Recent literature has suggested that bicycling may be associated with increases in serum PSA levels, a diagnostic and prognostic marker for prostate cancer. To further investigate this relationship, we conducted a systematic review and meta-analysis of current literature in this field.
MEDLINE, CENTRAL, CINAHL and SPORTDiscus were searched using MeSH terms and keywords for English publications related to bicycle riding and PSA. Studies were included if PSA was measured relative to cycling activity in healthy men who were free of any prostatic condition. Case studies were excluded.
Eight studies met our inclusion criteria, comprising 912 participants that engaged in, or self-reported, bicycling activity. Six studies investigated the acute pre-post change in PSA following bicycling activity that ranged from a single cycling bout of 15 min to a 4-day cycling event. Following cycling activity, two studies reported total PSA increased from baseline by up to 3.3-fold, free PSA increased in one study by 0.08±0.18 ng ml−1 and did not change in four studies. One study compared PSA in elite/professional cyclists versus non-cyclists and demonstrated no significant difference in PSA measurements between groups. Data from six studies were meta-analyzed and demonstrated no significant increase in PSA associated with cycling from pre to post (mean change +0.027 ng ml−1, s.e.m.=0.08, P=0.74, 95% confidence interval (CI)=−0.17–0.23).
Our findings suggest that there is no effect of cycling on PSA; however, the limited number of trials and the absence of randomized controlled trials limit the interpretation of our results. Additionally, the median sample size only consisted of 42 subjects. Therefore, our study may have low statistical power to detect a difference in PSA. Although, a higher sample size may demonstrate statistical significance, it may not be clinically significant. Studies of higher empirical quality are needed.
Bicycling is commonly used as transportation, fitness and recreation by individuals of all ages.1 Recent literature2, 3, 4 suggests that cycling may cause increases in serum PSA, a widely used biomarker for prostate cancer (PCa) detection and monitoring as well as other prostatic diseases.5 The positive relationship between cycling and PSA is hypothesized to arise due to the direct mechanical pressure and resultant microtrauma to the prostate.3 Others have suggested that cycling-related increases in PSA are due to contraction and movement of the pelvic muscles, squeezing the prostate gland and prostatic plexus. This manipulation of the prostate would then result in increased blood flow to the dorsal vein and releases blood from the tissue, including PSA, into circulation thereby increasing PSA concentration as much as threefold.2 To date, the few studies that have examined the relationship between cycling and PSA have yielded inconsistent findings.2, 3, 4, 6, 7, 8, 9, 10 For example, one study reported an increase in PSA from 2 to 3.3 times the baseline value in individuals who rode a bicycle ergometer for 15 min at a 100-W workload.2 Similarly, Mejak et al.4 reported an increase in PSA by 9.5% after a recreational 1-day cycling event where cyclists rode for distances of 55–160 km. Such increases could have important clinical implications related to when PSA measurements should be taken relative to a bout of cycling. However, Saka et al.10 observed no significant differences in PSA after a 300-km bicycle circuit in healthy athletes. Therefore, to better understand the effect of cycling on PSA, we conducted a systematic review and meta-analysis to determine the effect of cycling on PSA.
Materials and methods
A librarian-assisted search was conducted in MEDLINE, CINAHL, the Cochrane Central Register of Controlled Trials (CENTRAL) and SPORTDiscus using medical subject headings (MeSH), subject headings (SH), CINAHL headings (CH) and medical headings (MH) and additional keyword combinations to identify relevant studies. We used the following combination of search strategies to yield our results: (Motor activity OR physical activity OR sports [SH] OR cycling [SH] AND prostate-specific antigen; Motor activity OR physical activity OR sports [CH] OR cycling AND prostate-specific antigen [CH]); [MH Motor Activity] OR [MH Bicycling] OR [MH Sports] OR [MH Physical Exertion] AND [MH Prostate-Specific Antigen]; ‘Motor Activity’ [MeSH] OR ‘Bicycling’ [MeSH] OR ‘Sports’ [MeSH] OR ‘Physical Exertion’ [MeSH] AND ‘Prostate-Specific Antigen’ [MeSH]. Studies included were required to be published in English between 1 January 1950 and 29 July 2014. Case studies and gray literature (that is, abstracts, theses and dissertations, unpublished studies) were excluded. References of manuscripts were searched to identify additional literature related to PSA and cycling activity. Studies with missing information such as overall change in PSA were excluded from the meta-analysis (n=2). Authors of included studies were contacted by e-mail to identify additional relevant studies within the field for consideration.
The following eligibility criteria were used to identify studies for inclusion: (a) males older than 18 years of age; (b) participants were clinically healthy and free of prostatic disease; (c) participants engaged in, or self-reported, a volume of cycling activity; and (d) participants’ PSA was measured (serum total PSA (tPSA), serum free PSA (fPSA), %free PSA and complexed PSA). Cycling was operationally defined as activity that utilized either a stationary cycle or a bicycle for a reported duration or distance.
Two reviewers (DJ and AR) independently screened the titles and abstracts identified by the search. Full-text articles of relevant studies were obtained and reviewed to confirm inclusion. Data extraction summarized information about study design (randomized controlled trial, observational and so on), sample size, participant characteristics (age, pre-existing medical conditions and current level of fitness) and methods of cycling (long duration, short duration, laboratory setting and race setting). Disagreements were reconciled by discussion and, where necessary, by arbitration with a third investigator (DSM).
Risk of bias
The Cochrane Risk of Bias Assessment Tool11 was used to evaluate within-study risk of bias across six domains: sequence generation, concealed allocation, participant blinding, outcome assessment, incomplete data, selective outcome reporting and other potential sources of bias.
Pre-post values or the reported difference in PSA between groups as the measured effect of cycling and PSA were used for statistical analyses. Mixed-model repeated measure analysis was used to compare differences between groups with adjustments for age, cycling distance, time cycling and time to blood draw. Three subgroup analyses were conducted to determine the association between age, time to blood draw and the type of cycling (mountain biking versus on-road biking) on PSA following a bout of cycling. Two studies were excluded from our analysis due to missing overall mean PSA change. Therefore, we could not compare stationary biking with either mountain biking or on-road biking as only one of six included studies used stationary biking protocols. Our first subgroup analysis compared participants <50 and ⩾50 years adjusted for cycling distance, time cycled and time to blood draw. Our second subgroup analysis compared participants in terms of time to blood draw with individuals tested <2 h post race versus those tested >2 h post race. Our time to blood draw analysis was adjusted for age, distance and time cycled to control for the confounding effects of these variables on PSA changes. Our final subgroup analysis compared the type of cycling (mountain biking versus on-road cycling) on PSA adjusted for age, cycling distance, time cycled and time to blood draw. Study results were expressed as mean tPSA values (ng ml−1) as the included literature reported overall change in PSA and not individual PSA values of each participant that prevented comparison of median scores. Heterogeneity was assessed using the I2 statistic with values ranging from 0 to 100% (I2=0–40% is no heterogeneity; 30–60% is moderate heterogeneity; 50–90% is substantial heterogeneity; and 75–100% is considerable heterogeneity).11 The meta-analysis was conducted using Statistical Analysis System version 9.3 (SAS, Cary, NC, USA).
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram is presented in Figure 1. Overall, the database search yielded 111 citations, of which 68 were excluded based on title and abstract screening. Forty-three full-text articles were retrieved and reviewed in full, and an additional 35 were excluded for not meeting our inclusion criteria. A total of eight studies met the eligibility criteria and were included in the study. A total of eight reports were included in the systematic review.2, 3, 4, 6, 7, 8, 9, 10 Two studies2, 9 were excluded from our meta-analysis due to missing overall pre- and post-cycling PSA values. Therefore, six studies were included in the meta-analysis.3, 4, 6, 7, 8, 10
A summary of the included trials is presented in Table 1. Studies were published between 1996 and 2013. Studies were excluded from review for the following reasons: not including cycling as part of their intervention (n=30), not measuring PSA as part of their outcome (n=4) and for being a non-English study (n=1). A total of the eight studies were included in the review of which six were pre-post test studies,2, 3, 4, 6, 7, 8 one was a non-randomized controlled trial,10 and one was a cross-sectional study.9 A total of 912 participants were included in the eight studies. The median sample size of the included trials was 42, with a range of 20–301. In the pre-post test trials, pre-cycling PSA measurements ranged from 48 h to 15 min before cycling activity. The post-cycling PSA measurement most proximal to the conclusion of the cycling activity was taken between 5 min and 3 days post cycling. Three studies measured tPSA and five studies measured fPSA in relation to cycling behavior. Cycling activities included in our study consisted of low-to-moderate intensity. Low intensity activities include: ergometer cycling (15 min),2 long (160–400 km) and short distance biking (from 21 to 55 km).3, 4, 6, 7, 8, 9, 10 The majority of cycling activities were of low intensity where the participants were allowed to ride at their own pace; whereas a moderate intensity was defined as a time-cycling event where participants were required to complete a certain distance in a given time, or race against other participants (Figure 2).
Effect of cycling on PSA
Three of the eight studies compared the change in tPSA from pre-to-post cycling. Two studies reported no change in tPSA,3, 6 while one reported an increase in tPSA.4 Five studies examined the effect of cycling on fPSA; four reported no change7, 8, 9, 10 and one reported an increase.2
Risk of bias of individual studies assessing cycling and PSA
Table 2 presents the risk of bias scores for the included trials. The median risk of bias score was 3.5 out of 6, ranging from 34, 6, 9, 10 to 5.2 These studies are considered to have a high risk of bias.
Study heterogeneity of the six studies included in the quantitative analysis was calculated using the I2 statistic. There was significant heterogeneity in the included studies (I2=69%, P=0.011). A forest plot of the meta-analyzed studies is presented in Table 2. Two studies did not report pre and post PSA values or the mean change in PSA and they were excluded.2, 9 Overall, our analysis of the six included studies revealed no significant change in the total PSA levels after a bout of cycling (mean change +0.027 ng ml−1, s.e.m. 0.08, P=0.74, 95% confidence interval (CI)=−0.17–0.23). Three additional subgroup analyses were conducted to determine the association between age, time to blood draw and the type of cycling (mountain biking versus on-road cycling) to further assess their effect on PSA following a bout of cycling. The first subgroup analysis compared the mean change in tPSA between age groups (<50 years and ⩾50 years) and was adjusted for distance, time cycled and time to blood draw. Two studies included patients greater than 50 years and four studies included patients younger than 50 were compared. Our analysis revealed no differences in the change in tPSA between those <50 years versus ⩾50 years following a bout of cycling (mean change in tPSA of ⩾50 years compared with <50 years=+0.1027 ng ml−1; 95% CI =−1.80–2.00, P=0.62). The second subgroup analysis compared tPSA in studies that differed in duration between cycling and blood sampling and was adjusted for age, distance cycled and time cycled. Three studies collected blood samples <2 h post ride and were compared with three studies that collected blood samples ⩾2 h. There was no difference in change in PSA levels for those tested <2 versus ⩾2 h after the cycling bout (mean change in tPSA of ⩾2 h post cycling compared to <2 h=+0.00076 ng ml−1, 95% CI=−0.28–0.27, P=0.99). Our third analysis compared tPSA changes between the types of cycling activity and was adjusted for age, time cycled, distance cycled and time to blood draw. When comparing cycling activity, two studies reported mountain biking and were compared with three studies that reported on-road cycling. When comparing mountain bikers versus on-road cyclers, the type of cycling did not influence PSA levels (mean change in tPSA of mountain bikers compared with on-road cyclers=−0.00256 ng ml−1, 95% CI=−0.48–0.48, P=0.72).
Bicycle riding has been hypothesized to increase serum PSA levels; however, current findings are inconsistent.2, 3, 4, 6, 7, 9 PSA measurement is important as it is used to screen for PCa and other prostatic morbidities;12 therefore, potential PSA increases after cycling may be clinically relevant. To our knowledge, this is the first systematic review to examine the effect of cycling on PSA. The majority of existing literature indicates no effect of cycling on PSA; hence, the risk of a false-positive PSA change is relatively low. Additionally, this is substantiated by our meta-analysis, which reports no significant effect of cycling on PSA. Our study must be interpreted with caution, as there are only a limited number of studies exploring the effect of cycling on PSA, and most importantly, the evidence pool lacks randomized controlled trials and is burdened with other methodological weaknesses.
The previous literature has suggested that mechanical pressure to the prostate results in increases in PSA levels.3, 6, 13, 14 Orestein et al.15 reported that total PSA was elevated in men after a digital rectal examination (DRE) for at least a week before dropping down to normal levels. Cevik et al.13 also reported increases in PSA 30 min and 24 h after a DRE. Additionally, after 30 min of prostatic manipulation,16 both DRE and prostatic biopsy increased PSA by 1.12 and 2.43 ng ml−1, respectively (P<0.001). In contrast, studies by Glenski et al.17 and McAleer et al.18 reported no differences in PSA after DRE. Our findings suggest that cycling does not significantly alter PSA levels; and although the evidence relating PSA changes to direct pressure on the prostate (for example, via DRE) is equivocal, there may be specific cycling-related factors that mitigate pressure on the prostate. First, the saddle used in cycling may displace pressure across the perineal and gluteal region to effectively alleviate pressure on the prostate. Further, Hermann et al.8 suggests that the typical cycling position of an leaning forward and providing upper body support through the handlebars can displace upper body mass across the cycle, also reducing pressure on the perineal region and shifts the pressure from the posterior perineum to the anterior perineum and the symphysis. Additionally, these findings are also echoed by Lippi et al.9 who concluded that regardless of the bout of cycling (heavy or regular), compression of the prostatic region is not substantial enough to influence increases in PSA by the prostatic tissue. The lack of PSA increase could also be attributed to an efficient shock absorption of the connective tissue and muscle between the body’s surface and prostate.8 Results of our study suggest that neither age nor time to blood draw following a bout of cycling influenced post-cycling PSA.
Our results must be interpreted with caution. First, our study consisted of a few number of trials all with relatively small sample sizes; therefore, it is likely that the studies included in our analyses lacked the power to detect a statistically significant difference in PSA. However, it is important to note that even though increasing the sample size may improve statistical power to detect a statistical difference in PSA from pre- to post-cycling, the overall change in PSA we observed was not clinically significant. Additionally, studies included in our review had a high risk of bias with scores ranging from 3 to 5 of the Cochrane Risk of Bias assessment tool. Risk of bias was due to incomplete outcome data (improper reporting of attrition, performing a sub-group analysis on a select few participants without justification) and selective outcome reporting (not reporting the overall mean of PSA). For example, one2 of the three studies2, 4, 9 that observed increases in PSA after a bout of bicycle riding was excluded from our meta-analysis as they did not report the overall mean PSA values, making it difficult to interpret the results for our study. Moreover, the heterogeneity of methodologies and the absence of randomized controlled trials remain the primary limitation to this evidence base. On the other hand, our review also consists of several strengths. We included a comprehensive search strategy of five databases, and searched reference lists and contacted authors in the field; two investigators independently reviewed all studies identified during the search; and we conducted secondary analyses to further understand the relationship between factors that may impact PSA changes following cycling.
The trials included in this review suggest that cycling does not have any effect on PSA, regardless of age or time when PSA is measured. It is important to note significant limitations in study quantity and quality along with low statistical power may have influenced the outcome of the results in our review; thus, cautious interpretation is required. Future investigation into this relationship would greatly benefit from randomized controlled designs and other quality pre-post literature with consistent use of cycling protocols for enhanced generalizability across studies.
Leibovitch I, Mor Y . The vicious cycling: bicycling related urogenital disorders. Eur Urol 2005; 47: 277–l86; discussion 286–7.
Oremek GM, Seiffert UB . Physical activity releases prostate-specific antigen (PSA) from the prostate gland into blood and increases serum PSA concentrations. Clin Chem 1996; 42: 691–695.
Safford RH . The effect of bicycle riding on serum prostate specific antigen levels. Am Urol Assoc 1996; 156: 103–105.
Mejak SL, Bayliss J, Hanks SD . Long distance bicycle riding causes prostate-specific antigen to increase in men aged 50 years and over. PLoS ONE 2013; 8: e56030.
Nadler RB, Humphrey PA, Smith DS, Catalona WJ, Ratliff TL . Effect of inflammation and benign prostatic hyperplasia on elevated serum prostate specific antigen levels. J Urol 1995; 154: 407–413.
Swain RA, Montalto N, Ross D . The effect of long-distance cycling on the prostate-specific antigen level. Arch Fam Med 1997; 6: 500–502.
Luboldt H-J, Peck KD, Oberpenning F, Schmid H-P, Semjonow A . Bicycle riding has no important impact on total and free prostate-specific antigen serum levels in older men. Urology 2003; 61: 1177–1180.
Herrmann M, Scharhag J, Sand-Hill M, Kindermann W, Herrmann W . Long-distance mountain biking does not disturb the measurement of total, free or complexed prostate-specific antigen in healthy men. Clin Chem Lab Med 2004; 42: 347–349.
Lippi G, Corgnati a, Salvagno G, Schena F, Franchini M, Guidi G . Total and free PSA serum concentrations are not influenced by extensive physical exercise and bicycle riding. Int J Sports Med 2005; 26: 79–81.
Saka T, Sofikerim M, Demirtas A, Kulaksizoglu S, Caniklioglu M, Karacagil M . Rigorous bicycling does not increase serum levels of total and free prostate-specific antigen (PSA), the free/total PSA ratio, gonadotropin levels, or uroflowmetric parameters. Urology 2009; 74: 1325–1330.
Higgins JPT, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD et al. The Cochrane Collaboration's tool for assessing risk of bias in randomized trials. BMJ 2011; 343: d5928.
Arneth BM . Clinical significance of measuring prostate-specific antigen. Lab Med 2009; 40: 487–491.
Cevik I, Türkeri LN, Ozveri H, Ilker Y, Akdaş A . Short-term effect of digital rectal examination on serum prostate-specific antigen levels. A prospective study. Eur Urol 1996; 29: 403–406.
Ornstein DK, Smith DS, Humphrey Pa, Catalona WJ . The effect of prostate volume, age, total prostate specific antigen level and acute inflammation on the percentage of free serum prostate specific antigen levels in men without clinically detectable prostate cancer. J Urol 1998; 159: 1234–1237.
Ornstein DK, Rao G, Smith D . Effect of digital rectal examination and needle biopsy on serum total and percentage of free prostate specific antigen levels. J Urol 1997; 157: 195–198.
Collins GN, Martin PJ, Wynn-davies A, Brooman PJ, Reilly PHO . The effect of digital rectal examination, flexible cystoscopy and prostatic biopsy on free and total prostate. Eur Urol 1997; 157: 1744–1747.
Glenski WJ, Oesterling JE, Klee GG, Bergstralh EJ . Prostate-specific antigen: Establishment of the reference range for the clinically normal prostate gland and the effect of digital rectal examination, ejaculation, and time on serum concentrations. Prostate 1992; 21: 99–110.
McAleer JK, Gerson LW, McMahon D, Geller L . Effect of digital rectal examination (and ejaculation) on serum prostate-specific antigen after twenty-four hours. A randomized, prospective study. Urology 1993; 41: 111–112.
The authors declare no conflict of interest.
About this article
Cite this article
Jiandani, D., Randhawa, A., Brown, R. et al. The effect of bicycling on PSA levels: a systematic review and meta-analysis. Prostate Cancer Prostatic Dis 18, 208–212 (2015). https://doi.org/10.1038/pcan.2015.16
Journal of Clinical Urology (2020)
Raccomandazioni per la gestione di variabili preanalitiche legate al paziente nella determinazione del PSA in fase di screening e follow-up di cancro prostatico
La Rivista Italiana della Medicina di Laboratorio (2020)
International Journal of Clinical Practice (2016)
Tijdschrift voor Urologie (2016)