Abstract
and Objectives: The objectives of this review were as follows: to measure changes in renal biomarker levels before, immediately after, and 24 h post-marathon; to identify promising biomarkers for the diagnosis of acute kidney injury; and to describe the temporal patterns of biomarker dynamics in relation to the marathon.Materials and Methods: Studies of marathon runners reporting AKI-related biomarkers were included. Four databases (PubMed, EMBASE, Web of Science, and LILACS) were searched. Data on study design, participant characteristics, and biomarker values (pre-, post-, and 24 h post-race) were ex- tracted, and a random effects meta-analysis was performed. Risk of bias was assessed with the National Heart, Lung, and Blood Institute pre–post tool.Results: The study showed sig- nificant increases in most biomarkers immediately after the marathon compared to baseline values. The largest increases were observed in Tissue Inhibitor of Metalloproteinases-2* Insulin-like Growth Factor Binding Protein-7 (TIMP-2*IGFBP), copeptin, urinary Liver- type Fatty Acid Binding Protein (L-FABP), urinary Monocyte Chemoattractant Protein-1 (MCP-1), IGFBP-7, urinary Chitinase
with the National Heart, Lung, and Blood Institute pre–post tool.Results: The study showed sig- nificant increases in most biomarkers immediately after the marathon compared to baseline values. The largest increases were observed in Tissue Inhibitor of Metalloproteinases-2* Insulin-like Growth Factor Binding Protein-7 (TIMP-2*IGFBP), copeptin, urinary Liver- type Fatty Acid Binding Protein (L-FABP), urinary Monocyte Chemoattractant Protein-1 (MCP-1), IGFBP-7, urinary Chitinase 3-like Protein 1 (YKL-40), and TIMP-2, suggesting that these biomarkers are promising candidates for future research. Several patterns of biomarker evolution were observed: some increased without decreasing even at 24 h after the marathon; others increased post-marathon and decreased at 24 h while remaining above baseline; some increased after the marathon and then fell below baseline at 24 h. Conclusions: Marathon running causes significant increases in kidney injury biomarkers, with different patterns of evolution. Keywords:marathon; acute kidney injury; biomarkers 1. Introduction Recently, running has gained popularity [1–3]. Long-distance running is one of the forms of running, and can include half marathons, marathons, and ultramarathons. Physi- cal activity is acknowledged as a health-protective factor [2,4]. Nevertheless, long-distance running can induce high stress on different organs and systems [5,6]. Studies have found out that the kidneys can be affected during a marathon. Some runners may experience stage 1 acute kidney injury (AKI) [7,8]. AKI is defined as an unexpected and often temporary Medicina2025,61, 1775 https://doi.org/10.3390/medicina61101775
Medicina2025,61, 1775 2 of 20 decline in kidney function, as indicated by a rise in creatinine or a fall in urine volume [9]. Stage 1 AKI is defined by KDIGO as a rise in serum creatinine by≥0.3 mg/dL within 48 h, or a rise up to 1.5–1.9 times baseline within the prior 7 days, or a decrease in urine excretion <0.5 mL/kg/h for 6–12 h [10]. These episodes are mainly reversible, but when repeated, or maybe in cases with comorbidities or in the presence of other risk factors, a long-term renal injury can happen [11]. It is yet uncertain if recurrent renal insults that fit the criteria of AKI cause an accelerated development of long-term renal problems in long-distance running [12]. In addition to serum creatinine and urine output, there are many biomarkers for AKI. A classification by Oh et al. [13] identifies three main categories: functional biomarkers: serum creatinine, serum cystatin C (sCys C); tubular enzymes; damage biomarkers: urinary Neutrophil Gelatinase-Associated Lipocalin (uNGAL), and serum NGAL (sNGAL), urinary Kidney Injury Molecule-1 (uKIM-1), Interleukin 18, urinary liver-type fatty acid binding protein (uL-FABP); and pre-injury phase biomarkers: urinary Tissue Inhibitor of Metalloproteinases-2 (uTIMP-2) and urinary Insulin-like Growth Factor Binding Protein-7 (uIGFBP-7). Although many observational studies have shown biomarker modifications after a marathon, no systematic synthesis has assessed the biomarkers’ temporal patterns and diagnosis over time. The study aimed to investigate changes in urinary, serum, and plasma biomarkers indicative of acute kidney injury in individuals participating in marathons. The objectives were as follows: to measure changes in renal biomarker levels before, immediately after, and 24 h post-marathon; to identify promising biomarkers for the diagnosis of acute kidney injury; and to describe the temporal patterns of biomarker dynamics in relation to the marathon. 2. Materials and Methods The present manuscript followed the “Preferred Reporting Items for Systematic Re- views and Meta-analyses Protocols (PRISMA)” [14]. 2.1. Eligibility Criteria We included prospective and retrospective observational studies involving (P) indi- viduals who participated in (I) running a marathon, with the primary outcome (O) being the changes in urinary, serum, or plasma biomarkers that may be indicative
2. Materials and Methods The present manuscript followed the “Preferred Reporting Items for Systematic Re- views and Meta-analyses Protocols (PRISMA)” [14]. 2.1. Eligibility Criteria We included prospective and retrospective observational studies involving (P) indi- viduals who participated in (I) running a marathon, with the primary outcome (O) being the changes in urinary, serum, or plasma biomarkers that may be indicative of acute kidney injury (serum creatinine, urinary creatinine, BUN-to-creatinine ratio, serum urea, TIMP-2, IGFBP-7, TIMP-2*IGFBP, urinary L-FABP, urinary NGAL, plasma NGAL, serum cystatin C, plasma KIM-1, urinary KIM-1, plasma TNF-alpha, urinary TNF-alpha, plasma MCP-1, urinary MCP-1, plasma YKL-40, and urinary YKL-40). Secondary outcomes included serum C-reactive protein, copeptin, and serum creatine kinase. Studies involving other types of marathons (e.g., half-marathon, ultramarathon, or other distances), as well as reviews, meta-analyses, editorials, letters to the editor, and conference abstracts, were excluded. 2.2. Information Sources To identify studies that met our selection criteria, we searched four databases: PubMed, EMBASE, Web of Science, and LILACS. Additionally, the reference lists of the selected articles and reviews were examined to identify further relevant studies. 2.3. Search Strategy The search strategy included the following terms: acute kidney injury and marathon, as well as the relevant biomarkers, using MeSH terms, synonyms, singular and plural forms, and abbreviations. The search was conducted from inception until 29 May 2024. No
Medicina2025,61, 1775 3 of 20 language restrictions were applied in the search strategies or in the selection of articles. The complete search strategy for each database is provided in Supplementary Table S1. 2.4. Selection Process An initial semi-automated removal of duplicate studies was performed using Zotero version 7.0.11 (Corporation for Digital Scholarship, Vienna, VA, USA) [15]. Subsequently, titles and abstracts were manually screened to exclude articles that did not meet the eligibility criteria, as well as any remaining duplicates. Full-text versions of the remaining articles were then manually reviewed by two authors (D.-C.L., L.-I.T.), and irrelevant studies, non-eligible article types, and duplicates were excluded. 2.5. Data Collection Process From each selected study, data was manually extracted by several authors (D.-C.L., L.- I.T., O.A.,S,.L.P., and A.I.). The information included study characteristics, country, region, marathon location, study design, age, percentage of female marathon participants (sex), body mass index, the marathon event, previous marathons completed, weekly running distance (km/week), running history (years), AKI criteria, AKI stage 1 reporting, and values of urinary, serum, and plasma biomarkers, and indicators of acute kidney injury. 2.6. Effect Size For each outcome, the mean and standard deviation were extracted. When this data was unavailable, values were estimated from the median and interquartile range using formulas provided in the Cochrane Handbook [16]. The effect size of interest was represented by the mean values of the biomarkers. Measurements were collected at three time points: before the marathon, immediately after the marathon, and 24 h post-marathon. 2.7. Risk of Bias Assessment Several authors assessed the methodological quality (D.-C.L., O.A.,S,.L.P., and A.I.) of the selected articles with the quality assessment tool for before and after (pre–post) studies with no control group from the National Heart, Lung, and Blood Institute (NHLBI) [17]. In addition to the questions in the NHLBI assessment tool, we added 4 supplementary original questions to broaden and clarify the risk of bias assessment: Clear exclusion criteria for the following: alcohol or food abstinence 12 h before baseline measurement; exclusion of comorbidities that could alter measurements; exclusion of medication that could alter measurements; and presenting the use of water,
In addition to the questions in the NHLBI assessment tool, we added 4 supplementary original questions to broaden and clarify the risk of bias assessment: Clear exclusion criteria for the following: alcohol or food abstinence 12 h before baseline measurement; exclusion of comorbidities that could alter measurements; exclusion of medication that could alter measurements; and presenting the use of water, electrolytes, or food during the marathon. 2.8. Synthesis Methods Means and standard deviations were entered into meta-analyses using the meta pack- age in R [18]. Due to clinical heterogeneity among studies, the mean and 95% confidence intervals (CIs) for each variable were calculated using a random-effects model. Results were presented as forest plots. The chi-squared-based Q test and the I 2 statistic were used to assess statistical heterogeneity between studies. Statistical significance was defined as a p-value less than 0.05. All analyses were performed using R statistical software, version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria) [19]. 2.9. Assessment of Publication Bias Publication bias was assessed using Egger’s test. During the preparation of this manuscript, the authors used ChatGPT version 4.0 (OpenAI, San Francisco, CA, USA, accessed in June 2025) as a brainstorming tool to generate suggestions for perspectives to include in the Discussion section. Also, it was
Medicina2025,61, 1775 4 of 20 used to improve the scientific writing of the manuscript. The authors have reviewed and edited the output and take full responsibility for the content of this publication. A protocol for the publication can be found at protocols.io with the ID 225510 (regis- tered on 26 August 2025). 3. Results A total of 119 records were retrieved from the four databases searched, and one additional study was identified through reference screening of the selected articles. The identification and selection process is illustrated in Figure. After removing 37 duplicates, 83 records were screened based on title and abstract. Of these, 68 were excluded for not meeting the eligibility criteria or being duplicates. The full texts of the remaining 15 articles were then assessed, resulting in the inclusion of 9 studies in the final review and meta-analysis. Figure 1.Flow chart illustrating the identification, selection, and inclusion of articles in the review. 3.1. Study Characteristics The characteristics of the included studies are detailed in Tables. Five studies were conducted in Europe, and four in America. All studies employed a prospective
Medicina2025,61, 1775 5 of 20 cohort design. Acute kidney injury among marathon runners was diagnosed based on the RIFLE, AKIN, or KDIGO criteria. The mean age of participants varied across studies, with several reporting a mean age of around 40 years. However, one study included younger participants with a mean age of 23 years, while another reported a higher mean age of 50 years. Two studies included only male participants. In five studies, the proportions of male and female participants were approximately equal, while in two others, the percentage of female participants was 18% and 29%, respectively. Previous marathon experience, weekly running distance, and running history (in years) also varied across the studies. Table 1.Characteristics of the studies included in the review. Study Country Region Study Design Marathon Place Mean Age (Years) Females (%) BMI (kg/m 2 ) Leckie, 2023 [20] United Kingdom Europe prospective cohort Brighton 2019 41 ±10 18 NR Kosaki, 2022 [21] Japan Europe prospective cohort Tsukuba 2018 23 ±1 0 21.3 ±1.5 Atkins, 2021 [22] USA America prospective cohort Boston 2019 46 ±10 49 Nescolarde, 2020 [23] Spain Europe prospective cohort Barcelona 2017 41 ±4 0 24.0 ±2.1 Mansour, 2019 [8] USA America prospective cohort Hartford 2017 37 (35–44) 57 24 (22–25) Mansour, 2017 [7] USA America prospective cohort Hartford 2015 44.2±12.9 59 22.4 ±2.4 McCullough, 2011 [24] USA America prospective cohort Detroit 2008 38.7±9.0 52 23.0 ±2.6 Bekos, 2016 [25] Austria Europe prospective cohort Vienna 2012 36.83±7.56 29 22.29 ±2.16 Hewing, 2015 [26] Germany Europe prospective cohort Berlin 2006, 2007 50.3 [22–72] 53 22.4 ±2.1 Data are presented as mean±standard deviation or median (interquartile range) [minimum–maximum]; NR, not reported; BMI, body mass index. Table 2.Characteristics of the studies included in the review (continued). Study Previous Marathons (Number) km/Week Running History (Years) AKI Definition Criteria AKI Stage I * Measured Parameters Leckie, 2023 [20] 5±7 [0–39] 43±17 [15–72] NR KDIGO Yes sCr, uCr, TIMP-2, IGFBP-7, TIMP-2*IGFBP-7 Kosaki, 2022 [21] NR NR NR AKIN Yes sCr, L-FABP Atkins, 2021 [22] NR NR NR NR No sCr, uCr, sCys C, uNGAL Nescolarde, 2020 [23] NR NR 8.2 ±5.1
Marathons (Number) km/Week Running History (Years) AKI Definition Criteria AKI Stage I * Measured Parameters Leckie, 2023 [20] 5±7 [0–39] 43±17 [15–72] NR KDIGO Yes sCr, uCr, TIMP-2, IGFBP-7, TIMP-2*IGFBP-7 Kosaki, 2022 [21] NR NR NR AKIN Yes sCr, L-FABP Atkins, 2021 [22] NR NR NR NR No sCr, uCr, sCys C, uNGAL Nescolarde, 2020 [23] NR NR 8.2 ±5.1 AKIN Yes sCr, sCK, Uree serică, sPCR Mansour, 2019 [8] 3 (1–9) 47 (27–58) 9 (5, 12) AKIN Yes sCr, BUN/Cr, sCK, uNGAL, pNGAL, uKIM-1, pKIM-1, copeptină, TNF-alpha plasmatic, MCP-1 plasmatic, MCP-1 urinar, YKL-40 plasmatic, YKL-40 urinar
Medicina2025,61, 1775 6 of 20 Table 2.Cont. Study Previous Marathons (Number) km/Week Running History (Years) AKI Definition Criteria AKI Stage I * Measured Parameters Mansour, 2017 [7] 5 (2–16) 51 ±16 12.0 (5.0–15.0) AKIN Yes sCr, sCK, uNGAL, uKIM-1, TNF-alpha urinar, MCP-1 urinar, YKL-40 urinar McCullough, 2011 [24] 2.3±3.0 27 ±19 AKIN Yes sCr, sCys C, pNGAL, pKIM-1, sCK Bekos, 2016 [25] NR 61.85 ±20.58 NR AKIN Yes sCK, sCRP Hewing, 2015 [26] 6.0 [3.0–13.0] 50.0 [40.0–65.0] 10.0 [6.0–20.0] RIFLE Yes sCr, sCys C, sCRP Data are presented as mean±standard deviation or median (interquartile range) [minimum–maximum]; NR, not reported; AKI, acute kidney injury; KDIGO, Kidney Disease: Improving Global Outcomes; AKIN, Acute Kidney Injury Network; RIFLE, Risk, Injury, Failure, Loss, and End-stage Kidney Disease; sCr, serum creatinine; uCr, urinary creatinine; TIMP-2, Tissue Inhibitor of Metalloproteinases-2; IGFBP-7, Insulin-like Growth Factor Binding Protein-7; L-FABP, liver-type fatty acid binding protein; sCK, serum creatine kinase; sCys C, serum cystatin C; uNGAL, urinary Neutrophil Gelatinase-Associated Lipocalin; BUN/Cr, blood urea nitrogen/creatinine ratio; pNGAL, plasma Neutrophil Gelatinase-Associated Lipocalin; uKIM-1, urinary Kidney Injury Molecule-1; pKIM-1, plasma Kidney Injury Molecule-1; sCRP, serum C-reactive protein; TNF-alpha, tumor necrosis factor alpha; YKL-40, chitinase 3-like protein 1; MCP-1, monocyte chemoattractant protein-1; *, studies that reported the number of subjects with stage 1 AKI. 3.2. Most Frequently Reported Biomarkers Serum creatinine was the most frequently reported biomarker. Its mean value before the marathon was 0.88 mg/dL (95% CI: 0.85–0.92), it increased immediately after the marathon to 1.38 mg/dL (95% CI: 1.04–1.71), and decreased at 24 h post-marathon to 1.12 mg/dL (95% CI: 0.64–1.61) (Figure). Heterogeneity of the results was substantial, with I 2 inconsistency indices of 85%, 98%, and 98% for the three respective time points, all statistically significant. Another frequently reported biomarker was serum creatine kinase. The mean value before the marathon was 138.31 U/L (95% CI: 105.12–171.5), which increased threefold immediately after the marathon to 425.8 U/L (95% CI: 290.88–560.71), and rose even further at 24 h post-marathon—approximately an eightfold increase compared to baseline—to 1128.14 U/L (95% CI: 434.02–1822.26) (Figure). Heterogeneity was substantial, with I 2 inconsistency values of 75%, 89%, and
serum creatine kinase. The mean value before the marathon was 138.31 U/L (95% CI: 105.12–171.5), which increased threefold immediately after the marathon to 425.8 U/L (95% CI: 290.88–560.71), and rose even further at 24 h post-marathon—approximately an eightfold increase compared to baseline—to 1128.14 U/L (95% CI: 434.02–1822.26) (Figure). Heterogeneity was substantial, with I 2 inconsistency values of 75%, 89%, and 84% for the three respective time points, all statistically significant. The most frequently reported urinary biomarker was urinary Neutrophil Gelatinase- Associated Lipocalin (uNGAL). The mean value before the marathon was 8.92 ng/mL (95% CI: 6.11–11.73), which increased substantially—approximately fourfold—immediately after the marathon to 41.30 ng/mL (95% CI: 31.74–50.87), and decreased to 29.57 ng/mL (95% CI: –0.81 to 59.94) at 24 h post-marathon, remaining approximately three times higher than baseline (Figure). Heterogeneity was moderate before and immediately after the marathon (I 2 = 47% and 33%, respectively), and considerable at 24 h post-marathon (I 2 = 99%), with only the latter being statistically significant. Another reported urinary biomarker was urinary Kidney Injury Molecule-1 (uKIM- 1). The mean value before the marathon was 755.02 pg/mL (95% CI: –474.24 to 1984.29), which increased significantly—approximately threefold—immediately after the marathon to 2255.64 pg/mL (95% CI: 827.69 to 3683.6), and then decreased at 24 h post-marathon to 1630.76 pg/mL (95% CI: –367.39 to 3628.9), remaining about twice as high as baseline (Figure). Heterogeneity of the results was considerable at all three time points, with I 2 values of 97%, 93%, and 96%, respectively, all statistically significant.
Medicina2025,61, 1775 7 of 20 Figure 2.Forest plots of the mean serum creatinine values before the marathon, immediately after the marathon, and 24 h after the marathon. Before marathon [7,8,20,21,23,24,26], after marathon [7,8,20–24,26], 24 h after marathon [7,8,20–22,24]. Another reported biomarker was serum cystatin C. The mean value before the marathon was 0.39 mg/L (95% CI: –0.21 to 0.99), followed by a slight decrease immediately after the marathon to 0.36 mg/L (95% CI: –0.11 to 0.83), and a further, more pronounced decrease at 24 h post-marathon to 0.10 mg/L (95% CI: 0.06–0.15), representing roughly one-quarter of the baseline level (Figure). Heterogeneity was considerable at all three time points, with I 2 values of 100%, 100%, and 99%, respectively—all statistically significant. Another reported biomarker was C-reactive protein (CRP). The mean value before the marathon was 0.93 mg/L (95% CI: 0.03–1.82), which decreased immediately after the marathon to 0.59 mg/L (95% CI: 0.08–1.10) (Figure). Heterogeneity of the results was considerable for both measurements, with I 2 values of 96% and 95%, respectively, both statistically significant. 3.3. Other Biomarkers The mean urinary creatinine level before the marathon was 134.28 mg/dL (95% CI: 4.9–263.66) (Supplementary Table S2). It nearly doubled immediately after the marathon, reaching 283.09 mg/dL (95% CI: 127.91–438.27) (Supplementary Table S3), and then de- creased at 24 h post-marathon to 103.74 mg/dL (95% CI: 79.5–127.97), falling below baseline levels (Supplementary Table S4).
Description
This systematic review analyzes acute kidney injury biomarkers in marathon runners.