Secondary Logo

Journal Logo

Cytokine and Sclerostin Response to High-Intensity Interval Running versus Cycling


Medicine & Science in Sports & Exercise: December 2019 - Volume 51 - Issue 12 - p 2458–2464
doi: 10.1249/MSS.0000000000002076

Purpose This study examined whether the exercise-induced changes in inflammatory cytokines differ between impact and no-impact high-intensity interval exercise, and whether they are associated with postexercise changes in sclerostin.

Methods Thirty-eight females (n = 19, 22.6 ± 2.7 yr) and males (n = 19, 22.3 ± 2.4 yr) performed two high-intensity interval exercise trials in random order (crossover design): running on a treadmill and cycling on a cycle ergometer. Trials consisted of eight repetitions of 1 min running or cycling at ≥90% maximal heart rate, separated by 1 min passive recovery intervals. Blood was collected preexercise and 5 min, 1 h, 24 h, and 48 h postexercise, and it was analyzed for serum levels of interleukins (IL-1β, IL-6, and IL-10), tumor necrosis factor alpha (TNF-α), and sclerostin.

Results Inflammatory cytokines significantly increased over time in both sexes with some differences between trials. Specifically, IL-1β significantly increased from pre- to 5 min after both trials (23%, P < 0.05), IL-6 increased 1 h after both trials (39%, P < 0.05), IL-10 was elevated 5 min after running (20%, P < 0.05) and 1 h after both running and cycling (41% and 64%, respectively, P < 0.05), and TNF-α increased 5 min after running (10%, P < 0.05). Sclerostin increased 5 min after both trials, with a greater increase in males than that in females (62 vs 32 pg·mL−1 in running, P = 0.018; 63 vs 30 pg·mL−1 in cycling, P = 0.004). In addition, sclerostin was significantly correlated with the corresponding changes in inflammatory cytokines, and 34% of the variance in its postexercise gain score (Δ) was explained by sex and the corresponding gain scores in TNF-α, which was the strongest predictor.

Conclusion A single bout of either impact or no-impact high-intensity exercise induces changes in inflammatory cytokines, which are associated with the postexercise increase in sclerostin.

1Faculty of Applied Health Sciences, Department of Kinesiology, Brock University, St. Catharines, ON, CANADA

2Faculty of Applied Health Sciences, Centre for Bone and Muscle Health, Brock University, St. Catharines, ON, CANADA

3Faculty of Health, School of Kinesiology and Health Science, York University, Toronto, ON, CANADA

Address for correspondence: Panagiota Klentrou, Ph.D., Faculty of Applied Health Sciences, Department of Kinesiology, Brock University, 1812 Sir Isaac Brock Way, St. Catharines, ON, Canada L2S 3A1; E-mail:

Submitted for publication March 2019.

Accepted for publication June 2019.

Online date: June 26, 2019

The inflammatory response to exercise has been examined in numerous studies using different modes of exercise including cycling (1,2) and running (3,4). In fact, transient increases in both pro- and anti-inflammatory cytokines have been consistently shown after strenuous exercise in many studies (5–7). However, the response of inflammatory cytokines to exercise varies across studies depending on the mode, duration and intensity of exercise (6,8).

The role of inflammatory cytokines in bone metabolism has been highlighted in several studies (9–11). Pro-inflammatory cytokines such as tumor necrosis factor alpha (TNF-α) and interleukins 1β and 6 (IL-1β, IL-6) can affect bone remodeling through regulation of the activation of osteoclasts and osteoblasts (9–11). In particular, pro-inflammatory cytokines can negatively affect osteoblast activation via upregulation of sclerostin (9). Sclerostin, mainly produced by osteocytes, is a key osteokine that downregulates bone formation via inhibition of the Wnt-β catenin signaling pathway in osteoblasts (12,13). Specifically, TNF-α has been shown to increase sclerostin’s expression in vitro (14,15) and in vivo (animal) (16,17) studies. In humans, the relationship between circulating cytokines and sclerostin, either at rest or in response to exercise, has not been examined. However, sclerostin has consistently been shown to increase immediately following exercise (18–21). In the specific population of the current study, we recently demonstrated that sclerostin increases transiently following high-intensity interval exercise and that this increase was independent of impact and not directly related to changes in bone turnover markers (20,21). On the other hand, we previously reported parallel fluctuations in circulating sclerostin and both IL-6 and TNF-α during a 42-wk high-performance training cycle in elite female rowers (22). It is not clear whether these parallel changes also occur in response to a single exercise session.

Therefore, the purpose of this study was to examine the exercise-induced changes in inflammatory cytokines and sclerostin in young female and male adults after a high-intensity exercise session. Specifically, we examined whether exercise-induced cytokine changes (a) differ between impact and no-impact high-intensity interval exercise and (b) are associated with corresponding postexercise changes in circulating sclerostin. It was hypothesized that all inflammatory cytokines will increase postexercise, with and without impact, and that this response will be associated with corresponding increases in serum sclerostin.

Back to Top | Article Outline



This study involves further analysis of blood samples previously collected and analyzed, and sclerostin levels have been previously reported for females (20) and males (21). To examine the inflammatory response to the two exercise protocols, and their association with sclerostin, in both sexes, blood samples from 38 participants (19 females and 19 males) were further analyzed for IL-1β, IL-6, IL-10, and TNF-α. One female and one male participant from the original sample were excluded from the post hoc analysis as outliers because of their extremely high inflammatory cytokine levels.

Both studies included participants who signed the consent form and were 18–28 yr old, healthy, recreationally active (i.e., exercising 2 to 5 times per week), free of injuries or chronic conditions (e.g., knee/hip/lower back injuries, arthritis, neuromuscular diseases), having no fractures in the last year, nonsmokers, and not taking any medication or dietary supplements affecting bone health (e.g., protein, vitamin D, and calcium). Specifically, among the 38 participants included in the present analysis, females and males were of similar age (22.6 ± 2.7 vs 22.3 ± 2.4 yr, respectively; P = 0.78) and had similar leisure-time physical activity scores (56.5 ± 35.6 vs 53.6 ± 27.3, respectively; P = 0.78), energy intake (30.3 ± 8.1 vs 29.2 ± 13.7 kcal·kg−1·d−1, respectively; P = 0.77), protein intake (1.1 ± 0.3 vs 1.2 ± 0.7 g·kg−1·d−1, respectively; P = 0.61), and calcium intake (968 ± 286 vs 1148 ± 553 mg·d−1, respectively; P = 0.22). In terms of physical characteristics, females were of smaller stature (156 ± 37.9 vs 176 ± 8.2 cm, respectively; P < 0.05) and body mass (59.2 ± 9.2 vs 77.2 ± 12.3 kg, respectively; P < 0.001) but had higher relative body fat (27.1% ± 7.2% vs 19.4% ± 8.0%, respectively; P < 0.05). In addition, all female participants were on birth control. Both studies were conducted in accordance with the Declaration of Helsinki and were cleared by our institutional Research Ethics Board.

Back to Top | Article Outline

Experimental design and procedures

Both studies were conducted during the same season of 1 yr, in the same laboratory by the same investigators. As previously described, both studies used a crossover, within-subject design, where each participant performed two high-intensity exercise trials in random order: a high-intensity interval running (HIIR) trial on the treadmill and a high-intensity interval cycling (HIIC) trial on a cycle ergometer (20,21). Briefly, participants came to the laboratory for two preliminary visits, scheduled 1–3 d apart, where they were informed about the study and signed the consent form, completed the medical history questionnaire, and had their anthropometric and body composition measurements taken. Subsequently, they performed either a cycling or a running incremental exercise test to exhaustion in random order (one exercise test in each visit). The running test on the treadmill, started with a 4 min warm-up (incline: 2%, speed 4–6 mph for females and 6–8 mph for males based on their fitness level). The test proceeded with speed increments of 1.6 km·h−1 (1 mph) every minute up to ~12.9–14.5 km·h−1 (8–9 mph) for females and up to ~14.5–16 km·h−1 (9–10 mph) for males, followed by 1% incline increments every minute until exhaustion. The test finished with a 3-min cooldown similar to warm-up. The cycling test on the cycle ergometer begun with a 4-min warm-up (60–80 W for females and 80–100 W for males based on their fitness level), proceeded with power increments of 30 W every minute to exhaustion, and ended with a 3-min cooldown. These tests were used to assess participants’ maximal workload, i.e., maximal speed and incline (running) or watts (cycling) at volitional fatigue, when participants could no longer continue pedaling or running. The maximal heart rate and the perceived exertion (Borg scale) at the point of exhaustion were also recorded.

In the subsequent two visits, participants performed the two high-intensity exercise trials (HIIR and HIIC), where five blood samples were collected: preexercise and 5 min, 1 h, 24 h, and 48 h postexercise. All visits were scheduled in the morning between 1000 and 1200 h to control for diurnal variation in the biochemical markers. For consistency, before every visit to the laboratory, participants were instructed to consume the same standardized breakfast at home, which included one slice of whole grain bread with butter/margarine or peanut butter, one glass of 2% milk or one cup of 2% fat yogurt, one banana or apple, and one cup of coffee or tea. Likewise, they were instructed not to eat or drink (except water) for about 2 h after their breakfast and before their laboratory visits. In addition, participants were asked to avoid alcohol and exercise for 24 h before their two preliminary visits and 48 h prior and after the exercise trials. All the visits took place within three consecutive weeks (after the week of menstruation for females).

Back to Top | Article Outline

High-intensity interval exercise trials

For each participant, the two high-intensity interval exercise trials (HIIR and HIIC) were randomly assigned in a crossover manner and scheduled 1 wk apart. During each trial, participants performed eight repetitions of 1 min cycling or running separated by 1 min passive recovery intervals. The workload for the HIIR intervals was set using the maximum speed and incline achieved in the incremental running test. Similarly, the workload for the HIIC intervals was set using the maximum watts achieved in the incremental cycling test. To ensure high intensity, heart rate was recorded at the end of each interval. Mean heart rate and percentage of maximum heart rate for the eight intervals were subsequently calculated for each participant. During both trials, participants’ mean heart rate was >90% of maximum heart rate (92.3% ± 0.86% for HIIR and 90% ± 4.84% for HIIC). Borg RPE was recorded after each interval in both trials, with 19 and 20 being the mode values in HIIR and HIIC trials respectively.

Back to Top | Article Outline

Baseline measurements: anthropometrics, body composition, and questionnaires

Height was measured with a stadiometer to the nearest 0.1 cm with no shoes. Body composition was measured via air displacement plethysmography (BodPod; Life Measurement, Inc., Concord, CA) to get measures of body mass (kg), fat mass (kg), fat-free mass (kg), and percent body fat (%).

The Godin Shephard Leisure-Time Physical Activity Questionnaire was used to determine participants’ habitual physical activity levels by calculating their leisure score index. A food frequency questionnaire (Block 2014.1_6Mo, Nutrition Quest, Berkeley, CA) was used to assess habitual nutrient intake.

Back to Top | Article Outline

Biochemical analysis

Venous blood samples were collected from the median cubital vein in the antecubital fossa of each participant using a standard venipuncture technique. In each high-intensity interval exercise trial, approximately 10 mL of whole blood was collected from each participant at each time point (preexercise, 5 min, 1 h, 24 h, and 48 h postexercise) for a total of 10 blood draws per participant. All blood samples sat for 30 min at room temperature before being centrifuged at 3000g and 4°C for 15 min in a benchtop centrifuge (Allegra ZIR centrifuge; Beckman Coulter, Brea, CA). Serum was then aliquoted into microcentrifuge tubes and stored at −80°C until analysis. Plasma was also used to measure hematocrit, which was measured in triplicate by the same investigator, using microhematocrit tubes with heparin (VWR, Radnor, PA) for each blood sample, and was separated using an international micro capillary centrifuge (model MB; International Equipment Company, Needham, MA).

Serum levels of sclerostin were measured in duplicate using commercially available immunoassay (ELISA) kits (SCL, cat. no. DSST00; R&D Systems, Inc., Minneapolis, MN). Serum levels of inflammatory cytokines (IL-1β, IL-6, IL-10, and TNF-α) were also measured in duplicate using multiplex kits (human high-sensitivity T-cell magnetic bead panel, cat. no. HSTCMAG-28SK; EMD Millipore, Darmstadt, Germany) following the manual instructions. The average values of in-house inter- and intra-assay coefficients of variation (CV) for sclerostin were 8.2% and 3.7%, respectively, and for the four inflammatory cytokines 9% and 8.9%, respectively.

Further, serum levels of sclerostin and inflammatory cytokines at 5 min, 1 h, 24 h, and 48 h postexercise in both trials were corrected for exercise-induced plasma volume changes using the following formula: 100 + %ΔPV/100, where %ΔPV from pre- to postexercise was calculated for each participant using the following formula of Van Beaumont (23):

For both men and women, the mean percent change in plasma volume was highest at 5 min postexercise (HIIR: −3.82% in females and −3.76% in males; HIIC: −6.68% in females and −9.97% in males).

Back to Top | Article Outline

Statistical analysis

From a total of 380 blood samples used in this study (38 participants × 10 sampling times), there were two missing samples due to one (male) participant’s absence from the 24- and the 48-h post-HIIR blood draws. Missing values were replaced with the mean value for males at the corresponding time point. Data were then assessed for normality using the Shapiro–Wilk test, z-scores for skewness and kurtosis, and visual screening of histograms for symmetry. Log-transformations followed in cases of violations of normality, which was the case for all inflammatory cytokines. Independent t-tests were used to examine sex differences in age, anthropometrics, nutritional intake, and physical activity level.

A three-way repeated-measures ANOVA was used to examine the main effects of time and exercise mode (repeated-measures factors), and sex (between-group factor), as well as any interactions. In case of a significant time effect without other effects or interactions, pairwise comparisons with Bonferroni adjustment were performed (exercise modes and sexes combined). In the event of multiple significant main effects or a significant interaction, differences between means were assessed using either two-way and/or one-way ANOVA.

Pearson correlations were used to examine the relation between the exercise-induced changes (from pre- to 5 min postexercise) in each inflammatory cytokine and the corresponding changes in sclerostin (log-transformed gain (Δ) scores were used for all the variables). Multiple linear regression analysis was then used to explain the variance in sclerostin’s response to exercise with the log-transformed gain (Δ) score of sclerostin as the dependent variable and the corresponding log-transformed gain (Δ) scores of inflammatory cytokines (IL-1β, IL-6, IL-10, and TNF-α) and sex as the independent variables. All the multiple linear regression assumptions were checked and met.

Effect sizes were determined by the partial eta squared (ηp2) for ANOVA and were interpreted based on the Cohen criteria for partial η2: 0.01 = small, 0.06 = moderate, and 0.14 = large effect (24,25). Statistical significance was set at an alpha level of 0.05 and performed using IBM SPSS Statistics 24 (SPSS Inc., Chicago, IL).

Back to Top | Article Outline


In the case of IL-1β, the three-way ANOVA with repeated-measures showed a significant main effect for time, which reflects a significant increase (23%) in IL-1β 5 min postexercise for both sexes and exercise modes combined (Fig. 1). There was also a significant exercise mode–sex interaction. To further investigate this interaction, we ran a series of post hoc tests to determine differences between the following means: females during HIIR (collapsed across time points), females during HIIC (collapsed across time points), males during HIIR (collapsed across time points), and males during HIIC (collapsed across time points); however, no significant differences were found. By visual inspection of Figure 1, it appears that the interaction reflects higher IL-1β during running in females and higher IL-1β during cycling in males.



A significant main effect for time was also found for IL-6 with no significant effect for exercise mode or sex and no interactions (Fig. 2). Further, pairwise comparisons with exercise modes and sexes combined revealed a significant increase (39%) from pre- to 1 h postexercise. Additional significant differences were also observed between 5 min and 1 h, as well as between 1 h and 48 h postexercise in both trials and sexes combined (Fig. 2).



IL-10 showed a significant time–exercise mode interaction, but no main effect or interaction for sex (Fig. 3). Further, post hoc analysis with sexes combined showed that in the running trial, IL-10 levels increased significantly from pre- to 5 min and 1 h postexercise (20% and 41%, respectively) and returned to near baseline levels 24 h later. In the cycling trial, IL-10 levels were significantly elevated 1 h postexercise (64% for females and males combined). In fact, the 1 h post-HIIC concentration of IL-10 was significantly higher compared with the preexercise, 5 min, 24 h, and 48 h post-HIIC concentrations (Fig. 3).



In the case of TNF-α, we found a significant time–exercise mode interaction with no effect of sex and no interactions with sex (Fig. 4). Further, post hoc analysis with sexes combined showed a significant increase (10%) in TNF-α from pre- to 5 min post-HIIR, which subsequently decreased 1 and 24 h after HIIR. These changes, however, were not significant in HIIC (Fig. 4).



Sclerostin showed a significant time–sex interaction, but no effect of exercise mode and no interactions with exercise mode (Fig. 5). Specifically, sclerostin levels increased significantly from pre- to 5 min postexercise and returned to near baseline levels 1 h postexercise in both females and males (Fig. 5). In addition, the increase in sclerostin 5 min postexercise was significantly higher in males than females in both trials (62 vs 32 pg·mL−1 in HIIR, P = 0.018, and 63 vs 30 pg·mL−1 in HIIC, P = 0.004, for males vs females, respectively). Males had significantly higher circulating levels of sclerostin than women in all time points in both HIIR and HIIC trials.



There were significant correlations between sclerostin’s change from pre- to 5 min postexercise and the corresponding changes in the inflammatory cytokines (r = 0.487, P < 0.001, for sclerostin and TNF-α; r = 0.293, P = 0.010, for sclerostin and IL-1β; r = 0.447, P < 0.001, for sclerostin and IL-6; r = 0.404, P < 0.001, for sclerostin and IL-10). Multiple linear regression analysis showed that sclerostin’s significant gain (Δ) score from pre- to 5 min postexercise was explained by sex and the gain scores of TNF-α (adjusted R2 = 0.337, P < 0.05). Specifically, TNF-α was the strongest contributor to explain sclerostin’s change as shown by its highest beta coefficient, whereas IL-1β and IL-10 were not significant predictors. IL-6 was excluded from the model due to its low tolerance (0.015), and thus, severe multicollinearity (Table 1).



Back to Top | Article Outline


This study is a novel examination of the response of inflammatory cytokines and sclerostin to both impact and no-impact high-intensity interval exercise in females and males. The results demonstrated that in general, inflammatory cytokines and sclerostin significantly increased postexercise in both trials. The study also provides evidence that although sclerostin’s response to high-intensity exercise was independent of impact, there was a difference in the timing of the IL-10 response between trials and a greater TNF-α increase 5 min after running compared with cycling. In addition, sclerostin was higher in males than that in females across all time points and was significantly correlated with the corresponding changes in inflammatory cytokines, especially TNF-α, which along with sex, explained 34% of the variance in its postexercise elevation.

All inflammatory cytokines significantly increased over time (5 min and 1 h postexercise), with some differences between trials and sexes. Specifically, the response in pro-inflammatory cytokines was more evident 5 min after the running trial, as evidenced by a 23% overall significant increase in IL-1β for both sexes and trials combined and a similar 10% increase in TNF-α only post-HIIR in both sexes. On the other hand, IL-6 significantly increased by 39% 1 h after both trials and in both sexes. In addition, for both sexes combined, the anti-inflammatory IL-10 significantly increased 5 min after running (20%) and was significantly elevated 1 h postexercise in both trials (41% in HIIR and 64% in HIIC).

A previous study from our laboratory (2) also found a significant increase in serum levels of IL-1β, IL-6, and TNF-α, 5 min after HIIC in males, but with no changes in IL-10. Another previous study (3) also found a significant increase in plasma IL-6 levels after continuous treadmill running (45 min at 95% of anaerobic threshold and 7 min at 90% V˙O2max). Thus, the increase in inflammatory cytokines postexercise is consistently shown in human studies, but with some differences in the magnitude of change among the studies. These differences in the magnitude of change can be attributed to either exercise-related (i.e., mode, intensity, and duration) or analysis-related factors. For example, the difference in media used (plasma or serum) can account for some discrepancy in the inflammatory cytokine levels, as the latter can be affected by the anticoagulant used for the collection and process of biological samples from blood (26,27). In general, significant differences between human plasma and serum metabolite (i.e., proteins) profiles have been found (27). Further, in contrast to our study, most previous studies (2,3) have not corrected for plasma volume changes; thus, part of the changes in inflammatory cytokines might be attributed to hemoconcentration rather than exercise per se. Even without adjusting for plasma volume changes, our relative exercise-induced increases in cytokines were still lower than the previous studies. Another possible reason for the differences in the exercise-induced relative increase in inflammatory cytokines is the variety and sensitivity of assay kits used for the analysis.

Sclerostin increased 5 min postexercise to a similar degree after high-impact running and no-impact cycling (42% and 40.7%), although this increase was significantly higher in males than females (48% vs 36% in HIIR, and 50% vs 31.3% in HIIC, males vs females, respectfully). In fact, males had consistently higher sclerostin levels than females across time in both trials. Previous studies have demonstrated higher resting sclerostin levels than females, possibly because of their lower estrogen concentrations (28,29). However, none of the previous studies have examined whether there is a difference between sexes in the response of sclerostin to exercise. Sclerostin’s increase 5 min after both impact and no-impact high-intensity exercise is in agreement with findings previously reported from our laboratoryand others (18,19). In particular, because we corrected for exercise-induced plasma volume changes, the increase of sclerostin 5 min postexercise seems to be truly induced by exercise. It is possible, however, that part of this increase might be due to the release of previously synthesized sclerostin from osteocytes into the blood combined with the exercise-related increase in peripheral blood flow, rather than a direct increase in sclerostin’s gene expression in this short period (19).

The present study is the first to report an association between inflammatory cytokines and sclerostin’s response to one single bout of exercise, providing evidence that 34% of sclerostin’s postexercise increase can be explained by the corresponding change in TNF-α and sex (medium to large effect size (25)). It is interesting to note that TNF-α, which was the strongest predictor of the exercise-induced elevation in sclerostin, only increased after running and not after cycling, but this difference was not sufficient for a differential response in sclerostin. Only one previous study, examining the effects of 42 wk of training in elite female rowers, reported parallel changes in sclerostin and both IL-6 and TNF-α (22). This aligns with previous in vitro (14,15) and in vivo (animal) (16,17) studies, which have shown TNF-α causing an increase in sclerostin’s expression in osteocytes. Nevertheless, systemic changes in inflammatory cytokines seem to explain a significant part of sclerostin’s behavior during exercise of high-intensity warranting further examination in terms of the amount and type of exercise.

The crossover experimental design is one of the main strengths of this study. Because each participant was their own control by participating in both HIIR and HIIC trials, there was reduction in interindividual variability. The correction of exercise-induced plasma volume changes to all inflammatory cytokines and sclerostin responses is another strength of this study. Strenuous exercise can induce plasma volume changes that can cloud the interpretation of the observed responses of biochemical measurements in blood (30,31). However, when interpreting these results, one should also keep in mind the limitation of the accuracy of sclerostin measurements using currently available commercial ELISA kits, which has been previously questioned (32).

In conclusion, a single bout of either impact or no-impact high-intensity exercise induces changes in both pro- and anti-inflammatory inflammatory cytokines, which are associated with the postexercise increase in sclerostin, and with TNF-α significantly explaining part of the variance in the exercise-induced increase in sclerostin.

This study was funded by a National Science Engineer Research Council of Canada (NSERC) grant to P. Klentrou (grant no. 2015-04424). R. Kouvelioti holds an Ontario Trillium Scholarship. N. Kurgan holds an NSERC Doctoral Scholarship. W. Wards holds a Canada Research Chair in Bone and Muscle Development. The authors thank all the participants for participating in our study, all the volunteer undergraduates (D. Brown, R. Sweeney, D. Szkaradek, M. Nasato, and S. Pilakka), the phlebotomists (especially C. Watt), the laboratory coordinator (R. Dotan), and the laboratory technician (J. Gabrie) for their assistance with different parts of the study.

The authors have no conflict of interest to declare. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Steinberg JG, Ba A, Brégeon F, Delliaux S, Jammes Y. Cytokine and oxidative responses to maximal cycling exercise in sedentary subjects. Med Sci Sports Exerc. 2007;39(6):964–8.
2. Mezil YA, Allison D, Kish K, et al. Response of bone turnover markers and cytokines to high-intensity low-impact exercise. Med Sci Sports Exerc. 2015;47(7):1495–502.
3. Almada C, Cataldo LR, Smalley SV, et al. Plasma levels of interleukin-6 and interleukin-18 after an acute physical exercise: relation with post-exercise energy intake in twins. J Physiol Biochem. 2013;69(1):85–95.
4. Ostrowski K, Rohde T, Asp S, Schjerling P, Pedersen BK. Pro- and anti-inflammatory cytokine balance in strenuous exercise in humans. J Physiol. 1999;515(Pt 1):287–91.
5. Kasapis C, Thompson PD. The effects of physical activity on serum C-reactive protein and inflammatory markers. J Am Coll Cardiol. 2005;45(10):1563–9.
6. Shek PN, Shephard RJ. Physical exercise as a human model of limited inflammatory response. Can J Physiol Pharmacol. 1998;76(5):589–97.
7. Pedersen BK, Ostrowski K, Rohde T, Bruunsgaard H. The cytokine response to strenuous exercise. Can J Physiol Pharmacol. 1998;76(5):505–11.
8. Brown W, Davison GW, McClean CM, Murphy MH. A systematic review of the acute effects of exercise on immune and inflammatory indices in untrained adults. Sport Med Open. 2015;1(1):35.
9. Redlich K, Smolen JS. Inflammatory bone loss: pathogenesis and therapeutic intervention. Nat Rev Drug Discov. 2012;11(3):234–50.
10. Manolagas SC, Jilka RL. Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N Engl J Med. 1995;332(5):305–11.
11. Clarke B. Normal bone anatomy and physiology. Clin J Am Soc Nephrol. 2008;3(3 Suppl):S131–9.
12. Sapir-Koren R, Livshits G. Osteocyte control of bone remodeling: is sclerostin a key molecular coordinator of the balanced bone resorption–formation cycles? Osteoporos Int. 2014;25(12):2685–700.
13. Lin C, Jiang X, Dai Z, et al. Sclerostin mediates bone response to mechanical unloading through antagonizing Wnt/β-catenin signaling. J Bone Miner Res. 2009;24(10):1651–61.
14. Vincent C, Findlay DM, Welldon KJ, et al. Pro-inflammatory cytokines TNF-related weak inducer of apoptosis (TWEAK) and TNFalpha induce the mitogen-activated protein kinase (MAPK)-dependent expression of sclerostin in human osteoblasts. J Bone Miner Res. 2009;24(8):1434–49.
15. Kang J, Boonanantanasarn K, Baek K, et al. Hyperglycemia increases the expression levels of sclerostin in a reactive oxygen species- and tumor necrosis factor-alpha-dependent manner. J Periodontal Implant Sci. 2015;45(3):101.
16. Kim BJ, Bae SJ, Lee SY, et al. TNF-α mediates the stimulation of sclerostin expression in an estrogen-deficient condition. Biochem Biophys Res Commun. 2012;424(1):170–5.
17. Baek K, Hwang HR, Park H-J, et al. TNF-α upregulates sclerostin expression in obese mice fed a high-fat diet. J Cell Physiol. 2014;229(5):640–50.
18. Falk B, Haddad F, Klentrou P, et al. Differential sclerostin and parathyroid hormone response to exercise in boys and men. Osteoporos Int. 2016;27(3):1245–9.
19. Pickering M-E, Simon M, Sornay-Rendu E, et al. Serum sclerostin increases after acute physical activity. Calcif Tissue Int. 2017;101(2):170–3.
20. Kouvelioti R, Kurgan N, Falk B, Ward WE, Josse AR, Klentrou P. Response of sclerostin and bone turnover markers to high intensity interval exercise in young women: does impact matter? Biomed Res Int. 2018;2018:4864952.
21. Kouvelioti R, LeBlanc P, Falk B, Ward WE, Josse AR, Klentrou P. Effects of high intensity interval running versus cycling on sclerostin, and markers of bone turnover and oxidative stress in young men. Calc Tiss Inter. 2019;104(6):582–90.
22. Kurgan N, Logan-Sprenger H, Falk B, Klentrou P. Bone and inflammatory responses to training in female rowers over an olympic year. Med Sci Sports Exerc. 2018;50(9):1810–7.
23. Van Beaumont W. Evaluation of hemoconcentration from hematocrit measurements. J Appl Physiol. 1972;32(5):712–3.
24. Cohen J. Dedication. In: Statistical Power Analysis for the Behavioral Sciences. New York (NY): Routledge Academic; 1988.
25. Cohen J. Statistical power analysis. Curr Dir Psychol Sci. 1992;1(3):98–101.
26. Zhou X, Fragala MS, McElhaney JE, Kuchel GA. Conceptual and methodological issues relevant to cytokine and inflammatory marker measurements in clinical research. Curr Opin Clin Nutr Metab Care. 2010;13(5):541–7.
27. Yu Z, Kastenmüller G, He Y, et al. Differences between human plasma and serum metabolite profiles. PLoS One. 2011;6(7):e21230.
28. Amrein K, Amrein S, Drexler C, et al. Sclerostin and its association with physical activity, age, gender, body composition, and bone mineral content in healthy adults. J Clin Endocrinol Metab. 2012;97(1):148–54.
29. Mödder U, Clowes J, Hoey K, et al. Regulation of circulating sclerostin levels by sex steroids in women and in men. J Bone Miner Res. 2011;26(1):27–34.
30. Kargotich S, Goodman C, Keast D, Morton AR. The influence of exercise-induced plasma volume changes on the interpretation of biochemical parameters used for monitoring exercise, training and sport. Sports Med. 1998;26(2):101–17.
31. Weinstein Y, Bediz C, Dotan R, Falk B. Reliability of peak-lactate, heart rate, and plasma volume following the Wingate test. Med Sci Sports Exerc. 1998;30(9):1456–60.
32. Piec I, Washbourne C, Tang J, et al. How accurate is your sclerostin measurement? Comparison between three commercially available sclerostin ELISA kits. Calcif Tissue Int. 2016;98(6):546–55.


© 2019 American College of Sports Medicine