ANTIMUSCARINIC USE AMONG OLDER ADULTS WITH DEMENTIA AND OVERACTIVE BLADDER: A MEDICARE BENEFICIARIES STUDY

Abstract

OBJECTIVES:

This study examined the incidence and predictors of antimuscarinic medication use including non-selective antimuscarinics among older adults with dementia and overactive bladder (OAB).

METHODS:

The study used a new-user cohort design involving older adults (≥65 years) with dementia and OAB based on 2013–2015 Medicare Data. Antimuscarinics included non-selective (oxybutynin, tolterodine, trospium, fesoterodine) and selective (solifenacin, darifenacin) medications. Descriptive statistics and multivariable logistic regression models were used to determine the incidence and predictors of new antimuscarinic use including non-selective antimuscarinics, respectively.

RESULTS:

Of the 3.38 million Medicare beneficiaries with dementia, over one million (1.05) had OAB (31.03%). Of those, 287,612 (27.39%) were reported as prevalent antimuscarinics users. After applying continuous eligibility criteria, 21,848 (10.34%) incident antimuscarinic users were identified [77.6% non-selective; 22.4% selective]. Most frequently reported antimuscarinics were oxybutynin (56.3%) and solifenacin (21.4%). Multivariable analysis revealed that patients ≥75 years, of black race, and those with schizophrenia, epilepsy, delirium, and Elixhauser’s score were less likely to initiate antimuscarinics. Women, those with abnormal involuntary moments, bipolar disorder, gastroesophageal reflux disease, insomnia, irritable bowel syndrome, muscle spasm/low back pain, neuropathic pain, benign prostatic hyperplasia, falls/fractures, myasthenia gravis, narrow-angle glaucoma, Parkinson’s disease, syncope, urinary tract infection and vulvovaginitis were more likely to initiate antimuscarinics. Further, patients with muscle spasms/low back pain, benign prostatic hyperplasia and those taking higher anticholinergics had lower odds of receiving non-selective antimuscarinics, whereas white patients, black patients and those with schizophrenia and delirium were more likely to receive them.

CONCLUSIONS:

Nearly one-third of dementia patients had OAB and over one-fourth of them used antimuscarinics. Majority of the incident users were prescribed a non-selective antimuscarinic. Several demographic and clinical factors contributing to this behaviour. Given the high prevalence of OAB among dementia patients, there is a need to optimize their antimuscarinics use, considering their vulnerability for anticholinergic adverse effects.

INTRODUCTION

Overactive Bladder (OAB) often coexists among older adults with dementia patients. Resnick et al. conducted clinical and physiologic studies among dementia patients where the predominant cause of incontinence was reported to be detrusor overactivity in 61% patients.¹ Cystometry studies by Mori et al. also found 91% of the dementia patients with detrusor overactivity.² Dementia patients most often develop OAB with disease progression.³ Overactive bladder (OAB) - a major component of urinary incontinence (UI) is the most often cited cause of UI in dementia patients.¹ With an estimated prevalence ranging from 11–19% in both men and women, that increases with age, OAB carries a substantial economic burden.^4–8 In 2000, the total estimated cost of OAB in US was $12.02 billion, classified under direct costs ($11.18 billion) and indirect costs, that is, lost productivity ($0.84 billion).⁹ A prevalence-based model projected the total national costs for OAB to be $76.2 billion and $82.6 billion in 2015 and 2020, respectively.¹⁰ Besides presenting with bothersome symptoms, OAB can have significant detrimental effects on sleep, mental health, work productivity and overall health-related quality of life.^5,11

After behavioral approaches (bladder retraining, lifestyle changes or strategies to control urgency¹²) fail to provide satisfactory outcomes, antimuscarinic medications are the cornerstone of medical treatment for OAB.¹³ These muscarinic receptor antagonists block the muscarinic receptors (M2 and M3) on the detrusor and urothelium muscle, resulting in decreased bladder contractions and reduced sensation.¹⁴ Although all available antimuscarinics have comparable efficacy in reducing the symptoms of OAB,^15–16 there are differences in their safety and tolerability profiles. Safety and tolerability depend mainly on drug selectivity for the bladder over other organs, differential selectivity for muscarinic receptor subtypes (M1-M5) and ability to penetrate the blood brain barrier.¹⁷ Consequently, antimuscarinics such as oxybutynin, tolterodine, trospium and fesoterodine are non-selective as they have affinity for all muscarinic receptors (M1-M5), while others such as darifenacin and solifenacin are selective due to their high affinity for M3 receptors that are responsible for bladder contraction. This poses a significant threat in dementia patients as inhibition of the M1 receptors leads to adverse cognitive effects.¹⁸

Mounting evidence suggests that antimuscarinic agents are also associated with differential effects on cognitive and cardiac parameters.^19–21 These adverse effects are believed to be an outcome of the differential binding of antimuscarinic drugs to muscarinic receptor subtypes that have minimal or no involvement in bladder detrusor contractions, ie, M1, M2, M4 and M5 receptors.^{12,19, 22–32} Since the non-selective agents have affinity for all muscarinic receptors, they have the potential to lead to adverse effects of differing extents and affect the tolerability. That, in turn, may be a key reason for low adherence and persistence seen with OAB medications. Such differential receptor selectivity could cause selective antimuscarinics to offer advantages over non-selective agents with respect to adverse effects as also stated by prior research.^{24,30,33–37} Current literature also suggests that antimuscarinics with selectivity for M3 over M1 or M2 receptors, or limited CNS penetration, or both can potentially render a favorable balance of safety and efficacy in treating OAB, together with reduced cognitive adverse effects in the older population.³⁸

Little is known about the real-world use of antimuscarinics among older adults with dementia and OAB, and even limited data exists regarding the use of selective and non-selective antimuscarinics. As such, most of large scale epidemiological studies conducted on OAB, like the EPIC and NOBLE studies, focus on prevalence and burden in large populations and use patient surveys as the source of their information^5,8. The aims of this study are to evaluate incidence and predictors of antimuscarinic medication use in general, non-selective antimuscarinic use in particular, among older adults with dementia and OAB using national level Medicare claims data from 2013–2015.

METHODS

Data Source

Multiyear Medicare claims data from 2013–2015 were employed for this study. This data consisted of Part A, B, and D claims data files involving 100% of the national cohort of older patients with dementia and OAB. Medicare data files are available as Research Identifiable Files from the CMS upon request^39–40 and are restricted to claims submitted by Fee-for-Service (FFS) enrollees. Medicare Standard Analytical Files (SAF) are available on yearly basis from the CMS. For this study the MedPAR File, Outpatient SAF, Carrier SAF, Master Beneficiary Summary File (MBSF), and Prescription Drug Event (PDE) files were used. The MBSF section includes the demographics and beneficiary enrollment information (A/B/C/D). The PDE files includes events from all beneficiaries participating in the Part D program. The study was approved by the University of Houston Institutional Review Board Committee for the Protection of Human Subjects under the exempt category.

Study Design and Sample

A new-user retrospective cohort design was used to examine the patterns and predictors of antimuscarinic use among older patients with dementia and OAB. Figure 1 outlines the schematic presentation of the study design. The identification of patients with dementia and OAB was done via ICD-9-CM codes and/or medications, and ICD-9-CM codes, respectively (Supplementary Table S1). The study duration was from January 1, 2013 to December 31, 2015. The first claim (index date) of antimuscarinic medications for the older patients with dementia and OAB was defined as incident use of antimuscarinics in the patient identification period (January 1, 2014 – December 31, 2015) after 12 months of washout period (January 1, 2013 to December 31, 2013) from the study start date. This study design ensured that (i) only incident antimuscarinic users are included, avoiding survivor bias among prevalent users, and (ii) each patient had minimum 12 months of baseline period (12 months preceding the index date), allowing accurate identification of baseline comorbidities/co-medications that could influence treatment selection for OAB.

Figure 1.

Schematic Representation of the Study Design.

Antimuscarinic Use

The incident antimuscarinic medication classified as non-selective or selective was used to define the two treatment exposure groups. Antimuscarinic treatment was measured using PDE files via National Drug Codes and generic names (Supplementary Table S1). Incident users of solifenacin or darifenacin comprised the selective antimuscarinic cohort, whereas those of oxybutynin, fesoterodine, tolterodine or trospium comprised the non-selective antimuscarinic cohort. All of them were collectively defined as incident antimuscarinic users. The primary outcomes of this study were (i) antimuscarinic use [non-use as reference] and (ii) non-selective antimuscarinic use [selective antimuscarinic use as reference].

Covariates

To identify factors associated with incident antimuscarinic use, a large number of covariates were included in this study based on the conceptual framework of Andersen behavioral model (ABM). These covariates were identified during the one year baseline period. Per the ABM, an individual’s health service utilization is a function of three types of patient characteristics: predisposing, enabling, and need.^41–42 Predisposing characteristics describe the propensity of an individual to use health care services and included sociodemographic data such as age, gender, race. Enabling characteristics describe factors affecting the ability of patients to secure health care services such as included health insurance, metropolitan status area, or physician specialty; however, no enabling factors were used in this study. Need characteristics reflect the most immediate functional and health problems that generate the need for health care services and included several comorbid conditions. Conditions such as depression, urinary tract infections (UTIs), skin infections, and vulvovaginitis were selected based on existing literature evidence.^43–45 Additionally, given antimuscarinics are deemed inappropriate among dementia patients (Beers’ criteria), any diagnoses that necessitated the prescription of an anticholinergic (i.e., positively related including abnormal involuntary movements, anxiety, bipolar disorder, drug-induced acute dystonia, gastroesophageal reflux disease (GERD), insomnia, irritable bowel syndrome (IBS), muscle spasms/lower back pain, neuropathic pain, schizophrenia, secondary Parkinson’s disease) or any diagnoses that could be worsened by an anticholinergic (i.e., negatively related including benign prostatic hyperplasia (BPH), chronic constipation, epilepsy, delirium, falls/fractures, myasthenia gravis, narrow-angle glaucoma, Parkinson’s disease, myocardial infarction, and syncope) were included as study covariates addressing need characteristics.⁴⁶ Severity of illness, also considered an important predictor of treatment allocation,⁴⁷ was evaluated via Elixhauser’s score during the baseline period.⁴⁸ The anticholinergic drug scale (ADS) was used to evaluate and adjust for any baseline anticholinergic load due to other medications.⁴⁹ More information regarding the ADS and calculation of baseline anticholinergic load is provided under Supplementary Table S2.

Statistical Analysis

Descriptive statistics were calculated for all variables, including frequencies and percent responses for categorical variables and mean and standard deviation (SD) for continuous variables. Chi-squared tests and paired t-tests were used for the analyses of differences in categorical and continuous variables, respectively. Two separate multivariable logistic regression models were conducted to determine the predictors of antimuscarinic medication use and non-selective antimuscarinic use. The dependent variable was defined as antimuscarinic medication use (reference being no use of antimuscarinics) and non-selective antimuscarinic use (reference being selective antimuscarinic use). The predisposing, enabling or need factors defined as per the ABM constituted the independent variables. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC), with statistical significance set at an alpha level of 0.05.

RESULTS

Figures 2 show the attrition flowchart for identification of older adults with dementia and OAB who were incident users of antimuscarinic medications. Overall, between 2013 and 2015, there were 3,383,603 Medicare beneficiaries diagnosed with dementia. Over one million (1,049,867) of these older adults with dementia had OAB, resulting in an overall prevalence of 31.03%. Among OAB patients, 287,612 (27.39%) were prevalent antimuscarinics users during the study period; 79,397 (7.56%) were incident antimuscarinic users in 2014–2015. After applying other exclusion and continuous eligibility criteria, 21,848 incident antimuscarinic users were identified for further analyses.

Figure 2.

Flowchart for Identification of Older Adults with Dementia and Overactive Bladder

Among the incident antimuscarinic users, 16,955 (77.60%) were non-selective and 4,893 (22.40%) were selective antimuscarinic users. The most frequently reported non-selective antimuscarinic medications were oxybutynin (12,295, 72.52%) and tolterodine (2,868, 16.92%); among selective antimuscarinic agents, solifenacin (4,669, 95.42%) was most frequently used followed by darifenacin (224, 4.58%). Tables 1 and 2 report any differences in baseline characteristics between the two treatment groups.

Table 1.

Baseline Characteristics of Older Adults with Dementia and OAB across Non-users and Users of Antimuscarinic Medications

Characteristics		AntimuscarinicNon-Users (n = 1,89,257; 89.65%)	Antimuscarinic Users (n = 21,848; 10.35%)	Standardized Difference of Proportions (D)	P-value
Predisposing Characteristics
Age (in years)	≥ 65–75 years	42,880 (22.66)	6,287 (28.78)	0.21	<.0001*
	≥ 75–85 years	80,333 (42.45)	9,900 (45.31)
	≥ 85 years	66,044 (34.90)	5,661 (25.91)
Gender	Male	67,205 (35.51)	7,080 (32.41)	0.07	<.0001*
	Female	1,22,052 (64.49)	14,768 (67.59)
Race	White	1,58,334 (83.66)	18,753 (85.83)	0.09	<.0001*
	Black	17,605 (9.30)	1,498 (6.86)
	Other	13,318 (7.04)	1,597 (7.31)
Need Characteristics
Abnormal Involuntary Movements		7,956 (4.20)	1,264 (5.79)	0.07	<.0001*
Anxiety		50,753 (26.82)	6,765 (30.96)	0.09	<.0001*
Bipolar Disease		8,090 (4.27)	1,217 (5.57)	0.06	<.0001*
Drug Induced Acute Dystonia		54 (0.03)	6 (0.03)	−0.00	0.93
Gastro-esophageal Reflux Syndrome		61,427 (32.46)	8,218 (37.61)	0.11	<.0001*
Insomnia		24,018 (12.69)	3,691 (16.89)	0.12	<.0001*
Irritable Bowel Syndrome		5,083 (2.69)	841 (3.85)	0.07	<.0001*
Muscle Spasms/ Low Back Pain		39,747 (21.00)	6,606 (30.24)	0.21	<.0001*
Neuropathic Pain		3,001 (1.59)	598 (2.74)	0.08	<.0001*
Schizophrenia		6,534 (3.45)	745 (3.41)	−0.00	0.74
Secondary Parkinsonism		2,529 (1.34)	357 (1.63)	0.02	0.00*
Benign Prostatic Hyperplasia		32,793 (48.80ⱡ)	4,570 (64.55ⱡ)	0.09	<.0001*
Chronic Constipation		44,670 (23.60)	5,640 (25.81)	0.05	<.0001*
Epilepsy		12,327 (6.51)	1,316 (6.02)	−0.02	0.01*
Delirium		41,068 (21.70)	4,519 (20.68)	−0.02	0.00*
Falls/Fractures		85,255 (45.05)	10,872 (49.76)	0.09	<.0001*
Myasthenia Gravis		374 (0.20)	67 (0.31)	0.02	0.00*
Narrow-angle Glaucoma		29,519 (15.60)	3,593 (16.45)	0.02	0.00*
Parkinson’s Disease		17,392 (9.19)	2,406 (11.01)	0.06	<.0001*
Myocardial Infarction		7,662 (4.05)	910 (4.17)	0.01	0.41
Syncope		25,479 (13.46)	3,340 (15.29)	0.05	<.0001*
Urinary Tract Infection		92,139 (48.68)	12,306 (56.33)	0.15	<.0001*
Skin Infection		15,492 (8.19)	1,822 (8.34)	0.01	0.43
Vulvovaginitis		3,924 (2.07)	780 (3.57)	0.09	<.0001*
Baseline Anticholinergic Load		61,930 (32.72)	7,736 (35.41)	0.06	<.0001*
Elixhauser’s Score (Mean, SD)		11.79 (10.69)	11.66 (10.60)	−0.01	0.09

SD Standard Deviation,

^*Statistical significance at p-value < 0.05,

^ⱡDenominator is male population within non-users and users

Table 2.

Baseline Characteristics of Older Adults with Dementia and OAB across Selective and Non-Selective Antimuscarinic Medications

Characteristics		Selective Antimuscarinic (n = 4,893; 22.40%)	Non-selective Antimuscarinic (n = 16,955; 77.60%)	Standardized Difference of Proportions (D)	P-value
Predisposing Characteristics
Age (in years)	≥ 65–75 years	1,434 (29.31)	4,853 (28.62)	0.02	0.31
	≥ 75–85 years	2,231 (45.60)	7,669 (45.23)
	≥ 85 years	1,228 (25.10)	4,433 (26.15)
Gender	Male	1,760 (35.97)	5,320 (31.38)	0.10	<.0001*
	Female	3,133 (64.03)	11,635 (68.62)
Race	White	4,150 (84.82)	14,603 (86.13)	0.11	<.0001*
	Black	286 (5.85)	1,212 (7.15)
	Other	457 (9.34)	1,140 (6.72)
Need Characteristics
Abnormal Involuntary Movements		289 (5.91)	975 (5.75)	−0.01	0.68
Anxiety		1,450 (29.63)	5,315 (31.35)	0.04	0.02*
Bipolar Disease		244 (4.99)	973 (5.74)	0.03	0.04*
Drug Induced Acute Dystonia		2 (0.04)	4 (0.02)	−0.01	0.52
Gastro-esophageal Reflux Syndrome		1,869 (38.20)	6,349 (37.45)	−0.02	0.34
Insomnia		862 (17.62)	2,829 (16.69)	−0.02	0.13
Irritable Bowel Syndrome		209 (4.27)	632 (3.73)	−0.03	0.08
Muscle Spasms/ Low Back Pain		1,597 (32.64)	5,009 (29.54)	−0.07	<.0001*
Neuropathic Pain		137 (2.80)	461 (2.72)	−0.00	0.76
Schizophrenia		133 (2.72)	612 (3.61)	0.05	0.00*
Secondary Parkinsonism		73 (1.49)	284 (1.68)	0.01	0.37
Benign Prostatic Hyperplasia		1,206 (68.52ⱡ)	3,364 (63.23ⱡ)	−0.12	<.0001*
Chronic Constipation		1,221 (24.95)	4,419 (26.06)	0.03	0.12
Epilepsy		292 (5.97)	1,024 (6.04)	0.00	0.85
Delirium		913 (18.66)	3,606 (21.27)	0.07	<.0001*
Falls/Fractures		2,393 (48.91)	8,479 (50.01)	0.02	0.17
Myasthenia Gravis		15 (0.31)	52 (0.31)	0.00	0.99
Narrow-angle Glaucoma		850 (17.37)	2,743 (16.18)	−0.03	0.05*
Parkinson’s Disease		556 (11.36)	1,850 (10.91)	−0.01	0.37
Myocardial Infarction		185 (3.78)	725 (4.28)	0.03	0.13
Syncope		743 (15.18)	2,597 (15.32)	0.00	0.82
Urinary Tract Infection		2,704 (55.26)	9,602 (56.63)	0.03	0.09
Skin Infection		378 (7.73)	1,444 (8.52)	0.03	0.08
Vulvovaginitis		153 (3.13)	627 (3.70)	0.03	0.06
Baseline Anticholinergic Load		5,937 (35.02)	1,799 (36.77)	−0.04	0.02*
Elixhauser’s Score (Mean, SD)		11.42 [10.48]	11.73 [10.64]	0.03	0.07

SD Standard Deviation,

^*Statistical significance at p-value < 0.05,

^ⱡDenominator is male population within selective and non-selective antimuscarinic users

Predictors of Use versus No Use of Antimuscarinic Medication

Multivariable logistic regression analysis as shown in Table 3 revealed that patients 75 years or older [Odds ratio (OR) for ≥ 75–85 years 0.83; OR for ≥ 85 years 0.58], of black race [OR 0.76], and those with schizophrenia [OR 0.86], epilepsy [OR 0.86], delirium [OR 0.84], and Elixhauser’s score [OR 0.994] were less likely to initiate antimuscarinics. Women [OR 1.55], those with abnormal involuntary moments [OR 1.19], bipolar disorder [OR 1.10], gastroesophageal reflux disease [OR 1.11], insomnia [OR 1.22], irritable bowel syndrome [OR 1.14], muscle spasm/low back pain [OR 1.40], neuropathic pain [OR 1.35], benign prostatic hyperplasia [OR 1.76], falls/fractures [OR 1.11], myasthenia gravis [OR 1.37], narrow-angle glaucoma [OR 1.07], Parkinson’s disease [OR 1.15], syncope [OR 1.11], urinary tract infection [OR 1.31] and vulvovaginitis [OR 1.40] were more likely to initiate antimuscarinic medications.

Table 3.

Predictors of Antimuscarinic Use among Older Adults with Dementia and Overactive Bladder

Characteristics	Category	Odds Ratio	95% Confidence Interval
Predisposing Characteristics
Age (in years)	≥ 65–75	Reference	--
	≥ 75–85	0.83	0.80 – 0.86*
	≥ 85	0.58	0.56 – 0.60*
Gender	Male	Reference	--
	Female	1.55	1.48 – 1.62*
Race	Other	Reference	--
	White	1.00	0.95 – 1.06
	Black	0.76	0.70 – 0.82*
Need Characteristics
Abnormal Involuntary Movements	No	Reference	--
	Yes	1.19	1.12 – 1.27*
Bipolar Disorder	No	Reference	--
	Yes	1.10	1.03 – 1.18*
Gastro-esophageal Reflux Syndrome	No	Reference	--
	Yes	1.11	1.07 – 1.14*
Insomnia	No	Reference	--
	Yes	1.22	1.17 – 1.27*
Irritable Bowel Syndrome	No	Reference	--
	Yes	1.14	1.06 – 1.24*
Muscle Spasms/ Low back Pain	No	Reference	--
	Yes	1.40	1.36 – 1.45*
Neuropathic Pain	No	Reference	--
	Yes	1.35	1.23 – 1.48*
Schizophrenia	No	Reference	--
	Yes	0.86	0.79 – 0.93*
Benign Prostatic Hyperplasia	No	Reference	--
	Yes	1.76	1.67 – 1.85*
Epilepsy	No	Reference	--
	Yes	0.86	0.81 – 0.92*
Delirium	No	Reference	--
	Yes	0.84	0.81 – 0.87*
Falls/Fractures	No	Reference	--
	Yes	1.11	1.08 – 1.14*
Myasthenia Gravis	No	Reference	--
	Yes	1.37	1.05 – 1.79*
Narrow-angle Glaucoma	No	Reference	--
	Yes	1.07	1.03 – 1.11*
Parkinson’s Disease	No	Reference	--
	Yes	1.15	1.10 – 1.21*
Syncope	No	Reference	--
	Yes	1.11	1.07 – 1.16*
Urinary Tract Infection	No	Reference	--
	Yes	1.31	1.27 – 1.35*
Vulvovaginitis	No	Reference	--
	Yes	1.40	1.29 – 1.52*
Elixhauser’s Score	-	0.99	0.992 – 0.995*

Reference: non-use

^*Statistical significance at p-value < 0.05; Adjusted all baseline predisposing, enabling, and need factors

Predictors of Non-Selective versus Selective Antimuscarinic Medication Use

Multivariable logistic regression analysis as shown in Table 4 revealed that patients with muscle spasms/low back pain [OR 0.87], benign prostatic hyperplasia [OR 0.79] and those taking higher anticholinergic medications [OR 0.91] had lower odds of receiving non-selective antimuscarinics. White patients [OR 1.33], black patients [OR 1.59] and those with schizophrenia [OR 1.31] and delirium [OR 1.15] were more likely to receive non-selective antimuscarinics.

Table 4.

Predictors of Non-selective Antimuscarinic Use among Older Adults with Dementia and Overactive Bladder using Antimuscarinics

Characteristics	Category	Odds Ratio	95% Confidence Interval
Predisposing Characteristics
Race	Other	Reference	--
	White	1.33	1.18 – 1.49*
	Black	1.59	1.34 – 1.89*
Need Characteristics
Muscle Spasms/ Low back Pain	No	Reference	--
	Yes	0.87	0.81 – 0.93*
Schizophrenia	No	Reference	--
	Yes	1.31	1.07 – 1.60*
Benign Prostatic Hyperplasia	No	Reference	--
	Yes	0.80	0.71 – 0.89*
Delirium	No	Reference	--
	Yes	1.15	1.06 – 1.26*
Baseline Anticholinergic Load	No	Reference	--
	Yes	0.91	0.85 – 0.98*

Reference: Selective

^*Statistical significance at p-value < 0.05; Adjusted all baseline predisposing, enabling, and need factors

DISCUSSION

In a national cohort of older adults with dementia, about one-third (31.03%) were diagnosed with OAB. Antimuscarinics were used in over one-fourth (27.01%) of older adults with dementia and OAB. A range of estimates have emerged from the existing literature. Odeyemi et al. found 28.2% (19,444) of patients with OAB symptoms received at least one prescription for OAB medication;⁵⁰ Moga et al. found 10% of residents 65 years of age and older admitted to VA nursing homes for long-term care used antimuscarinic medications;⁵¹ Ju et al. evaluated outpatient office visits by women with OAB and found 8.1 million (1.6 %) visits to be associated with antimuscarinic prescription;⁵² Goldman et al. reported 34% of women and 19% of men diagnosed with OAB were prescribed medication.⁵³ Weighted analyses on 2009–2010 NAMCS and NHAMCS revealed that 41.43 % of older adults with OAB received antimuscarinic medication prescription.⁴⁵ However, none of these can be compared directly against the estimates from the current study, primarily due to the fact that the current study addresses antimuscarinic medication use among older adults with dementia with OAB. Besides, methodological differences, different study populations/settings or unit of analysis further contribute to variation in the estimates. However, the current estimates are comparable to those from one of our previous researches which reported use of potentially inappropriate anticholinergic medications in 26.95% of older adults with dementia; 11.49% of this was contributed by selected antimuscarinics⁴⁶

The study found that that among incent users 78% involved non-selective and 22% involved selective antimuscarinic agents. To the author’s knowledge, this is the first study that focused on the prevalence and predictors of antimuscarinic medication use, especially non-selective antimuscarinics among older adults with dementia and OAB. The most frequently used antimuscarinics found in the study were oxybutynin, solifenacin, tolterodine, followed by fesoterodine, trospium and darifenacin. In total, the non-selective agents were more frequently used that the selective counterparts (78% vs 22%). This trend reflects concerns regarding the prescribing pattern because (i) safer and more tolerable selective agents are available and (ii) safer nonanticholinergic alternatives such as mirabegron are available.^54–55

As previously mentioned, the differential binding of non-selective antimuscarinics to different receptor subtypes that have minimal or no involvement in bladder detrusor contractions (M1, M2, M4 and M5 receptors) leads to the various central and peripheral adverse events such as increased heart rate, decreased secretions, pneumonia, sedation, visual disturbances.^{12,19,22–32} Of particular importance is the inhibition of M1 receptors that leads to adverse cognitive effects and makes it a grave concern for the cognitively-compromised dementia patients.³⁰ In addition, the pro-arrhythmic and pro-ischemic effects of non-selective antimuscarinics leads to their increased cardiovascular risk.^56–57 Prior research found that use of non-selective antimuscarinics was associated with a 50% increased mortality risk among older adults with dementia and OAB.⁵⁸ Given the safety concerns regarding non-selective antimuscarinic agents, there is a significant need to increase the use selective agents in the management of OAB for older patients with dementia.

Multivariable logistic regression revealed multiple significant predictors for antimuscarinic use. Increasing age was associated with lower odds of incident antimuscarinic use among older adults with dementia and OAB. In contrast, one previous research suggested that physicians were 3.5-times more likely to use pharmacotherapy for OAB patients aged ≥85 years versus those aged 65–74 years.⁴⁵ This is encouraging as it reflects physician awareness regarding the cautious use of anticholinergic medications including antimuscarinics in dementia population. Black patients are 24% less likely than others to receive a new antimuscarinic, as also supported by Felton et al., where patients of black race were 24–45% less likely to use any anticholinergics compared to those of white race over 10 years (all p<0.05).⁵⁹ Some studies also suggest that older patients of black race are at less risk than same aged patients of white race for potentially inappropriate medication use.^60–63

Those with comorbid schizophrenia, epilepsy and delirium were also less likely to use antimuscarinics, which is a reassuring observation, considering the central adverse effects associated with antimuscarinics (impaired cognition and memory). Neurocognitive impairment is a core feature in schizophrenia^64–65 and has been shown to be worsened by exposure to muscarinic receptor antagonists.^66–68 Possible association between antimuscarinic use and epilepsy may be supported by some preclinical evidence on antimuscarinic-induced convulsions in fasted animals after food intake.^69–70 Besides, anticholinergics were referred as the second most common category of drugs implicated for seizures after antidepressants.^71–73 A recent study focusing on bladder antimuscarinic use in dementia patients found that the majority of serious falls and delirium among people with dementia and bladder antimuscarinic use - 76%, were followed by bladder antimuscarinic prescription fills within 12 months after the event.⁷⁴

The study also found that female patients and those with abnormal involuntary moments, bipolar disorder, gastroesophageal reflux disease, insomnia, irritable bowel syndrome, muscle spasm/low back pain, neuropathic pain, benign prostatic hyperplasia, falls/fractures, myasthenia gravis, narrow-angle glaucoma, Parkinson’s disease, syncope, urinary tract infection and vulvovaginitis are more likely to use antimuscarinic medications. While this increased use is encouraging in patients with some comorbidities, it raises concerns for some other diagnoses. Contraindications for the use of antimuscarinic agents include patients with narrow-angle glaucoma or myasthenia gravis because of their anticholinergic effects on the bowel.^75–76 Conditions such as BPD or Parkinson’s disease are associated with cognitive impairment, thereby making the use of antimuscarinics further concerning due to the associated central adverse events.^77–78 Recent literature confirms that OAB often coexists with other medical conditions, such as falls/fractures, urinary tract infections and vulvovaginitis;^8,43–44 prevalence of these comorbidities was reported to be significantly higher (p<0.0001) among OAB patients than the controls.⁴³ The increased odds of antimuscarinic use is a possible indication of a vigorous treatment pathway for OAB cure to improve the quality of life.

Generally, antimuscarinics cause urinary retention and tend to exacerbate or lead to significant bladder outflow obstruction among BPH patients. Hence, they are contraindicated in this condition. However, often BPH symptoms such as frequency and urgency, especially at night (nocturia), are not completely relieved with combination of alpha blockers and 5-alpha reductase inhibitors. In such situations, antimuscarinics can help to relieve the symptoms of hyperactive bladder.^79–80 As high as 50% of patients with benign prostatic hyperplasia (BPH) experience storage symptoms,⁸¹ antimuscarinic drugs can be considered to help to ameliorate these symptoms.⁸² Due to their anticholinergic nature, antimuscarinics decrease the lower esophageal sphincter pressure, resulting in increased esophageal exposure to acid and worsening GERD.^83–84 The smooth muscle relaxation mediated by M3 receptor stimulation may be useful in irritable bowel syndrome. Traditional IBS treatments include M3 specific receptor antagonists such as darifenacin that reduce intestinal motility.⁸⁵ OAB has also been cited as one of the risk factors for low back pain.^86–87

Conventional pathology describing the association between muscle spasms/low back pain and OAB or increased use of antimuscarinics is still unclear. However, evidence from clinical and epidemiological studies supports a possible association between low back pain and bowel^88–89 and bladder dysfunction.^90–91 One exploratory research to estimate the relationship between urinary incontinence and back problems among Canadian adults reported that having incontinence increased the risk of also having back problems in both men (OR = 2.45; 95% CI = 2.06–2.91) and women (OR = 2.97; 95% CI = 2.64–3.35) compared to those without incontinence.⁹² Conversely, a cross-sectional study conducted on data from Kentucky Women’s Health Registry reported that stress urinary incontinence was higher in women with chronic back pain than those without (49.0% vs. 35.2%, p<0.01); this trend remained consistent after adjusting for potential confounders ( OR=1.44; 95% CI = 1.11 – 1.86).⁹³ OAB patients may experience frequent urination at night (nocturia) and that can be associated with insomnia or other sleep disturbances.^5,8,44 Increased use of antimuscarinics could be indicative of aggressive treatment route for OAB. More research is needed to identify the association between increased use of antimuscarinics and conditions like syncope or neuropathic pain.

Regarding predictors of non-selective antimuscarinic use - patients with muscle spasms/low back pain, benign prostatic hyperplasia and those taking higher anticholinergic medications had 13%, 21% and 9% lower odds of receiving non-selective antimuscarinics, respectively. As explained above, antimuscarinic use could be beneficial for these comorbid conditions and in this situation, preferential use of non-selective agents is rather a reassuring observation due to a safer adverse effect profile. Conversely, increased non-selective antimuscarinic use among patients with schizophrenia and delirium is very troubling because the significant cognitive impairment in these patients can be further precipitated due to the central adverse effects associated with anticholinergics. Patients of white and black race have 33% and 59% higher odds of receiving non-selective agents, respectively. Although this difference remains unexplained, there is existing literature to support racial differences in antimuscarinic use among elderly.^94–95

Strengths and Limitations

This study had several strengths, including the study design and the analytical approach used. The study also used relevant covariates, including anticholinergic load to control for baseline anticholinergic burden and Elixhauser’s Index, a widely used risk adjustment tool proven to be statistically superior for predicting various outcomes. However, the study results must be interpreted in light of certain limitations too. Primarily, as with any claims database, the Medicare data are subject to data coding limitations (such as data entry errors). Any exposure or outcome measurements were based on diagnostic data available in medical claims submitted by the health providers and could further be subject to coding issues.

The medication information comes from the claims data; hence the actual use of prescribed medications could not be ascertained. While statistical analysis was conducted to adjust for patients’ characteristics, systemic differences may still exist between the exposure groups. Despite being based on previous literature, this adjustment was limited to those predisposing, enabling, and need characteristics that were available within the claims data. Other variables such as belief constructs or patient perceptions, prescriber characteristics, enabling factors (metropolitan status area, physician specialty, family support, supply of health services) and other local-area characteristics could not be studied because of database limitations. Claims data did not include the use of over-the-counter drugs and prescription drugs in inpatient settings. Lastly, the study is limited to Medicare beneficiaries in community settings; therefore, future studies are needed to assess the predictors of antimuscarinics among patients aged <65 years and other diverse settings.

CONCLUSIONS

The study found that nearly one-third of older adults with dementia had OAB. Antimuscarinics were used in over one-fourth of older adults with dementia and OAB. Among incident antimuscarinic users, majority of them using non-selective agents (77.6%). Oxybutynin and solifenacin were found to be the most frequently used antimuscarinics in non-selective and selective categories, respectively. Several predisposing and need factors were found associated with the use of antimuscarinic medications and non-selective agents. Given the high prevalence of OAB and the importance of OAB management among older dementia patients, there is a need to optimize antimuscarinic use given their vulnerability for central anticholinergic adverse effects.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are not publicly available as, per government regulations and the data use agreement, data from the Centers for Medicare and Medicaid Services cannot be shared.