Skip to main content
Download PDF

Causal association between non-steroidal anti-inflammatory drugs use and the risk of benign prostatic hyperplasia: a univariable and multivariable Mendelian randomization study

BMC Medical Genomics volume 18, Article number: 60 (2025) Cite this article

Abstract

Background

The results of earlier observational research on the relationships between the usage of non-steroidal anti-inflammatory medicines (NSAIDs) and the risk of benign prostatic hyperplasia (BPH) have been inconsistent.

Methods

To assess these associations, we performed both univariable and multivariable Mendelian randomization (MR) studies. Instrumental variables (IVs) associated with exposures at the significance level (p < 5 × 10–6) were selected from a comprehensive meta-analysis conducted by the United Kingdom Biobank (UKB). Summary data for BPH were obtained from the FinnGen consortium, which comprised 30,066 cases and 119,297 controls. Sensitivity analyses were performed to evaluate heterogeneity and pleiotropy.

Results

We found evidence by univariable MR (UVMR) that genetically predicted NSAIDs use increased the risk of BPH (odds ratio [OR] per unit increase in log odds NSAIDs use: 1.164, 95% confidence interval [CI]: 1.041–1.302, p = 0.008). After controlling for inflammation in multivariable MR (MVMR), the link persisted (OR: 1.165, 95% CI: 1.049–1.293, p = 0.004). There were no indications of potential heterogeneity and pleiotropy in UVMR and MVMR analyses.

Conclusion

The results of the MR estimates suggest that genetically predicted NSAIDs use may elevate the risk of BPH. This outcome prompts the imperative for deeper exploration into potential underlying mechanisms.

Peer Review reports

Introduction

Benign prostatic hyperplasia (BPH) manifests as a prevalent affliction among the middle-aged and elderly male populace. Meticulous investigations into prostate histopathology through post-mortem samples have unveiled an annual surge, ranging from 41 to 90%, in the cohort of men bearing a histological diagnosis of BPH. Additionally, a noteworthy 50% of men aged between 51 and 60 years evince pathological attributes concomitant with BPH [1]. Globally, the year 2019 bore witness to a staggering 94.0 million prevalent cases of BPH, in stark contrast to the 51.1 million cases recorded in 2000 [2]. Evidently, the absolute burden of BPH is escalating at a disquieting pace throughout most regions of the world. Neglecting to address this condition could have a significant impact on one’s quality of life [3].

Numerous studies have elucidated inflammation as one of the perilous factors contributing to the onset of BPH [4,5,6]. Non-steroidal anti-inflammatory drugs (NSAIDs) are primarily harnessed for their antipyretic, analgesic, anti-inflammatory, and anti-rheumatic properties [7]. The quelling of inflammatory pathways by NSAIDs holds the potential to abate and arrest the progression of BPH [8]. Given the common prevalence of both BPH and NSAIDs usage among elderly males, fathoming their interrelationship assumes paramount significance. Nevertheless, the extant corpus of literature concerning this domain remains somewhat limited, yielding findings that have engendered discordant outcomes [9, 10]. Several investigations have illuminated the potential benefits of NSAIDs usage in ameliorating BPH or lower urinary tract symptoms (LUTS) [11,12,13,14]. Conversely, some observational studies have indicated a higher likelihood of BPH development among male individuals using NSAIDs [15,16,17]. Additionally, certain inquiries have been unable to establish a significant correlation between NSAIDs use and BPH or LUTS [18,19,20]. The incongruous nature of these findings highlights the potential limitations of observational studies in establishing causal inferences, encompassing concerns such as reverse causality, unobserved confounding, and various biases [21]. Furthermore, a significant challenge in these studies involves disentangling the direct impact of the drug on the onset of BPH or LUTS from the underlying, presumably systemic inflammatory process that triggers the utilization of NSAIDs [10]. Consequently, this has compelled us to embark on pioneering methodologies, in quest of a profound comprehension of veritable causality.

The application of Mendelian randomization (MR) has garnered widespread recognition in unveiling plausible causal connections between exposures and outcomes [22]. This sophisticated approach leverages genetic variations linked to environmental exposures as instrumental variables (IVs) to meticulously examine the intricate relationship between such exposures and their respective outcomes. Genetic variants, being randomly allocated during conception, are preferred as candidate IVs owing to their inherent immunity to potential confounding effects arising from environmental factors [23]. Given the prevailing uncertainty surrounding the causal linkage between NSAIDs use and BPH, the present study adopts the univariable MR (UVMR) design, delving into the potential causal effect of NSAIDs, salicylic acid, and its derivatives, as well as anilides use on BPH, while making judicious use of comprehensive genomewide association study (GWAS) data. In order to address concerns regarding systemic inflammation acting as a potential confounding factor, the analysis strategically incorporates multivariate MR (MVMR), thereby effectively mitigating bias attributable to inflammation.

Materials and methods

Study design

The study design overview is depicted in Fig. 1. Employing a two-sample MR design, we systematically evaluated the causal association between NSAIDs usage and BPH. To ensure a compelling MR design, three fundamental assumptions must be met:(1) IVs exhibit robust associations with the exposures; (2) IVs are not associated with any potential confounding factors; (3) genetic instruments exert their sole impact on the outcome through the relevant exposures. The verification of the last two assumptions, jointly known as horizontal pleiotropy, can be accomplished through diverse statistical methodologies [24]. This study adheres to the reporting guidelines of Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian Randomization (STROBE-MR) (Additional file 1: STROBE MR checklist).

Fig. 1
figure 1

The study design overview for analyzing the causal effects of NSAIDs use on the risk of BPH using Mendelian randomization. Assumption 1: instrumental variables (IVs) exhibit robust associations with the exposures; Assumption 2: IVs are not associated with any potential confounding factors; Assumption 3: genetic instruments exert their sole impact on the outcome through the relevant exposures. SNP: single nucleotide polymorphisms, NSAIDs: non-steroidal anti-inflammatory drugs, LD: linkage disequilibrium, MR: Mendelian randomization, IVW: inverse variance weighted

Full size image

Data sources

Summary GWAS data concerning the utilization of NSAIDs were extracted from a comprehensive meta-analysis conducted by the United Kingdom Biobank (UKB) [25]. The medication categories were determined by classifying them according to the active ingredient, utilizing the Anatomical Therapeutic Chemical (ATC) Classification System. Subsequently assigned to 23 medication-taking traits by their respective active ingredients. Anti-inflammatory medications were categorized into three groups, encompassing NSAIDs (M01A), acetylsalicylic acid (N02BA), and acetaminophen (N02BE) (Additional file 3: Table S1). The UKB assessment encompassed a total of 502,616 participants who possessed medication records, and approximately 73% of them contained non-blank medication information. The average age of participants upon attending the assessment center was 56.53 years, while the mean body mass index (BMI) of the participants was 27.43. The proportion of individuals using medication demonstrated an increasing trend with age. We used Interleukin-17 (IL-17) as a blood biomarker of inflammation. The GWAS data for plasma protein IL-17 were obtained from the study by Zhao et al. (n = 14,824, of European descent) [26]. Plasma proteins were quantified using the Olink Target-96 Inflammation immunoassay panel. Normalized Protein eXpression (NPX) values represent Olink’s normalized relative units on a log2 scale [26].

In order to address sample overlap, BPH summary statistics from the FinnGen consortium’s R9 release were obtained [27]. The N40 code in the International Classification of Diseases-10th Revision (ICD-10), as well as the 600 code in ICD-8 and ICD-9, were used within the FinnGen consortium to identify BPH patients. 30,066 BPH cases and 119,297 persons categorized as controls were included in the R9 release of the FinnGen consortium data. For additional information pertaining to the utilized data sources, definitions, units, participant details, population characteristics, and adjusted covariates, please refer to Additional file 3: Table S1.

Univariable Mendelian randomization analysis

We put in place a rigorous screening process for IVs to guarantee the reliability of the results of the MR study. Initially, we changed the threshold to p-value < 5 × 10–6 due to the small number of single nucleotide polymorphisms (SNPs) achieving genome-wide significance for medication, following the approach employed by Rosoff et al. [28]. In order to reduce linkage disequilibrium (LD), we simultaneously defined a threshold of r2 = 0.001 and the breadth of the LD region = 10,000 kb. Furthermore, the F-statistic for each SNP was determined using the (F = β2/SE2), and those with an F-statistic of less than 10 were excluded to lessen the effects of weak instrumental bias. Subsequently, the exposure SNPs were then retrieved from the outcome statistics, and those related to the BPH (p < 1 × 10–5) were disregarded. To find and exclude SNPs associated with confounding variables, we utilized the PhenoScanner V2 database (http://www.phenoscanner.medschl.cam.ac.uk, accessed on 25 June 2023). Previous literature has demonstrated that metabolic syndrome, inflammation, and sex steroid hormones are among the main risk factors for BPH [29, 30]. Finally, data harmonization was performed to ensure compatibility. Incompatible alleles and palindromic SNPs with intermediate allele frequencies were discarded from this study.

For UVMR analysis, the causal relationship between the use of NSAIDs and BPH was assessed using a variety of analyses. The main MR analyses involved utilizing the random effects inverse variance weighted (IVW) method, which yields highly precise estimates while assuming all SNPs are valid instruments. In cases where the results exhibit homogeneity, the fixed effects IVW method offers a more reliable causal evaluation. MR-Egger regression can be estimated while accounting for horizontal pleiotropy, albeit with a slight reduction in precision [31, 32]. The weighted median method yields a valid causal assessment even when up to 50% of the information is derived from erroneous IVs [33].

Heterogeneity was examined using Cochran’s Q statistic and Rucker’s Q statistic, with a significance threshold of p-value > 0.05 indicating the absence of heterogeneity [34]. The intercept term from the MR-Egger regression approach was used to evaluate directional pleiotropy [31]. Additionally, the MR pleiotropy residual sum and outlier (MR-PRESSO) test were utilized to identify outliers and potential pleiotropy, with a global test p-value > 0.05 indicating no potential pleiotropy [35]. Finally, a leave-one-out analysis was conducted to investigate the influence of a single SNP on the outcomes of the genetic evaluation and assess the robustness of the MR findings. Furthermore, we conducted the Steiger directionality test to verify if the observed causal relationships were affected by reversed causation [36]. A Steiger p-value > 0.05 suggests a potential bias in the direction of causal inference. The risk of BPH per unit increase in the log odds of using NSAIDs was calculated using the MR estimations, which were shown as odds ratios (OR). We calculated the statistical power of our primary MR estimates at a significance level of 0.05 using an on-line calculator (https://shiny.cnsgenomics.com/mRnd/) [37].

Multivariable Mendelian randomization analysis

In examining whether there was an independent relationship between the genetic prediction of NSAIDs use and the risk of BPH, we conducted MVMR analyses, adjusting for genetically predicted IL-17 levels. We identified IVs that were associated with IL-17 levels at a significance threshold of p-value < 5 × 10–6. We next eliminated duplicate and correlated SNPs (LD region = 10,000 kb; r2 = 0.001). Following this, we combined the SNPs on medication use and IL-17 levels, extracting SNP effects and accompanying standard errors (SE) from the exposure and outcome GWAS data. Then we used the “MVMR” package to calculate the overall F-statistic of the IVs. Finally, data harmonization was performed.

For the MVMR analysis, we used the random effects IVW, weighted median, and MR-Egger analyses [38, 39]. Cochran’s Q statistic and Rucker’s Q statistic were used to assess the heterogeneity of the MVMR analysis, while the Egger intercept was used to assess pleiotropy. The MVMR-PRESSO test was also utilized to find outliers and possible pleiotropy.

A Bonferroni-corrected criterion of p-value < 0.0167 (α = 0.05/3 exposure factors) was utilized in this research to account for multiple tests. Significance was attributed to associations with p-value < 0.0167, whereas those with p-value ≥ 0.0167 and < 0.05 were considered suggestive. All analyses were two-sided and carried out utilizing “TwoSampleMR” package (version 0.5.7), “MendelianRandomization” package (version 0.8.0), “MVMR” package (version 0.4), and “MRPRESSO” package (version 1.0) within the R software (version 4.3.1).

Results

In the UVMR analysis, 37, 32 and 36 SNPs were identified to be IVs for NSAIDs, salicylic acid and its derivatives and anilines, respectively, following the stringent screening process for IVs. The F-statistics for these SNPs varied from 21 to 65, indicating the absence of weak IVs (Additional file 3: Table S2-4). In the MVMR analysis, 47, 44 and 49 SNPs were identified with F-statistics > 10, respectively. The harmonized data was presented in Additional file 3: Table S5-7 for reference.

Univariable Mendelian randomization analysis

The random effects IVW method revealed that the genetically predicted NSAIDs use was linked to an elevated risk of BPH (OR: 1.164, 95% CI: 1.041–1.302, p = 0.008) (Fig. 2, Additional file 2: Fig. S1A). Similar results were obtained with the fixed effects IVW method (OR: 1.164, 95% CI: 1.053–1.287, p = 0.003). The MR-Egger regression and weighted median methods provided consistent estimates in terms of direction and magnitude, providing further support for the robustness of the causal association. Additionally, a suggestive positive relationship between the genetically predicted anilides usage and the risk of BPH was found (Fig. 2, Additional file 2: Fig. S2A). The MR estimates of the causal relationship between anilides use and BPH remained consistent using the MR models of the random effects IVW method (OR: 1.127, 95% CI: 1.014–1.253, p = 0.027), fixed effects IVW method (OR: 1.127, 95% CI: 1.014–1.253, p = 0.027), and weighted median (OR: 1.168, 95% CI: 1.005–1.358, p = 0.043). MR-Egger regression estimates were also consistent in terms of direction and magnitude. However, we could not discover any proof that genetically predicted salicylic acid and derivatives increased the risk of BPH (random effects IVW method, OR: 0.987, 95% CI: 0.892–1.093, p = 0.801).

Fig. 2
figure 2

UVMR analysis results of medications use (NSAIDs, salicylic acid, and anilides) and BPH. OR: odds ratio, CI: confidence interval

Full size image

Comprehensive sensitivity analyses, as presented in Table 1, were performed to assess the reliability of the MR analyses. Heterogeneity was evaluated using Cochran’s Q statistic (MR-IVW) and Rucker’s Q statistic (MR Egger), with p-value > 0.05 (Table 1), along with the assessment of funnel plot symmetry (Additional file 2: Fig. S1B, Fig. S2B), indicating the absence of heterogeneity. The MR-Egger intercept analysis yielded a p-value > 0.05, suggesting no evidence of horizontal pleiotropy. The MR-PRESSO test demonstrated that the MR analyses for medication use and BPH was not affected by potential pleiotropy or outliers (Global Test p > 0.05). Furthermore, the leave-one-out analysis indicated that the MR results remained robust in the absence of any high-influence SNP (Additional file 2: Fig. S1C, Fig. S2C). Additionally, to further establish the direction of the association between medication use and BPH, a Steiger test was performed (Table 2). The Steiger p-value showed that the detected causal correlations were not skewed by reverse causation. Overall, the thorough sensitivity studies demonstrated that neither pleiotropy nor heterogeneity had an impact on the MR estimations.

Table 1 Sensitivity analysis of the MR analysis results of medications use and BPH
Full size table
Table 2 Steiger direction test from medications use to BPH
Full size table

Multivariable Mendelian randomization analysis

In the context of MVMR analysis, we investigated whether the association between medication and a reduced risk of BPH was affected by inflammation. After adjusting for IL-17 levels, the causal effect of genetically predicted NSAIDs use on BPH persisted, with estimates observed in the IVW analysis (OR: 1.165, 95% CI: 1.049–1.293, p = 0.004) (Fig. 3). MR-Egger and Weighted median method had the same direction, although the p-value is not significant. When IL-17 levels were taken into account, the causal estimates of genetically predicted salicylic acid and derivatives use and anilides use on BPH remained statistically non-significant.

Fig. 3
figure 3

MVMR analysis results of medications use (NSAIDs, salicylic acid, and anilides) and BPH

Full size image

MVMR analysis showed no evidence of heterogeneity (Additional file 3: Table S8). The MR-Egger intercept analysis once again did not indicate the existence of horizontal pleiotropy. Moreover, the MVMR-PRESSO test in MVMR did not find any outliers or potential pleiotropy.

Discussion

Our study revealed a positive association between genetically predicted NSAIDs use and an elevated risk of BPH, even after adjusting for inflammation. In contrast, no substantial indications were discovered, connecting the usage of salicylic acid and derivatives, as well as anilides, to the risk of BPH. Remarkably, this MR study stands as the inaugural exploration into the causal influence of NSAIDs use on BPH, to the utmost extent of our cognizance.

Previous investigations concerning NSAIDs usage and BPH risk have been scarce, and each study exhibits distinct characteristics that hinder direct comparisons with our findings, thereby leading to limited interpretability. Notably, our outcomes diverge from those of earlier conducted randomized controlled trials (RCTs), which found no connection with the risk of BPH [20]. According to a different meta-analysis, COX-2 inhibitors may provide temporary relief for male LUTS-related symptoms [40]. However, no significant differences in changes to the total prostate volume were observed among the patients [40]. It is imperative to acknowledge that RCTs usually entail a meticulously selected study population, potentially posing challenges in generalizing the outcomes to broader populations. Furthermore, the primary endpoints in their study focused on the International Prostate Symptom Score (IPSS) or maximum urinary flow (QMax), which differed from the primary outcomes in our present study. St. Sauver et al. [11] found a noteworthy negative correlation between NSAIDs use, particularly aspirin, and multiple indirect and direct indicators of BPH, including IPSS, QMax, prostate volume, and prostate-specific antigen (PSA) level. In their research, NSAIDs utilization was ascertained through structured interviews at baseline and questionnaires during follow-up, with only men who reported using NSAIDs at the start were deemed exposed. Kang et al. [17] conducted a cross-sectional study, wherein they found that regular use of aspirin and ibuprofen in the preceding year was linked to a higher likelihood of having a history of BPH that has been medically diagnosed (OR: 1.2, 95% CI: 1.1–1.3). However, the temporal relationships between NSAIDs usage and BPH endpoints remained unclear in their study, and the slight increase in risk observed might potentially be attributed to uncontrolled confounding variables. In another cohort study, the use of any NSAIDs, aspirin, and non-aspirin NSAIDs was shown to be strongly related to an elevated risk of BPH [15]. These relationships were marginally lessened by controls for NSAIDs indications and baseline IPSS, but they all lost their significance. In this study, incident BPH was referred to as prolonged, clinically significant BPH symptoms or as surgical or medicinal therapy. Only the beginning of NSAIDs usage was considered, so continued use was assumed. Conventional observational studies suffer from limitations such as reverse causality, residual confounding, and limited sample sizes, which restrict the extent to which we can understand the impact of NSAIDs use on BPH to date. As a consequence, the potential influence of NSAIDs use on BPH remains inadequately elucidated.

Our discoveries have broadened the comprehension of the association between NSAIDs usage and BPH in various dimensions. Principally, our findings enhance the existing observational literature, indicating that the utilization of NSAIDs potentially heightens the risk for BPH. Specifically, our investigation delves into the connection between the usage of NSAIDs (excluding aspirin and acetaminophen), salicylic acid and its derivatives (primarily aspirin), and anilides (mainly acetaminophen) in the context of BPH. Secondly, our study, conducted within the framework of MR, provides compelling evidence supporting a causal association between genetically predicted NSAIDs use (excluding aspirin and acetaminophen) and an elevated risk of BPH. The strength of this causal inference was reinforced by consistently aligned effect estimates obtained through various MR methods, including IVW, MR-Egger, and weighted median. Although the precise mechanism underlying this phenomenon remains unclear, a case-control study discovered that NSAIDs use was linked to a greater incidence of acute urinary retention (AUR) even after adjusting for AUR risk factors (OR: 2.02, 95% CI: 1.23–3.31) [41]. However, the risk of getting AUR was not shown to be higher in patients who were already using acetylsalicylic acid. The Italian spontaneous reporting system (SRS) database also noted a connection between NSAIDs usage and the incidence of AUR [42]. It is possible that prostaglandins may trigger slow tonic contractions of bladder muscle strips, and pretreatment with prostaglandin inhibitors could reduce muscle tone and contractility [43]. An alternative perspective is that NSAIDs may be prescribed for the initial symptoms of AUR, potentially introducing a protopathic bias [41]. Notably, there was no correlation between acetylsalicylic acid usage and the risk of developing AUR, which could be related to the fact that it is often used at low cardioprotective dosages [41]. Lastly, distinguishing between the use of medication for managing systemically symptomatic inflammatory processes that might be causative of BPH and any potential direct effect of the drug itself poses a challenge [10]. The likelihood of causal bias arising from indications for NSAIDs use has been previously discussed in the literature [15, 16]. However, in our study, even after controlling for inflammation in the MVMR analysis, genetically predicted NSAIDs use remained strongly associated with BPH, suggesting that causal bias due to an individual’s systemic inflammation is unlikely. This points to a potential direct contribution of NSAIDs use to an increased risk of BPH. In conclusion, further investigations are necessary to validate the effects of NSAIDs on BPH and to delve deeper into the underlying mechanisms.

There are several significant advantages to this two-sample MR study examining the relationship between genetically predicted NSAID use and BPH. Firstly, by sourcing exposure data from the UKB and outcome data from the FinnGen consortium, we effectively avoid any overlapping sample sizes, ensuring the validity of our analyses. Secondly, to ensure robustness, we employed complementary MR methods to thoroughly scrutinize potential violations of MR assumptions. Notably, the Steiger test provided support for the notion that IVs influence exposure before the outcome, rather than the opposite direction of effect. Additionally, a salient virtue of this study emanates from the remarkable homogeneity characterizing the participants in the GWAS, all sharing European heritage. This approach minimizes the susceptibility of introducing biased results due to demographic stratification biases. The utilization of summary genetic associations sourced from the largest GWAS further amplifies measurement precision through the integration of larger sample sizes. Lastly, a notable strength in this work lies in the implementation of MVMR. By adjusting for systemic inflammation, MVMR allows us to derive a direct effect of genetically predicted NSAIDs use on BPH, while simultaneously accounting for potential confounding factors arising from inflammation.

This study is subject to several limitations that warrant consideration. Firstly, despite utilizing the largest GWAS on NSAIDs use, we only identify a small number of SNPs that met genome-wide significance, possibly leading to potentially weak IVs. In order to counteract this, we lowered the statistical cutoff (p < 5 × 10− 6) and included more SNPs while guaranteeing that all of them had F-statistics > 10. Secondly, owing to the representation of each drug usage phenotype as binary variables (use/non-use), the capacity to discern potential dose-dependent alterations in the risk associated with NSAIDs use remains constrained. Thirdly, it is possible that the genetic variants associated with NSAIDs medication use could also be related to underlying diseases or inflammation, which may influence BPH risk, thereby potentially impacting the study results. Fourthly, data-driven IVs selection may introduce bias into the causal estimate in the summary statistics for drug usage from specific GWAS studies, which may restrict their ability to identify genetic connections [44]. Fifthly, our MR analyses may currently lack sufficient statistical power. As shown in Additional File 2: Fig. S3, the statistical power for a causal inference with an odds ratio (OR) of 1.164 was approximately 0.5. Although the power increases as the OR approaches 1.488, it remains below the commonly accepted threshold of 80%. Future studies leveraging larger GWAS datasets will be essential to validate these findings. Sixthly, the exclusive inclusion of individuals of European ancestry necessitates cautious extrapolation of the findings to other demographic groups. Finally, although the MR method demonstrates excellent performance in causal inference, we caution that the findings of this MR study should be further validated through robust RCTs to confirm the existence of causal relationships.

Conclusions

In summary, our MR analyses present compelling evidence for a causal effect of the genetically predicted NSAIDs use on the heightened risk of BPH, and this causal relationship remains evident even after excluding inflammation. These findings impart novel perspectives on the underlying correlation between NSAIDs use and BPH, thereby potentially prompting clinicians to exercise heightened vigilance in the monitoring and management of BPH among patients employing NSAIDs.

Data availability

The datasets for IL-17 levels can be found in GWAS Catalog (https://www.ebi.ac.uk/gwas/publications/37563310). The summary statistics for drug use are available in GWAS Catalog (https://www.ebi.ac.uk/gwas/publications/31015401). The summary statistics for BPH are available at the FinnGen consortium (https://www.finngen.fi/fi).

Abbreviations

BPH:

Benign prostatic hyperplasia

NSAIDs:

Non-steroidal anti-inflammatory drugs

LUTS:

lower urinary tract symptoms

MR:

Mendelian randomization

IVs:

Instrumental variables

UVMR:

Univariable Mendelian randomization

GWAS:

Genomewide association study

MVMR:

Multivariate MR

UKB:

United Kingdom Biobank

ATC:

Anatomical Therapeutic Chemical

Il-17:

Interleukin-17

ICD-10:

International Classification of Diseases—10th Revision

SNPs:

Single nucleotide polymorphisms

LD:

linkage disequilibrium

IVW:

Inverse variance weighted

MR-PRESSO:

MR pleiotropy residual sum and outlier

OR:

Odds ratio

SE:

Standard errors

RCTs:

Randomized controlled trials

IPSS:

International Prostate Symptom Score

PSA:

Prostate-specific antigen

AUR:

Acute urinary retention

SRS:

Spontaneous reporting system

References

  1. Devlin CM, Simms MS, Maitland NJ. Benign prostatic hyperplasia - what do we know? BJU Int. 2021;127(4):389–99.

    Article  CAS  PubMed  Google Scholar 

  2. Awedew AF, Han H, Abbasi B, Abbasi-Kangevari M, Ahmed MB, Almidani O, et al. The global, regional, and National burden of benign prostatic hyperplasia in 204 countries and territories from 2000 to 2019: a systematic analysis for the global burden of disease study 2019. Lancet Healthy Longev. 2022;3(11):e754–76.

    Article  Google Scholar 

  3. Miernik A, Gratzke C. Current treatment for benign prostatic hyperplasia. Deutsches Ärzteblatt international; 2020.

  4. Lim KB. Epidemiology of clinical benign prostatic hyperplasia. Asian J Urol. 2017;4(3):148–51.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Gandaglia G, Zaffuto E, Fossati N, Cucchiara V, Mirone V, Montorsi F, et al. The role of prostatic inflammation in the development and progression of benign and malignant diseases. Curr Opin Urol. 2017;27(2):99–106.

    Article  PubMed  Google Scholar 

  6. Dobrek Ł, Thor PJ. Benign prostatic hyperplasia - progress in pathophysiology and management. Pol Merkur Lekarski. 2015;39(233):263–70.

    PubMed  Google Scholar 

  7. Bindu S, Mazumder S, Bandyopadhyay U. Non-steroidal anti-inflammatory drugs (NSAIDs) and organ damage: A current perspective. Biochem Pharmacol. 2020;180:114147.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Minnery CH, Getzenberg RH. Benign prostatic hyperplasia cell line viability and modulation of jm-27 by Doxazosin and ibuprofen. J Urol. 2005;174(1):375–9.

    Article  CAS  PubMed  Google Scholar 

  9. Lloyd GL, Ricke WA, McVary KT. Inflammation, voiding and benign prostatic hyperplasia progression. J Urol. 2019;201(5):868–70.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Lloyd GL, Marks JM, Ricke WA. Benign prostatic hyperplasia and lower urinary tract symptoms: what is the role and significance of inflammation?? Curr Urol Rep. 2019;20(9):54.

    Article  PubMed  PubMed Central  Google Scholar 

  11. St. Sauver JL, Jacobson DJ, McGree ME, Lieber MM, Jacobsen SJ. Protective association between nonsteroidal antiinflammatory drug use and measures of benign prostatic hyperplasia. Am J Epidemiol. 2006;164(8):760–8.

    Article  PubMed  Google Scholar 

  12. Kahokehr A, Vather R, Nixon A, Hill AG. Non-steroidal anti-inflammatory drugs for lower urinary tract symptoms in benign prostatic hyperplasia: systematic review and meta-analysis of randomized controlled trials. BJU Int. 2013;111(2):304–11.

    Article  CAS  PubMed  Google Scholar 

  13. Falahatkar S, Mokhtari G, Pourreza F, Asgari SA, Kamran AN. Celecoxib for treatment of nocturia caused by benign prostatic hyperplasia: a prospective, randomized, double-blind, placebo-controlled study. Urology. 2008;72(4):813–6.

    Article  PubMed  Google Scholar 

  14. Di Silverio F, Bosman C, Salvatori M, Albanesi L, Proietti Pannunzi L, Ciccariello M, et al. Combination therapy with rofecoxib and finasteride in the treatment of men with lower urinary tract symptoms (LUTS) and benign prostatic hyperplasia (BPH). Eur Urol. 2005;47(1):72–8. discussion 8–9.

    Article  PubMed  Google Scholar 

  15. Schenk JM, Calip GS, Tangen CM, Goodman P, Parsons JK, Thompson IM, et al. Indications for and use of nonsteroidal antiinflammatory drugs and the risk of incident, symptomatic benign prostatic hyperplasia: results from the prostate cancer prevention trial. Am J Epidemiol. 2012;176(2):156–63.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Nygård LH, Talala K, Taari K, Tammela TLJ, Auvinen A, Murtola TJ. The effect of non-steroidal anti-inflammatory drugs on risk of benign prostatic hyperplasia. Prostate. 2017;77(9):1029–35.

    Article  PubMed  Google Scholar 

  17. Kang D, Andriole GL, Van De Vooren RC, Crawford D, Chia D, Urban DA, et al. Risk behaviours and benign prostatic hyperplasia. BJU Int. 2004;93(9):1241–5.

    Article  CAS  PubMed  Google Scholar 

  18. Sutcliffe S, Grubb Iii RL, Platz EA, Ragard LR, Riley TL, Kazin SS, et al. Non-steroidal anti-inflammatory drug use and the risk of benign prostatic hyperplasia-related outcomes and nocturia in the prostate, lung, colorectal, and ovarian cancer screening trial. BJU Int. 2012;110(7):1050–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Meigs JB, Mohr B, Barry MJ, Collins MM, McKinlay JB. Risk factors for clinical benign prostatic hyperplasia in a community-based population of healthy aging men. J Clin Epidemiol. 2001;54(9):935–44.

    Article  CAS  PubMed  Google Scholar 

  20. Ozdemir I, Bozkurt O, Demir O, Aslan G, Esen AA. Combination therapy with Doxazosin and tenoxicam for the management of lower urinary tract symptoms. Urology. 2009;74(2):431–5.

    Article  PubMed  Google Scholar 

  21. Sekula P, Del Greco MF, Pattaro C, Köttgen A. Mendelian randomization as an approach to assess causality using observational data. J Am Soc Nephrol. 2016;27(11):3253–65.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Davey Smith G, Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet. 2014;23(R1):R89–98.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Pierce BL, Burgess S. Efficient design for Mendelian randomization studies: subsample and 2-sample instrumental variable estimators. Am J Epidemiol. 2013;178(7):1177–84.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Emdin CA, Khera AV, Kathiresan S. Mendelian Randomization Jama. 2017;318(19):1925–6.

  25. Wu Y, Byrne EM, Zheng Z, Kemper KE, Yengo L, Mallett AJ, et al. Genome-wide association study of medication-use and associated disease in the UK biobank. Nat Commun. 2019;10(1):1891.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Zhao JH, Stacey D, Eriksson N, Macdonald-Dunlop E, Hedman ÅK, Kalnapenkis A et al. Genetics of Circulating inflammatory proteins identifies drivers of immune-mediated disease risk and therapeutic targets. Nat Immunol. 2023.

  27. Kurki MI, Karjalainen J, Palta P, Sipila TP, Kristiansson K, Donner KM, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613(7944):508–18.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Rosoff DB, Smith GD, Lohoff FW. Prescription opioid use and risk for major depressive disorder and anxiety and Stress-Related disorders: A multivariable Mendelian randomization analysis. JAMA Psychiatry. 2021;78(2):151–60.

    Article  PubMed  Google Scholar 

  29. Corona G, Vignozzi L, Rastrelli G, Lotti F, Cipriani S, Maggi M. Benign prostatic hyperplasia: a new metabolic disease of the aging male and its correlation with sexual dysfunctions. Int J Endocrinol. 2014;2014:329456.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Patel ND, Parsons JK. Epidemiology and etiology of benign prostatic hyperplasia and bladder outlet obstruction. Indian J Urol. 2014;30(2):170–6.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32(5):377–89.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect Estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent Estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304–14.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Greco MF, Minelli C, Sheehan NA, Thompson JR. Detecting Pleiotropy in Mendelian randomisation studies with summary data and a continuous outcome. Stat Med. 2015;34(21):2926–40.

    Article  Google Scholar 

  35. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal Pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Hemani G, Tilling K, Davey Smith G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet. 2017;13(11):e1007081.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Brion MJ, Shakhbazov K, Visscher PM. Calculating statistical power in Mendelian randomization studies. Int J Epidemiol. 2013;42(5):1497–501.

    Article  PubMed  Google Scholar 

  38. Rees JMB, Wood AM, Burgess S. Extending the MR-Egger method for multivariable Mendelian randomization to correct for both measured and unmeasured Pleiotropy. Stat Med. 2017;36(29):4705–18.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Burgess S, Thompson SG. Multivariable Mendelian randomization: the use of pleiotropic genetic variants to estimate causal effects. Am J Epidemiol. 2015;181(4):251–60.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Tang G, Liu M, Ding G, Liu S, Chu Y, Cui Y, et al. The efficacy of Cyclooxygenase-2 inhibitors for the male treatment of lower urinary tract symptoms: A systematic review and Meta-Analysis. Am J Mens Health. 2023;17(3):15579883231176667.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Verhamme KM, Dieleman JP, Van Wijk MA, van der Lei J, Bosch JL, Stricker BH, et al. Nonsteroidal anti-inflammatory drugs and increased risk of acute urinary retention. Arch Intern Med. 2005;165(13):1547–51.

    Article  CAS  PubMed  Google Scholar 

  42. Crisafulli S, Cutroneo PM, Verhamme K, Ferrajolo C, Ficarra V, Sottosanti L, et al. Drug-induced urinary retention: an analysis of a National spontaneous adverse drug reaction reporting database. Eur Urol Focus. 2022;8(5):1424–32.

    Article  PubMed  Google Scholar 

  43. Gruenenfelder J, McGuire EJ, Faerber GJ. Acute urinary retention associated with the use of cyclooxygenase-2 inhibitors. J Urol. 2002;168(3):1106.

    Article  PubMed  Google Scholar 

  44. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–65.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We want to acknowledge the participants and investigators of the studies used in this research.

Funding

This research was funded by the National Natural Science Foundation of China (No. 81272846).

Author information

Authors and Affiliations

  1. Department of Urology, The Second Affiliated Hospital of Xi’an Jiaotong University, Xi’an, 710004, Shaanxi, China

    Zi-He Peng, Ming-Rui Li, Min-Xin He, Yao Dong, Chong Yan, Zi-Hao Li, Tie Chong & Zhao-Lun Li

  2. Health Science Center, Xi’an Jiaotong University, Xi’an, 710061, Shaanxi, China

    Zi-He Peng, Ming-Rui Li, Min-Xin He, Jing Liu, Jia-Hao Dou, Ya-Wen Wang, Yao Dong, Chong Yan & Zi-Hao Li

  3. Department of Pediatrics, Jiangxi Hospital Affiliated to Children’s Hospital of Chongqing Medical University (Jiangxi Provincial Children’s Medical Center), Nanchang, 330000, Jiangxi, China

    Jing Liu

Authors
  1. Zi-He Peng
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Ming-Rui Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Min-Xin He
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Jing Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Jia-Hao Dou
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Ya-Wen Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Yao Dong
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Chong Yan
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Zi-Hao Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Tie Chong
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Zhao-Lun Li
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Conceptualization: Z.H.P, T.C and Z.L.L; Formal analysis: Z.H.P, M.R.L and M.X.H; Funding acquisition: Z.L.L; Investigation: J.L, J.H.D and Y.W.W; Methodology: M.R.L and M.X.H; Project administration: T.C and Z.L.L; Resources: Y.D, C.Y and Z.H.L; Software: Z.H.P, M.R.L and M.X.H; Supervision: T.C and Z.L.L; Validation: M.R.L; Visualization: Z.H.P; Writing– original draft: Z.H.P; Writing– review & editing: M.R.L and M.X.H. All authors have read and agreed to the published version of the manuscript.

Corresponding authors

Correspondence to Tie Chong or Zhao-Lun Li.

Ethics declarations

Ethics approval and consent to participate

All data sources were derived from publicly available summary-level data, so no informed patient consent was required to release the summary-level data. The ethics committee of the Second Affiliated Hospital of Xi’an Jiaotong University waived the need for ethical approval and informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Additional file 1: STROBE MR checklist.

12920_2025_2128_MOESM2_ESM.pdf

Additional file 2: figure S1. Association between genetically predicted non-steroidal anti-inflammatory drugs use and benign prostatic hyperplasia presented in (A) scatter plot, (B) funnel plot, (C) leave-one-out sensitivity analysis, and (D) forest plot. Figure S2. Association between genetically predicted anilides use and benign prostatic hyperplasia presented in (A) scatter plot, (B) funnel plot, (C) leave-one-out sensitivity analysis, and (D) forest plot. Figure S3. Power calculation of Mendelian randomization analysis.

12920_2025_2128_MOESM3_ESM.xlsx

Additional file 3: table S1. Detailed information on used studies. Table S2. Characteristics of the genetic variants used as the instrumental variables for NSAIDs. Table S3. Characteristics of the genetic variants used as the instrumental variables for salicylic acid and derivatives. Table S4. Characteristics of the genetic variants used as the instrumental variables for anilides. Table S5. Characteristics of the genetic variants in multivariate MR analysis used as the instrumental variables for NSAIDs. Table S6. Characteristics of the genetic variants in multivariate MR analysis used as the instrumental variables for salicylic acid and derivatives. Table S7. Characteristics of the genetic variants in multivariate MR analysis used as the instrumental variables for anilides. Table S8. Sensitivity analysis of the multivariate MR results of medications use and BPH.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Peng, ZH., Li, MR., He, MX. et al. Causal association between non-steroidal anti-inflammatory drugs use and the risk of benign prostatic hyperplasia: a univariable and multivariable Mendelian randomization study. BMC Med Genomics 18, 60 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12920-025-02128-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12920-025-02128-1

Keywords

BMC Medical Genomics

ISSN: 1755-8794

Contact us