Use of bladder antimuscarinics is associated with an increased risk of dementia: a retrospective population-based case–control study

The association between bladder antimuscarinic use and dementia development is unclear. We used data from the Taiwan National Health Insurance Research Database to determine the association between the exposure dose and duration of bladder antimuscarinics and the subsequent dementia risk. We enrolled participants aged 55 years or more and defined a dementia cohort (International Classification of Diseases, Ninth Revision, Clinical Modification codes 290, 294.1, and 331.0). We used a propensity score matching method, and randomly enrolled two controls without dementia. We evaluated dementia risk with respect to the exposure dose and duration of treatment with seven bladder antimuscarinics (oxybutynin, propiverine, tolterodine, solifenacin, trospium, darifenacin, and fesoterodine) used for at least 1 year before the index date, after adjusting for age, sex, comorbidities, and medications. The dementia risk was 2.46-fold (95% confidence interval: 2.22–2.73) higher in Taiwanese patients who used bladder antimuscarinics for ≥ 1 year than in those who were not exposed to this treatment. The risk proportionally increased with increasing doses of antimuscarinics for less than 4 years. Taiwanese patients aged 55 years or more on bladder antimuscarinics exhibited a higher risk of dementia. Additional studies in other countries are required to determine whether this result is valid worldwide.

www.nature.com/scientificreports/ There are considerable differences in brain penetration between the bladder antimuscarinics and ACs used for the central nervous system disorders. It is uncertain whether the urogenital use of antimuscarinic ACs would increase the risk of developing dementia due to the apparent pharmacodynamic differences between urogenital and central nervous ACs. For further investigation, we aimed to study the correlation between the exposure of bladder antimuscarinics and the risk of developing subsequent dementia. We used data from a nationwide, population-based database in Taiwan to analyze their possible relationships.

Methods
Data resource. The dataset used in this study was derived from the National Health Insurance Research Database (NHIRD) in Taiwan, which covers approximately 99% of the entire population of 23 million people in Taiwan. The Longitudinal Health Insurance Database (LHID) includes all original claims data and registration files from 2000 to 2013 for one million individuals randomly sampled from the Registry for Beneficiaries of the NHIRD program in 2000 in Taiwan. This database has been validated by several studies [11][12][13] , to prove the correct coding of different diseases. This study was approved by the Institutional Review Board of China Medical University and the Hospital Research Ethics Committee (IRB permit number: CMUH-104-REC2-115) and is in compliance with institutional guidelines. Written informed consent from patients was waived due to low risk, and the study was approved by the institutional IRB of China Medical University and the Hospital Research Ethics Committee.

Study subjects.
In this case-control study, we aimed to examine the effects of bladder antimuscarinics on the development of dementia. Study subjects comprised patients with dementia coded with ICD-9-CM 290, 294.1, and 331.0, and diagnosed by a neurologist or a general physician in the medical care system of Taiwan during 2000-2013. The first diagnosed date of dementia was defined as the index date. Disease diagnosis in the LHID was defined according to the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM). The controls were subjects without dementia during 2000-2013 and were randomly selected from the LHID. We excluded individuals younger than 55 years and individuals with missing data of age and sex. Considering that such a short-term exposure to antimuscarinics was less likely to cause dementia, subjects who used antimuscarinics for less than 1 year were excluded. For each dementia case, we used a propensity score matching method and randomly selected two controls from the non-dementia group. The controls were assigned the same index year as their matched cases with respect to age, sex, comorbidities, and medications mentioned below.
Exposure assessment and covariates. The use of bladder antimuscarinics was evaluated before the index date. For those who used ACs for at least 1 year before the index date, we obtained data of seven types of bladder antimuscarinics (oxybutynin, G04BD04; propiverine, G04BD06; tolterodine, G04BD07; solifenacin, G04BD08; trospium, G04BD09; darifenacin, G04BD10; fesoterodine, G04BD11) based on World Health Organization ATC codes 14 . Patients without any prescription of ACs during the study period were classified as AC non-users. The duration of AC use was categorized as medium (1-3 years), long (4-7 years), and prolonged (> 7 years) durations. The cumulative dose of AC use during the study period was quantified for each patient using the World Health Organization Defined Daily Dose (DDD) 14 , and graded as follows: non-use, low-dose (≤ 207), medium-dose (207-3271), and high-dose (> 3271) users.
Furthermore, we adjusted for the potentially confounding effects of other drugs, including aspirin, nonsteroidal anti-inflammatory drugs, antihypertensives, statins, anxiolytics, hypnotics, antidepressants, and anti-Parkinson's disease and antipsychotic medications. Treatment with these drugs before the index date was evaluated as a part of the analysis. Statistical analysis. Propensity score matching was used to optimize comparability between the dementia and non-dementia groups using a non-parsimonious multivariable logistic regression model, with dementia as the dependent variable. Age, sex, comorbidities, medications, and index year were used as independent variables to match cases between the two groups. Descriptive statistics for the cases of dementia and non-dementia groups were reported, including demographic characteristics, comorbid disease, and medications. The standardized difference was used to test the differences in continuous and categorical matching variables. A standardized mean difference of ≤ 0.10 indicates a negligible difference between the groups.
We used conditional logistic regression to assess the risk of dementia associated with bladder antimuscarinics. The odds ratio (OR) and 95% confidence interval (CI) for dementia were calculated and subsequently adjusted www.nature.com/scientificreports/ for covariates including age, sex, comorbidities, and medications. The covariates adjusted for in the analytical models were listed as adjusted OR (aOR). To assess the dose-effect relationship, we analyzed the risks of dementia according to the cumulative DDD of bladder antimuscarinics (≤ 207 DDD, 207-3271 DDD, and > 3271 DDD) relative to non-users and stratified by 1-3, 4-7, and > 7 exposure years. We used SAS statistical software (Version 9.4 for Windows; SAS Institute, Inc., Cary, NC, USA) for data analysis. Results with a P-value of less than 0.05 were considered to be statistically significant. Table 1 shows the demographic and clinical characteristics of the study population. A total of 20,246 patients with dementia and 40,394 patients without dementia were enrolled in this study between January 1, 2000 and December 31, 2013 in the propensity score-matched population. The mean (SD) age was 77.3 (8.54) and 77.3 (10.3) years in the dementia and non-dementia groups, respectively. Among the subjects, females aged 75 to 84 years were dominant (Table 1). After propensity score matching, distribution of age, sex, comorbidities, and medications did not significantly differ between the groups. The detailed flow chart for the identification of the study subjects is shown in Fig. 1. Table 2 presents the association between bladder antimuscarinics and the risk of developing dementia. After adjusting for potential confounders, antimuscarinic users exhibited a 2.46-fold increased risk of dementia compared with that in non-users (95% CI = 2.22-2.73). With respect to comorbidities, subjects with hypertension (aOR = 0.93, 95% CI = 0.89-0.98), asthma (aOR = 0.91, 95% CI = 0.89-0.97), and COPD (aOR = 0.86, 95% CI = 0.82-0.90) exhibited a lower risk of developing dementia. As for medication use, patients using aspirin, nonsteroidal anti-inflammatory drugs, statin, anxiolytics, and hypnotics exhibited a lower risk of developing dementia. On the contrary, those using anti-depressants drug, anti-Parkinson's disease drug, and anti-psychotic drug showed a higher risk of developing dementia ( Table 2). Table 3 presents the association between the cumulative DDD of bladder antimuscarinics and the risk of dementia by stratification according to the exposure to antimuscarinics. In patients who had been taking antimuscarinics for less than 4 years before the index date, an increased DDD was proportionally associated with the increased risk of developing dementia (aOR = 2.23, 95% CI = 1.12-4.44 for ≤ 207 DDD; aOR = 2.35, 95% CI = 0.87-6.32 for 207-3271 DDD; aOR = 12.8, 95% CI = 5.15-32.1 for > 3271 DDD) compared with that in the controls. In individuals exposed for 4-7 years, those who used ≤ 207 DDD (aOR = 2.82, 95% CI = 1.68-4.75), 207-3271 DDD (aOR = 2.23, 95% CI = 1.10-4.53), and > 3271 DDD (aOR = 1.90, 95% CI = 0.94-3.81) of antimuscarinics did not exhibit the trend of proportional increase in the risk of developing dementia. In patients with an exposure duration greater than 7 years, only those taking antimuscarinics at 207-3271 DDD (aOR = 1.19, 95% CI = 1.00-1.41) presented an equal risk of developing dementia compared with that in the controls (Table 3). Table 4 presents the duration (years) of dementia identified in the dementia group and presents the year of study entry in the non-dementia group. We calculated the duration of exposure in both groups. The mean (SD) duration of exposure was 5.87 (3.96) and 5.93 (3.47) years in the dementia and non-dementia groups (Table 4).

Discussion
In this retrospective nation-wide population-based case-control study, we noted that the risk of dementia increased 2.46-fold in Taiwanese patients aged 55 years or older who had been previously using bladder antimuscarinics for 1 year or more. Specifically, the risk proportionally increased with increasing dosage in patients taking antimuscarinics for less than 4 years. Richardson et al. reported that dementia development was associated with an increased use of antidepressant, urological, and anti-Parkinson agents 6 . They suggested that prior exposure to ACs of up to 20 years before the diagnosis of incident dementia could be detected. However, in our study, we noted that dementia development might be associated with an increasing dose of antimuscarinics for medium exposure duration. We believe that antidepressants, antihistamines, and bladder antimuscarinics should be started at different ages in patients. A medium exposure duration (< 4 years) would be related to less death claims in elderly subjects on antimuscarinics than in those with long or prolonged use in this study. We carefully adjusted the possible confounding effects from central nervous system disorders and ACs used for purpose other than urogenital organs in the study design. To our knowledge, this is the first study to report the dose-response effects on dementia development in patients using bladder antimuscarinics for less than 4 years in an Asian population.
We further observed a different relationship between the risk of dementia development and the duration of antimuscarinics exposure than that in other studies. The risk of dementia development seemed to decrease by time in patients on a high dose of antimuscarinics (> 3271 DDD). Those on antimuscarinics and aged 75-84 years seemed to be resistant to a higher risk of developing dementia. It implied a higher mortality rate for the first few years in aged patients with urogenital disorders. Once they crossed the high-risk period of urogenital disorders, the very aged survivors show a trend to resist developing another disorder. As in any observational study, the positive association that we observed between bladder antimuscarinics and dementia risk should be considered a part of numerous factors involved in dementia patho-mechanism. The directionality of the association can only be hypothesized by current data. To maintain the quality of life of patients, we suggest that primary caregivers should not limit the use of antimuscarinics for urogenital disorders before the symptoms of dementia are observed in the patients.
In this study, we analyzed the correlations between seven bladder antimuscarinics (oxybutynin, propiverine, tolterodine, solifenacin, trospium, darifenacin, and fesoterodine) and the subsequent dementia development. There are five subtypes (M 1 -M 5 ) of muscarinic ACh receptors, and the M2 and M3 muscarinic Ach receptors are the major receptors that mediate smooth muscle contraction, proliferation, and remodeling of the bladder 17,18 . In contrast, evidence from recent postmortem human brain studies have implicated the involvement of the  19,20 . Some studies have further demonstrated the dominant functional distribution of the M 1 muscarinic receptors for ACh uptake in the human brain 21,22 . Our study results implied that bladder antimuscarinics might possibly produce subtle effects on M 1 activity, in addition to their known antagonistic effects on the M 2 and M 3 receptors. Because of the potential effect of bladder antimuscarinics on the M1 muscarinic receptor, bladder antimuscarinics has been supposed to cause dementia development. Furthermore, we cannot completely rule out the role of other subtypes of muscarinic receptors besides the M 1 receptor that exhibits less expressed in the brain, and the brain penetration of various bladder antimuscarinics might be different. More laboratory studies could help clarify the detailed mechanism of different antimuscarinics and expression of different ACh subtypes in dementia development.
Globally, approximately 47 million people suffer from dementia with an estimated global cost of 818 billion US dollar in 2015, and the patient number would triple by 2050 1 . Older adults who develop dementia are less likely to return to their ordinary lives than those who do not. The mechanisms of dementia development are too complicated to be fully understood in older adults with co-existing chronic disorders. First, older adults with bladder disorders might have sedentary lifestyles with poor sleep and personal hygiene or even with alcohol abuse and drug addiction. Sedentary lifestyles would potentially increase the risk of dementia development 23,24 . Second, bladder disorders often exist with local or systemic inflammation to associate with an increased risk of dementia development 15,16 . It is difficult to design a study that can distinguish the effects of inflammatory disorders from that of antimuscarinics on dementia development. However, the relationship between the risk of dementia and the dose of bladder antimuscarinics provides supports our hypothesis 25 . These results indicate that bladder antimuscarinics increase predisposition to dementia development.
The major limitation of this study was that dementia cases diagnosed using the ICD-9-CM coding system are often underestimated 26 . Although the NHI program covers nearly 99% of Taiwanese citizens and guarantees equality of access to medical services for everyone throughout the country, some dementia cases might be outside the scope of our study. With a higher rate of dementia diagnosis in older adults with bladder disorders, there might be some patients who did not receive bladder antimuscarinics before the incident dementia diagnosis. Second, we could not directly contact the patients because their identities were anonymized in the accessible LHID. Therefore, we could not analyze all confounding factors for dementia development within the patients' families or the psychological burden on patients. Third, the poor adherence to bladder antimuscarinics in patients was another potential limitation of this study. However, our study demonstrates a statistically significant increase in the risk of dementia development in patients using bladder antimuscarinics. These results highlight the need to further explore bladder antimuscarinics as a predisposing factor for dementia development.

Conclusions
Taiwanese patients aged 55 years or more undergoing treatment with bladder antimuscarinics exhibited a higher risk for dementia development. Additional studies in other countries are required to determine whether this result is valid worldwide.

Data availability
The original data are available at NHI, and we are not allowed to release despite reasonable application. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.