Association between serum urea nitrogen levels and prostate-specific antigens (NHANES 2003–2010)

Abstract

Background

Increasing evidence suggests that serum urea nitrogen may be a risk factor for prostate cancer (PCa) and influence serum prostate-specific antigen (PSA) concentrations, but direct evidence of a relationship between PSA and serum urea nitrogen levels in the general population is still lacking. The aim of this study was to demonstrate the relationship between serum urea nitrogen levels and prostate-specific antigen (PSA) and prostate cancer.

Methods

We conducted a cross-sectional study using the National Health and Nutrition Examination Survey (NHANES, 2003–2010) database. We performed a multifactorial regression analysis of the association between serum urea nitrogen levels and PSA and PCa, followed by subgroup analyses.

Results

A total of 5256 subjects were included in this study, and after adjusting for demographic, biological, and immunologic variables, we found that there was a threshold for blood urea nitrogen concentration below which every 1 ng/mL increase in serum urea nitrogen was associated with a 0.0325 ng/mL increase in PSA concentration (log2 transformed) (95% CI: 0.0064, 0.0586), and the P trend was was less than 0.05 and the difference was statistically significant. Sensitivity analyses using the generalized additive model (GAM) showed a linear relationship between serum urea nitrogen and serum PSA concentrations when blood urea nitrogen concentrations ranged from 0 ng/ml to 6.78 ng/ml.

Conclusion

Serum urea nitrogen was independently and positively correlated with serum PSA concentration when the concentration of serum urea nitrogen ranged from 0 ng/ml to 6.78 ng/ml.

Introduction

Prostate cancer (PCa) is a major disease affecting men’s health worldwide. It is the second most common cancer in men. The incidence of prostate cancer has been on the rise globally in recent years, and many putative risk factors, including androgens, diet, physical activity, sexual factors, inflammation and obesity, and serum urea nitrogen, have been implicated in the etiology of PCa, although their roles are unclear [1,2,3].

Advanced prostate cancer is the leading cause of most PCa-related deaths [4, 5], and the 5-year relative survival rate for patients diagnosed with distant metastatic PCa is 28%, whereas the survival rate for patients with localized or regional PCa approaches 100% [6]. Therefore, early diagnosis and detection of limited or progressive tumors may allow patients to receive more personalized treatment, which is conducive to improving their survival and quality of life. Serum PSA testing is currently the most useful biomarker for early detection of PCa, and a prostate-specific antigen (PSA value) of ≥ 4-ng/ml is the currently recommended biopsy value, but there is still no clear threshold for accurate cancer diagnosis [7]. In addition, because PSA is secreted by normal prostate epithelial cells, PSA levels may still be elevated in many benign prostate diseases, leaving some possibility of a false-negative diagnosis [8, 9]. Therefore, in order for screening to be successful and to reduce the rate of leakage and misdiagnosis, it is important to identify the factors that influence PSA. Recent studies have shown that serum urea nitrogen is an independent marker of overall cancer survival and a prognostic marker for related diseases [10, 11] and that markers of PCa may be associated with the urea cycle, tricarboxylic acid cycle (TCA), fatty acid metabolism and glycine cleavage [12]. Blood urea nitrogen and prostate-specific antigen levels were found to be significantly higher in PCa cancer patients than in BPH [13]. Understanding the relationship between PSA and specific oncogenic mechanisms could improve future screening methods. To date, direct evidence of the relationship between PSA and serum urea nitrogen levels in the general population is still lacking. Therefore, we performed a data analysis based on the available NHANES data. We aimed to explore the relationship between serum urea nitrogen levels and PSA levels. In addition, we assessed whether an increase or change in serum urea nitrogen would affect PSA levels.

Materials and methods

Data sources

We included data from the NHANES dataset from 2003 to 2010 (4 2-year cycles). NHANES is an ongoing cross-sectional observational study that collects health-related information from a nationally representative sample of civilian, noninstitutionalized populations in the U.S. The official NHANES website (https://wwwn.cdc.gov/nchs/nhanes/Default.aspx) data are available for free download. The NHANES protocol was reviewed and approved by the National Center for Health Statistics Research Ethics Review Committee. Written informed consent was obtained from all participants. More detailed information about NHANES can be found on the official website.

Study population

We collected a total of 41,156 participants from the NHANES database who participated in the National Health and Nutrition Examination Survey for the first time between 2003 and 2010. After a series of screenings, 5256 men were finally selected for data analysis. We listed the screening procedures as follows: (1) women (n = 20785); (2) men < 40 years of age (n = 13231); (3) men with oncology (n = 951); (4) men with prostatitis, or recent prostate manipulation (i.e., rectal examination within 1 week, prostate biopsy, surgery, or cystoscopy within 1 month) (n = 677); (5) missing PSA data (n = 244); and (6) missing blood urea nitrogen data (n = 12). Ultimately, 5256 subjects were included in the study (see Fig. 1 for details of the screening process).

Fig. 1

Screening flow chart for male participants (≥ 40 years) in the US National Health and Nutrition Examination Survey (2003–2010)

Variables

In the current study, the target independent variable was blood urea nitrogen (ng/ml), and the method used to measure blood urea nitrogen concentration on LX20 was the two-color digital endpoint method. In the reaction, urea nitrogen is bound to the bromocresol violet (BCP) reagent to form a complex. The system monitors the change in absorbance at 600 nm. The change in absorbance is proportional to the concentration of urea nitrogen in the sample. The dependent variable is PSA (ng/mL). serum samples from participants collected by NHANES physicians were recorded on Beckman Access using the Hybritech PSA method to record total serum PSA concentrations (ng/mL) (https://wwwn.cdc.gov/Nchs/Nhanes/2001-2002/ L11P_2_B .htm). In our analysis, continuous variable PSA data were used as outcome variables. Covariates were selected based on previous studies that demonstrated an association between these covariates and prostate cancer/PSA [13,14,15]. Covariates included demographic, biological and immunological variables. The following variables were included in the database file: continuous variables included BMI, C-reactive protein, glycosylated hemoglobin, HDL-cholesterol, age, triglycerides, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase, albumin, total serum calcium, cholesterol, bicarbonate, serum glucose, lactate dehydrogenase and blood phosphorus and total bilirubin. Categorical variables included race, lifetime smoking of at least 100 cigarettes, history of hypertension, history of diabetes, history of coronary artery disease, history of stroke, education level, and marital status. A more detailed explanation of these variables can be found on the official NHANES website.

Data analysis

We performed statistical analyses according to the criteria of the CDC guidelines (https://wwwn.cdc.gov/nchs/nhanes/tutorials/default.aspx)). Serum urea nitrogen was normally distributed, expressed as mean ± standard deviation, and continuous variables were used for data analysis. Since PSA is skewed, we used log2 transformation and used the transformed data as independent variables for data analysis. Continuous variables were expressed as mean ± standard deviation (normal distribution) or median (quartiles) (skewed distribution), and categorical variables were expressed as frequencies or percentages. To investigate whether serum was associated with PSA levels in selected participants with urea nitrogen, our statistical analysis consisted of three main steps. First, serum urea nitrogen was divided into four groups according to quartile levels and the distribution of baseline data in the different serum urea nitrogen groups (Quartile) of patients included in this study was presented. The chi-square test (categorical variables), one-way ANOVA (normal distribution) or Kruskal-Wallis test (skewed distribution) were used to demonstrate differences between the four quartile groups. In the second step of data analysis, weighted univariate and multivariate linear regression models were used. Four statistical models were constructed: model I, without adjusting for covariates; model II, adjusted for socio-demographic data only; model III, model II + other covariates shown in the table or Kruskal-Wallis test (skewed distribution) was used to demonstrate differences between the four quartiles. The third step of data analysis was to perform GAM model and smoothed curve fitting (penalized spline method) to explore the non-linear association between serum urea nitrogen and PSA levels. If the GBM model detected nonlinearity, we first calculated the inflection points using a recursive algorithm and then constructed weighted two-stage linear regression models on both sides of the inflection points. To ensure the robustness of the data analysis, we performed the following sensitivity analyses: (1) we converted serum urea nitrogen to a categorical variable by quartiles and calculated the trend P. The purpose was to validate the results of serum urea nitrogen as a categorical variable and to observe the possibility of nonlinearity; (2) we used a weighted GAM model to adjust for continuous variables in Model III. All analyses were performed using the statistical software R (http://www.r-project.org, R Foundation) and EmpowerStats (http:/www.empower-stats.com, X&Y Solutions, Inc., Boston, MA). A p value less than 0.05 (both sides) was considered statistically significant.

Results

Baseline characteristics of participants: Table 1 shows the weighted distributions of demographic characteristics and other covariates selected from the NHANES database for selective participants from 2003 to 2010. The distributions of triglycerides, total calcium, and direct HDL cholesterol were not statistically significantly different in the different serum urea nitrogen groups (quartiles, Q1-Q4) (all P values > 0.05). Compared to the Q4 group, participants in Q1-Q3 had younger age, serum glucose, lactate dehydrogenase, and glycated hemoglobin, and higher alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, and cholesterol, and participants in Q1-Q3 had higher rates of hypertension, diabetes, coronary heart disease, and stroke; in contrast, participants in Q4 had higher serum glucose as well as total In contrast, Q4 participants had higher serum glucose and total bilirubin, and Q4 participants were more educated and the majority of participants were non-Hispanic white.

Table 1 Baseline characteristics of selected participants

Serum urea nitrogen concentration and prostate-specific antigen (PSA) levels: univariate and multiple linear regression model results are presented in Table 2 for the magnitude of the correlation between serum urea nitrogen values and PSA levels. Model 1 is the unadjusted model. Model 1 showed that for each 1 mol/L increase in serum urea nitrogen, PSA levels increased by 0.0859 (0.0693, 0.1025) (log2 transformed), p < 0.05. In model 2, after adjusting for sociodemographic variables (race/ethnicity, age, marital status, education level), PSA levels increased by 0.0207 (0.0207, 0.1025) for each 1 mol/L increase in serum urea nitrogen. 0.0207 (0.0040, 0.0373) (log2 transformed), P < 0.05. In the fully adjusted model, adjusting for age; race; education; marriage; alanine aminotransferase; aspartate aminotransferase; alkaline phosphatase; serum albumin; total calcium; cholesterol; bicarbonate; serum glucose; lactate dehydrogenase; blood phosphorus; total bilirubin; triglycerides; C Reactive protein; glycated hemoglobin; BMI; direct HDL cholesterol; history of hypertension; history of coronary heart disease; history of diabetes mellitus; history of stroke; after smoking more than 100 cigarettes in a lifetime, each 1ng/ml increase in serum urea nitrogen was associated with a 0.0325ng/ mL (log2 transformed) increase in PSA concentration (95% CI: 0.0064, 0.0586), P < 0.05,a statistically significant difference. To resolve the linearity we also used GAM to adjust for continuous variables in the covariates. Despite these transformations (fitting continuous variables to smoothing), the results did not change significantly from the fully adjusted model direction.

Table 2 Univariate and multivariate analysis by weighted linear regression model and GAM model

To confirm the stability of the results, a series of sensitivity analyses can be performed. First, total dietary sugar intake was transformed from a continuous variable to a categorical variable in the quartiles and a p-value was calculated, which was found to be consistent with the results for serum urea nitrogen as a continuous variable. In addition, GAM sensitivity analysis was used because the generalized linear model could not resolve nonlinearities. The fully adjusted model was observed to be consistent with the GAM model. The possibility of a nonlinear relationship between serum urea nitrogen levels and PSA concentrations was investigated using a smoothed curve-fitting model. After adjusting for other covariates (using the adjustment strategy of the fully adjusted model), we observed that the relationship between blood urea nitrogen and PSA concentration was linear (Fig. 2). In addition to this, a linear regression model and a two-segment linear regression model were compared. This result suggests that a two-segment linear regression model should be used to fit the model. With the two-segment linear regression model and the recursive algorithm, we calculated the inflection point of 6.78 mol/L (Table 3). On the left side of the inflection point, a positive correlation between serum urea nitrogen and PSA was observed with an effect size, 95% CI and P value of 0.0325 (0.0064, 0.0586), P < 0.05, a statistically significant difference. The threshold value for serum urea nitrogen level was 6.78 mol/L. These findings suggest a threshold effect between serum albumin and PSA levels. On the right side of the inflection point, the effect size, 95% CI and P value were 0.0119 (-0.0202, 0.0439) P = 0.4692, respectively.

Fig. 2

The relationship between blood urea nitrogen and prostate-specific antigen (PSA) concentrations

Table 3 Result of nonlinearity addressing by weighted two-piecewise linear model

Discussion

Prostate cancer is the most common cancer in American men and the second leading cause of cancer-related deaths [16]. Therefore, early screening for PCa can help in early detection and early treatment to reduce mortality. Current screening of the PCa population is mainly based on PSA [17], and clarifying the factors affecting PSA is beneficial to improve the quality of screening. We used a nationally representative sample of US adult men to define the relationship between serum urea nitrogen and PSA. In addition, serum urea nitrogen has been reported in the literature to be associated with prostate cancer [10]. We hypothesized that serum urea nitrogen would also affect PSA levels. To test our hypothesis, the NHANES database was used to explore the relationship between serum urea nitrogen and PSA in US adults without a history of prostate tumors. By analyzing 5256 NHANES participants, we found that each 1ng/ml increase in serum urea nitrogen was associated with a 0.0325ng/ mL (log2 transformed) increase in PSA concentration (95% CI: 0.0064, 0.0586) with a P trend of less than 0.05. This result was confirmed by sensitivity analysis and is plausible.

Currently, an association between serum urea nitrogen and PCa has been documented. In the study of Tsutomu Nishiyama et al. on the effect of androgen deprivation therapy (ADT) on the metabolism of prostate cancer patients, it was found that serum urea nitrogen was measured before and 6 months after ADT treatment, and a significant increase in serum urea nitrogen (P = 0.03) was found, so it can be concluded that there is an essential association and role in androgen deprivation for prostate cancer [18]. In addition to this, in the results of Ugur Uyeturk et al. who studied serum reticulin levels in prostate cancer patients, it was found that there was a significant difference in serum blood urea nitrogen in the prostate cancer group and benign prostatic hyperplasia, and the results showed that blood urea nitrogen (p < 0.001) levels were significantly higher in the prostate cancer group than in the prostatic hyperplasia group, and from the results of this article, it can be speculated whether blood urea nitrogen plays a role in the development of prostate cancer plays an important role [19]. Some articles have reported that serum urea nitrogen (BUN) can be used as a prognostic indicator for cardiopulmonary vascular diseases, such as Jihong Fang et al. who found that the ratio of urea nitrogen to serum albumin may be a simple and useful prognostic tool for predicting mortality in critically ill patients with acute pulmonary embolism [11, 20], and BUN levels have been reported to be positively associated with the risk of type 2 diabetes mellitus in Chinese adults [21]. event risk in Chinese adults has been reported to be positively correlated [21]. Some scholars have even found that blood urea nitrogen can be used as a prognostic indicator for tumors, such as Kaiming Zhang and others who found that blood urea nitrogen as a biomarker in systemic oxidative stress can independently predict the prognosis of patients with surgical breast cancer in a prognostic study of breast cancer patients [22]. As for prostate cancer, it is worthwhile to further investigate and explore whether blood urea nitrogen can be used as its biomarker to provide new therapeutic tools for the treatment of prostate cancer.

There are regional differences in the incidence of prostate cancer, which may be due to differences in dietary habits. Nutrients, including fat, protein, carbohydrates, vitamins (vitamins A, D and E) and polyphenols, may influence the pathogenesis and progression of PCa by mechanisms including: inflammation, antioxidant effects and the action of sex hormones. Urea nitrogen is the main end product of protein metabolism in the body. Excessive protein catabolism or intake can lead to an increase in blood urea nitrogen in the body, and cancer is a wasting disease that increases protein consumption in the body, which increases blood urea nitrogen levels [ [23]. An animal study using a mouse model of PCa reported that high milk consumption showed a slight protective effect against PCa progression by decreasing the expression of Ki-67 and G protein-coupled receptor family C group 6 member A [24]. Reactive oxygen species (ROS) and reactive nitrogen species (RNS) are not only by-products of normal cellular metabolism but also play an important role in cellular signaling. However, when ROS and RNS levels are elevated, cells are exposed to oxidative stress, which activates various mechanisms that allow them to cope with these changes. Studies have shown that oxidative stress conditions play an important role in the development and progression of prostate cancer by regulating molecules such as DNA, enhancers, transcription factors and cell cycle regulators, and one study reported that total bilirubin (TBIL), lactate dehydrogenase (LDH), creatinine (CRE) and blood urea nitrogen (BUN) were significantly elevated in a mouse model of systemic oxidative stress [25]. After analysis, blood urea nitrogen is used as a systemic oxidative stress biomarker to assess the development and prognosis of prostate cancer.

The present study demonstrates several advantages. First, the highlight of this study is the large sample size. The study included a large number of 5256 participants, which provides a high statistical power to quantitatively assess the association between serum urea nitrogen and PSA levels. Secondly, we not only had to deal with different types of missing data, but also to consider the impact of missing data on the results, and finally, sensitivity analyses were performed on the missing data and the effect sizes were evaluated, and a generalized additive model (GAM) was used to verify that the linear relationship was accurate.

The current research work presents several limitations that must be considered when it comes to the results. First, the design of our study is cross-sectional in nature. Due to its inherent limitations, we were unable to derive a causal relationship between serum urea nitrogen and PSA, and it was difficult to distinguish between them causally. Second, the study population was limited to U.S. adults only, and thus the generalizability was geographically limited. Third, this study was based on a secondary analysis of published data and therefore could not adjust for variables not included in the dataset, such as dihydrotestosterone concentration. Fourth, we excluded users with male prostatitis, or recent prostate manipulation (i.e., rectal examination within 1 week, prostate biopsy, surgery, or cystoscopy within 1 month) and participants with malignancy because these special populations have a significant effect on PSA concentrations. Therefore, the results included in this study are not applicable to the aforementioned populations.

Conclusion

Adjusting for demographic, biological, and immunological variables, in adult American men without a history of malignancy, we found an independent and positive correlation between serum urea nitrogen and serum PSA concentrations below a threshold. It is unclear whether blood urea nitrogen is further involved in the development and progression of PCa. This is a difficult question and needs to be evaluated by further research studies.

Data Availability

This study analyzed publicly available datasets. The data can be found here: https://www.cdc.gov/nchs/nhanes/index.htm.