The Influence of Dietary Habits on Male Lower Urinary Tract Symptoms: Differential Influence of Water Intake on Voiding Symptoms and Storage Symptoms

Abstract

Purpose

To investigate the dietary factors affecting male lower urinary tract symptoms (LUTS).

Materials and Methods

This retrospective study analyzed men who underwent health check-ups. The men who completed the International Prostate Symptom Score (IPSS) and a dietary questionnaire with 19 items were included in the study. Men with a history of medication for LUTS were excluded from the study. The influences of each dietary habit on total IPSS, voiding symptoms, storage symptoms and quality of life were evaluated separately. Dietary risk group was defined by the number of significant dietary risk factors.

Results

The mean age of 28,463 men was 52.2±8.7 years. After adjusting for age and other conditions, 13 dietary habits (for examples; overeating, preference for salty food, frequently eating sweets, and infrequently eating vegetables, etc.) were the significant risk factors related to worse total IPSS scores. For voiding symptoms, storage symptoms, and quality of life scores, 14, 12, and 12 dietary habits were identified as independent risk factors, respectively. Drinking four or more cups of water per day was related to worse storage symptoms (odds ratio [OR]=1.12, 95% confidence interval [95% CI]=1.09–1.22), but better voiding symptoms (OR=0.90, 95% CI=0.86–0.95). The newly developed dietary risk group showed that total and subtotal IPSS scores increased by the number of bad dietary habits in all age groups, respectively.

Conclusions

This study showed that dietary habits had a significant impact on LUTS. The amount of water consumed had a differential influence on each subdomain symptom.

INTRODUCTION

Lower urinary tract symptoms (LUTS) related to benign prostate hyperplasia (BPH) are highly prevalent in the general population [1,2]. The results of the Korean Epidemiology of Prostate Conditions study showed that the prevalence of LUTS in Korean men was 53.7% and increased with age [1]. LUTS related to BPH are affected by dietary habits [3,4,5,6].

Previous studies investigated the influence of dietary habits on LUTS, however, the dietary risk factors of LUTS were different from study to study [3,4,5,6]. It might be explained by the lack of standardized methods for assessing dietary habits and the different methods for evaluating LUTS across studies.

Recent studies have shown a strong association between metabolic syndrome and male LUTS, and diet and lifestyle have been recognized as important factors in this association [7]. Therefore, when assessing the impact of diet on LUTS, it is reasonable to use a dietary questionnaire that can reflect metabolic syndrome.

The International Prostate Symptom Score (IPSS) questionnaire is the most widely used tool to evaluate LUTS related to BPH [8,9]. IPSS items are further divided into items on voiding symptoms and items on storage symptoms [9,10]. However, most previous studies have not used the IPSS to assess LUTS. Furthermore, few studies have evaluated the difference between the impact of dietary habits on voiding and storage symptoms.

The influence of dietary habits on voiding symptoms and storage symptoms might be different [6]. For example, considering the amount of drinking water consumed, the urine stream improves when the voided volume is large, but a high 24-hours urine volume increases the voiding frequency and could worsen storage symptoms [11,12].

The aim of this study was to evaluate the dietary risk factors that affect LUTS and identify whether there are differences in dietary risk factors between voiding symptoms and storage symptoms, using dietary questionnaire reflecting metabolic syndrome and standardized IPSS questionnaire.

MATERIALS AND METHODS

1. Ethics statement

This study was approved by the Institutional Review Board of Seoul National University Hospital (H-1811-153-989). IRB or ethics committee waived the requirement for informed consent.

2. Subjects

The study included 36,347 men who underwent routine check-ups at Seoul National University Hospital Gangnam Healthcare Center between August 2011 and October 2018. The data were collected at March 2019. The initial screening data were used if the patients underwent a health check-up more than twice. Men who submitted completed IPSS questionnaires for eight items and the dietary questionnaire for 19 items were included. The exclusion criteria were as follows: missing data for any item of the IPSS questionnaire; missing data for any item of the dietary questionnaire; previous or current medication for LUTS (alpha-adrenoceptor inhibitors, five alpha-reductase inhibitors, and antimuscarinic agents); history of urologic disease including prostatitis, neurogenic bladder, and urethral strictures; history of any surgical treatment for the prostate or kidney; prostate cancer or other malignancy; history of renal disease or dialysis; history of cerebrovascular disease; history of heart failure or angina; history of liver cirrhosis neurologic disease; serum creatinine level of more than 2.0 mg/dL; or serum prostate-specific antigen (PSA) level of more than 10.0 ng/mL. After exclusion, a total of 28,463 men were included in the final analysis (Fig. 1). Response rate of each variables were presented in Supplement Table 1.

Fig. 1. Flowchart of study population enrollment. IPSS: International Prostate Symptom Score, LUTS: lower urinary tract symptoms, PSA: prostate-specific antigen, sCr: serum creatinine.

3. Data collection

The demographic details including age, height, weight, waist circumference, history of medical diseases, medication, smoking, drinking, educational level, and economic status were obtained. The health check-up evaluations included blood tests for PSA, serum creatinine, and lipid profiles. IPSS questionnaire and dietary questionnaire data were also collected. All data were fully anonymized before assessment.

4. Dietary questionnaires

The dietary questionnaire consists of 19 items (Supplement Fig. 1). The 19 items were selected from a dietary questionnaire in the guidebook of the Seoul Metabolic Syndrome Management project and a questionnaire used in the Dietary Guidelines for Better Nutrition by the Korea Health Industry Development Institute [13,14]. The developed questionnaire consists of three domains including regular diet, balanced diet, and healthy diet. All questions are multiple-choice with three or four options for the response.

5. Grouping of dietary status

The scoring system of dietary habits was developed to show the influence of dietary habits on LUTS. First, significant dietary risk factors for each IPSS domain, respectively, with a p-value of ≤0.001 after adjusting age were selected. Among the selected dietary risk factors, the number of habits that had adverse effects on the total IPSS score was counted individually in the responses on each patient’s questionnaire. Based on the individually counted number, 4 groups with 2 points interval were created. The interval between groups was decided by the most discriminative value for each IPSS domain scores.

6. Statistical analysis

All statistical analyses were performed using PASW statistics version 17.0 (SPSS Inc.). The data are presented as the mean±standard deviation. All hypotheses were evaluated in a two-sided manner, and a p-value of <0.05 was considered statistically significant. To evaluate the influence of parameters on LUTS, linear regression analysis with backward elimination was performed. To discriminate the influence of dietary habits on total and subdomain scores, the analysis was performed with each dependent variable of the total IPSS score, voiding score sum, storage score sum, and quality of life score, respectively. The significant dietary factors in univariate analysis, demographic details, medical disease, smoking, drinking, and social history, were included in the multivariate analysis.

RESULTS

The baseline characteristics of the study population are presented in Table 1. The mean age of the patients was 52.2±8.7 years. The mean total IPSS score and quality of life (Qol) score were 6.7±5.9 points and 2.1±1.4 points, respectively.

Table 1. Baseline characteristics.

Variable		Baseline
Age (y)		52.2±8.7
Body mass index (kg/m2)		24.5±2.7
Waist circumference (cm)		88.0±7.4
Medical disease
	Hypertension	6,755 (23.7)
	Diabetes	2,326 (8.2)
	Hyperlipidemia	5,344 (18.8)
	Asthma	571 (2.0)
	Allergic rhinitis	3,917 (13.8)
Presence of metabolic syndrome		10,218 (35.9)
Alcohol consumption amount
	<30 g/day	20,086 (70.6)
	≥30 g/day	6,643 (23.3)
Smoking
	None	6,868 (24.1)
	Ex-smoker	12,709 (44.7)
	Current smoker	8,071 (28.4)
PSA (ng/mL)		1.1±0.9
Serum creatinine (mg/dL)		0.9±0.1
Total cholesterol (mg/dL)		194.6±34.2
LDL (mg/dL)		124.8±30.4
HDL (mg/dL)		49.5±10.4
TG (mg/dL)		132.5±86.9
Glucose (mg/dL)		102.8±20
HbA1c (%)		5.7±0.7
Prostate volume by TRUSa
	Total volume (mL)	33.1±10.4
	Transitional zone volume (mL)	11.1±6.0
IPSS score
	IPSS item 1	1.0±1.3
	IPSS item 2	1.2±1.2
	IPSS item 3	0.8±1.1
	IPSS item 4	0.9±1.2
	IPSS item 5	1.4±1.5
	IPSS item 6	0.7±1.1
	IPSS item 7	0.8±0.8
	QoL	2.1±1.4
	Voiding symptom sum	3.9±3.8
	Storage symptom sum	2.8±2.4
	Total IPSS scores	6.7±5.9

Values are presented as mean±standard deviation or number (%).

TRUS: transrectal ultrasonography, PSA: prostate-specific antigen, LDL: low-density lipoprotein, HDL: high-density lipoprotein, TG: triglycerides, HbA1c: hemoglobin A1c, IPSS: International Prostate Symptom Score, Qol: quality of life.

^aTRUS were not evaluated in all patients. The number of patients who performed TRUS was 11,572.

1. Dietary factors that influenced total IPSS scores

Among the 19 items on the dietary risk factor questionnaire, 13 items were identified as having a significant impact on total IPSS scores in multivariate analysis (Table 2). In the analysis, an increase in age was significantly correlated with an increase in total IPSS scores. After adjusting for age and other conditions, irregular meals, overeating, frequently eating out, eating between-meal snacks, eating grains at every meal, frequently eating fried food, frequently eating fat-rich meat, frequently eating instant and processed food, the habit of adding salt to food, preferring salty food, frequently eating sweets, and frequently drinking beverages were significant risk factors that were related to worse total IPSS scores. In contrast, frequently eating vegetables was significantly related to better total IPSS scores.

Table 2. Linear regression analysis of dietary factors influencing total IPSS scores.

Variable	Multivariate
	B	95% CI	p-value
Age (y)	0.189	0.179–0.198	<0.001
Body mass index (kg/m2)	-0.196	-0.248 to -0.145	<0.001
Waist circumference (cm)	0.041	0.022–0.061	<0.001
Hypertension	0.229	0.053–0.405	0.011
Diabetes	0.635	0.374–0.896	<0.001
Hyperlipidemia	0.425	0.223–0.627	<0.001
Asthma	1.340	0.837–1.842	<0.001
Allergic rhinitis	0.702	0.505–0.898	<0.001
Smoking (current or previous)	0.494	0.330–0.657	<0.001
Alcohol consumption ≥ 30 g/day	0.440	0.274–0.606	<0.001
Prostate specific antigen	0.357	0.276–0.437	<0.001
Presence of metabolic syndrome	-0.362	-0.547 to -0.176	<0.001
Irregular meal	0.261	0.071–0.451	0.007
Overeating	0.874	0.640–1.107	<0.001
Dinning out ≥3 times per week	0.289	0.145–0.434	<0.001
Eating between meals every day	0.382	0.186–0.578	<0.001
Eating >1 bowl of cooked rice per meal	0.075	-0.110–0.261	0.426
Eating cereal at every meal	0.186	0.039–0.334	0.013
Eating high-protein food every day	-0.103	-0.274–0.068	0.239
Eating vegetables every day	-0.472	-0.632 to -0.312	<0.001
Eating ≥1 cup of dairy products per day	-0.125	-0.290–0.040	0.137
Eating fruit ≥3 times per week	-0.042	-0.214–0.131	0.635
Drinking ≥4 cups of water per day	0.161	0.004–0.317	0.044
Eating fried food ≥3 times per week	0.160	-0.016–0.336	0.075
Eating fatty food ≥3 times per week	0.310	0.151–0.470	<0.001
Eating instant food ≥3 times per week	0.322	0.132–0.511	0.001
Adding salt to food frequently	0.558	0.411–0.705	<0.001
Preference for salty food	0.532	0.318–0.746	<0.001
Eating sugary food frequently	0.616	0.374–0.858	<0.001
Drinking beverages ≥1 time per day	0.325	0.128–0.521	0.001
Drinking ≥3 cups of coffee per day	0.040	-0.104–0.184	0.583

IPSS: International Prostate Symptom Score, 95% CI: 95% confidence interval.

Included variables are age, height, weight, body mass index, waist circumference, 17 medical conditions, alcohol consumption, smoking, educational level, income, prostate specific antigen, serum creatinine, lipid profiles, glucose level, HbA1c, presence of metabolic syndrome and 19 dietary habits. All dietary variables are presented. Other presented variables are only significant factors after backward elimination.

2. Dietary factors influenced the voiding symptom score sum on the IPSS

For the voiding symptom score sum, 14 dietary items were significant factors in multivariate analysis (Table 3). Compared to risk factors for total IPSS, frequently eating dairy products and drinking four or more cups of water per day were significant factors related to lower voiding symptoms scores. Frequently eating fried foods was not a significant factor in the voiding symptom analysis. Men who drink 4 cups or more of water have lower risk of worse voiding symptom (voiding symptom sum ≥4 points) than the others (age adjusted odds ratio [OR]=0.90, 95% confidence interval [95% CI]=0.86–0.95).

Table 3. Linear regression analysis of the dietary factors influencing the IPSS voiding score sum.

Variable	Multivariate
	B	95% CI	p-value
Age (y)	0.114	0.108–0.120	<0.001
Body mass index (kg/m2)	-0.118	-0.152 to -0.085	<0.001
Waist circumference (cm)	0.024	0.011–0.037	<0.001
Diabetes	0.369	0.198–0.539	<0.001
Hyperlipidemia	0.281	0.149–0.413	<0.001
Asthma	0.836	0.507–1.165	<0.001
Allergic rhinitis	0.436	0.307–0.564	<0.001
Smoking (current or previous)	0.394	0.287–0.500	<0.001
Alcohol consumption ≥30 g/day	0.308	0.200–0.417	<0.001
Prostate specific antigen	0.238	0.185–0.290	<0.001
Presence of metabolic syndrome	-0.193	-0.312 to -0.074	0.001
Irregular Meal	0.259	0.134–0.383	<0.001
Overeating	0.564	0.411–0.716	<0.001
Dinning out ≥3 times per week	0.163	0.069–0.258	0.001
Eating between meals every day	0.257	0.129–0.386	<0.001
Eating >1 bowl of cooked rice per meal	0.010	-0.111–0.132	0.867
Eating cereal at every meal	0.144	0.047–0.24	0.003
Eating high-protein food every day	-0.062	-0.174–0.050	0.278
Eating vegetables every day	-0.304	-0.409 to -0.199	<0.001
Eating ≥1 cup of dairy products per day	-0.134	-0.242 to -0.026	0.015
Eating fruit ≥3 times per week	-0.049	-0.162–0.063	0.390
Drinking ≥4 cups of water per day	-0.155	-0.258 to -0.053	0.003
Eating fried food ≥3 times per week	0.045	-0.070–0.161	0.439
Eating fatty food ≥3 times per week	0.203	0.098–0.307	0.000
Eating instant food ≥3 times per week	0.220	0.096–0.343	0.001
Adding salt to food frequently	0.303	0.207–0.399	<0.001
Preference for salty food	0.384	0.244–0.524	<0.001
Eating sugary food frequently	0.394	0.236–0.553	<0.001
Drinking beverages ≥1 time per day	0.175	0.046–0.304	0.008
Drinking ≥3 cups of coffee per day	0.024	-0.070–0.118	0.618

IPSS: International Prostate Symptom Score, 95% CI: 95% confidence interval.

Included variables are age, height, weight, body mass index, waist circumference, 17 medical conditions, alcohol consumption, smoking, educational level, income, prostate specific antigen, serum creatinine, lipid profiles, glucose level, HbA1c, presence of metabolic syndrome and 19 dietary habits. All dietary variables are presented. Other presented variables are only significant factors after backward elimination.

3. Dietary factors that influenced the storage symptom score sum on the IPSS

Among 19 dietary risk items on the questionnaire, 12 items were identified that were significantly related to storage symptom scores (Table 4). In a comparison of dietary risk factors for the total IPSS score, eating grains every day was not a significant factor for storage symptoms. Drinking water (≥4 cups per day) was a significant risk factor that was related to higher storage symptom scores, in contrast to the voiding symptom analysis results. Risk for storage symptom sum ≥4 was higher in men who drink 4 cups or more of water than the others (age adjusted OR=1.12, 95% CI=1.09–1.22).

Table 4. Linear regression analysis of the dietary factors influencing the IPSS storage score sum.

Variable		Multivariate
	B	95% CI	p-value
Age (y)	0.076	0.072–0.080	<0.001
Body mass index (kg/m2)	-0.076	-0.098 to -0.055	<0.001
Waist circumference (cm)	0.016	0.008–0.024	<0.001
Hypertension	0.158	0.085–0.232	<0.001
Diabetes	0.290	0.181–0.398	<0.001
Asthma	0.508	0.299–0.718	<0.001
Allergic rhinitis	0.271	0.189–0.353	<0.001
Smoking (current or previous)	0.100	0.032–0.168	0.004
Alcohol consumption ≥30 g/day	0.133	0.064–0.203	<0.001
Prostate specific antigen	0.118	0.085–0.152	<0.001
Presence of metabolic syndrome	-0.091	-0.16 to -0.022	0.010
Irregular meal	0.002	-0.077–0.081	0.963
Overeating	0.312	0.214–0.409	<0.001
Dinning out ≥3 times per week	0.127	0.067–0.188	<0.001
Eating between meals every day	0.124	0.042–0.206	0.003
Eating >1 bowl of cooked rice per meal	0.064	-0.014–0.141	0.108
Eating cereal at every meal	0.040	-0.021–0.102	0.197
Eating high-protein food every day	-0.039	-0.111–0.032	0.279
Eating vegetables every day	-0.169	-0.236 to -0.103	<0.001
Eating ≥1 cup of dairy products per day	0.009	-0.060–0.078	0.802
Eating fruit ≥3 times per week	0.007	-0.065–0.079	0.855
Drinking ≥4 cups of water per day	0.315	0.250–0.380	<0.001
Eating fried food ≥3 times per week	0.118	0.045–0.192	0.002
Eating fatty food ≥3 times per week	0.106	0.039–0.172	0.002
Eating instant food ≥3 times per week	0.102	0.023–0.181	0.011
Adding salt to food frequently	0.254	0.193–0.316	<0.001
Preference for salty food	0.151	0.061–0.240	0.001
Eating sugary food frequently	0.222	0.121–0.323	<0.001
Drinking beverages ≥1 time per day	0.146	0.064–0.228	<0.001
Drinking ≥3 cups of coffee per day	0.015	-0.045–0.075	0.629

IPSS: International Prostate Symptom Score, 95% CI: 95% confidence interval.

Included variables are age, height, weight, body mass index, waist circumference, 17 medical conditions, alcohol consumption, smoking, educational level, income, prostate specific antigen, serum creatinine, lipid profiles, glucose level, HbA1c, presence of metabolic syndrome and 19 dietary habits. All dietary variables are presented. Other presented variables are only significant factors after backward elimination.

4. Dietary factors that influenced the storage symptom score sum for quality of life

In multivariate analysis for Qol scores, 12 dietary items were significant factors (Supplement Table 2). Frequently eating fat-rich meat was a dietary habit negatively affecting Qol scores, and frequently eating fruits was a dietary habit positively affecting Qol scores. Drinking water or beverages was not a significant factor.

5. Influence of metabolic syndrome on LUTS

The presence of metabolic syndrome increased total IPSS (B=0.369, 95% CI=0.226–0.513, p<0.001), voiding symptom (B=0.210, 95% CI=0.117–0.303, p<0.001), storage symptom (B=0.160, 95% CI=0.100–0.219, p<0.001) and quality of life scores (B=0.105, 95% CI=0.071–0.140, p<0.001) on univariate regression analysis. After adjusting age the presence of metabolic syndrome was correlated with all IPSS domain scores. However, on multivariate analysis adjusting all risk factors including dietary habits, metabolic syndrome was negatively correlated to each IPSS domain scores (Table 2, 3, 4 and Supplement Table 2).

6. Development of a dietary risk group for each IPSS domain

To develop risk group for total IPSS scores, 10 significant dietary risk factor with p-value <0.001 were selected, and the number of dietary habits that had adverse effects on the total IPSS score was counted. The subjects were divided into four risk groups according to the sum number of bad dietary habits; group 1 (0 or 1), group 2 (2 or 3), group 3 (4 or 5), group 4 (6 or more risk factors). Risk group for voiding symptom scores, storage symptom scores, and Qol scores were also developed by 11, 9, and 9 selected significant risk factors, respectively, with same method of total IPSS risk group development. The mean score of each IPSS domain by age group according to risk group is presented in Fig. 2 (detailed value, number and p-value were presented in Supplement Table 3). The newly developed dietary risk group showed that total and subtotal IPSS scores increased by the number of bad dietary habits in all age groups, respectively.

Fig. 2. Scores for each IPSS domain by risk group and age group. (A) Total IPSS score, (B) voiding symptom score sum, (C) storage symptom score sum, and (D) Qol score. Group 1 (0 or 1 risk factors), Group 2 (2 or 3 risk factors), Group 3 (4 or 5 risk factors), Group 4 (6 or more risk factors). Detailed value, number and p-value were presented in Supplement Table 3. IPSS: International Prostate Symptom Score, Qol: quality of life.

7. Differential Impact of the Amount of Water Intake between voiding symptom and storage symptom

Fig. 3 showed the difference of mean IPSS total, voiding symptom and storage symptom scores between men who drinks water 4 or more cups and the others. Among men under 55 years of age, those who drink less than 4 cups of water per day had significantly higher voiding symptom scores. Among men under 65 years of age, those who drink 4 or more cups of water had higher storage symptom scores. Total IPSS did not show significant differences across all age groups.

Fig. 3. Scores for each IPSS domain by drinking water. IPSS: International Prostate Symptom Score. ^*p<0.05.

DISCUSSION

This study showed that dietary habits influenced LUTS. The total IPSS score, as well as voiding symptom, storage symptoms, and quality of life scores, were influenced by dietary habits. Furthermore, the dietary habits for good and bad voiding symptoms and storage symptoms were different. Drinking enough water was related to a lower voiding symptom score sum, whereas it was related to a higher storage symptom score sum. This study also showed a cumulative effect of increasing numbers of bad dietary habits on LUTS. As the number of bad dietary habits increased, each IPSS domain score increased in all age groups. A comparison of the results of this study and previous representative studies on dietary habits affecting LUTS is presented in Table 5.

Table 5. Summary of results from previous studies and this study, evaluating the association of diet with LUTS/BPH.

Author	Year	Subject	Evaluation of diet	Evaluation of LUTS	Risk factor	Protective factor	Neutral	Nation, remark
Erickson et al [3]	2012	1,385 men ≥40 y	Healthy Eating Index	Developed questionnaire	Overall poor diet with little dietary variety	Diary, protein and alcohol intake	Cholesterol, fat, fruit, grain, Saturated fat, sodium, vegetable	US, NHANES 2000–2001
Bravi et al [4]	2006	1,369 patients with BPH and 1451 controls, <75 y	Food frequency questionnaire	Surgery for BPH	Starch	Polyunsaturated fatty acids, linoleic acid and linolenic acid	Protein, sugar, fat	Italia
Suzuki et al [16]	2002	3,523 men with BPH, 40–75 y	Food frequency questionnaire	BPH surgery and IPSS	Total energy intake, total protein intake, eicosapentaenoic, docosahexaenoic, and arachidonic acids		Total fat intake, carbohydrate	US
Chen et al [22]	2016	4,208 men, ≥50 y	Developed questionnaire	IPSS	Diary	Meat, fish and egg	Vegetable	China
Kristal et al [23]	2008	18,880 men without BPH, ≥50 y	Developed food frequency questionnaire	IPSS, treatment for BPH	Total fat, polyunsaturated fat and red meat	Protein, alcoholic beverages, vegetables, lycopene, zinc, and supplemental vitamin D	Antioxidants, Total energy, Saturated fat, Carbohydrates, vitamin C, Carotenoids and Calcium	US, PCPT trial, 7 years longitudinal study
This study	2024	28,463 men without BPH treatment, ≥40 y	Developed questionnaire	IPSS	Overeating, out-dinning, snack, fat, instant food, salt, sugar and beverage	Vegetables	Protein, diary, fruit, coffee	Korea

LUTS: lower urinary tract symptoms, BPH: benign prostate hyperplasia, IPSS: International Prostate Symptom Score, NHANES: National Health and Nutrition Examination Survey, PCPT: Prostate Cancer Prevention Trial.

The mechanism of how dietary habits affect LUTS is not clearly understood, however there are potential mechanisms that have been proposed. First, oxidative stress within the prostate has been shown to increase the development of BPH [15]. Therefore, healthier foods with higher levels of antioxidants may reduce the development of BPH. Previous observational study showed that the antioxidants obtained from vegetables, namely β-carotene, lutein, and vitamin C, protected participants from LUTS/BPH [16]. Second, dietary habit can affect your autonomic nervous system, and sympathetic hyperactivity has been linked to the development of LUTS [17]. The activation of the sympathetic nervous system by a high energy intake may cause the prostate smooth muscle to contract, resulting in a worsening of LUTS [17]. One study showed that sympathetic tone directly correlated with increasing IPSS scores [18]. Third, androgens which are necessary for prostate growth can be influenced by diet. A decrease in the bioavailability of testosterone relative to estrogen levels has been shown to influence prostate growth. And last, metabolic syndrome which is strongly related to dietary habit influences the development and progression of BPH [19]. A meta-analysis to evaluate the relationship between metabolic syndrome and LUTS showed that the components of metabolic syndrome, such as hypertriglyceridemia, elevated fasting glucose and diabetes were significantly associated with a higher risk of LUTS [19].

In this study, age was the most independent factor influencing LUTS beyond dietary habits. It is well known that the incidence of LUTS increases with age [20]. The proportion of men with each dietary risk factors decreased. However, the impact of dietary habits was significant after adjusting for age. And the dietary habits were independent significant risk factors in multivariate analysis after adjusting age and medical conditions. The statistical significance of dietary habits still remained even when participant were divided into age groups. Each IPSS domain score in men who did not have bad dietary habits showed similar results as 5- to 10-year younger men with 6 or more bad dietary habits.

A previous study showed that an increased total energy intake was associated with LUTS [16]. In a longitudinal study that followed men aged 40 to 75 years from 1986 to 1994, the total energy intake was related to BPH, and total protein intake was positively associated with BPH. However, total fat intake was not associated with the risk of BPH after adjusting for energy intake. The Boston Area Community Health study (2002–2005), a randomized population-based study also showed that a greater total energy intake was associated with higher LUTS symptom scores (p for trend <0.01) and an increased likelihood of storage symptoms [5]. In our study, overeating (dietary questionnaire item 2) was also a significantly bad habit in regard to total LUTS, voiding symptoms, storage symptoms, and Qol scores. Other significant dietary risk factors such as eating between-meal snacks every day, eating grains at every meal, and frequently eating sweets in our study might be related to increased total energy intake. The mechanisms how which high energy diets adversely affect LUTS are still unclear. Increased oxidative stress by high energy diets might be one possible cause. And high energy diets are a risk factor for metabolic syndrome that could influence LUTS [21]. In this study, metabolic syndrome was positively correlated with LUTS in univariate analysis. However, the correlation was negative after adjusting dietary habits. It suggests that diet habit has a significant effect on metabolic syndrome itself.

Previous studies reported the influence of high-protein or high-fat diet on LUTS with inconsistent results. Chen et al. also investigated the relationship between dietary habits and LUTS and reported that more frequent meat, fish, and egg intake (p<0.05) decreased the total IPSS scores, and frequent milk and dairy product intake increased the total IPSS scores in Chinese men [22]. In the Prostate Cancer Prevention Trial, BPH incidence was higher in men with high total fat and polyunsaturated fat consumption, and lower in high consumers of vegetables and red meat [23]. The National Health and Nutrition Examination Survey survey 2000 to 2001 also investigated the association of single nutrients or food factors on LUTS. Poor overall diet quality, lower intake of fruits or vegetables, and a higher intake of red meats or fat was associated with LUTS [3]. In our study, eating a high-protein diet showed a weak association with Qol scores. However, it was not related to total IPSS, voiding symptom scores, and storage symptom scores. In contrast, frequently eating fat-rich meat was strongly related to all LUTS. This difference might be attributed to different methods of assessing protein intake. There are many different types of protein, and the perception of protein-rich foods varies from person to person. Therefore, it is more difficult to assess high-protein diet, compared to assess high-fat diet. However, there is still no validated standard questionnaire to assess protein intake.

Frequently eating vegetables was consistently a good factor for LUTS in previous studies [5,23]. A study to identify the modifiable risk factors for BPH and LUTS reported that vegetables (particularly carotenoids such as tomatoes and carrots), and fruits potentially decreased the risk of BPH [9]. One recent review also noted that high intake of fruits and vegetable, lycopene and zinc has protective effect for BPH [24]. Also in our study, eating vegetables every day was a strong factor that reduced all IPSS domain scores, whereas eating fruit showed a weak correlation with low Qol scores. One potential explanation is that vegetables and fruits contain high levels of antioxidants, polyphenols, vitamins, minerals, and fiber, which may play important roles in altering inflammatory pathways associated with the pathogenesis of BPH [25].

In this study, the amount of salt intake was identified as a strong factor that worsened all LUTS. It is consistent with the results of the Boston Area Community Health study, which showed that sodium intake had positive linear associations with LUTS and storage symptom scores [6]. The cause of how the high salt intake influences the severity of LUTS is not well known. High salt intake might be related to hypertension and other co-morbidities [26,27] and was related to water intake and nocturnal urine production [28,29,30]. A cross-sectional study showed that a daily salt intake of more than 9.2 g negatively affected frequency and nocturia [28]. Another study reported that an estimated high salt intake was positively correlated with the nocturnal urine volume, but not with the diurnal urine volume [29]. Moreover, nocturia and the polyuria index improved in patients who were successful in reducing their daily salt intake in a prospective study [30]. Excessive salt intake is one of the chief causative factors of hypertension [27]. Essential hypertension is related to nocturnal polyuria in elderly persons [26]. Therefore, salt intake might influence comorbidity and urine production, especially nocturnal urine production, and subsequently, influence LUTS.

One interesting finding in our study was the relationship between LUTS and the amount of water intake. Generally, adequate water intake is known as necessary for maintaining health [31]. However, excessive water intake has a negative effect on overactive bladder symptoms [12]. It is still unclear whether there is an association between fluid intake and LUTS. In one randomized controlled trial, 138 men with BPH were randomized to a group with their fluid intake increased by 1.5 L per day or a placebo group [32]. There was no difference in the IPSS total, voiding, or Qol scores between the groups at six months. However, the IPSS storage scores were worse in those who increased their fluid intake. In a non-randomized, uncontrolled study, IPSS increased after increasing fluid intake by 2 L per day for eight weeks [11]. A recent systematic review concluded that fluid intake worsened storage symptoms with no impact on voiding symptoms and most studies showed a positive association between fluid intake and frequency and/or urgency symptoms, but the evidence on whether fluid restriction reduced nocturia was inconclusive [33]. In our study, water intake had a positive effect on voiding symptoms, but a negative impact on storage symptoms. Therefore, the effects were offset in the total IPSS scores (Fig. 3). This could be one explanation for the inconsistent results in the previous studies.

The findings of this study can be used to help individualized counseling for patients with LUTS. By examining the patient's dietary habits, physician could explain which dietary risk factors are currently contributing to their LUTS and explain the need to change them. Especially, if patients who have predominant voiding symptom is drinking less water, it could be explained that drinking more water may help improve voiding symptoms. And conversely, patients with mainly storage symptoms may benefit from drinking less water. It can also be used as a reference to show that reducing overall bad eating habits is necessary to maintain prostate health. Dietary recommendations based on the findings of this study are summarized in Table 6.

Table 6. Dietary recommendations for men with lower urinary tract symptoms.

Recommendation for overall LUTS
	Avoid overeating
	Reduce out-dinning
	Avoiding Eating between meals
	Eat vegetables every day
	Reduce fatty food
	Reduce instant food
	Do not adding salt to food frequently
	Avoid salty food
	Reduce sugary food
	Avoid drinking beverages
Particular recommendation for men with mainly voiding symptom
	Eat regularly
	Reduce eating cereal
	Eating dairy products every day
	Drink enough water
Particular recommendation for men with mainly storage symptom
	Avoid fried food
	Limit excessive water intake

LUTS: lower urinary tract symptoms.

One of the main limitations of this study were the representativeness of study population and its retrospective study design. The subjects of this study were men who underwent heath check-ups, which is not representative of the general population. This study found a relationship between dietary habits and LUTS, but it could not clarify a cause-and-effect relationship. The second limitation of this study was the tool used to evaluate dietary habits. However, there are no standardized evaluation tools for assessing the effect of dietary habits on LUTS. Furthermore, the types of methods used to evaluate dietary habits differed from study to study. Third, the prostate volume was not included as a variable in this analysis. Prostate volume by transrectal prostate ultrasonography was examined in a subset of men. And, urodynamic parameters such as uroflowmetry or post-void residual urine volumes were not included in this study. However, the significant dietary factors for LUTS did not change in the subgroup analysis including prostate volume in the patients who underwent the test. Despite these limitations, our study showed the differential effects of dietary habits on voiding symptoms and storage symptoms. Furthermore, we validated increases in symptom scores with increasing numbers of bad dietary habits in all groups. This finding could be useful for producing behavioral recommendations for men with LUTS.

CONCLUSIONS

This study showed that various dietary habits had a significant impact on LUTS. Dietary habits were still independent significant risk factors after adjusting age and medical conditions. The risk factors for total IPSS, voiding symptoms, storage symptoms, and Qol scores were not identical. Especially, the amount of water consumed had opposite effects on voiding symptoms and storage symptoms. The domain scores increased in all age groups according to the number of risk factors in each IPSS domain.

Data Sharing Statement

The data required to reproduce these findings cannot be shared at this time due to legal and ethical reasons.

Supplementary Materials

Supplementary materials can be found via https://doi.org/10.5534/wjmh.230361.

Supplement Table 1

Response rate of each variables

Supplement Table 2

Linear regression analysis of the dietary factors influencing IPSS quality of life scores

Supplement Table 3

Comparison of International Prostate Symptom Score (IPSS) domain scores by each risk group, presented in Fig. 2

Supplement Fig. 1

The dietary questionnaire consists of 19 items.