Received: 13 September 2018 | Revised: 19 August 2019 | Accepted: 14 October 2019 DOI: 10.1111/ecc.13186 ORIGINAL ARTICLE The effects of exercise on insulin, glucose, IGF-axis and CRP in cancer survivors: Meta-analysis and meta-regression of randomised controlled trials Yafeng Wang1 | Ben Jin2 | Raheem J. Paxton3 | Weili Yang4 | Xirui Wang5 | Yurui Jiao6 | Chuanhua Yu1 | Xiong Chen7 1 Department of Epidemiology and Biostatistics, School of Health Sciences, Wuhan University, Wuhan, China 2 Department of Urology, The First Affiliated Hospital of Xi'an Jiaotong University, Xi'an, China 3 Department of Community Medicine and Population Health, The University of Alabama, Tuscaloosa, AL, USA 4 Department of Pediatric Surgery, The Second Affiliated Hospital of Xi'an Jiaotong University, Xi'an, China 5 Department of Epidemiology and Biostatistics, Xi'an Jiaotong University, Xi'an, China 6 Department of endocrinology, The Second Hospital of Shanxi Medical University, Taiyuan, China 7 Department of Endocrinology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, China Correspondence Chuanhua Yu, School of Health Sciences, Wuhan University, Wuhan, Hubei, China. Email: YuCHua@whu.edu.cn Abstract Background: The purpose of this study was to investigate the relationship between physical activity and biological mediators of cancer recurrence and survival. Methods: We conducted a comprehensive literature search of PubMed, ScienceDirect and Web of Science for randomised controlled trials examining the association between physical activity and C-reactive protein (CRP), glucose, insulin, insulin resistance and insulin growth factor-one (IGF-1) up to December 2017. Standardised mean difference (SMD) scores were calculated, and meta-regression was performed. Results: The meta-analysis indicated that survivors randomised to physical activity conditions experienced greater improvements in Insulin (SMD = −0.59; 95% CI, −1.05 to −0.14), CRP (SMD = −0.52; 95% CI, −0.87 to −0.17), insulin resistance (SMD = −0.20; 95% CI, −0.41 to −0.003) and glucose (SMD = −0.19; 95% CI, −0.35 to −0.02) than survivors randomised to control conditions. The meta-regression showed that study duration was positively, albeit marginally related (p = .056) to change in CRP levels among survivors in the physical activity conditions. Furthermore, higher baseline insulin levels in the physical activity conditions were associated with improving insulin levels throughout the intervention (p = .007). Conclusions: Promoting physical activity throughout the survivorship continuum is an effective intervention strategy for improving levels of insulin, glucose control, in- Xiong Chen, Department of Endocrinology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang, China. Email: chasecx@126.com sulin resistance and CRP among cancer survivors. KEYWORDS cancer survivors, CRP, exercise, meta-analysis, glucose, IGF-axis, insulin, meta-regression 1 | I NTRO D U C TI O N completion (Hill et al., 2014; Schover et al., 2014; Seth, Singh, Seth, & Sapra, 2017). Encouragingly, engaging in healthy lifestyle behaviours Scientific advancements in cancer treatment during the past 30 years my shield survivors from late effects and improve their overall health have resulted in improvements in cancer-specific and overall survival and well-being (Pierce et al., 2007; Turner et al., 2018). rates (Torre et al., 2015). Despite significant improvements, survivors Despite the benefits of a healthy lifestyle, many fail to meet the experience significant deficits in overall health and well-being (Brown health guidelines proposed by the American Cancer Society (Rock et al., 2003). In fact, many experiences late effects (e.g., infertility, car- et al., 2012). Of all the guidelines, the health benefits of physical diovascular injury and fatigue) persist for years following treatment activity have been the most consistently documented (US Cancer Eur J Cancer Care. 2019;00:e13186. https://doi.org/10.1111/ecc.13186 wileyonlinelibrary.com/journal/ecc © 2019 John Wiley & Sons Ltd | 1 of 10 2 of 10 | WANG et Al. Statistics Working Group, 2007). In particular, several meta-analytic cancer; the outcomes included HOMA, glucose, insulin, IGF-1 or CRP studies have confirmed the relationship between physical activity with corresponding mean and standard deviations for each condition. and improved survival (Spei et al., 2019; Wu et al., 2016; Yeganeh, The exclusion criteria included participants were ≤18 years at study Harrison, Vincent, Teede, & Boyle, 2018). Studies have speculated enrolment; the outcome reported only median and range values; that the same biological mediators that escalated one's risk for cancer and whether multiple studies reported results from the same study initiations and metastasis (i.e., glucose tolerance, insulin sensitivity population. Two reviewers independently carried out the literature and inflammation) may be the same culprits in determining cancer search and study selection. Any disagreements on the inclusion of recurrence and survival rates (Bastard et al., 2000; Frank et al., 2005). studies were resolved by a third reviewer. However, limited data exist on the biological mechanisms underlying the relationship between physical activity, recurrence and survival. In cancer survivors, several studies have investigated the exercise-in- 2.3 | Data extraction and quality assessment duced changes in physiological biomarkers. However, the results have been inconsistent and the exercise-induced changes appear to A standardised form was used to extract data from each study. The differ among cancer sites (Fong et al., 2012; Kang et al., 2017). following extracted information included characteristics of the The purpose of this study was to address the gaps in the literature study (first author, year of publication, study location, sample size, by conducting a meta-analysis to evaluate the effect of physical activ- the mean age, gender, mean body mass index (BMI), %body fat, the ity interventions on insulin, insulin resistance, glucose, insulin growth baseline insulin level, cancer type and cancer stage), characteristic of factors (IGF) and related binding proteins (BP), and C-reactive protein the physical activity intervention and control condition (i.e., exercise (CRP) in cancer survivors. The proposed study expands upon prior me- type, duration, frequency, intensity and treatment for the control ta-analyses by addressing some of the methodological limitations that group) and results (changes in insulin, HOMA, glucose, IGF-axis and may have contributed to inconsistent findings (Fong et al., 2012; Kang CRP). Two researchers independently extracted relevant data using a et al., 2017; Meneses-Echávez et al., 2016). In particular, we propose standardised excel template. Any discrepancies between two reviewers to adjust for potential confounding factors such as lifestyle and study were discussed, and a consensus was achieved. The two reviewers also characteristics, which may influence the results. In addition, we pro- assessed risk of bias by using Cochrane Collaboration risk assessment pose to conduct a meta-regression model to comprehend the indepen- tools for randomised controlled trials (Higgins & Green, 2011). dent effects of physical activity on each individual biological marker. 2 | M E TH O DS 2.4 | Statistical analysis Given that different measurements were used for each outcome, 2.1 | Search strategy we calculated standardised mean difference (SMD) scores with 95% confidence intervals (CIs). Statistical heterogeneity was assessed using The meta-analysis was performed in accordance with the preferred the I2 test and Cochran's Q statistic (I2 values > 50% indicated the reporting items for systematic reviews and meta-analyses criteria existence of heterogeneity and p < .10 indicated a statistical significance). described elsewhere (Moher, Liberati, Tetzlaff, & Altman, 2010). A When there was no significant heterogeneity, a fixed-effects model comprehensive literature search was conducted for all relevant stud- was used; otherwise, a random-effects model was used. To explore the ies published in PubMed, ScienceDirect and Web of Science before potential for heterogeneity, meta-regression and subgroup analyses January 2017 using the following keywords: “cancer OR neoplasm were carried out. Subgroup analysis was also performed according to OR tumor” AND “physical activity OR exercise OR resistance OR cancer type. In addition, we conducted a sensitivity analysis to assess strength OR stretching” AND “randomized controlled trial OR con- the stability of the pooled results by removing one study at a time. trolled clinical trial OR randomized OR trial.” A detailed description of Publication bias was evaluated with the Begg's and Egger's tests. All the Boolean Search criteria was included in Appendix S1. To ensure statistical analyses were performed with Stata version 11.0 (STATA). A that we captured additional articles, we expanded the comprehen- two-sided p value of ≤.05 was considered statistically significant. sive literature search to December 2017. Furthermore, no language restrictions were imposed, and the reference list of retrieved articles was also searched for relevant studies. 2.2 | Selection criteria 3 | R E S U LT S 3.1 | Characteristics of included studies A total of 11,057 potentially eligible articles were identified via The included studies fulfilled the following inclusion criteria: study our online search. After eliminating duplicates, 9,960 articles were design was limited to RCT; the study compared a physical activity selected for further screening, with 123 articles being thoroughly intervention with a control group; adults were diagnosed with any assessed for potential eligibility. After reviewing the full text of these | WANG et Al. 3 of 10 articles, only 15 articles were deemed eligible for our meta-analysis when compared to the control condition (SMD = −0.59; 95% CI, (Bower et al., 2014; Cormie et al., 2015; Fairey et al., 2003; Galvão, −1.05 to −0.14; p = .011; see Figure 2a). The analysis also indicated Taaffe, Spry, Joseph, & Newton, 2010; Guinan et al., 2013; Irwin et significant between-study heterogeneity (I2 = 86.0%; see Table 2), al., 2009; Janelsins et al., 2011; Jones et al., 2013; Lahart, Metsios, but no publication bias. We also observed greater reductions in Nevill, Kitas, & Carmichael, 2016; Ligibel et al., 2008; Nuri et al., glucose (SMD = −0.19; 95% CI, −0.35 to −0.02; p = .024) and insulin 2012; Nuri, Moghaddasi, Darvishi, & Izadpanah, 2016; Schmitz, resistance (HOMA: SMD = −0.20; 95% CI, −0.41 to −0.003; p = .046) Ahmed, Hannan, & Yee, 2005; Sprod et al., 2012; Thomas, Alvarez- among survivors in the intervention conditions when compared Reeves, Lu, Yu, & Irwin, 2013) (Figure 1). to the control groups. No between-study heterogeneity or bias The characteristics of the included studies are summarised was observed for glucose (I2 = 33.8%; see Figure 2b) nor insulin in Table 1. Among the included studies, there were a total of 793 resistance (I2 = 46.7%; see Figure 2c). Clinically and statistically participants distributed roughly proportional in the intervention significant mean differences in CRP (SMD = −0.52; 95% CI, −0.87 to (n = 403) and control (n = 390) conditions. Studies were conducted −0.17; Figure 2d) were observed between intervention and control in United States (n = 8), Australia (n = 2), Iran (n = 2), the United conditions, with significant between-study heterogeneity. However, Kingdom (n = 1), Canada (n = 1) and Ireland (n = 1). The studies fo- it should be noted that survivors in the physical activity conditions cused primarily on breast cancer survivors (n = 12), with few focus- did not differ from the control conditions with respect to IGF-1 ing on prostate (n = 2) or colorectal cancer (n = 1) survivors. The (SMD = −0.09; 95% CI, −1.34 to 1.16; p = .89) with no evidence of age of survivors in the intervention conditions ranged from 52.4 to heterogeneity (I2 = 94.9%; see Table 2) or bias. 69.6 years old, whereas the age range of survivors in the control conditions ranged from 52.7 to 70.1 years old. Study duration varied widely from 6 weeks to 12 months. 3.3 | Exploratory heterogeneity analysis The meta-regression showed that baseline insulin levels may be the 3.2 | Changes in insulin, HOMA, glucose, CRP and IGF-1 main source of heterogeneity in the meta-analysis (Figure 3b). In particular, higher baseline insulin levels in physical activity conditions were associated with a greater decrease in insulin levels throughout The results of the meta-analysis indicated that survivors in the the study (slope = −0.12, 95% CI, −0.20 to −0.04; p = .007; Figure 3b). physical activity condition experienced greater reductions in insulin Body size (i.e., body fat and BMI) and intervention duration were not F I G U R E 1 Flow chart for the selection of eligible studies 4 of 10 TA B L E 1 Characteristics of 15 randomised controlled trials included in meta-analysis of effect of exercise in cancer survivors Sample Age Intervention | Author Cancer type Country Janelsins et al. (2011) Breast cancer USA Lahart et al. (2016) Breast cancer UK Exercise group Control group Exercise group Control group 9 10 54.3 ± 10.6 40 40 52.4 ± 10.3 Study quality Exercise group Control group Duration Outcome 52.7 ± 6.7 Tai chi chuan exercise Non-physical activity control 3 months Insulin, IGF−1, HOMA A 54.7 ± 8.3 Home-based PA intervention Usual care 6 months Insulin A Irwin et al. (2009) Breast cancer USA 37 38 56.4 ± 9.5 55.6 ± 7.7 Aerobic exercise Usual care 6 months Insulin, IGF−1 A Fairey et al. (2003) Breast cancer Canada 25 28 59 ± 5 58 ± 6 The exercise group The delayed control group 15 weeks Insulin, Glucose, IGF−1 A Schmitz et al. (2005) Breast cancer USA 39 40 53.3 ± 8.7 52.8 ± 7.6 Weight training The delayed control group 6 months Insulin, Glucose, IGF−1, HOMA A Ligibel et al. (2008) Breast cancer USA 51 49 52 ± 9 53 ± 9 Strength training exercise intervention Usual care 4 months Insulin, Glucose, HOMA A Jones et al. (2013) Breast cancer USA 36 31 56.4 ± 9.6 55.4 ± 7.6 Aerobic exercise intervention Usual care 6 months CRP A Guinan et al. (2013) Breast cancer Ireland 16 10 50.1 ± 8.3 45.1 ± 9.0 Aerobic programme Control group 3 months Insulin, Glucose, CRP, HOMA A Nuri et al. (2012) Breast cancer Iran 14 15 NA NA Combined training Control group 15 weeks Insulin, Glucose, HOMA A Sprod et al. (2012) Breast cancer USA 9 10 54.3 ± 3.6 52.7 ± 2.1 Tai chi chuan exercise Control group 3 months Insulin, Glucose, IGF−1 A Bower et al. (2014) Breast cancer USA 16 15 NA NA Yoga Control group 3 months CRP A Thomas et al. (2013) Breast cancer USA 35 30 56.5 ± 9.8 55.1 ± 7.6 Moderate-intensity aerobic exercise Usual care 6 months Glucose A Nuri et al. (2016) Colorectal cancer Iran 15 15 NA NA Training group Control group 2 months Insulin, Glucose, HOMA A Cormie et al. (2015) Prostate cancer Australia 32 31 69.6 ± 6.5 67.1 ± 7.5 Supervised exercise Usual care 3 months Insulin, Glucose, CRP A Galvão et al. (2010) Prostate cancer Australia 29 28 69.5 ± 7.3 70.1 ± 7.3 The exercise group Usual care 3 months Insulin, Glucose, CRP A Abbreviations: CRP, c-reactive protein; HOMA, homeostasis model assessment; IGF-1, insulin-like growth factor-1; NA, not available. WANG et Al. | WANG et Al. 5 of 10 significantly related to changes in insulin, HOMA, glucose and IGF-1 compared to survivors in controlled conditions. In our subgroup Figure 3a, c, and d). Results of the sensitivity analysis indicated that analysis, it appeared that changes in CRP were more pronounced insulin level change remained stable (Figure 4). In addition, we found in breast and prostate cancer survivors. The sensitivity analysis that the Schmitz et al. study was a primary source of heterogeneity indicated that caution is warranted when interpreting the impact of and excluding this study resulted in stabilising the results (see physical activity intervention on levels IGF-1, HOMA and glucose Figure 5). The meta-regression indicated that the duration of the in cancer survivors. Furthermore, our meta-regression analysis intervention was marginally accounted for the relationship between indicated that having high baseline insulin levels and study duration the physical activity and change in CRP levels (slope = 0.26, 95% CI, influenced the impact of the intervention on insulin and CRP levels, −0.01 to 0.53; p = .056; Figure 6). The subgroup analysis showed respectively. Overall, these data provide evidence compelling that significant mean changes in CRP levels were more pronounced evidence physical activity intervention may influence markers in prostate cancer survivors (SMD = −0.63; 95% CI, −1.00 to −0.26; associated with the IGF-axis and inflammatory pathways. p = .004) than it was for breast cancer survivors (SMD = −0.49; 95% CI, −1.04 to 0.07; p = .087; see Table 2). In our study, we observed that survivors enrolled in physical activity interventions experienced greater improvements in insulin, glucose and insulin resistance than participants in the control conditions and that many of these changes may be related to having 4 | D I S CU S S I O N high insulin levels at baseline. We were somewhat surprised that null findings were observed with IGF-1. These findings add to the un- The present meta-analysis evaluated the role that participating in certainty of using plasm and serum levels of IGF-1 as an indicative physical activity intervention had on biological mediators of cancer marker (Fong et al., 2012; Kang et al., 2017; Meneses-Echávez et al., recurrence and survival. Results indicated that survivors who 2016; Zhu et al., 2016). In addition, significant heterogeneity was participated in physical activity intervention experienced greater observed for IGF-1, which was minimised after removing a study. improvements in insulin, glucose, insulin resistance and CRP when Therefore, exercise might improve IGF-1 in breast cancer survivors, FIGURE 2 CRP Meta-analysis for the effect estimate of exercise on insulin, glucose, HOMA and CRP. (a) insulin; (b) glucose; (c) HOMA; (d) 6 of 10 | TA B L E 2 WANG et Al. Subgroup analysis in meta-analysis of effect of exercise in cancer survivors Subgroup N SMD Lower Upper PSMD Q I2 PH HOMA All 7 −0.20 −0.41 0.00 0.046 11.26 46.70% 0.081 Breast cancer 6 −0.19 −0.40 0.02 0.073 11.06 54.80% 0.050 Colorectal cancer 1 −0.36 −1.08 0.36 0.328 11 −0.19 −0.35 −0.02 0.024 15.10 33.80% 0.128 0 Glucose All Breast cancer 8 −0.15 −0.34 0.04 0.119 14.17 50.60% 0.048 Prostate cancer 2 −0.33 −0.69 0.03 0.075 0.19 0.00% 0.663 Colorectal cancer 1 −0.17 −0.89 0.54 0.635 0 Insulin All Breast cancer 12 −0.59 −1.05 −0.14 0.011 78.80 86.00% <0.001 9 −0.81 −1.41 −0.21 0.008 71.04 88.70% <0.001 0.29 0.00% 0.591 Prostate cancer 2 0.09 −0.27 0.45 0.628 Colorectal cancer 1 −0.31 −1.03 0.41 0.396 All 5 −0.09 −1.34 1.16 0.89 77.68 94.90% <0.001 Excluding one study 4 −0.76 −1.08 −0.44 <0.001 4.31 30.40% 0.230 IGF-I CRP All 6 −0.52 −0.87 −0.17 0.004 10.71 53.30% 0.058 Breast cancer 4 −0.49 −1.04 0.07 0.087 9.00 66.70% 0.029 Prostate cancer 2 −0.63 −1.00 −0.26 0.001 0.47 0.00% 0.493 Abbreviations: CRP, c-reactive protein; HOMA, homeostasis model assessment; IGF-1, insulin-like growth factor-1. but further studies were required to draw a definitive conclusion. Prior systematic review and meta-analytic studies have indicated The results we observed with markers associated with the IGF- that participating in regular physical activity is associated with a re- axis resemble those observed in non-cancer populations (Kasapis duction in CRP (Fedewa, Hathaway, & Ward-Ritacco, 2017; Kasapis & Thompson, 2005) and a recent study promoting aerobic and re- & Thompson, 2005), which correspond with reduction in body size. sistance training (Kang et al., 2017; Ligibel et al., 2008; Meneses- Similarly, Kang et al. (2017) observed a meaningful reduction of CRP Echávez et al., 2016). It should be noted that our data differs from levels in response to the exercise interventions. The results of our that of a prior meta-analysis performed in breast cancer survivors meta-analysis confirm the results of prior review and meta-analytic (Löf, Bergström, & Weiderpass, 2012). Some have suggested that studies with respect to reduction in CRP. Importantly, the results physical activity-induced changes in biological markers are depen- appeared to be more pronounced in prostate cancer survivors. In ad- dent on changes in body size or high baseline insulin levels (Gunter dition, longer duration will weaken the effect of physical activity on et al., 2009; Ligibel et al., 2008; Pasanisi et al., 2006). However, our CRP levels throughout the study. Unfortunately, we were not able meta-regression only partially confirmed these suggestions, with to perform the meta-regression of effect of body size in the physical greater support for the higher baseline insulin levels, but not for activity and CRP relationships because no more than three studies higher baseline body size (i.e., BMI and per cent body fat). Overall, provided sufficient data on change in body size. Therefore, the fur- results of the meta-analysis, subsequent meta-regression and test ther RCTs investigating these associations are required to confirm, for heterogeneity provide strong evidence indicating that physi- whether these relationships differ by body size. cal activity intervention may improve markers associated with the There were a number of limitations that should be discussed. IGF-axis and that survivors with high insulin levels may receive the First, heterogeneity between intervention and control groups was greatest benefits from these types of interventions. Subsequent me- observed, which may affect the robustness of our results. In par- ta-analyses with a larger variety of studies with relevant variables ticular, there was significant heterogeneity between groups for are required to confirm our results. Inconsistent with a previous certain biomarkers, preventing us from moving further with the me- meta-analysis, our results showed that physical activity can improve ta-analysis for these variables. To minimise this limitation with the insulin resistance (Fong et al., 2012). In our study, we included three remaining variables, we performed the subgroup, sensitivity and me- additional included studies, and smaller sample size in their study ta-regression analyses to explore factors associated with heteroge- can lead to the wider CIs. neity. Secondly, our analysis was based on unadjusted data, without WANG et Al. | 7 of 10 F I G U R E 3 Meta-regression of the BMI change, %body fat change, baseline insulin levels and intervention duration and insulin. (a) change of the BMI; (b) baseline insulin levels; (c) %Body Fat change; (d) intervention duration F I G U R E 4 Sensitivity analysis for the effect estimate of exercise on insulin 8 of 10 | WANG et Al. F I G U R E 5 Sensitivity analysis for the effect estimate of exercise on IGF-1 adjusting for the confounding factors that might influence outcome and duration, as well as baseline health characteristics are sources of variables. Lastly, we were unable to examine the impact of exercise heterogeneity in future meta-analyses. Furthermore, future meta- type, intensity and duration of the results due to limitations of the analytic work should consider adopting our analytic framework to included studies. Despite these limitations, we were able to conduct increase the robustness of their findings. a number of statistical analyses to identify sources of heterogeneity and bias resulting in a robust analytic framework that can be repli- AC K N OW L E D G E M E N T S cated in future studies. None. 5 | CO N C LU S I O N C O N FL I C T O F I N T E R E S T The authors declare no conflict of interest. In conclusion, physical activity is an effective intervention for AU T H O R C O N T R I B U T I O N S improving levels of insulin, insulin resistance, glucose and CRP Conception and design of the study: YFW and CHY. Acquisition of among cancer survivors. The results of our meta-analyses support data and analysis: BJ, WLY, XRW and YRJ. Writing and revision of the strong and consistent relationships between these variables, but manuscript: YFW, RJP, CHY and XC. All authors read and approved also identify variables that should be consistently reported on and the final manuscript. explored further. For example, more data are needed to determine whether factors such as cancer treatment, exercise type, intensity E T H I C A L A P P R OVA L Every study included in the meta-analysis has been published after the approval of an ethics committee, and each patient enrolled signed a written consent. For this reason, the present meta-analysis did not require further ethics committee approval. REFERENCES FIGURE 6 Meta-regression of intervention duration and CRP Bastard, J. P., Jardel, C., Bruckert, E., Blondy, P., Capeau, J., Laville, M., … Hainque, B. (2000). Elevated levels of interleukin 6 are reduced in serum and subcutaneous adipose tissue of obese women after weight loss. The Journal of Clinical Endocrinology & Metabolism, 85, 3338–3342. https://doi.org/10.1210/jcem.85.9.6839 Bower, J. E., Greendale, G., Crosswell, A. D., Garet, D., Sternlieb, B., Ganz, P. A., … Cole, S. W. (2014). Yoga reduces inflammatory signaling in fatigued breast cancer survivors: A randomized controlled trial. Psychoneuroendocrinology, 43, 20–29. https://doi.org/10.1016/j. psyneuen.2014.01.019 Brown, J. K., Byers, T., Doyle, C., Coumeya, K. S., Demark-Wahnefried, W., Kushi, L. H., … Sawyer, K. A. (2003). Nutrition and physical WANG et Al. activity during and after cancer treatment: An American Cancer Society guide for informed choices. CA A Cancer Journal for Clinicians, 53, 268–291. https://doi.org/10.3322/canjclin.53.5.268 Cormie, P., Galvão, D. A., Spry, N., Joseph, D., Chee, R., Taaffe, D. R., … Newton, R. U. (2015). Can supervised exercise prevent treatment toxicity in patients with prostate cancer initiating androgen-deprivation therapy: A randomised controlled trial. BJU International, 115, 256–266. https://doi.org/10.1111/bju.12646 Fairey, A. S., Courneya, K. S., Field, C. J., Bell, G. J., Jones, L. W., & Mackey, J. R. (2003). Effects of exercise training on fasting insulin, insulin resistance, insulin-like growth factors, and insulin-like growth factor binding proteins in postmenopausal breast cancer survivors: A randomized controlled trial. Cancer Epidemiology Biomarkers & Prevention, 12, 721–727. https://cebp.aacrjournals.org/content/12/8/721.long Fedewa, M. V., Hathaway, E. D., & Ward-Ritacco, C. L. (2017). Effect of exercise training on C-reactive protein: A systematic review and meta-analysis of randomised and non-randomised controlled trials. British Journal of Sports Medicine, 51, 670–676. https://doi. org/10.1136/bjsports-2016-095999 Fong, D. Y., Ho, J. W., Hui, B. P., Lee, A. M., Macfarlane, D. J., Leung, S. S., … Cheng, K. K. (2012). Physical activity for cancer survivors: Metaanalysis of randomised controlled trials. BMJ, 30, 344:e70. https:// doi.org/10.1136/bmj.e70 Frank, L. L., Sorensen, B. E., Yasui, Y., Tworoger, S. S., Schwartz, R. S., Ulrich, C. M., … McTiernan, A. (2005). Effects of exercise on metabolic risk variables in overweight postmenopausal women: A randomized clinical trial. Obesity Research, 13, 615–625. https://doi. org/10.1038/oby.2005.66 Galvão, D. A., Taaffe, D. R., Spry, N., Joseph, D., & Newton, R. U. (2010). Combined resistance and aerobic exercise program reverses muscle loss in men undergoing androgen suppression therapy for prostate cancer without bone metastases: A randomized controlled trial. Journal of Clinical Oncology, 28, 340–347. https://doi.org/10.1200/ JCO.2009.23.2488 Guinan, E., Hussey, J., Broderick, J. M., Lithander, F. E., O'Donnell, D., Kennedy, M. J., & Connolly, E. M. (2013). The effect of aerobic exercise on metabolic and inflammatory markers in breast cancer survivors—a pilot study. Supportive Care in Cancer, 21, 1983–1992. https:// doi.org/10.1007/s00520-013-1743-5 Gunter, M. J., Hoover, D. R., Yu, H., Wassertheil-Smoller, S., Rohan, T. E., Manson, J. E., … Strickler, H. D. (2009). Insulin, insulin-like growth factor-I, and risk of breast cancer in postmenopausal women. JNCI Journal of the National Cancer Institute, 101, 48–60. https://doi. org/10.1093/jnci/djn415 Higgins, J. P. T., & Green, S. (2011). Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. The Cochrane Collaboration. http://handbook-5-1.cochrane.org Hill, D. A., Horick, N. K., Isaacs, C., Domchek, S. M., Tomlinson, G. E., Lowery, J. T., … Finkelstein, D. M. (2014). Long-term risk of medical conditions associated with breast cancer treatment. Breast Cancer Research and Treatment, 145, 233–243. https://doi.org/10.1007/ s10549-014-2928-4 Irwin, M. L., Varma, K., Alvarez-Reeves, M., Cadmus, L., Wiley, A., Chung, G. G., … Yu, H. (2009). Randomized controlled trial of aerobic exercise on insulin and insulin-like growth factors in breast cancer survivors: the Yale exercise and survivorship study. Cancer Epidemiology Biomarkers & Prevention, 18, 306–313. https://doi.org/10.1158/10559965.EPI-08-0531 Janelsins, M. C., Davis, P. G., Wideman, L., Katula, J. A., Sprod, L. K., Peppone, L. J., … Mustian, K. M. (2011). Effects of Tai Chi Chuan on insulin and cytokine levels in a randomized controlled pilot study on breast cancer survivors. Clinical Breast Cancer, 11, 161–170. https:// doi.org/10.1016/j.clbc.2011.03.013 Jones, S. B., Thomas, G. A., Hesselsweet, S. D., Alvarez-Reeves, M., Yu, H., & Irwin, M. L. (2013). Effect of exercise on markers of | 9 of 10 inflammation in breast cancer survivors: The Yale exercise and survivorship study. Cancer Prevention Research, 6, 109–118. https://doi. org/10.1158/1940-6207.CAPR-12-0278 Kang, D. W., Lee, J., Suh, S. H., Ligibel, J., Courneya, K. S., & Jeon, J. Y. (2017). Effects of exercise on insulin, IGF-axis, adipocytokines, and inflammatory markers in breast cancer survivors: A systematic review and meta-analysis. Cancer Epidemiology Biomarkers & Prevention, 26, 355–365. https://doi.org/10.1158/1055-9965.EPI-16-0602 Kasapis, C., & Thompson, P. D. (2005). The effects of physical activity on serum C-reactive protein and inflammatory markers: A systematic review. Journal of the American College of Cardiology, 45, 1563–1569. https://doi.org/10.1016/j.jacc.2004.12.077 Lahart, I. M., Metsios, G. S., Nevill, A. M., Kitas, G. D., & Carmichael, A. R. (2016). Randomised controlled trial of a home-based physical activity intervention in breast cancer survivors. BMC Cancer, 17, 234. https:// doi.org/10.1186/s12885-016-2258-5 Ligibel, J. A., Campbell, N., Partridge, A., Chen, W. Y., Salinardi, T., Chen, H., … Winer, E. P. (2008). Impact of a mixed strength and endurance exercise intervention on insulin levels in breast cancer survivors. Journal of Clinical Oncology, 26, 907–912. https://doi.org/10.1200/ JCO.2007.12.7357 Löf, M., Bergström, K., & Weiderpass, E. (2012). Physical activity and biomarkers in breast cancer survivors: A systematic review. Maturitas, 73, 134–142. https://doi.org/10.1016/j.maturitas.2012.07.002 Meneses-Echávez, J. F., Jiménez, E. G., Río-Valle, J. S., Correa-Bautista, J. E., Izquierdo, M., & Ramírez-Vélez, R. (2016). The insulin-like growth factor system is modulated by exercise in breast cancer survivors: A systematic review and meta-analysis. BMC Cancer, 16, 682. https:// doi.org/10.1186/s12885-016-2733-z Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G., PRISMA Group (2010). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. International Journal of Surgery, 8, 336–341. https://doi.org/10.1016/j.jcms.2010.02.007 Nuri, R., Kordi, M. R., Moghaddasi, M., Rahnama, N., Damirchi, A., Rahmani-Nia, F., & Emami, H. (2012). Effect of combination exercise training on metabolic syndrome parameters in postmenopausal women with breast cancer. Journal of Cancer Research and Therapeutics, 8, 238–242. https://doi.org/10.4103/0973-1482. 98977 Nuri, R., Moghaddasi, M., Darvishi, H., & Izadpanah, A. (2016). Effect of aerobic exercise on leptin and ghrelin in patients with colorectal cancer. Journal of Cancer Research and Therapeutics, 12, 169–174. https:// doi.org/10.4103/0973-1482.155982 Pasanisi, P., Berrino, F., De Petris, M., Venturelli, E., Mastroianni, A., & Panico, S. (2006). Metabolic syndrome as a prognostic factor for breast cancer recurrences. International Journal of Cancer, 119, 236– 238. https://doi.org/10.1002/ijc.21812 Pierce, J. P., Stefanick, M. L., Flatt, S. W., Natarajan, L., Sternfeld, B., Madlensky, L., … Rock, C. L. (2007). Greater survival after breast cancer in physically active women with high vegetable-fruit intake regardless of obesity. Journal of Clinical Oncology, 25, 2345–2351. https://doi.org/10.1200/JCO.2006.08.6819 Rock, C. L., Doyle, C., Demark-Wahnefried, W., Meyerhardt, J., Courneya, K. S., Schwartz, A. L., … Gansler, T. (2012). Nutrition and physical activity guidelines for cancer survivors. CA: A Cancer Journal for Clinicians, 62, 243–274. https://doi.org/10.3322/caac.21142 Schmitz, K. H., Ahmed, R. L., Hannan, P. J., & Yee, D. (2005). Safety and efficacy of weight training in recent breast cancer survivors to alter body composition, insulin, and insulin-like growth factor axis proteins. Cancer Epidemiology Biomarkers & Prevention, 14, 1672–1680. https://doi.org/10.1158/1055-9965.EPI-04-0736 Schover, L. R., van der Kaaij, M., van Dorst, E., Creutzberg, C., Huyghe, E., & Kiserud, C. E. (2014). Sexual dysfunction and infertility as late effects of cancer treatment. European Journal of Cancer Supplements, 12, 41–53. https://doi.org/10.1016/j.ejcsup.2014.03.004 10 of 10 | Seth, R., Singh, A., Seth, S., & Sapra, S. (2017). Late effects of treatment in survivors of childhood cancers: A single-center experience. Indian Journal of Medical Research, 146, 216–223. https://doi.org/10.4103/ ijmr.IJMR_196_16 Spei, M. E., Samoli, E., Bravi, F., La Vecchia, C., Bamia, C., & Benetou, V. (2019). Physical activity in breast cancer survivors: A systematic review and meta-analysis on overall and breast cancer survival. Breast, 44, 144–152. https://doi.org/10.1007/s10549-014-2928-4 Sprod, L. K., Janelsins, M. C., Palesh, O. G., Carroll, J. K., Heckler, C. E., Peppone, L. J., … Mustian, K. M. (2012). Health-related quality of life and biomarkers in breast cancer survivors participating in Tai Chi Chuan. Journal of cancer survivorship. Journal of Cancer Survivorship, 6, 146–154. https://doi.org/10.1007/s11764-011-0205-7 Thomas, G. A., Alvarez-Reeves, M., Lu, L., Yu, H., & Irwin, M. L. (2013). Effect of exercise on metabolic syndrome variables in breast cancer survivors. International Journal of Endocrinology, 2013, 168797. https://doi.org/10.1155/2013/168797 Torre, L. A., Bray, F., Siegel, R. L., Ferlay, J., Lortet-Tieulent, J., & Jemal, A. (2015). Global cancer statistics, 2012. CA: A Cancer Journal for Clinicians, 65, 87–108. https://doi.org/10.3322/caac.21262 Turner, R. R., Steed, L., Quirk, H., Greasley, R. U., Saxton, J. M., Taylor, S. J., … Bourke, L. (2018). Interventions for promoting habitual exercise in people living with and beyond cancer. Cochrane Database of Systematic Reviews, 9, CD010192. https://doi.org/10.1002/14651 858.CD010192.pub3 US Cancer Statistics Working Group (2007). United States cancer statistics: 2004 Incidence and Mortality (pp. 1–22). Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention, and National Cancer Institute [Document file]. Retrived from https://stacks.cdc.gov/view/cdc/5415 WANG et Al. Wu, W., Guo, F., Ye, J., Li, Y., Shi, D., Fang, D., … Li, L. (2016). Pre- and post-diagnosis physical activity is associated with survival benefits of colorectal cancer patients: A systematic review and meta-analysis. Oncotarget, 7, 52095–52103. https://doi.org/10.18632/oncot arget.10603 Yeganeh, L., Harrison, C., Vincent, A. J., Teede, H., & Boyle, J. A. (2018). Effects of lifestyle modification on cancer recurrence, overall survival and quality of life in gynaecological cancer survivors: A systematic review and meta-analysis. Maturitas, 111, 82–89. https://doi. org/10.1016/j.maturitas.2018.03.001 Zhu, G., Zhang, X., Wang, Y., Xiong, H., Zhao, Y., & Sun, F. (2016). Effects of exercise intervention in breast cancer survivors: A meta-analysis of 33 randomized controlled trails. OncoTargets and Therapy, 9, 2153– 2168. https://doi.org/10.2147/OTT.S97864 S U P P O R T I N G I N FO R M AT I O N Additional supporting information may be found online in the Supporting Information section. How to cite this article: Wang Y, Jin B, Paxton RJ, et al. The effects of exercise on insulin, glucose, IGF-axis and CRP in cancer survivors: meta-analysis and meta-regression of randomised controlled trials. Eur J Cancer Care. 2019;00:e13186. https://doi.org/10.1111/ecc.13186