Introduction
Overactive bladder (OAB) is a syndrome of urinary urgency, usually with frequency and nocturia. It is common in Japan. In a nationwide community survey in 2023, the prevalence of OAB was 11.9% in people aged 20 years or older and 13.8% in those aged 40 years or older [1].
The prevalence rises with age, and reaches about 37% in people aged 80 years or older [2]. Many patients with OAB are treated not only by urologists but also by primary care physicians [3, 4]. For many years, anticholinergics were the main drug treatment for OAB. β3‐adrenoceptor agonists then became available in Japan: mirabegron (Betanis) was launched in September 2011, and vibegron (Beova) in November 2018 [5].
Their use has since increased.
Anticholinergics work by blocking muscarinic receptors in the bladder, but they also act on the central nervous system. This raises several concerns in older people. First, they impair cognition, and cumulative use of strong anticholinergics has been linked to a higher risk of dementia in large cohort and case–control studies [6, 7].
Second, they increase the risk of falls and fractures. Third, they cause peripheral adverse effects such as dry mouth, constipation, and urinary retention, which reduce quality of life and adherence. For these reasons, prescribing guidelines for older adults, such as the Beers criteria, advise caution with anticholinergics. β3‐adrenoceptor agonists do not act on muscarinic receptors and may therefore avoid these problems.
However, β3‐adrenoceptor agonists have not fully replaced anticholinergics in daily practice. Anticholinergics are still widely used, and β3‐adrenoceptor agonists are also given as add‐on therapy to patients who are already taking an anticholinergic [8]. The increasing use of β3‐adrenoceptor agonists may therefore not translate into a lower anticholinergic burden in older people.
It is not known how far this change has actually progressed in the patients who are most vulnerable to the risks of anticholinergics. In particular, no nationwide study in Japan has examined whether this change differs by age or sex. We therefore used the NDB Open Data to describe the use of anticholinergics and β3‐adrenoceptor agonists for OAB in Japan from FY2017 to FY2024, and to examine how this change progressed by age and sex.
We also quantified the anticholinergic burden using the anticholinergic cognitive burden (ACB) score, and assessed changes in the ACB‐weighted prescription volume by sex and age group.
Materials and Methods
This was a descriptive study using the National Database (NDB) Open Data, which are published by the Ministry of Health, Labour and Welfare of Japan. The NDB Open Data are aggregated tables derived from health insurance claims submitted under Japan's universal health insurance system, and they cover the whole population. They contain no individual records.
We used the sex‐ and age‐stratified tables for oral medications for the eight fiscal years from FY2017 to FY2024 (each fiscal year runs from April to March). We analyzed outpatient dispensing and combined in‐hospital and out‐of‐hospital prescriptions. Inpatient data are not published for oral medications, so only outpatient data were used.
This study was approved by the Ethics Committee of Yokohama City University Medical Center, Yokohama, Japan (approval number: F260700015), and was conducted in accordance with the Declaration of Helsinki. Because we used only published aggregate data that contain no identifiable information, the requirement for informed consent was waived.
We analyzed two β3‐adrenoceptor agonists (mirabegron and vibegron) and six anticholinergics (solifenacin, imidafenacin, fesoterodine, propiverine, oxybutynin, and tolterodine) that are used orally for OAB. All brand‐name and generic products were included and were grouped by generic name (Table S1). Because the number of tablets taken per day differs between drugs, we converted the number of dispensed tablets into patient‐days by dividing by the standard number of tablets per day for each drug, based on the usual dose in the package insert.
We calculated the β3‐adrenoceptor agonist share as the volume of β3‐adrenoceptor agonists divided by the combined volume of β3‐adrenoceptor agonists and anticholinergics. We analyzed 5‐year age groups and also grouped patients into three age categories: 65–74, 75–84, and ≥ 85 years.
To measure anticholinergic burden, we calculated ACB‐weighted prescription volume as the sum of patient‐days multiplied by the ACB score of each drug. We used the ACB scale of Boustani et al. [9]. Solifenacin, fesoterodine, tolterodine, propiverine, and oxybutynin are all listed on this scale with a score of 3, which denotes a definite anticholinergic that may cause delirium. β3‐adrenoceptor agonists do not act on muscarinic receptors and were therefore scored 0.
Imidafenacin is used mainly in Japan and other Asian countries and is not listed on the ACB scale. However, it is an antimuscarinic drug, so a score of 0 would not be appropriate. Because every other OAB anticholinergic that is listed on the scale carries a score of 3, we assigned imidafenacin a score of 3 in the primary analysis, and we repeated the analysis with scores of 1 and 2 as sensitivity analyses (Table S2).
Because the NDB Open Data are aggregated tables rather than patient‐level records, this ACB‐weighted prescription volume is a population‐level index that reflects the dispensed volume of OAB antimuscarinics weighted by the ACB score; it does not represent the comprehensive anticholinergic burden of an individual patient across all drug classes, and it cannot capture the concurrent prescription of two anticholinergics in the same patient. To distinguish demographic change from change in prescribing, we also expressed ACB‐weighted prescription volume per 1000 population of the same sex and age group; annual population denominators were obtained by linear interpolation between the FY2018 and FY2024 populations (Table S6).
We observed a temporary fluctuation in anticholinergic volumes in FY2021 and FY2022. During the same period, Japan experienced a nationwide shortage of generic medicines that followed regulatory sanctions against generic manufacturers. This fluctuation was seen only in anticholinergics for which a generic was available, and not in mirabegron, which was brand‐name only at that time.
The size of the fluctuation also did not correspond to whether a new generic had been listed (Figure 2). We therefore judged that the fluctuation reflected the supply shortage, and we defined FY2021–FY2022 as a period of supply disruption. Data from these years are shown in all figures and tables, but when we estimated change over time we used FY2018 and FY2024 as endpoints, because both lie outside the disruption.
In the NDB Open Data, cells with values below 1000 are suppressed; we treated these as missing and excluded them. The proportion of the total dispensed volume that was captured by the sex‐ and age‐stratified cells (cell coverage) was ≥ 99.9% in FY2018–FY2024, but only 88.3% in FY2017, so we used FY2018 as the starting point for change over time (Table S3).
This was a descriptive study of population‐wide data, and we did not perform hypothesis testing. The NDB Open Data reflect the whole population rather than a sample, so sampling error does not apply and statistical inference is not required. We report observed values and relative changes.
Data ion were performed using Python 3.12 (pandas, NumPy, and Matplotlib).
Use of Artificial Intelligence
Generative artificial intelligence (Claude, Anthropic) was used solely to assist with English language editing and proofreading of the manuscript. All AI‐assisted edits were reviewed and approved by the authors, who take full responsibility for the final content of the manuscript.
Results
Prescribing Trends and the FY2021–FY2022 Fluctuation
We used the NDB Open Data to examine the use of anticholinergics and β3‐adrenoceptor agonists for OAB from FY2017 to FY2024. The drugs analyzed and their ACB scores are listed in Table S1. The total volume of OAB medications rose each year, from 4.85 × 108 tablets in FY2017 to 8.37 × 108 tablets in FY2024, a 1.7‐fold increase (Figure 1A; Table S3).
This increase was driven by β3‐adrenoceptor agonists, which rose 3.1‐fold from 1.76 × 108 to 5.43 × 108 tablets. Anticholinergic use fell only slightly, from 3.10 × 108 to 2.94 × 108 tablets (−5.0%). The fall in anticholinergic use was therefore much smaller than the rise in β3‐adrenoceptor agonist use, and β3‐adrenoceptor agonists did not simply replace anticholinergics (Figure 1B).
The β3‐adrenoceptor agonist share rose from 36.2% to 64.9% and exceeded that of anticholinergics from FY2020 (Table S3). During this period, anticholinergic use fell temporarily in FY2021 (2.73 × 108 tablets; −7.2% from the previous year). Taking FY2020 as the reference (=100), anticholinergic volumes fell to 73.4–90.2 in FY2021 and then rose sharply to 99.6–253.9 in FY2022 (Figure 2A).
The change from FY2021 to FY2022 was largest for oxybutynin (+245.7%), followed by propiverine (+91.6%) and solifenacin (+10.4%). Mirabegron, which was available only as a brand‐name product at that time, showed no such fluctuation (+12.8% in FY2021 and −4.4% in FY2022) (Figure 2B; Table S4). Solifenacin was the only drug with a newly listed generic during this period (June 18, 2021), so this pattern cannot be explained by price.
A nationwide shortage of generic medicines occurred in Japan at the same time, and we consider that the fluctuation reflects this shortage. We therefore treated FY2021–FY2022 as a period of supply disruption. Data from these years are shown throughout, but when we estimated change over time we used FY2018 and FY2024 as endpoints, because both lie outside the disruption.
Sex Differences in the β3‐Adrenoceptor Agonist Share
We next examined the β3‐adrenoceptor agonist share by age and sex in FY2024 (Figure 3A). In men, the share rose with age and peaked at 73.0% at 75–79 years, and then stayed at a similar level (72.5% at 85–89 years and 72.1% at ≥ 90 years). In women, the share also peaked at 75–79 years (62.7%), but then fell (60.5% at 85–89 years and 58.0% at ≥ 90 years).
As a result, the difference between men and women widened with age, from +1.2 percentage points at 50–54 years to +14.2 percentage points at ≥ 90 years. Over time, the β3‐adrenoceptor agonist share rose in every age group, from 39.0% to 66.5% at 65–74 years, from 38.7% to 67.4% at 75–84 years, and from 33.3% to 64.0% at ≥ 85 years between FY2017 and FY2024. The share was higher in men than in women in every age group and in every year.
The sex difference narrowed slightly over time (from 17.2 to 13.0 percentage points at ≥ 85 years), but it remained larger in older age groups (in FY2024: +10.4 percentage points at 65–74 years, +10.6 at 75–84 years, and + 13.0 at ≥ 85 years) (Figure 3B; Table 1).
ACB‐Weighted Prescription Volume by Sex and Age
We then calculated ACB‐weighted prescription volume (patient‐days multiplied by the ACB score) by sex and age group (Figure 4). Exposure was higher in women than in men in every age group, and the female‐to‐male ratio increased with age: in FY2024 it was 1.1 at 65–74 years, 1.6 at 75–84 years, and 2.8 (range 2.8–2.9) at ≥ 85 years (Table 1). Between FY2018 and FY2024, prescription volume fell by about 20% at 65–74 years (−19.3% in men and −22.8% in women) and at 75–84 years (−21.3% in men and −21.1% in women).
In contrast, the fall was smallest in women aged ≥ 85 years (−11.3%; range −9.1% to −11.3%), and was also smaller than in men of the same age (−15.8%) (Figure 4C; Table S5). These findings did not change when imidafenacin was given an ACB score of 1, 2, or 3 (Table S2). In summary, the increase in the β3‐adrenoceptor agonist share was smaller in older women, and as a result the fall in ACB‐weighted prescription volume was also smaller in this group.
After standardization to population size, the same pattern persisted but was attenuated: in FY2024 the female‐to‐male ratio at ≥ 85 years was 1.33 (vs. 2.8 for the crude value) and still rose with age (1.04 at 65–74 years, 1.20 at 75–84 years, and 1.33 at ≥ 85 years). The decline between FY2018 and FY2024 remained smallest in women aged ≥ 85 years and was larger in magnitude after standardization (e.g., −23.6% vs. −11.3% for the crude value), because the population aged ≥ 85 years grew over the period (Figure 4B–D; Tables S5 and S6).
Discussion
In this study, we found that prescribing patterns for OAB in Japan are changing, with an increasing share of β3‐adrenoceptor agonists relative to anticholinergics, and that the β3‐adrenoceptor agonist share rose from 36.2% to 64.9%. However, this change differed clearly between men and women. In men, the β3‐adrenoceptor agonist share peaked at 75–79 years (73.0%) and then stayed high.
In women, it also peaked at 75–79 years (62.7%), but then fell, to 58.0% at ≥ 90 years. As a result, the difference between men and women widened with age and reached 14.2 percentage points at ≥ 90 years. Older women are therefore the group in which the increase in the β3‐adrenoceptor agonist share was smallest.
One possible explanation for this sex difference is that men often have benign prostatic hyperplasia (BPH). Anticholinergics reduce detrusor contractility, and they may cause acute urinary retention in men with bladder outlet obstruction. For this reason, they have long been regarded as contraindicated in men with lower urinary tract symptoms due to BPH [10, 11].
The risk of acute urinary retention also rises with age. This clinical constraint may have contributed to the greater use of β3‐adrenoceptor agonists in men. β3‐adrenoceptor agonists do not affect detrusor contractility, and they can be used safely in men who are already receiving drug treatment for BPH [12]. In women, this concern about urinary retention is much smaller, and anticholinergics are therefore prescribed with less hesitation.
In other words, the avoidance of anticholinergics in men may not reflect a deliberate concern about cognition, but may instead be a by‐product of a coexisting condition. If this interpretation is correct, then prescribing that actively takes cognitive risk into account is not yet common in either sex. We should also note that recent studies suggest that anticholinergics can be used safely in carefully selected men with BPH who have a low post‐void residual volume, often in combination with an α1‐blocker, and the idea of an absolute contraindication is being re‐examined [10].
Even so, our findings suggest that prescribing behavior in daily practice is still strongly shaped by this concern. Because the NDB Open Data contain no diagnoses, we could not test this mechanism directly, and it remains an inference. Other factors that we could not measure, such as physician specialty, care setting, prescribing preferences, and differences in patient characteristics, may also contribute, so BPH should be regarded as one of several possible explanations rather than the principal one.
As an indirect, ecological check, we compared age‐group patterns in men and found that the ratio of urinary‐selective BPH‐drug volume to OAB‐drug volume rose in parallel with the β3‐adrenoceptor agonist share across age groups (Figure S1); however, because the data are aggregated, we could not determine at the patient level how many men received OAB drugs together with BPH treatment or OAB drugs alone.
We also found that anticholinergic use fell temporarily in FY2021 and then rose sharply in FY2022. A likely explanation is that this fluctuation reflects the nationwide shortage of generic medicines that occurred in Japan in 2021. The fluctuation was seen only in anticholinergics for which a generic was available, and not in mirabegron, which was brand‐name only at that time.
The size of the change also differed greatly between drugs: from FY2021 to FY2022, oxybutynin rose by 245.7% and propiverine by 91.6%, whereas solifenacin, the only drug with a newly listed generic in that period, rose by only 10.4%. This pattern is not consistent with an explanation based on drug price. Our findings therefore show that the use of OAB drugs can be strongly affected not only by clinical decisions but also by the drug supply system.
In Japan, regulatory sanctions against several generic manufacturers in 2021 led to a widespread shortage, and by the end of 2022 about 40% of generic products were on stop or restricted supply. Pharmacoepidemiological studies that use Japanese claims data from FY2021 and FY2022 therefore risk mistaking a supply‐driven change for a change in prescribing behavior or for a price effect. We examined this point explicitly and treated these 2 years as a period of supply disruption.
This interpretation is based on the temporal association between the fluctuation and the supply disruption and does not represent a directly verified causal relationship. This methodological point should also be considered in other pharmacoepidemiological studies that use Japanese data from the same period.
A rise in the β3‐adrenoceptor agonist share does not necessarily mean that the anticholinergic burden in older people has fallen. We therefore quantified the ACB‐weighted prescription volume using the ACB score. ACB‐weighted prescription volume was higher in women than in men in every age group, and the female‐to‐male ratio rose with age, reaching 2.8 at ≥ 85 years.
Between FY2018 and FY2024, prescription volume fell by about 20% at 65–74 years and at 75–84 years, but the fall was smallest in women aged ≥ 85 years (−11.3%). These sex differences persisted after standardization for population size, although the female‐to‐male ratio was smaller (1.33 at ≥ 85 years in FY2024; Figure 4B–D). This measure is based on national dispensed volume weighted by the ACB score rather than on individually measured burden, and it is restricted to OAB drugs.
Cumulative anticholinergic exposure has been linked to cognitive decline and to a higher risk of dementia in several large studies. Gray et al. followed 3434 people and found a higher risk of dementia in those with the greatest cumulative anticholinergic exposure over 10 years [6]. Coupland et al., in a case–control study of 58 769 patients with dementia, found an association between dementia and the use of strong anticholinergics, including bladder antimuscarinics [7].
A higher risk of falls and fractures has also been reported. Our findings show that women aged ≥ 85 years, who are the most vulnerable to these risks, have both the highest anticholinergic exposure and the smallest improvement over time.
These findings suggest that efforts to review medication and reduce anticholinergic exposure in older people may need to focus in particular on women. In men, a coexisting condition (BPH) indirectly encourages avoidance of anticholinergics, but women have no such coexisting clinical constraint. In women, therefore, prescribers may wish to review the anticholinergic burden and, where appropriate, consider switching to a β3‐adrenoceptor agonist.
Switching may also be supported by better adherence: in a real‐world claims analysis, patients receiving vibegron had higher adherence (proportion of days covered 0.67 vs. 0.58) and longer persistence than those receiving anticholinergics [13]. Adverse effects of anticholinergics, such as dry mouth and constipation, are a common reason for stopping treatment in older people.
We also note that the total volume of OAB drugs rose 1.7‐fold, whereas anticholinergic use fell by only 5.0%. We cannot tell from these data whether the rise in β3‐adrenoceptor agonist use reflects switching from an anticholinergic, new prescriptions in patients who were not previously treated, or add‐on therapy in patients who continue to take an anticholinergic. Combination therapy with an anticholinergic and a β3‐adrenoceptor agonist is effective in refractory OAB and is used in Japan [8].
If add‐on therapy accounts for a large share of the increase, then the anticholinergic burden may not fall at all, while the number of drugs increases. This would also be a concern from the viewpoint of polypharmacy.
This study has several limitations. First, it is based on aggregate national prescription data and does not follow individual patients over time. We therefore cannot tell whether the observed changes reflect switching between drugs, add‐on therapy, or an increase in new prescriptions.
Patient‐level data are needed to answer this question. Second, dispensed volume does not reflect actual drug intake. The conversion into patient‐days is an approximation based on the standard daily dose, and adherence was not taken into account.
Third, the ACB scale was developed in Western populations and does not include imidafenacin. We assigned imidafenacin a score of 3 in the primary analysis and confirmed in sensitivity analyses that the conclusions did not change across scores of 1–3. Fourth, the ACB‐weighted prescription volume was derived from aggregated national dispensing data and was restricted to OAB drugs; it therefore reflects population‐level prescription volume weighted by the ACB score rather than the comprehensive anticholinergic burden of individual patients, and it does not capture co‐prescribed anticholinergic drugs from other classes or the concurrent use of two anticholinergics in the same patient.
Fifth, we did not measure cognitive function, falls, or other clinical outcomes; the concern about these outcomes is based on previous studies rather than on the present data. Despite these limitations, this is the first study to use nationwide, population‐wide data from Japan to show trends in the use of OAB drugs, the sex difference in these trends, and the change in anticholinergic burden over an eight‐year period.
Conclusion
Prescribing patterns for OAB in Japan are changing, with an increasing share of β3‐adrenoceptor agonists relative to anticholinergics; however, this increase is smaller in older women, in whom anticholinergic exposure remains high. In men, coexisting BPH may be one factor that indirectly leads to lower use of anticholinergics, whereas women have no such coexisting clinical constraint. Because cumulative anticholinergic exposure has been associated with cognitive decline and falls in previous studies, these findings support careful review of anticholinergic exposure and, where appropriate, consideration of alternatives in older people, and particularly in older women.
Because these aggregate data cannot distinguish treatment switching from new treatment initiation, add‐on therapy, or expansion of the treated population, we describe these findings as changes in the relative prescription share rather than as a confirmed shift from anticholinergics to β3‐adrenoceptor agonists.
Funding
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information