Abstract
Objectives
Benign prostatic hyperplasia (BPH) refers to nonmalignant hyperplasia of prostate tissue, which causes lower urinary tract symptoms and has become a global public health concern in the aging population. The purpose of this study is to identify modifiable factors, which would prevent or delay BPH development.
Methods
The association between BPH marker drugs and climate-, socioeconomic-, health condition-, and lifestyle habits-related variables was investigated by analyzing nationwide datasets which were collected in 2018, aggregated by prefecture (administrative unit), and published by Japanese ministries. Uroselective α1 receptor blockers and dutasteride were used as marker drugs referring to BPH prevalence. Correlation analysis, multiple linear regression analysis, and binomial logistic regression analysis were conducted with 47 Japanese prefectures as the unit.
Results
The variables which showed |r| > 0.5 by correlation analysis were exercise habits (r = −0.5696), smoking habits (r = 0.6116), and daily drinking (r = 0.6001) for uroselective α1 receptor blockers, and antihypertensive medication (r = 0.5971), smoking habits (r = 0.6598), a small amount of drinking (r = −0.5292), and serum alanine aminotransferase (r = 0.6814) for dutasteride. Multiple linear regression equations were constructed by including these variables (R2 = 0.5453 for uroselective α1 receptor blockers and R2 = 0.5673 for dutasteride). Binomial logistic regression analysis found a significant association between climate in the resident area and BPH development.
Conclusion
This ecological study, analyzing Japanese nationwide datasets, demonstrates that healthy lifestyle habits, especially avoidance of smoking, implementation of exercise in daily life, and a small amount of alcohol consumption, are important to prevent or delay BPH development. High blood pressure and high serum alanine aminotransferase are suggested as risk factors of BPH development.
Keywords: Benign prostatic hyperplasia, Climate, Health condition, Lifestyle habits, Socioeconomic variable
Abbreviations: ALT, serum alanine aminotransferase; BMI, body mass index; BP, blood pressure; BPH, benign prostatic hyperplasia; Ccr, creatinine clearance; eGFR, estimated glomerular filtration rate; HbA1c, hemoglbinA1c; HDL-C, high density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; MHLW, Ministry of Health, Labour and Welfare; NDB, National Database of Health Insurance Claims and Specific Health Checkups; VIF, variance inflation factor
1. Introduction
Benign prostatic hyperplasia (BPH) refers to nonmalignant hyperplasia of prostate tissue1 and is the most common condition of the prostate in aging men worldwide2. BPH causes urethra compression and bladder outflow obstruction, resulting in clinical manifestations of symptoms associated with voiding and/or storage disturbances3. A study showed that there were 11.26 million new cases of BPH globally in 20194. The treatment options include pharmacotherapy, whereas surgical therapies are recommended if moderate to severe symptoms remain even after enough nonsurgical treatments5, 6, 7. Owing to the rapid increase in the world's aging population8, BPH has become a global public health concern1,2,9. Although influences of health conditions and lifestyles on BPH development have not attracted sufficient attention10, it is warranted to identify modifiable factors to prevent or delay BPH development11. Although the health condition- and lifestyle habits-related factors are considered to influence BPH development12, no definitive conclusions have been made. For example, a study demonstrated that metabolic syndrome was associated with prostatic volume (morphology) but not BPH symptoms13. A meta-analysis on smoking habits did not find a significant association with BPH development14 although other studies have demonstrated a significant impact of smoking habits on BPH development15. Regarding alcohol drinking habits, some studies have demonstrated that moderate alcohol consumption reduces BPH development15,16.
Japan is an island country with a bow-shaped land that is wide in both east-west and north-south directions. It is divided into 47 administrative units (prefectures), and each prefecture has its own unique climate, socioeconomic status, health condition, and lifestyle habits based on geographical feature. Japanese ministries have annually collected data on various categories, including demographics, climate, socioeconomics, health conditions, and lifestyle habits over the country and aggregated them by prefecture. Among them, National Database of Health Insurance Claims and Specific Health Checkups (NDB) Open Data Japan contains almost all insurance claim data from all over the nation under the universal health insurance system17,18. It also contains millions of health check-up records of adults aged 40–74 years19. These datasets are published by the Japanese government and have increasingly been utilized for researches including clinical epidemiology18. Thus, analysis of the nationwide datasets aggregated by prefecture is an efficient research approach to explore factors associated with the target of interest health outcome. For example, a recent study using the datasets including NDB Open Data Japan has revealed that the prevalence of chronic constipation is closely associated with low outside temperature and use of antihypertensive drugs20.
The purpose of this study is to identify modifiable factors, which would prevent or delay BPH development, by analyzing climate-, socioeconomic-, health condition-, and lifestyle habits-related datasets published by Japanese ministries.
2. Materials and methods
2.1. Study design
This cross-sectional ecological study investigated the associations between climate-, socioeconomic-, health condition-, and lifestyle habits-related variables and BPH drugs variable referring to BPH prevalence, using Pearson's correlation analysis, multiple linear regression analysis, and binomial logistic regression analysis with 47 Japanese prefectures as the unit. The research protocol was approved by the Research Ethics Committee of the International University of Health and Welfare (approval number: 21-Ifh-045).
2.2. Marker drugs referring to BPH prevalence
The number of claims for reimbursement of drugs exclusively used for BPH treatment, per 1,000 male population, was used as the variable referring to BPH prevalence. Two classes of drugs, which are at the position of the highest grade recommendation in the clinical guidelines for BPH by the Japanese Urological Association5 and approved exclusively for treating BPH by the Japanese authority, were used as marker drugs. One of them was the uroselective α1 receptor blocker (i.e., tamsulosin, naftopidil, and silodosin), which relieves outlet obstruction by inhibiting contractions mediated by prostatic α1-adrenaline receptors5. The other was the 5α-reductase inhibitor (i.e., dutasteride [brand name: Avolve]), which inhibits the conversion of testosterone to dihydrotestosterone and hinders prostatic growth, shrinks prostate volumes, and improves BPH-related lower urinary tract symptoms5,21. It is particularly beneficial for patients with an enlarged prostate (≥30 mL)22. Finasteride, another 5α-reductase inhibitor, was not employed as a marker drug because it was not approved for the treatment of BPH in Japan.
2.3. NDB Open Data Japan
NDB Open Data Japan was constructed based on the universal health coverage system covering most of the medical care in Japan. All individuals who have resided in Japan for 3 months or more need to be enrolled in the public health insurance program. Under this universal health coverage, patients pay coinsurance to a medical provider for medical service, including the prescription of drugs, and the provider submits the claims to the insurer for reimbursement of the remaining portion of the service fee. Nationwide exhaustive data on the number of claims for reimbursement of individual drugs can be collected during the abovementioned procedures and stored in the NDB Open Data Japan23. The number of claims of marker drugs per 1,000 male population can thus be used as the variable, which refers to BPH prevalence. NDB Open Data Japan also contains data on specific health check-ups and answers to lifestyle habits-related questions. In Japan, public health insurers are obliged to provide a specific health checkup, including questions regarding lifestyle habits, for insured individuals aged 40–74 years24. More than 29 million eligible adults participated in the check-up in 2018, and the results were stored in the 6th NDB Open Data Japan19.
2.4. Variables and data sources
All datasets used in the present study were collected in 2018 and have been published on the webpages managed by Japanese ministries. Data on BPH prevalence, i.e., number of claims for reimbursement of the marker drugs for outpatients per 1,000 male population, were acquired from “the 5th NDB Open Data Japan,” which contains a total number of 1.9 billion claims23. Data on demographic-, climate-, and socioeconomics-related variables were obtained from the System of Social and Demographic Statistics, which is a government statistics portal site where Japanese statistics in various fields are stored and can be browsed25. Data corresponding to health condition-related variables, i.e., the prefecture level average values of health check-up results, were obtained from the 6th NDB Open Data Japan, which contains approximately 29.4 million health check-up records of adults aged 40–74 years19. Data corresponding to lifestyle habits-related variables, i.e., the percentage of respondents who selected each answer option in the lifestyle habits-related questions, were also obtained from the 6th NDB Open Data Japan19. Since answers to questions were collected as part of the health check-up, approximately 29.4 million responses (same as the health check-up records) were included in it19.
2.5. Standardization of health condition- and lifestyle habits-related variables by age
Since health condition- and lifestyle habits-related data aggregated by prefecture are largely affected by the age structure of prefecture residents, they are standardized by age using the following formula: standardized data = (∑ age-specific raw data in a 5-year age group × standard population in that age group)/(total population in standard population), where the Japanese demographic composition [14] is used as the standard population. The variables and data sources used in this study are summarized in Table 1.
Table 1.
Variables for analysis and data sources
| Category | Variables | Abbreviation | Unit | |
|---|---|---|---|---|
| BPH drugs | Number of claims for reimbursement of uroselective α1 receptor blockers/male population | Uroselective α1 receptor blockers | n/1,000 men | 20 |
| Number of claims for reimbursement of dutasteride/male population | Dutasteride | n/1,000 men | ||
| Demography | Average age | Age | age | 22 |
| Climate | Annual average temperature | Temperature | °C | |
| Annual precipitation | Precipitation | mm | ||
| Annual average relative humidity | Humidity | % | ||
| Socioeconomics | Average population density | Population density | persons/km2 | |
| Average monthly pretax income per two or more persons households | Household income | Yen | ||
| Enrollment rate of male high school graduate | Enrollment rate | % | ||
| Number of doctors per population | Number of doctors | n/1,000 persons | ||
| Health condition | Average body mass index (BMI) | BMI | kg/m2 | 17 |
| Average serum HbA1c | Serum HbA1c | % | ||
| Average serum high density lipoprotein cholesterol (HDL-C) | Serum HDL-C | mg/dL | ||
| Average serum low density lipoprotein cholesterol (LDL-C) | Serum LDL-C | mg/dL | ||
| Average serum alanine aminotransferase (ALT) | Serum ALT | U/L | ||
| Average creatinine clearance (Ccr) | Ccr | mL/min | ||
| Average estimated glomerular filtration rate (eGFR) | eGFR | mL/min/1.73 m2 | ||
| Positive result in urine protein test | Urine protein | % of men | ||
| Average serum hemoglobin | Serum hemoglobin | g/dL | ||
| Average systolic blood pressure (BP) | Systolic BP | mmHg | ||
| Abnormal finding in electrocardiogram test | ECG | % of men | ||
| Lifestyle habits-related questions | Are you under any medication for high blood pressure? (Y/N) | Antihypertensive medication | % Yes | |
| Are you under insulin injections or other medications to reduce blood glucose level? (Y/N) | Antidiabetic medication | % Yes | ||
| Are you under medication to reduce cholesterol level? (Y/N) | Antihyperlipidemic drug | % Yes | ||
| Have you ever been diagnosed as stroke (e.g., cerebral hemorrhage, cerebral infarction) or got treated for it? (Y/N) | History of stroke | % Yes | ||
| Have you ever been diagnosed as heart disease (e.g., angina, myocardial infarction) or got treated for it? (Y/N) | History of heart disease | % Yes | ||
| Have you ever been diagnosed with chronic kidney failure or got treated (e.g., dialysis) for it? (Y/N) | History of kidney failure | % Yes | ||
| Have you ever been diagnosed as anemic? (Y/N) | History of anemia | % Yes | ||
| Are you a habitual cigarette smoker (a total of over 100 cigarettes or for over six months) at present? (Y/N) | Smoking habits | % Yes | ||
| Have you gained 10 kg or more compared to your body weight when you were 20 years-old? (Y/N) | Body weight gain | % Yes | ||
| Have you been doing sweaty exercise (30 min or more per session and two days or more per week) for over a year? (Y/N) | Exercise habits | % Yes | ||
| Have you been walking or doing any equivalent amount of physical activity over an hour per day in everyday life? (Y/N) | Walking habits | % Yes | ||
| Is your walking speed faster than the speed of those of almost the same age and of the same gender? (Y/N) | Walking speed | % Yes | ||
| What is your condition when you chew your meal? (Chew anything/Sometimes hard to chew/Hardly chew) | Chewing condition | % Chew anything | ||
| Is your eating speed faster than others? (Faster/Ordinary/Slower) | Fast eating speed | % Faster | ||
| Do you eat snacks other than three meals? (Y/N) | Snack habits | % Yes | ||
| Do you skip breakfast three times or more per week? (Y/N) | Skipping breakfast | % Yes | ||
| How often do you drink alcohol? (Every day/Sometimes/Rarely or never) | Daily drinking | % Every day | ||
| How much do you drink in terms of sake per day? 180 mL of sake is equivalent to 500 mL of beer. (<180 mL/180–360 mL/360–540 mL/540 mL <) | A small amount of drinking | % < 180 mL of sake |
All data were collected in 2018 and have been made public by Japanese ministries. Health condition- and lifestyle habits-related variables are standardized by age using the Japanese demographic composition14 as the standard population. Y: Yes, N: No, n: number.
2.6. Statistical analysis
The association between socioeconomic-, health condition-, and lifestyle habits-related variables and BPH drugs variable was investigated by Pearson's correlation analysis. Coefficients (r) for all possible pairs of variables were first calculated, and their statistical significances were tested using the t-test. Variables with a correlation coefficient of 0.5 or larger as the absolute value were put into the multiple regression equations, where uroselective α1 receptor blockers claims and dutasteride claim were used as response variables. Binomial logistic regression analysis was conducted for investigating the relationship between BPH marker drugs and climate categorical variables, where the median in the sorted climate values of 47 prefectures (16.8 °C for temperature, 1,760 mm for precipitation, and 71% for relative humidity) was used as cut-off value to categorize climate data into high and low categories. Statistical significance was set at P < 0.05. Statistical analysis was performed using BellCurve for Excel version 3.20 (Social Survey Research Information Co., Ltd., Tokyo, Japan).
3. Results
3.1. Correlation analysis
The correlations between the BPH drug variables (uroselective α1 receptor blockers and dutasteride) and age-, socioeconomics-, health condition-, and lifestyle habits-related variables are shown in Table 2. The variables with |r| > 0.5 by Pearson's correlation analysis are marked with bold face. Among health condition-related variables, serum alanine aminotransferase (ALT) represented a positive correlation with dutasteride (r = 0.6814). Among lifestyle-related variables, smoking habits were positively correlated with both uroselective α1 receptor blockers (r = 0.6116) and dutasteride (r = 0.6598). Exercise habits and a small amount of drinking were inversely correlated with uroselective α1 receptor blockers (r = −0.5696) and dutasteride (r = −0.5292), respectively. Daily drinking and antihypertensive medication were positively correlated with uroselective α1 receptor blockers (r = 0.6001) and dutasteride (r = 0.5971), respectively. Other socioeconomics-, health-condition-, or lifestyle habits-related variables did not show |r| larger than 0.5. While age, a demography variable, was closely correlated with both marker drugs, the r was larger for uroselective α1 receptor blockers (r = 0.8961) than for dutasteride (r = 0.6469).
Table 2.
Correlations between BPH drugs variables and demography-, socioeconomics-, health condition-, and lifestyle habits-related variables
| Category | Variables | Uroselective α1 receptor blockers |
Dutasteride |
|---|---|---|---|
| r with 95% CI | r with 95% CI | ||
| BPH drugs | α1 receptor blockers claim | – | 0.7615 (0.6071 ∼ 0.8605)∗∗ |
| Dutasteride claim | 0.7615 (0.6071 ∼ 0.8605)∗∗ | – | |
| Demography | Age | 0.8961 (0.8199 ∼ 0.9411)∗∗ | 0.6469 (0.4419 ∼ 0.7878)∗∗ |
| Socioeconomics | Population density | −0.4420 (−0.8402 ∼ −0.1774)∗∗ | −0.3359 (−0.7658 ∼ −0.0540)∗ |
| Enrollment rate | −0.4021 (−0.6179 ∼ −0.1300)∗∗ | −0.4545 (−0.6560 ∼ −0.1924)∗∗ | |
| Household income | 0.0830 (−0.2091 ∼ 0.3616) | −0.0617 (−0.3428 ∼ 0.2295) | |
| Number of doctors | −0.0409 (−0.3243 ∼ 0.2492) | −0.1765 (−0.4413 ∼ 0.1165) | |
| Health conditiona | BMI | −0.0947 (−0.3718 - 0.1978) | 0.2769 (−0.0111 ∼ 0.5225) |
| Serum HbA1c | −0.0585 (−0.3400 ∼ 0.2325) | 0.0786 (−0.2134 ∼ 0.3577) | |
| Serum HDL-C | 0.1877 (−0.1051 ∼ 0.4506) | 0.1785 (−0.1145 ∼ 0.4430) | |
| Serum LDL-C | −0.3772 (−0.5994 ∼ −0.1010)∗∗ | −0.2682 (−0.5157 ∼ 0.0205) | |
| Serum ALT | 0.4889 (0.2346 ∼ 0.6805)∗∗ | 0.6814 (0.4902 ∼ 0.8101)∗∗ | |
| Ccr | −0.0347 (−0.3187 ∼ 0.2550) | 0.0406 (−0.2495 ∼ 0.3240) | |
| eGFR | 0.0397 (−0.2503 ∼ 0.3232) | 0.0643 (−0.2271 ∼ 0.3451) | |
| Urine protein | −0.3435 (−0.5740 ∼ −0.0625)∗ | −0.2391 (−0.4924 ∼ 0.0516) | |
| Serum hemoglobin | 0.1749 (−0.1182 ∼ 0.4400) | 0.3336 (0.0513 ∼ 0.5665)∗ | |
| Systolic BP | 0.4772 (0.2202 ∼ 0.6723)∗∗ | 0.3604 (0.0817 ∼ 0.5868)∗ | |
| ECG | −0.0152 (−0.3010 ∼ 0.2732) | 0.1945 (−0.0982 ∼ 0.4562) | |
| Lifestyle habits-related questionsa | Antihypertensive medication | 0.2956 (0.0092 ∼ 0.5372)∗ | 0.5971 (0.3740 ∼ 0.7548)∗∗ |
| Antidiabetic medication | 0.1826 (−0.1104 ∼ 0.4463) | 0.4356 (0.1696 ∼ 0.6424) ∗∗ | |
| Antihyperlipidemic drug | −0.0683 (−0.3487 ∼ 0.2232) | 0.0022 (−0.2851 ∼ 0.2892) | |
| History of stroke | −0.3417 (−0.5727 ∼ −0.0604)∗ | −0.1062 (−0.3817 ∼ 0.1867) | |
| History of heart disease | −0.0041 (−0.2909 ∼ 0.2834) | 0.2781 (−0.0098 ∼ 0.5235) | |
| History of kidney failure | −0.4215 (−0.6322 ∼ −0.1528)∗∗ | −0.2167 (−0.4744 ∼ 0.0751) | |
| History of anemia | −0.3371 (−0.5692 ∼ −0.0553)∗ | −0.3489 (−0.5781 ∼ −0.0686)∗ | |
| Smoking habits | 0.6116 (0.3936 ∼ 0.7645)∗∗ | 0.6598 (0.4598 ∼ 0.7961)∗∗ | |
| Body weight gain | −0.1262 (−0.3989 ∼ 0.1671) | 0.0797 (−0.2123 ∼ 0.3587) | |
| Exercise habits | −0.5696 (−0.7364 ∼ −0.3377)∗∗ | −0.3026 (−0.5426 ∼ −0.0170) ∗ | |
| Walking habits | −0.1450 (−0.4149 ∼ 0.1484) | 0.1573 (−0.1360 ∼ 0.4252) | |
| Walking speed | −0.3412 (−0.5723 ∼ −0.0599)∗ | −0.1222 (−0.3955 ∼ 0.1709) | |
| Chewing condition | −0.1758 (−0.4407 ∼ 0.1173) | 0.2039 (−0.4639 ∼ 0.0885) | |
| Fast eating speed | −0.4091 (−0.6231 ∼ −0.1382)∗∗ | −0.4218 (−0.6324 ∼ −0.1532)∗∗ | |
| Snack habits | −0.1205 (−0.3940 ∼ 0.1727) | −0.3959 (−0.6133 ∼ −0.1226) ∗∗ | |
| Skipping breakfast habits | −0.4116 (−0.6249 ∼ −0.1411)∗∗ | −0.0814 (−0.3601 ∼ 0.2107) | |
| Daily drinking | 0.6001 (0.3781 ∼ 0.7568)∗∗ | 0.4083 (0.1372 ∼ 0.6225) ∗∗ | |
| A small amount of drinking | −0.3725 (−0.5959 ∼ −0.0956)∗∗ | −0.5292 (−0.7087 ∼ −0.2854)∗∗ |
Bold face represents |r| > 0.5 by Pearson's correlation analysis. CI, confidence interval; ∗p < 0.05; ∗∗p < 0.01 by t-test.
Values, that are age standardized for the age range of 40–74 years, are used as health condition- and lifestyle-related variables.
3.2. Multiple linear regression analysis
The variables with |r| > 0.5 were put into the multiple linear regression equations. Age was excluded from the equations because the main purpose of this study is to identify modifiable factors to prevent BPH development. Multiple linear regression equations were developed using the remaining variables with |r| > 0.5. Table 3 presents the regression coefficients of the equations. Exercise habits, smoking habits, and daily drinking were included in the equation for uroselective α1 receptor blockers (R2 = 0.5453, P < 0.001). Multicollinearity was not suspected, as variance inflation factor (VIF) was 1.5451 or less for these variables. Antihypertensive medication, smoking habits, a small amount of drinking, and serum ALT were included in the equation for dutasteride (R2 = 0.5673, P < 0.001). Multicollinearity was not suspected because VIF was 3.6226 or less for these variables. Smoking habits showed a positive impact on both uroselective α1 receptor blockers and dutasteride. Exercise habits and daily dinking influenced uroselective α1 receptor blockers inversely and positively, respectively. Antihypertensive medication and serum ALT positively affected dutasteride, while a small amount of drinking inversely influenced it.
Table 3.
Multiple regression equations for BPH marker drugs
| Response variables | Explanatory variables and constant term | B with 95% CI | P | Β | VIF | R2 |
|---|---|---|---|---|---|---|
| Uroselective α1 receptor blockers | Exercise habits | −255.6 | 0.0253 | −0.2813 | 1.3928 | 0.5453 p < 0.001 |
| (−477.9 ∼ −33.2) | ||||||
| Smoking habits | 321.3 | 0.0209 | 0.3065 | 1.5451 | ||
| (51.1 ∼ 591.6) | ||||||
| Daily drinking | 244.7 | 0.0099 | 0.3292 | 1.4074 | ||
| (61.8 ∼ 427.5) | ||||||
| Constant term | 2167.8 | 0.7543 | ||||
| (−11709.2 ∼ 16044.9) | ||||||
| Dutasteride | Antihypertensive medication | 53.8 | 0.2220 | 0.1872 | 2.2128 | 0.5673 p < 0.001 |
| (−33.8 ∼ 141.3) | ||||||
| Smoking habits | 86.5 | 0.0072 | 0.3721 | 1.6822 | ||
| (24.7 ∼ 148.2) | ||||||
| A small amount of drinking | −1.7 | 0.9232 | −0.0147 | 2.2262 | ||
| (−37.9 ∼ 34.4) | ||||||
| Serum ALT | 219.6 | |||||
| (−68.8 ∼ 508.0) | 0.1318 | 0.2969 | 3.6226 | |||
| Constant term | 7937.9 | 0.0378 | ||||
| (−15407.6 ∼ −468.2) |
CI, confidence interval; B, partial regression coefficient; β, standardized partial regression coefficient indicating the impact of one individual variable, independent of the scales of the variables; VIF, variance inflation factor; R2, coefficient of determination; Negative B and β values indicated an inverse correlation.
3.3. Binomial logistic regression analysis
Binomial logistic regression equations were constructed using the categorical climate data (high or low) as response variables and the marker drugs as explanatory variables. The result was summarized in Table 4. The regression coefficient for the temperature variable was negative and statistically significant (P = 0.0084) in the equation for uroselective α1 receptor blockers (likelihood ratio = 10.0925, P = 0.0015), indicating that temperature is negatively associated with the BPH prevalence. The regression coefficient for the relative humidity variable was positive and statistically significant (P = 0.0333) in the equation for dutasteride (likelihood ratio = 5.6262, P = 0.0177), suggesting that humidity is associated with the increase of BPH prevalence. The regression coefficient for the precipitation was not statistically significant in the equation for either BPH marker drug.
Table 4.
Binomial logistic regression analysis between BPH marker drugs and climate variables
| Categorical climate variables | BPH marker drugs | McFadden R2 | Likelihood ratio | B with 95% CI | Wald tatistics | Constant term with 95% CI | Odds ratio with 95% CI |
|---|---|---|---|---|---|---|---|
| Temperature (0 or 1) | Uroselective α1 receptor blockers | 0.1549 | 10.0925 | −0.0005 | 6.9404 | 7.1405 | 0.9995 |
| p = 0.0015∗∗ | (−0.0008 ∼−0.0001) | p = 0.0084∗∗ | (1.8307 ∼ 12.4503) | (0.9992 ∼ 0.9999) | |||
| Precipitation (0 or 1) | 0.0006 | 0.0379 | 0.00002 | 0.0379 | −0.3021 | 1.0000 | |
| p = 0.8457 | (−0.0002 ∼0.0002) | p = 0.8457 | (−3.8239 ∼ 3.2196) | (0.9998 ∼ 1.0002) | |||
| Relative humidity (0 or 1) | 0.0585 | 3.8007 | 0.0002 | 3.8007 | −3.6524 | 1.0002 | |
| p = 0.0512 | (−0.00002 ∼0.0005) | p = 0.0512 | (−7.8103 ∼ 0.5055) | (1.0000 ∼ 1.0005) | |||
| Temperature (0 or 1) | Dutasteride | 0.0541 | 3.5233 | −0.0010 | 3.0943 | 2.4367 | 0.9990 |
| p = 0.0605 | (−0.0022 ∼0.0001) | p = 0.0786 | (−0.2746 ∼ 5.1481) | (0.9978 ∼ 1.0001) | |||
| Precipitation (0 or 1) | 0.0081 | 0.5259 | −0.0004 | 0.5159 | 0.9191 | 0.9996 | |
| p = 0.4684 | (−0.0014 ∼0.0007) | p = 0.4726 | (−1.5383 ∼ 3.3764) | (0.9986 ∼ 1.0007) | |||
| Relative humidity (0 or 1) | 0.0866 | 5.6262 | 0.0014 | 4.5293 | −3.0598 | 1.0014 | |
| p = 0.0177∗ | (0.0001 ∼ 0.0027) | p = 0.0333∗ | (−6.0078 ∼ −0.1119) | (1.0001 ∼ 1.0027) |
CI, confidence interval; B, regression coefficient; The median in the sorted climate values of 47 prefectures was used as cut-off value to categorize climate data into low (0) and high (1) categories (16.8°C for temperature, 1,760 mm for precipitation, and 71% for relative humidity).
4. Discussion
The analysis of NDB Open Data Japan suggests that smoking habits have a precipitative impact on BPH development. This result supports the previous studies demonstrating a significant correlation between smoking habits and BPH development. For example, a study reported that those who smoked 35 or more cigarettes per day had a higher risk of BPH development than nonsmokers15. Previous studies investigating changes in molecular expression have shown that inflammation26, generation of reactive oxygen species27, and downregulation of glucocorticoid receptor28 are induced in the prostate by exposure to cigarette smoking. These changes could underlie the disruption of proliferative and apoptotic balance of the prostate cell and the development of prostate proliferative diseases. Regarding alcohol drinking habits, the inverse correlation between a small amount of drinking and dutasteride in the present study is consistent with those in a large-scale prospective study that demonstrated that individuals who moderately consume alcohol are at a lower risk of BPH development than nondrinkers15,29. A previous study has shown that moderate alcohol consumption for a month lowers the plasma testosterone level via decreased production and increased metabolism in men30. Given that proliferation of prostate is definitely dependent on the androgen, the reduction of plasma androgen is possibly involved in the preventive effect of alcohol on BPH development31. Importantly, the present study also demonstrates that the beneficial influence of alcohol intake is limited to only modest consumers (<180 mL of sake/day). The positive correlation between daily drinking and uroselective α1 receptor blockers suggest that excessive alcohol consumption contrarily increases the risk of BPH development. It has become clear that excessive alcohol intake damages structure of organelles (e.g., Golgi body) of the prostate cell, which could cause abnormal prostate proliferation32. The beneficial effect which modest consumers obtain would be counteracted by the cellular damage occurring at high concentration of alcohol in binge drinkers. Next, the inverse correlation between exercise habits and uroselective α1 receptor blockers demonstrates that implementation of exercise in daily life reduces the risk of BPH development. This result coincides with a meta-analysis which demonstrates that lifestyle with moderate to high intensity exercise significantly reduces the risk of BPH development33. Another study also demonstrates that increased physical activity reduces the risk of development of lower urinary tract symptoms34. Previous studies have shown that physical exercise causes reduction of androgen receptors35 and enhancement of apoptosis35 in the prostate, and decrease of testosterone36 and inflammation marker proteins37 in the plasma. Those biological reactions induced by physical exercise could be associated with the prevention of BPH development. Above all, the findings in the present study strongly suggest that lifestyle habits are tightly associated with BPH development. Healthy lifestyle habits, especially avoidance of smoking, implementation of exercise in daily life, and moderate alcohol consumption, would prevent or delay BPH development.
A significant association between metabolic syndrome (a cluster of conditions including hypertension, hyperglycemia, and obesity) and BPH development has been suggested38 although the impact of metabolic syndrome-related factors on BPH development has been inconclusive. The present study detected hypertension, among variables associated with metabolic syndrome, as a stimulatory factor on BPH development. This finding is consistent with a previous study demonstrating that the prostate volume positively correlates with systolic BP in patient with BPH39. Another literature suggests that cardiovascular disease risk factors, including hypertension, are associated with BPH development10. The present study also detected that serum ALT (a deviation enzyme from hepatocytes) was strongly associated with dutasteride. This result is in accordance with a recent study which demonstrates a positive association between development of nonalcoholic fatty liver disease, a metabolic aberration-associated disease, and prostate volume40. Of note, a recent systematic review demonstrates that nonalcoholic fatty liver disease should be considered as a risk factor for BPH development41. On the other hand, the present study did not detect a tight correlation between BPH drugs and other metabolic syndrome-related variables including BMI, HbA1c, and serum cholesterols. The impact of metabolic syndrome-related factors on BPH development seems inconclusive based on previous studies. Although a literature suggests that obesity is a predictive factor42, another study demonstrates no significant relationship between BMI and BPH development43. A study demonstrated that an increased amount of fat mass and serum LDL-C levels even reduced the risk of BPH development within 5 years in men who were healthy at baseline44. Another study reported a precipitative influence of metabolic syndrome on prostate volume (morphology) but not BPH symptoms13. Importantly, the relationship between metabolic syndrome and BPH development has been inconclusive, and the mechanism underlying the relationship is far from elucidated. The present study suggests especially high blood pressure and serum ALT as risk factors of BPH among metabolic syndrome-associated factors. Other factors are, however, possibly involved, and further studies are warranted.
The present study shows that age is firmly associated with BPH development, which is consistent with previous studies45. Of note, the r referring to the strength of association between age and BPH prevalence is larger for the uroselective α1 receptor blockers than for dutasteride. Considering this result, some previous studies suggest that the prostate growth rate declines at advanced ages. For example, a study reported that the prostate growth rate peaked at 55–65 years of age and then declined46. The slow growth rate of the prostate at advanced ages might be responsible for the smaller r for dutasteride compared to uroselective α1 receptor blockers because the recommendation of dutasteride use in Japan is based on the volume of the prostate (≥30 mL).
The association between socioeconomics-related variables and BPH development in correlation analysis was weak although previous studies suggest a significant association47. This difference would be largely dependent on the region where the study was conducted because the socioeconomic status varies substantially by country or region. Regarding climate-related variables, the present study suggests that climate (temperature and humidity) in the resident area is associated with BPH prevalence. While the cause of this association remains to be elucidated, climate possibly affects residents’ lifestyle habits that are firmly associated with BPH development.
This study has several limitations. First, a snapshot obtained in the study with a cross-sectional design does not establish a causal association. Second, because this study used data aggregated by prefecture but not data at individual level, it should be noted that the association between the explanatory variables and BPH prevalence at the prefecture level is not necessarily applicable to the individual level. Third, the number of claims of marker drugs, which was used as the variable referring to BPH prevalence, does not provide detailed information of patients such as the severity of symptoms. Fourth, stratified analysis for subdivided lifestyles (e.g., cigarettes/e-cigarettes and current smokers/ex-smokers) has not been performed. Further studies with improved study design (e.g., longitudinal study, use of individual level data, and stratified analysis) are warranted to obtain more detailed information regarding the lifestyle habits which affect BPH development.
In conclusion, this ecological study, analyzing Japanese nationwide datasets, demonstrates that lifestyle habits- and health condition-related variables are closely associated with BPH development. Healthy lifestyle habits, especially avoidance of smoking, implementation of exercise in daily life, and a small amount of alcohol consumption, are important to prevent or delay BPH development. Among metabolic syndrome-associated factors, high blood pressure and high serum ALT are suggested as risk factors of BPH development.
Conflicts of interest
The authors have no conflicts of interest to declare.
References
- 1.Roehrborn C.G. Benign prostatic hyperplasia: an overview. Rev Urol. 2005;7(Suppl 9):S3–S14. Suppl 9. [PMC free article] [PubMed] [Google Scholar]
- 2.Lee S.W.H., Chan E.M.C., Lai Y.K. The global burden of lower urinary tract symptoms suggestive of benign prostatic hyperplasia: a systematic review and meta-analysis. Sci Rep. 2017;7(1):7984. doi: 10.1038/s41598-017-06628-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Abrams P. LUTS, BPH, BPE, BPO: a plea for the logical use of correct terms. Rev Urol. 1999;1(2):65. [PMC free article] [PubMed] [Google Scholar]
- 4.Xu X.F., Liu G.X., Guo Y.S., Zhu H.Y., He D.L., Qiao X.M., et al. Global, regional, and national incidence and year lived with disability for Benign Prostatic Hyperplasia from 1990 to 2019. Am J Men’s Health. 2021;15(4) doi: 10.1177/15579883211036786. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Homma Y., Gotoh M., Yokoyama O., Masumori N., Kawauchi A., Yamanishi T., et al. Outline of JUA clinical guidelines for benign prostatic hyperplasia. Int J Urol. 2011;18(11):741–756. doi: 10.1111/j.1442-2042.2011.02860.x. [DOI] [PubMed] [Google Scholar]
- 6.Kim D.K., Park J.J., Yang W.J., Doo S.W., Kim J.H., Song Y.S. Changes in diagnosis rate and treatment trends of benign prostatic hyperplasia in Korea: a nationwide population-based cohort study. Prostate Int. 2021;9(4):215–220. doi: 10.1016/j.prnil.2021.08.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cho J.M., Moon K.T., Lee J.H., Choi J.D., Kang J.Y., Yoo T.K. Open simple prostatectomy and robotic simple prostatectomy for large benign prostatic hyperplasia: comparison of safety and efficacy. Prostate Int. 2021;9(2):101–106. doi: 10.1016/j.prnil.2020.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ellen M.E., Panisset U., Araujo de Carvalho I., Goodwin J., Beard J. A knowledge translation framework on ageing and health. Health Pol. 2017;121(3):282–291. doi: 10.1016/j.healthpol.2016.12.009. [DOI] [PubMed] [Google Scholar]
- 9.Speakman M., Kirby R., Doyle S., Ioannou C. Burden of male lower urinary tract symptoms (LUTS) suggestive of benign prostatic hyperplasia (BPH) - focus on the UK. BJU Int. 2015;115(4):508–519. doi: 10.1111/bju.12745. [DOI] [PubMed] [Google Scholar]
- 10.Moyad M.A., Lowe F.C. Educating patients about lifestyle modifications for prostate health. Am J Med. 2008;121(8 Suppl 2):S34–S42. doi: 10.1016/j.amjmed.2008.05.025. [DOI] [PubMed] [Google Scholar]
- 11.Bientinesi R., Gandi C., Vaccarella L., Sacco E. Lifestyle in urology: Benign diseases. Urologia. 2021;88(3):163–174. doi: 10.1177/0391560321994386. [DOI] [PubMed] [Google Scholar]
- 12.Wang W., Guo Y., Zhang D., Tian Y., Zhang X. The prevalence of benign prostatic hyperplasia in mainland China: evidence from epidemiological surveys. Sci Rep. 2015;5 doi: 10.1038/srep13546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gacci M., Corona G., Vignozzi L., Salvi M., Serni S, De Nunzio C., et al. Metabolic syndrome and benign prostatic enlargement: a systematic review and meta-analysis. BJU Int. 2015;115(1):24–31. doi: 10.1111/bju.12728. [DOI] [PubMed] [Google Scholar]
- 14.Xu H., Fu S., Chen Y., Chen Q., Gu M., Wang Z. Smoking habits and benign prostatic hyperplasia: a systematic review and meta-analysis of observational studies. Medicine (Baltimore) 2016;95(32):e4565. doi: 10.1097/MD.0000000000004565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Platz E.A., Rimm E.B., Kawachi I., Colditz G.A., Stampfer M.J., Willett W.C., et al. Alcohol consumption, cigarette smoking, and risk of benign prostatic hyperplasia. Am J Epidemiol. 1999;149(2):106–115. doi: 10.1093/oxfordjournals.aje.a009775. [DOI] [PubMed] [Google Scholar]
- 16.Parsons J.K., Mougey J., Lambert L., Wilt T.J., Fink H.A., Garzotto M., et al. Lower urinary tract symptoms increase the risk of falls in older men. BJU Int. 2009;104(1):63–68. doi: 10.1111/j.1464-410X.2008.08317.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Reich M.R., Ikegami N., Shibuya K., Takemi K. 50 years of pursuing a healthy society in Japan. Lancet. 2011;378(9796):1051–1053. doi: 10.1016/S0140-6736(11)60274-2. [DOI] [PubMed] [Google Scholar]
- 18.Hirose N., Ishimaru, M., Morita K., Yasunaga H. A review of studies using the Japanese National Database of Health Insurance Claims and Specific Health Checkups. Ann Clin Epidemiol. 2020;2(1):13–26. [Google Scholar]
- 19.MHLW Japan The 6th NDB Open Data. 2021. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000177221_00010.html
- 20.Mihara H., Murayama A., Nanjo S., Ando T., Tajiri K., Fujinami H., et al. Factors correlated with drug use for constipation: perspectives from the 2016 open Japanese National Database. BMC Gastroenterol. 2020;20(1):284. doi: 10.1186/s12876-020-01425-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kosilov K.V., Kuzina I.G., Kuznetsov V., Kosilova E.K. Improvement of the symptoms of lower urinary tract and sexual dysfunction with tadalafil and solifenacin after the treatment of benign prostatic hyperplasia with dutasteride. Prostate Int. 2020;8(2):78–84. doi: 10.1016/j.prnil.2019.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Füllhase C., Schneider M.P. 5-alpha-reductase inhibitors and combination therapy. Urol Clin. 2016;43(3):325–336. doi: 10.1016/j.ucl.2016.04.003. [DOI] [PubMed] [Google Scholar]
- 23.MHLW Japan . 2020. The 5th NDB Open Data.https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000177221_00008.html [Google Scholar]
- 24.MLHW Japan Specific health checkups and specific health guidance. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000161103.html
- 25.Statistics Bureau MoIAaC Regional statistics database (system of social and demographic statistics) 2021. https://www.e-stat.go.jp/en/regional-statistics/ssdsview
- 26.Dwivedi S., Goel A., Mandhani A., Khattri S., Pant K.K. Tobacco exposure may enhance inflammation in prostate carcinoma patients: an explorative study in north Indian population. Toxicol Int. 2012;19(3):310–318. doi: 10.4103/0971-6580.103681. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Lee J., Taneja V., Vassallo R. Cigarette smoking and inflammation: cellular and molecular mechanisms. J Dent Res. 2012;91(2):142–149. doi: 10.1177/0022034511421200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Veras A.S.C., Baptista D.B., Dos Santos N.J., Thorpe H.H.A., Seraphim P.M., Florido Neto A.R., et al. Impact of cigarette smoke and aerobic physical training on histological and molecular markers of prostate health in rats. Braz J Med Biol Res. 2020;53(5) doi: 10.1590/1414-431X20209108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Parsons J.K., Im R. Alcohol consumption is associated with a decreased risk of benign prostatic hyperplasia. J Urol. 2009;182(4):1463–1468. doi: 10.1016/j.juro.2009.06.038. [DOI] [PubMed] [Google Scholar]
- 30.Gordon G.G., Altman K., Southren A.L., Rubin E., Lieber C.S. Effect of alcohol (ethanol) administration on sex-hormone metabolism in normal men. N Engl J Med. 1976;295(15):793–797. doi: 10.1056/NEJM197610072951501. [DOI] [PubMed] [Google Scholar]
- 31.Crispo A., Talamini R., Gallus S., Negri E., Gallo A., Bosetti C., et al. Alcohol and the risk of prostate cancer and benign prostatic hyperplasia. Urology. 2004;64(4):717–722. doi: 10.1016/j.urology.2004.05.002. [DOI] [PubMed] [Google Scholar]
- 32.Macke A.J., Petrosyan A. Alcohol and prostate cancer: time to draw conclusions. Biomolecules. 2022;12(3) doi: 10.3390/biom12030375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Parsons J.K., Bergstrom J., Barrett-Connor E. Lipids, lipoproteins and the risk of benign prostatic hyperplasia in community-dwelling men. BJU Int. 2008;101(3):313–318. doi: 10.1111/j.1464-410X.2007.07332.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Parsons J.K. Modifiable risk factors for benign prostatic hyperplasia and lower urinary tract symptoms: new approaches to old problems. J Urol. 2007;178(2):395–401. doi: 10.1016/j.juro.2007.03.103. [DOI] [PubMed] [Google Scholar]
- 35.Teixeira G.R., Fávaro W.J., Pinheiro P.F., Chuffa L.G.A., Amorim J.P.A., Mendes L.O., et al. Physical exercise on the rat ventral prostate: steroid hormone receptors, apoptosis and cell proliferation. Scand J Med Sci Sports. 2012;22(5):e86–e92. doi: 10.1111/j.1600-0838.2012.01501.x. [DOI] [PubMed] [Google Scholar]
- 36.Hackney A.C. Endurance exercise training and reproductive endocrine dysfunction in men: alterations in the hypothalamic-pituitary-testicular axis. Curr Pharmaceut Des. 2001;7(4):261–273. doi: 10.2174/1381612013398103. [DOI] [PubMed] [Google Scholar]
- 37.Abramson J.L., Vaccarino V. Relationship between physical activity and inflammation among apparently healthy middle-aged and older US adults. Arch Intern Med. 2002;162(11):1286–1292. doi: 10.1001/archinte.162.11.1286. [DOI] [PubMed] [Google Scholar]
- 38.Ngai H.Y., Yuen K.S., Ng C.M., Cheng C.H., Chu S.P. Metabolic syndrome and benign prostatic hyperplasia: An update. Asian J Urol. 2017;4(3):164–173. doi: 10.1016/j.ajur.2017.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Hammarsten J., Högstedt B., Holthuis N., Mellström D. Components of the metabolic syndrome-risk factors for the development of benign prostatic hyperplasia. Prostate Cancer Prostatic Dis. 1998;1(3):157–162. doi: 10.1038/sj.pcan.4500221. [DOI] [PubMed] [Google Scholar]
- 40.Chung G.E., Yim J.Y., Kim D., Kwak M.S., Yang J.I., Park B., et al. Nonalcoholic Fatty Liver Disease Is Associated with Benign Prostate Hyperplasia. J Korean Med Sci. 2020;35(22):e164. doi: 10.3346/jkms.2020.35.e164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Zhao S., Wang Y., Wu W., Yang S., Feng L., Tao F., et al. Nonalcoholic fatty liver disease and risk of prostatic diseases: Roles of insulin resistance. Andrologia. 2021;53(6) doi: 10.1111/and.14060. [DOI] [PubMed] [Google Scholar]
- 42.Dahle S.E., Chokkalingam A.P., Gao Y.T., Deng J., Stanczyk F.Z., Hsing A.W. Body size and serum levels of insulin and leptin in relation to the risk of benign prostatic hyperplasia. J Urol. 2002;168(2):599–604. [PubMed] [Google Scholar]
- 43.Meigs J.B., Mohr B., Barry M.J., Collins M.M., McKinlay J.B. Risk factors for clinical benign prostatic hyperplasia in a community-based population of healthy aging men. J Clin Epidemiol. 2001;54(9):935–944. doi: 10.1016/s0895-4356(01)00351-1. [DOI] [PubMed] [Google Scholar]
- 44.Park J.S., Koo K.C., Kim H.K., Chung B.H., Lee K.S. Impact of metabolic syndrome-related factors on the development of benign prostatic hyperplasia and lower urinary tract symptoms in Asian population. Medicine (Baltimore) 2019;98(42) doi: 10.1097/MD.0000000000017635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Berry S.J., Coffey D.S., Walsh P.C., Ewing L.L. The development of human benign prostatic hyperplasia with age. J Urol. 1984;132(3):474–479. doi: 10.1016/s0022-5347(17)49698-4. [DOI] [PubMed] [Google Scholar]
- 46.Williams A.M., Simon I., Landis P.K., Moser C., Christens-Barry W., Carter H.B., et al. Prostatic growth rate determined from MRI data: age-related longitudinal changes. J Androl. 1999;20(4):474–480. [PubMed] [Google Scholar]
- 47.Zhang W., Zhang X., Li H., Wu F., Wang H., Zhao M., et al. Prevalence of lower urinary tract symptoms suggestive of benign prostatic hyperplasia (LUTS/BPH) in China: results from the China Health and Retirement Longitudinal Study. BMJ Open. 2019;9(6) doi: 10.1136/bmjopen-2018-022792. [DOI] [PMC free article] [PubMed] [Google Scholar]