical symptoms, there was a significant correlation between self-reported adherence to BD medications (TRQ-BD) and antihypertensives (TRQ-HTN) (r_s = .44, P < .001, N = 79) at screen as well as baseline (r_s = .28, P < .01, N = 64). There was also a significant correlation between TRQ-HTN and eCAP at baseline (r_s = 0.351, p < = .05, N = 50) yet eCAP indicated more days with missed doses (mean = 35.43% ± 33.69) than TRQ-HTN (mean = 13.76% ± 18.77) [z = −4.12, P < 0.001, N = 50]. Additionally, TRQ-BD was positively correlated with BPRS at baseline (r_s = 0.30, P < .05, N = 64; r_s = 0.249, P < .05, N = 64, respectively) such that worse adherence was associated with more symptoms. No such relationship was seen for TRQ-BD and MADRS at either screen or baseline (P > .05). TRQ-HTN was positively correlated with SBP at screen (r_s = 0.311, P < 0.01, N = 83) such that worse adherence was associated with higher SBP. At baseline, SBP was not correlated with TRQ-HTN (P > .05) but there was a trend with eCAP (r_s = 0.23, P = 0.10, N = 50).

With regard to change in self-reported adherence, there was a significant improvement in TRQ-HTN from screen to baseline (z = 4.51 P < 0.01) and a trend for TRQ-BD (z = 1.88, P = .06). In parallel, SBP significantly decreased from screen to baseline (z = −2.76, P < 0.01, N = 67). There was no such decrease in symptoms of depression from screen to baseline.

Discussion

Patients with BD have high rates of suboptimal-medication adherence, which may lead to relapse and complications secondary to poorly managed medical illnesses.^1,35 This analysis, using screening and baseline data from an ongoing RCT, provides information on this at-risk group of individuals. Strengths of the study include a racially diverse sample and inclusion of patients with BD who are often excluded from standard RCTs by virtue of being poorly adherent with prescribed medication. Additionally, given that all procedures were remote, participants were recruited nationally across the United States.

The study results suggest that adherence levels for both BD and HTN medication treatments varied widely and that there are significant differences between adherence levels assessed using different measurement approaches. Despite a strong association between self-report and objective measurement of antihypertensive adherence, objective pill openings identified rates of missed medications approximately 22% higher than self-report. This is consistent with other reports in the literature for antihypertensive adherence.³⁶ Adherence behaviors related to taking prescribed medications for BD and for HTN were roughly similar, and for both BD and HTN medications, self-reported adherence improved shortly after study enrollment, likely due to participation in an adherence study and the fact that adherence behaviors were being closely monitored.

Consistent with the BD adherence literature, and supporting our hypotheses, worse self-reported BD adherence was associated with worse BD symptoms and longer illness duration.^1,37,38 No such relationship was found for BD Type, most likely because the large majority of the sample had BD Type I. Also consistent with our hypotheses, worse self-reported antihypertensive adherence was associated with higher SBP. Aligned with studies showing worse adherence among younger psychiatrically ill individuals with medical morbidities,³⁹ in our sample of patients with BD and poorly controlled hypertension, older age was associated with better antihypertensive adherence. No such relationship was found, however, for adherence to mood stabilizers.

With respect to the relationship between sex and race variables and adherence, we found significant results. Inconsistent with some of the BD literature,^40,41 in our sample, worse self-reported adherence to BD medications was reported by women than men at screen. With regard to race, White participants self-reported worse adherence to both antihypertensive and BD medications than other racial groups. Potential explanatory factors include the medical complexity of the sample, a relatively large percentage of non-White participants, and differences in willingness to acknowledge the extent of adherence challenges.

Our findings also speak to the impact of adherence monitoring on medication-taking behaviors. We found a notable Hawthorne effect, with a large improvement in self-reported adherence between the screening and baseline assessments, a timeframe of approximately 50 days. Perhaps, participating in an adherence-focused research study improves medication taking behavior. This is consistent with other medication adherence trials both in BD and HTN.²¹ The duration of the Hawthorne effect could not be determined based on the current analyses as we were looking only at screening and baseline data. Furthermore, there is some evidence that electronic pill monitoring alone changed adherence behavior yet improved adherence was not always associated with improved symptoms.⁴² In our analysis, self-reported BD adherence was related to the severity of global psychiatric symptoms, including items related to manic but not depressive symptoms. We also found that while antihypertensive adherence was related to lower SBP at screen, this was not the case at baseline.

There are several clinical and research implications of our findings. First, greater overall BD symptom severity may be a clinical indicator to assess for adherence problems. A narrative review by Jawad et al found that level of adherence to medication may fluctuate with changes in psychiatric status.³⁵ This review also found that sub-optimal adherence was associated with rapid cycling, suicide attempts, earlier onset of illness, active anxiety or alcohol use disorder, a greater number and severity of affective symptoms, incomplete response to treatment, comorbid obsessive compulsive disorder, and a recent manic or hypomanic episode.^27,43-45 We did not find an association between adherence to BD medications and depressive symptoms. However, it must be noted that based on MADRS norms,²⁹ our sample had relatively mild levels of depressive symptoms with a mean total MADRS score of 18.95 (SD 11.01) at screening and 16.90 (9.42) at baseline.²⁹ It is important to note that patients may have had manic symptoms but this aspect of the illness was only evaluated with 2 items from the BPRS, assessing grandiosity and excitement. Perhaps if our sample had been more symptomatic with respect to depressive symptoms, we would have found an association between depressive symptoms and adherence scores. Another factor that may have been at play is that measures were conducted remotely using videoconferencing. As such, it was more difficult to assess observable aspects of depression, such as psychomotor slowing, which may have contributed to lower severity scores. While previous research reports high reliability between remote and in-person assessment using the MADRS structured interview,⁴⁶ further study is warranted given the increasing use of telehealth in both psychiatric and medical care settings.

Another finding of interest is the improvement in SBP from screening to baseline with no active intervention beyond electronic pill and BP monitoring. This finding is consistent with the literature on the value of home BP monitoring as one aspect of HTN management.^47,48 However, further study is needed to understand the extent and duration of the monitoring effect as it relates to long-term home-based BP monitoring.

It is also important to note that our sample of patients were prescribed approximately 2 medications to treat their BD and 2 medications to treat their HTN, a substantial medication burden. In other areas of medicine, increased medication regime complexity is associated with worse adherence. For example, a systematic review of 76 studies (mainly in cardiovascular disease) that used electronic monitoring devices found significantly higher adherence with once daily dosing compared with three times a day.⁴⁹ Polypharmacy, combined with having 2 chronic health conditions that each require substantial life-style changes to optimize health, may set the stage for ongoing poor adherence and both physical and mental health complications. The U.K.’s National Institute for Health and Care Excellence (NICE) guidelines for BD Assessment and Management advise clinicians to take into account a patient’s preference and other clinical factors such as comorbid physical health to inform medication treatment.^50,51 Coordinating care to lower polypharmacy may minimize medication burden and optimize future adherence. Shared electronic health records may also allow for more such collaboration and coordination.

Study limitations

While our study has some findings with clinically relevant implications, results should be interpreted with caution given the methodological limitations. First, the sample size is relatively small and the selection of individuals who agree to participate in an RCT may not represent all patients with BD and HTN, thus limiting generalizability. Also, BD symptom levels were relatively mild in this sample, providing a limited opportunity to assess effects of symptom severity and adherence. Finally, an estimation of longer-term adherence trajectory could not be determined as the analyses were restricted to screening and baseline time points. Despite these limitations, our findings may help to guide treatment approaches and shape avenues for future research in the area of medication adherence in patients with multimorbid conditions.

Conclusions

In this interim analysis of an ongoing clinical trial, we investigated medication adherence among individuals with BD and HTN, highlighting the complex interplay between these comorbidities. We found that adherence to both BD and HTN medications varied widely, with objective measures revealing worse adherence than self-report. BD medication adherence was positively correlated with global psychopathology and antihypertensive adherence was negatively correlated with SBP at screening. Antihypertensive adherence improved and SBP decreased between screening and baseline, during which time medication and BP were being monitored. Limitations include a small sample size, low BD symptom severity, and restriction to screening and baseline data, which may limit generalizability. Nonetheless, our findings provide valuable insights into managing medication adherence in this vulnerable population, suggesting potential avenues for future research and improved clinical care.

Footnotes

Author Notes

Portions of this manuscript were presented at the American Heart Association EPI/Lifestyles Scientific Sessions, Chicago, IL, March, 2024.

Author Contributions

The authors have all made substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND the drafting the work or revising it critically for important intellectual content; AND give their final approval of the version to be published; AND agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The specific contributions of the authors are as follows: Jennifer B. Levin - design of study, conceptualization, critical review of results, substantial writing; David J. Moore - conceptualization, critical review of results and implications. Farren Briggs - data analysis and interpretation of data; development of models. Mahboob Rahman - methodological expertise related to hypertension and blood pressure management; Jessica Montoya - conceptualization, critical review of results and implications. Colin Depp - conceptualization, critical review of results and implications. Douglas Einstadter -methodological expertise related to hypertension and blood pressure management; Kurt C. Stange - methodological expertise related to hypertension and blood pressure management; Celeste Weise - implementation of study procedures; Taylor Maniglia - implementation of study procedures; Richard Barigye - data analysis and oversight of data integrity; Gracie Howard Griggs - implementation of study procedures; Clara Adeniyi - project administration, oversight of implementation of protocol; Joy Yala - oversight of data integrity and analysis; Martha Sajatovic - design of study, conceptualization, critical review of results and implications.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Author M.S.: Research grants within past 3 years: Intra-Cellular, Merck, Otsuka, Alkermes, International Society for Bipolar Disorders (ISBD), National Institutes of Health (NIH), Centers for Disease Control and Prevention (CDC), Patient-Centered Outcomes Research Institute (PCORI). Consultant in the past year: Alkermes, Otsuka, Janssen, Lundbeck, Teva, Neurelis. Royalties: Springer Press, Johns Hopkins University Press, Oxford Press, UpToDate.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research reported in this publication was supported by the National Heart, Lung, And Blood Institute of the National Institutes of Health under Award Number R01HL149409. Support was also received from the Clinical and Translational Science of Cleveland, UL1TR000439 from the National Center for Advancing Translational Sciences (NCATS) component of the National Institutes of Health and NIH roadmap for Medical Research. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

ORCID iD

Jennifer B. Levin

References

Levin

Krivenko

Howland

Schlachet

Sajatovic

. Medication adherence in patients with bipolar disorder: a comprehensive review. CNS Drugs. 2016;30:819-835. doi:10.1007/s40263-016-0368-x.

Goldstein

Schaffer

Wang

Blanco

. Excessive and premature new-onset cardiovascular disease among adults with bipolar disorder in the US NESARC cohort. J Clin Psychiatry. 2015;76:163-169. doi:10.4088/JCP.14m09300.

Weiner

Warren

Fiedorowicz

. Cardiovascular morbidity and mortality in bipolar disorder. Ann Clin Psychiatr. 2011;23:40-47.

Corrao

Parodi

Nicotra

, et al. Better compliance to antihypertensive medications reduces cardiovascular risk. J Hypertens. 2011;29:610-618. doi:10.1097/HJH.0b013e328342ca97. Research Support, Non-U.S. Gov’t 2010/12/16.

Perreault

Dragomir

Roy

, et al. Adherence level of antihypertensive agents in coronary artery disease. Br J Clin Pharmacol. 2010;69:74-84. doi:10.1111/j.1365-2125.2009.03547.x.

Chakrabarti

. Treatment attitudes and adherence among patients with bipolar disorder: a systematic review of quantitative and qualitative studies. Harv Rev Psychiatr. 2019;27:290-302. doi:10.1097/HRP.0000000000000228.

McVoy

Levin

. Updated strategies for the management of poor medication adherence in patients with bipolar disorder. Expert Rev Neurother. 2023;23:365-376. doi:10.1080/14737175.2023.2198704.

Chauhan

Chakrabarti

Grover

. Identifying poor adherence in outpatients with bipolar disorder: a comparison of different measures. J Neurosci Rural Pract. 2022;13:12-22. doi:10.1055/s-0041-1736155.

Choudhry

Kronish

Vongpatanasin

, et al. Medication adherence and blood pressure control: a scientific statement from the American heart association. Hypertension. 2021;79:e1-e14. doi:10.1161/HYP.0000000000000203.

10.

Algabbani

. Treatment adherence among patients with hypertension: findings from a cross-sectional study. Clin Hypertens. 2020;26:18. doi:10.1186/s40885-020-00151-1.

11.

Kronish

Moise

McGinn

, et al. An electronic adherence measurement intervention to reduce clinical inertia in the treatment of uncontrolled hypertension: the MATCH cluster randomized clinical trial. J Gen Intern Med. 2016;31:1294-1300. doi:10.1007/s11606-016-3757-4.

12.

Levin

Aebi

Howland

, et al. The relationship between medication attitudes and medication adherence behavior in adults with bipolar disorder. J Nerv Ment Dis. 2020;208:87-93. doi:10.1097/NMD.0000000000001083.

13.

Klein

Aebi

Sajatovic

, et al. Differential medication attitudes to antihypertensive and mood stabilizing agents in response to an automated text-messaging adherence enhancement intervention. J Behav Cogn Ther. 2020;30:57-64. doi:10.1016/j.jbct.2020.03.015.

14.

Klein

Aebi

Sajatovic

15.

Mukhtar

Weinman

Jackson

. Intentional non-adherence to medications by older adults. Drugs Aging. 2014;31:149-157. doi:10.1007/s40266-014-0153-9.

16.

Kassaoui

Bouhlel

Nakhli

Ben Mahmoud

Ben Nasr

Ben Hadj Ali

. Factors related to treatment compliance in bipolar disorder. Tunis Med. 2016;94:60-65.

17.

Bogner

de Vries

O'Donnell

Morales

. Measuring concurrent oral hypoglycemic and antidepressant adherence and clinical outcomes. Am J Manag Care. 2013;19:e85-e92. Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov't 2013/03/29.

18.

Zogg

Woods

Sauceda

Wiebe

Simoni

. The role of prospective memory in medication adherence: a review of an emerging literature. J Behav Med. 2012;35:47-62. doi:10.1007/s10865-011-9341-9.

19.

Levin

Krivenko

Bukach

Tatsuoka

Cassidy

Sajatovic

. A reexamination of nonpsychiatric medication adherence in individuals with bipolar disorder and medical comorbidities. J Nerv Ment Dis. 2016;205:182-187. doi:10.1097/NMD.0000000000000544.

20.

Levin

Aebi

Tatsuoka

Cassidy

Sajatovic

. Adherence to psychotropic and nonpsychotropic medication among patients with bipolar disorder and general medical conditions. Psychiatr Serv. 2016;67:342-345. doi:10.1176/appi.ps.201500010.

21.

Levin

Sajatovic

Rahman

, et al. Outcomes of psychoeducation and a text messaging adherence intervention among individuals with hypertension and bipolar disorder. Psychiatr Serv. 2019;70:608-612. doi:10.1176/appi.ps.201800482.

22.

Levin

Moore

Depp

, et al. Using mHealth to improve adherence and reduce blood pressure in individuals with hypertension and bipolar disorder (iTAB-CV): study protocol for a 2-stage randomized clinical trial. Trials. 2022;23:539. doi:10.1186/s13063-022-06449-9.

23.

Sheehan

Lecrubier

Sheehan

, et al. The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry. 1998;59(Suppl 20):22-57. 22-33;quiz 34-57. Comparative Study Historical Article Research Support, Non-U.S. Gov't Review 1999/01/09.

24.

Peet

Harvey

. Lithium maintenance: 1. A standard education programme for patients. Br J Psychiatry. 1991;158:197-200.

25.

Scott

Pope

. Nonadherence with mood stabilizers: prevalence and predictors. J Clin Psychiatry. 2002;63:384-390.

26.

Corporation

. eCAP … The smart cap that enhances medication adherence. 2015. https://www.informationmediary.com/ecap

27.

Sajatovic

Levin

Sams

, et al. Symptom severity, self-reported adherence, and electronic pill monitoring in poorly adherent patients with bipolar disorder. Bipolar Disord. 2015;17:653-661. doi:10.1111/bdi.12326.

28.

Montgomery

Asberg

. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979;134:382-389.

29.

Müller

Himmerich

Kienzle

Szegedi

. Differentiating moderate and severe depression using the Montgomery-Asberg depression rating scale (MADRS). J Affect Disord. 2003;77:255-260. doi:10.1016/s0165-0327(02)00120-9.

30.

Leentjens

Verhey

Lousberg

Spitsbergen

Wilmink

. The validity of the Hamilton and Montgomery-Asberg depression rating scales as screening and diagnostic tools for depression in Parkinson’s disease. Int J Geriatr Psychiatr. 2000;15:644-649.

31.

Leucht

Kane

Kissling

Hamann

Etschel

Engel

. Clinical implications of brief psychiatric rating scale scores. Br J Psychiatry. 2005;187:366-371. doi:10.1192/bjp.187.4.366.

32.

Leucht

Engel

Davis

, et al. Equipercentile linking of the brief psychiatric rating scale and the clinical global impression scale in a catchment area. Eur Neuropsychopharmacol. 2012;22:501-505. doi:10.1016/j.euroneuro.2011.11.007.

33.

Shafer

. Meta-analysis of the brief psychiatric rating scale factor structure. Psychol Assess. 2005;17:324-335. doi:10.1037/1040-3590.17.3.324.

34.

Harris

Taylor

Thielke

Payne

Gonzalez

Conde

. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inf. 2009;42:377-381. doi:10.1016/j.jbi.2008.08.010.

35.

Jawad

Watson

Haddad

Talbot

McAllister-Williams

. Medication nonadherence in bipolar disorder: a narrative review. Ther Adv Psychopharmacol. 2018;8:349-363. doi:10.1177/2045125318804364.

36.

Ayoade

Oladipo

. Evaluation of the correlation between self-report and electronic monitoring of adherence to hypertension therapy. Blood Pres. 2012;21:161-166. doi:10.3109/08037051.2012.679029.

37.

Leclerc

Mansur

Brietzke

. Determinants of adherence to treatment in bipolar disorder: a comprehensive review. J Affect Disord. 2013;149:247-252. doi:10.1016/j.jad.2013.01.036.

38.

Gianfrancesco

Rajagopalan

Sajatovic

Wang

. Treatment adherence among patients with bipolar or manic disorder taking atypical and typical antipsychotics. J Clin Psychiatry. 2006;67:222-232. Comparative Study Research Support, Non-U.S. Gov’t 2006/03/29.

39.

Casaletto

Kwan

Montoya

, et al. Predictors of psychotropic medication adherence among HIV+ individuals living with bipolar disorder. Int J Psychiatr Med. 2016;51:69-83. doi:10.1177/0091217415621267.

40.

Belzeaux

Correard

Boyer

, et al. Depressive residual symptoms are associated with lower adherence to medication in bipolar patients without substance use disorder: results from the FACE-BD cohort. J Affect Disord. 2013;151:1009-1015. doi:10.1016/j.jad.2013.08.028.

41.

Zeber

Miller

Copeland

, et al. Medication adherence, ethnicity, and the influence of multiple psychosocial and financial barriers. Adm Policy Ment Health. 2011;38:86-95. doi:10.1007/s10488-010-0304-1.

42.

Chan

AHY

Foot

Pearce

Horne

Foster

Harrison

. Effect of electronic adherence monitoring on adherence and outcomes in chronic conditions: a systematic review and meta-analysis. PLoS One. 2022;17:e0265715. doi:10.1371/journal.pone.0265715.

43.

Perlis

Ostacher

Miklowitz

, et al. Clinical features associated with poor pharmacologic adherence in bipolar disorder: results from the STEP-BD study. J Clin Psychiatry. 2010;71:296-303. doi:10.4088/JCP.09m05514yel.

44.

Baldessarini

Perry

Pike

. Factors associated with treatment nonadherence among US bipolar disorder patients. Hum Psychopharmacol. 2008;23:95-105. doi:10.1002/hup.908.

45.

Perlis

Ostacher

Goldberg

, et al. Transition to mania during treatment of bipolar depression. Neuropsychopharmacology. 2010;35:2545-2552. doi:10.1038/npp.2010.122.

46.

Williams

Kobak

. Development and reliability of a structured interview guide for the Montgomery Asberg depression rating scale (SIGMA). Br J Psychiatry. 2008;192:52-58. doi:10.1192/bjp.bp.106.032532.

47.

Tucker

Sheppard

Stevens

, et al. Self-monitoring of blood pressure in hypertension: a systematic review and individual patient data meta-analysis. PLoS Med. 2017;14:e1002389. doi:10.1371/journal.pmed.1002389.

48.

Agarwal

Bills

Hecht

Light

. Role of home blood pressure monitoring in overcoming therapeutic inertia and improving hypertension control: a systematic review and meta-analysis. Hypertension. 2011;57:29-38. doi:10.1161/HYPERTENSIONAHA.110.160911.

49.

Claxton

Cramer

Pierce

. A systematic review of the associations between dose regimens and medication compliance. Clin Ther. 2001;23(8):1296-1310. doi:10.1016/s0149-2918(01)80109-0

50.

Kendall

Morriss

Mayo-Wilson

, et al. Assessment and management of bipolar disorder: summary of updated NICE guidance. BMJ. 2014;349:g5673. doi:10.1136/bmj.g5673.

51.

(UK) NCCfMH . Bipolar disorder: the NICE guideline on the assessment and management of bipolar disorder in adults, children and young people in primary and secondary care. 2014.