Association Between Use of Anti-gout Preparations and Dementia: Nested Case–Control Nationwide Population-Based Cohort Study

Objectives: Gout is the most common form of inflammatory arthritis and was found to be independently associated with incident dementia in the elderly. However, the associations between anti-gout preparations and dementia were not well-studied. Methods: Data were collected from Taiwan's National Health Insurance Research Database (NHIRD). A 2005–2013 retrospective cohort study was conducted, and all investigated subjects were identified by International Statistical Classification of Diseases and Related Health Problems, 9th Revision, Clinical Modification. Conditional logistic regression was used to evaluate the odds ratio of dementia in relation to different gout preparations (benzbromarone, allopurinol, sulfinpyrazone, probenecid) and number of days of anti-gout preparation use, after adjustment for potential confounding variables. Results: A total of 3,242 gout patients with and without dementia were selected from the NHIRD and included in the final analysis after 1:1 matching for age, gender, and diagnosis year of gout. In the anti-gout preparations, only use of Benzbromarone decreased the risk of dementia (adjusted OR, 0.81; 95% CI, 0.68–0.97). The result of the subgroup analysis revealed a trend toward a lower risk of dementia with longer use of benzbromarone. Use of benzbromarone for ≥180 days showed a significantly lower risk of dementia (adjusted OR, 0.72; 95% CI, 0.58–0.89). Moreover, the protective effect was more pronounced in males compared with females. Conclusion: This cohort study reveals that gout patients taking benzbromarone are at a decreased risk of developing incident dementia, especially with longer use and in male. Further prospective trials are warranted to confirm our findings.


INTRODUCTION
with renal insufficiency because it is cleared mainly via the liver (17).
The primary uricosuric drugs include probenecid, benzbromarone, and sulfinpyrazone. Benzbromarone increases urinary excretion of uric acid and the starting dose is 25-50 mg/d in adults, which should be adjusted to 75 or 100 mg/d according to the serum UA level (18). Benzbromarone can be used in patients with mild-to-moderate renal dysfunction. The recommended dose is 50 mg/d in patients with an eGFR of 20-60 ml·min −1 ·1.73 m −2 and is contraindicated in patients with an eGFR <20 ml·min −1 ·1.73 m −2 or uric acid nephrolithiasis. The most concern adverse reaction of benzbromarone is hepatic toxicity (19). Probenecid can be started by 250 mg twice daily for 1 week and may increase to a maximum of 2 g/day. Moreover, Probenecid is avoided use in patients with eGFR <30 mL/min (20,21). Sulfinpyrazone is started at a dose of 50 mg twice daily, with increments over several weeks to 100 to 200 mg three or four times daily as needed. The maximum effective dose of sulfinpyrazone is 800 mg/day, and avoid be used in persons with CKD or a history of uric acid kidney stones (22).
Like the description above, gout may be associated with various comorbidities including dementia; however, it is unclear whether anti-gout preparations can reduce the incidence of dementia. Hence, in the present study, we enrolled gout patients (newly diagnosed from 2000 to 2008) from the Taiwan Nation Health Insurance Research Database (NHIRD) to investigate the correlations of ULT and further dementia after at least 5 years from gout diagnosis.

Data Source
In 1995, Taiwan developed a health insurance system based on a single-payer government-run National Health Insurance (NHI) program, which currently includes comprehensive health care data for almost all Taiwanese citizens. The large, computerized databases that became the NHIRD were derived from this system by the Bureau of National Health Insurance and were maintained by the National Health Research Institutes, and include includes information about hospitalization, emergency care, and medical visits. These databases were used for research purposes, with a coverage rate >99% in 2010, and thus contains data for ∼23 million beneficiaries (23).
The Longitudinal Health Insurance Research Dataset 2000 contains the clinical information of one million beneficiaries randomly selected from the NHIRD during the period from 2000 to 2013 (24). Moreover, the International Statistical Classification of Diseases and Related Health Problems, 9th Revision, Clinical Modification (ICD-9-CM) is used for the diagnosis by physicians. Patients' medical data include drug items, NHI code, dosage, frequency of use, and number of days prescribed. This information can be used to detecting drug interactions and potential duplicate medications when patients visit multiple hospitals. The NHI code corresponds to a code in the five-level Anatomical Therapeutic Chemical classification system recommended by the World Health Organization for studies on drug utilization (25).

Data Collection
We identified one million people from the database. We selected patients aged ≥50 years with newly diagnosed gout based on the ICD-9-CM code 274 and use of antigout preparations, allopurinol (NHI code: M04AA01), benzbromarone (NHI code: M04AB03), sulfinpyrazone (NHI code: M04AB02), or probenecid (NHI code: M04AB01), within 6 months from 2000 to 2008. Dementia was defined as a diagnosis with one or more of the following ICD-9-CM codes: 290.0-290.4, 294.1, and 331.0-331.2. Moreover, the diagnosis of dementia before the first gout diagnosis date was excluded.
Individuals were divided into two groups. The dementia group comprised individuals newly diagnosed with dementia at following ≥2 outpatient assessments or a single inpatient admission after diagnosis of gout and 5 years apart. The nondementia (control) group comprised of individuals who had never been diagnosed with dementia after being diagnosed with gout. The index date was set at 5 years after taking anti-gout preparations. The ratio of patients in dementia to non-dementia groups was 1:1, with individual matching for age, gender, and gout diagnosis year.

Patient Subgroups
The age of the patients was defined on the index date. For the sex variable, zero represented female and one represented male. Baseline disease was diagnosed from two outpatient or one inpatient assessment 5 years before dementia diagnosis and Patients with no comorbidities were represented by 0, and those who developed comorbidity were represented by 1. Moreover, use of warfarin (NHI code: B01AA03) or statins (NHI codes: C10AA01-05, 07-08) 5 years before dementia diagnosis were also included in the baseline characteristics.

Statistical Analysis
All analyses were performed using SAS version 9.1.3 for Windows (SAS Institute, Inc., Cary, NC). P-values < 0.05 were considered statistically significant.
Kaplan-Meier analysis was applied to evaluate the cumulative incidence of dementia in the subgroups and log-rank test was to test for significance. Conditional logistic regression was used to evaluate the odds ratio (OR) of dementia in relation to different gout preparations, and number of days of anti-gout preparation use, after adjustment for potential confounding variables (allopurinol, benzbromarone, sulfinpyrazone, probenecid, hypertension, hyperlipidemia, CLD, CKD, DM, COPD, AD, CVD, stroke, warfarin, and statin). Subgroup analyses were performed by age (50-64 and ≥65 years) and gender.

RESULTS
A total of 3,242 gout patients with and without dementia were selected from the NHIRD and included in the final analysis after 1:1 matching for age, gender, and diagnosis year of gout. A flowchart of the study population selection protocol is shown in Figure 1. Table 1 shows the demographic characteristics of the dementia and non-dementia group. The dementia group had higher prevalence of underlying disease include hypertension, DM, COPD, CVD, stroke, depression, and PD. Moreover, the dementia group had a higher prevalence of warfarin use and a similar prevalence of statin use compared with the non-dementia group. The most common anti-gout preparation used was benzbromarone (73.7% in the dementia group, 77.5% in the nondementia group), followed by allopurinol (29.4% in the dementia group, 27% in the non-dementia group), sulfinpyrazone, and probenecid. Use of benzbromarone was significantly higher in the non-dementia group. Table 2 shows the conditional logistic regression of risk of dementia. After adjusting for confunding factors, patients with underlying DM (OR, 1.21; 95% CI, 1.03-1.43), COPD (OR, 1.31; 95% CI, 1.10-1.55), and stroke (OR, 2.76; (95% CI, 2.35-3.24) showed an increased risk of dementia. Moreover, use of benzbromarone decreased the risk of dementia (adjusted OR, 0.81; 95% CI, 0.68-0.97).
The result of the subgroup analysis of anti-gout preparations divided by the number of days of use revealed a trend toward a lower risk of dementia with longer use of benzbromarone. Use of benzbromarone for ≥180 days showed a significantly lower risk of dementia (adjusted OR, 0.72; 95% CI, 0.58-0.89) ( Table 3). Moreover, the protective effect of benzbromarone on dementia differed among the different age groups ( Table 4). In adults aged 50-64 years, benzbromarone use lowered the risk of dementia with use for <90 days (adjusted OR, 0.21; 95% CI, 0.06-0.76), and use for ≥180 days (adjusted OR, 0.21; 95% CI, 0.06-0.76). In the elderly (≥65 years), benzbromarone use for ≥180 days decreased the risk of dementia (adjusted OR, 0.73; 95% CI, 0.59-0.92). Table 5 shows the different odd ratios in males and females. Both males and females showed a trend toward a decreased risk of dementia with longer use of benzbromarone. However, only use for ≥180 days showed a significant OR. Moreover, the protective effect was more pronounced in males (adjusted OR, 0.68) compared with females (adjusted OR, 0.72).

DISCUSSION
The present study was a large population real-world study investigating the risk of dementia in gout patients using anti-gout preparations. Our findings indicated that use of benzbromarone to treat hyperuricemia and prevent recurrent gout flares in gout patients decreased the risk of incident dementia. Moreover, the protective effect was more pronounced in younger patients.
The ULT for gout patients may not be the same in different countries. In the United States and Europe, gout patients were treated with allopurinol as the preferred first-line agent and are strongly recommended for all patients with normal renal function (26,27). However, allopurinol can cause skin allergic reactions and fatal exfoliative dermatitis and other hypersensitivity syndromes might develop in severe cases.
Human leukocyte antigen (HLA)-B * 5801 allele positivity, which is significantly higher in Han Chinese than in Caucasians, is a risk factor for the development of adverse reactions to allopurinol, and screening for this gene before initiating allopurinol treatment is important (28,29). Since Allopurinol is contraindicated in HLA-B * 5801 allele-positive patients, in Asian areas, allopurinol is usually not the most commonly used ULT (30)(31)(32).
Moreover, febuxostat, which can be used in patients with mild-to-moderate CKD, is approved for the treatment of gout in the United States in doses of 40 or 80 mg; by contrast, European approval of febuxostat is for doses of 80 or 120 mg (26,27). For the choice of uricosuric drugs, probenecid is the only available drug in the United States, but it is neither as efficacious as benzbromarone, which is available in several other countries but not in the United States (26,33). Benzbromarone, a potent uricosuric drug, was introduced in the 1970s and was registered in about 20 countries throughout Asia, South America, and Europe. In 2003, the drug was withdrawn by Sanofi-Synthélabo, after reports of serious hepatotoxicity and had been not used in American. However, it is still marketed in several countries by other drug companies (33). For example, In Asia (including China and Japan) and Europe, benzbromarone is still the most use uricosuric drugs in gout patients, which is the same as in Taiwan (31,32,34). Hence, the use of benzbromarone associated with a reduced risk of future dementia can be a reference in most areas of the world. In Taiwan, a nationwide population study conducted from 2005 to 2010 reported that the prevalence of gout may be as high as 6.24%, with only 22.93% of patients with gout being prescribed urate-lowering treatment (allopurinol, benzbromarone, probenecid, or sulfinpyrazone) (35). Moreover, among patients who received treatment, 60.08% received uricosuric agents alone (benzbromarone, probenecid, or sulfinpyrazone), 28.54% received xanthine oxidase inhibitor (allopurinol), and 11.38% received combined therapy. In this study, we enrolled gout patients from 2000 to 2008 and is like the epidemiology investigation, most patients use uricosuric drugs than xanthine oxidase (Allopurinol). However, the febuxostat is approved by the U.S. Food and Drug Administration in 2009 and not listed until 2012 in Taiwan. Hence, in the study period, few patients have taken the drug and cannot be enrolled in the analysis.
Risk factors for Alzheimer's disease and vascular dementia are similar to those for CVD (36,37). Hyperuricemia and gout are also positively associated with CVD (38, 39); therefore, it is rational to postulate with common causality reasoning that gout and hyperuricemia are positively correlated with dementia and that anti-gout preparations have a protective effect. However, previous studies have shown these associations to be controversial.
In 2008, Euser et al. conducted a prospective cohort study investigating the association between serum UA and cognitive function and dementia in patients aged ≥55 years. The data showed higher serum UA levels were associated with a decreased risk of dementia and better cognitive function (40). Previous cohort studies have shown that patients with gout have a lower risk of developing dementia and that this phenomenon exists for both Alzheimer's disease and vascular dementia (41,42). These studies indicate that the protective effect may be due to the antioxidative effects of UA, which can reduce biological oxidants such as peroxynitrite radicals and protect from neurodegenerative diseases (43). However, this hypothesis remains controversial, as Hershfield et al. did not observe any changes in oxidative stress markers in patients receiving pegloticase, which greatly lowers serum UA levels (44). Furthermore, Desideri et al. reported that UA increased oxidative stress and potentiated the neurotoxic effects of amyloid-β in neuronal cells (45). In contrast, Singh et al. revealed that gout was independently associated with a 15% higher risk of incident dementia (5), and Latourte et al. found that high serum UA levels may increase the risk of incident dementia in the elderly, particularly vascular or mixed type (46). However, while antigout preparations are known to lower serum UA levels and  prevent gout flare, few studies have focused on the association between anti-gout preparations and dementia.
A recent case-control study showed a slight reduction in the risk of dementia in patients with hyperuricemia, both with and without anti-hyperuricemic treatment (47). In this study, allopurinol was the most frequently prescribed drug (98.4%), followed by benzbromarone (1.8%) and febuxostat (0.2%). Hence, the association between anti-gout preparations and dementia was mostly affected by allopurinol. However, In the present study, the most used drug was benzbromarone, followed by allopurinol. Use of allopurinol was not associated with incident dementia. However, the use of benzbromarone in patients with gout was associated with a lower risk of incident dementia (OR, 0.81).
Benzbromarone is a uricosuric drug that has been widely used since the 1970s as a therapeutic agent for hyperuricemia. Its mechanism of action was identified as suppression of UA resorption via inhibition of urate transporter 1 (18). At standard doses (100 mg/day), the hypouricaemic activity of benzbromarone is more efficacious than that of allopurinol (300 mg/day) and probenecid (1,000 mg/day) and leads to a mean reduction in plasma urate between 25 and 50% (48-51). Moreover, benzbromarone is more effective than probenecid and sulfinpyrazone in lowering UA and can be used in patients with moderate renal insufficiency (52). As expected from any marked reduction in plasma urate concentrations, long-term treatment with benzbromarone is more effective in reducing the frequency of acute gout attacks by 35-100% after 2 years of treatment (53,54). Besides, patients taking benzbromarone alone achieved a faster reduction of tophi than patients taking allopurinol alone (55). Hence, since hyperuricemia and gout may be risk factors for dementia (5,46), the more potent effects of benzbromarone in reducing serum UA level and preventing acute gout attacks may explain the protective effect from incident dementia. A recent study by Muraya et al. suggested that benzbromarone had direct antioxidant effects in vivo (56), and long-term use of antioxidant supplements can reduce the risk of dementia (57,58). Hence, this antioxidant effect can also explain why the use of benzbromarone was more effective in reducing incident dementia than other anti-gout preparations. The present study had some limitations. First, all diagnoses in the NHIRD were made by physicians using ICD-9-CM codes and were mainly registered by general practitioners rather than rheumatologists. Therefore, the definition of gout may have resulted in non-differential misclassification. However, patients were diagnosed with gout following at least two outpatient visits or one admission with the use of anti-gout preparations to ensure that only patients with an accurate diagnosis were selected. Second, information about lifestyles, such as smoking habits and alcohol consumption, was not collected in the insurance database. To reduce this bias, we adjusted for COPD to gauge smoking habits, CLD to reflect alcohol consumption, and other comorbidities. Third, as our data were limited to insurance claims and did not provide further information on treatment and test results of serum UA levels or gout flare, we could not attribute the results to the effect of the drugs. However, previous studies showed that gout patients treated with benzbromarone may have relatively lower serum UA compared with those using other agents.

CONCLUSIONS
The present study investigated patients with gout who were treated with ULT and showed that the use of benzbromarone was associated with a reduced risk of dementia. This effect may be due to the potent UA-lowering and antioxidant effects of benzbromarone. Further prospective trials are warranted to confirm our findings.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary materials, further inquiries can be directed to the corresponding author/s.

ETHICS STATEMENT
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.