Effectiveness of Resistance Exercise on Inflammatory Biomarkers in Patients with Type 2 Diabetes Mellitus: A Systematic Review with Meta-Analysis
Article information
Abstract
Background
Type 2 diabetes mellitus (T2DM) is related to increased inflammatory processes. The effects of resistance exercise on inflammatory biomarkers in T2DM are controversial. Our purpose was to determine the effectiveness of resistance exercise on inflammatory biomarkers in patients diagnosed with T2DM.
Methods
We searched four databases until September 2021. We included randomized clinical trials (RCTs) of the effects of resistance exercise on inflammatory biomarkers (C-reactive protein [CRP], tumor necrosis factor alpha, interleukin-6, and interleukin-10) in patients with T2DM. A random effects meta-analysis was conducted to determine the standardized mean difference (SMD) and the raw mean difference (MD) for CRP.
Results
Thirteen RCTs were included in the review, and 11 in the meta-analysis for CRP. Lower CRP levels were observed when resistance exercise was compared with the control groups (SMD=–0.20; 95% confidence interval [CI], –0.37 to –0.02). When conducting the MD meta-analysis, resistance exercise showed a significant decrease in CRP of –0.59 mg/dL (95% CI, –0.88 to –0.30); otherwise, in the control groups, the CRP values increased 0.19 mg/dL (95% CI, 0.17 to 0.21).
Conclusion
Evidence supports resistance exercise as an effective strategy to manage systemic inflammation by decreasing CRP levels in patients with T2DM. The evidence is still inconclusive for other inflammatory biomarkers.
INTRODUCTION
In recent years, type 2 diabetes mellitus (T2DM) has been increasing exponentially [1], ranking among the 10 leading causes of death in adults, having a worldwide prevalence of over 9% and affecting approximately 463 million people [1]. Consequently, developing preventive measures to delay the onset and early treatment strategies to slow the progression of T2DM is a major concern among clinical health professionals and public health researchers [2].
T2DM is caused by chronic inadequate insulin production by pancreatic β-cells, leading to hyperglycemia [3]. This causes an immune response that results in a chronic low-grade inflammatory status [3], which includes increased levels of inflammatory biomarkers such as C-reactive protein (CRP), tumor necrosis factor alpha (TNF-α), or interleukin-6 (IL-6) [4]. Low-grade chronic inflammation has great implications for the onset and progression of T2DM [5]. For instance, these biomarkers affect insulin production by progressive damage to pancreatic β-cells and inflammation [3], along with other factors (e.g., aging, physical inactivity, obesity, etc.) can be involved in the development of insulin resistance [3] and, therefore, promote the inefficient use of insulin by the body’s cells. Thus, both mechanisms contribute to chronic hyperglycemia [3]. Additionally, it is well established that patients with T2DM generally have higher adiposity levels [6], particularly visceral adiposity, which is also associated with chronic inflammation and insulin resistance [7].
In this context, exercise could be a nonpharmacological intervention that may delay the progression of the disease and improve the management and quality of life of people with T2DM [2]. Accordingly, it has been suggested that exercise could delay the progression of insulin resistance [8] because of its effect on reducing circulating levels of inflammatory biomarkers such as CRP, TNF-α, and IL-6 [9,10]. In this sense, resistance exercise has gained importance for patients with T2DM [11] due to its multisystemic and specific musculoskeletal benefits [12]. Additionally, resistance exercise may be a useful exercise strategy in patients with a diagnosis of T2DM due to their exposure to accelerated muscle loss [13], which would be related to an increased risk of mortality and other comorbidities [14].
There is increasing evidence of resistance exercise in different populations, and recently, international guidelines, including the American Diabetes Association Standards of Medical Care in Diabetes (2022) recommend resistance exercise of any intensity to improve glycemic control as well as strength, balance, and activities of daily living in patients with T2DM [15]. Despite this, although some studies have shown benefits on several health parameters, including inflammatory biomarkers [11,16], other studies have questioned its effectiveness [9]. Therefore, our purpose was to synthesize the current evidence and determine the effectiveness of resistance exercise on inflammatory biomarkers in patients with T2DM.
METHODS
Ethical statement
This systematic review and meta-analysis were conducted by collecting data from primary studies in which informed consent had been obtained by the respective original authors; thus, our review was exempt from ethics approval.
Search strategy and study selection
The present systematic review and meta-analysis were conducted according to the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions [17] and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Appendix 1) [18]. We registered this review in the PROSPERO database (registration number: CRD42021261-762).
A systematic search was conducted in the MEDLINE (via PubMed), Scopus, Cochrane CENTRAL, and Web of Science databases from inception until September 2021 to identify randomized controlled trials (RCTs) aimed at determining the effectiveness of resistance exercise on inflammatory biomarkers in adults diagnosed with T2DM. The search strategy combined the following medical subject headings with free terms and matching synonyms: ‘type 2 diabetes,’ ‘noninsulin-dependent diabetes mellitus,’ ‘resistance training,’ ‘resistance exercise,’ ‘strength training,’ ‘strength exercise,’ ‘strengthening,’ ‘inflammation,’ ‘inflammatory markers,’ ‘inflammatory cytokines,’ ‘inflammatory biomarkers,’ ‘c reactive protein,’ ‘tumor necrosis factor alpha,’ and ‘interleukin 6.’ The complete search strategy for each database is available in the Supplementary Table 1.
Eligibility criteria
The titles and abstracts of the retrieved articles were examined by two independent reviewers (R.F.R., S.M.C.) to identify suitable studies. Articles related to this systematic review were selected for full text screening and evaluated according to the eligibility criteria. The inclusion criteria were as follows: (1) type of participants: adults (≥18 years) with a medical diagnosis of T2DM; (2) type of intervention: at least one arm trial had to be related to resistance exercise; (3) control condition with nonexercise intervention (i.e., usual care, advice); (4) outcome: inflammatory biomarkers such as CRP, TNF-α, IL-6, or IL-10; (5) type of studies: RCTs. Moreover, the studies were excluded when (1) some participants were non-clinically diagnosed with T2DM and (2) resistance exercise was not the only type of exercise performed (i.e., multimodality, concurrent training). A third reviewer (A.E.M.) was consulted to resolve disagreements between reviewers. No language restrictions were applied. Excluded studies with the reason for exclusion are available in Supplementary Table 2.
Data extraction and risk of bias assessment
Two independent reviewers (R.F.R., S.M.C.) extracted the following information from the included studies: first author’s name and year of publication; country; study design; characteristics of the study population (mean age, women’s percentage, baseline body mass index [BMI], comorbidities), total sample size and sample size by group, intervention characteristics (exercise protocol), medication, comparison characteristics, outcome measures and main results. A third reviewer (A.E.M.) was consulted to resolve disagreements between reviewers.
Two reviewers (R.F.R., S.M.C.) independently assessed the risk of bias of the included studies using the Cochrane risk-of-bias tool for randomized trials [19]. The following six domains were assessed: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. As it would be impossible to truly blind patients to treatment allocation in exercise trials, this specific item of the risk of bias was not included to generate the overall risk of bias assessment. In this sense, each domain was assessed as ‘low risk of bias,’ ‘some concerns,’ or ‘high risk of bias,’ and the overall risk of bias for each study was classified as (1) ‘low risk of bias’ when a low risk of bias was determined for all domains; (2) ‘some concerns’ when at least one domain was assessed as raising some concerns, but no single domain was assessed as high risk of bias; or (3) ‘high risk of bias’ when high risk of bias was reached for at least one domain or some concerns in multiple domains.
Disagreements between initial reviewers were solved by a third coauthor (A.E.M.).
Quality of evidence
The “Grades of Recommendations, Assessment, Development, and Evaluation” (GRADE) tool was used to evaluate and summarize the quality of the evidence [20]. Based on the design of the studies, the inflammatory biomarker outcome measure was rated as high-, moderate-, low-or very low-quality evidence considering the following domains: (1) risk of bias (–1 when <75% of the analyzed studies were at low risk of bias); (2) inconsistency (–1 when I2 >50%); (3) indirect evidence (from population, intervention, control or outcomes); (4) imprecision displayed in wide confidence intervals (CIs); and (5) publication bias, which downgraded the quality of evidence risk of bias. The GRADE tool was used for those outcomes with enough data for the meta-analysis.
Data analysis
When at least five studies reported valid data for the outcome, we extracted the primary data from each study, including pre-post mean inflammatory biomarker values, standard deviations and sample sizes of intervention and control groups (CG). Therefore, meta-analyses were conducted for CRP (11 studies), TNF-α (six studies), and IL-6 (five studies) but not for IL-10 (two studies) as the statistical analysis would not be able to translate the potential effect of resistance exercise. The standardized mean difference (SMD) with its 95% CI was calculated for each study using the DerSimonian and Laird random-effects method [21]. Then, the pooled SMDs were estimated for the effect of resistance exercise versus the CG. Furthermore, to show the clinical change in outcome units of measurement (mg/dL), we computed the pooled raw mean difference (MD) after transforming all outcome data into the same unit. Additionally, the heterogeneity was evaluated with the I2 statistic as follows: I2 values of 0%–40% were considered to be ‘not important’ heterogeneity, 30%–60% indicated ‘moderate’ heterogeneity, 50%–90% indicated ‘substantial’ heterogeneity, and 75%–100% indicated ‘considerable’ heterogeneity, taking into account the corresponding P values and 95% CIs [22].
We conducted a sensitivity analysis to determine the robustness of the summary estimates by removing each included study from the analysis one by one. Furthermore, subgroup analyses based on reported comorbidities, as well as meta-regression models considering mean age, sample size, length of the intervention, T2DM duration (years), total body fat percentage, glycosylated hemoglobin (HbA1c, %), blood glucose levels (mg/dL), and percentage of women, to determine their influence on the SMD estimates and on the raw MD for CRP levels were conducted. Moreover, we explore whether the MD on CRP levels (mg/dL) between resistance exercise and CG could be influenced by baseline HbA1c levels (%) considering as a cut point the median value of HbA1c among the studies (7.5%). Finally, we evaluated publication bias through visual inspection of funnel plots and Egger’s regression asymmetry test to assess small study effects [23]. We performed all statistical analyses using Stata SE version15 (StataCorp., College Station, TX, USA).
RESULTS
Study selection
After duplicated articles were removed and analyzed by title and abstract, a total of 57 full-text articles were assessed for eligibility, of which 13 [24-36] were included in the systematic review, and 11 RCTs [25-34,36] were included for the meta-analysis on CRP levels (Fig. 1). The excluded studies with reasons for exclusion after full-text reading are available in Supplementary Table 2.
Flow diagram of the studies through the review.
Characteristics of studies
The 13 RCTs included with a parallel design were conducted between 2006 and 2021. The studies were conducted in different countries such as Australia, Brazil, Grecia, Iran, Spain, Sri Lanka, Taiwan, and the USA. Further details are available in Table 1.
Characteristics of the studies included in the systematic review
Participants
A total of 568 adults with T2DM (mean age between 48 and 72 years) were included. Among the participants, 261 adults were in the resistance exercise groups, and 256 were in the control conditions. Most studies had similar rates of men and women, except for one study in which only males were included [36]. The baseline BMI values of the participants ranged from 25 to 35 kg/m2 (Table 1). Moreover, six studies reported comorbidities across the included population, such as overweight and obesity [27,29,30,32,34,35]. The most reported medications prescribed were hypoglycemic, antihypertensive, and lipidlowering drugs. Further details of the covariates are available in Supplementary Table 3.
Interventions
Although the resistance exercise protocols were different across the intervention groups, the training length was set between 60 and 75 minutes, with a frequency of two to three sessions per week for 12 to 16 weeks. Eight studies performed resistance exercise with global protocols (i.e., upper limbs, core, and lower limbs) [26-29,31,33,34,36], four studies only included lower or upper limbs [24,25,30,32], and one study did not report the protocol [35]. Resistance exercise was performed using calisthenic exercises [26-28,34,36], dumbbells [24-27,29-33,35,36], and machines [24-28,30,34-36]. However, it was not reported to the rest of the groups [29,31-33].
Among the CG there were five studies in which participants received usual care and advice through medical visits or by telephone [24,25,27,28,33,34]; four studies in which participants received recommendations about physical activity [30,32,33,35] and three studies that included a general stretching protocol [26,29,31].
Outcome
The CRP values were evaluated in 11 [25-34,36] out 13 included studies. Six studies measured TNF-α values [24,26,28,29,35,36], five studies measured IL-6 [26,28,29,35,36], and two studies reported IL-10 values [26,35]. Overall, inflammatory biomarkers were analyzed according to the clinical standards of the laboratory or the manufacturer’s guidelines with enzyme-linked immunosorbent assay (ELISA).
Risk of bias
When the RoB2 tool was used to assess the risk of bias, nine out of 11 studies scored at ‘low risk of bias’ [24-27,29,31,33,35,36], and four scored at ‘some concerns’ [28,30,32,34]. The risk of bias assessment is displayed in Supplementary Fig. 1.
Quality of evidence
The quality of evidence was rated as high for CRP outcomes, since the certainty assessment showed low concerns regarding the risk of bias, inconsistency, and imprecision. A table with a summary of the findings is available in Supplementary Table 4.
Data synthesis
Meta-analysis
The SMDs for the effect on CRP of resistance exercise versus control was –0.20 (95% CI, –0.37 to –0.02; I2=0%) (Fig. 2). Considering the mean raw difference in CRP levels, after resistance exercise, there was a significant reduction of -0.59 mg/dL (95% CI, –0.88 to –0.30; I2=55%); otherwise, the CRP values showed an increase in the CG of 0.19 mg/dL (95% CI, 0.17 to 0.21; I2=0%) (Fig. 3).
Standardized mean difference meta-analysis for resistance exercise compared to control groups on C-reactive protein levels. n, sample size; T2DM (yr), years since the first diagnosis of type 2 diabetes mellitus; HbA1c, glycosylated hemoglobin; ES, effect size; CI, confidence interval; NR, not reported.
Raw mean difference (MD) meta-analysis for resistance exercise and control groups on C-reactive protein levels (mg/dL). n, sample size; CI, confidence interval.
The SMDs for the effect on TNF-α of resistance exercise versus CG showed a slight nonsignificant effect of –0.28 (95% CI, –0.85 to 0.28; I2=69.9%) (Fig. 4), as well as when considering IL-6 (SMD, –0.24; 95% CI, –1.43 to 0.96; I2=89.2%) (Fig. 5). Additionally, the effect plot without pooling the SMD for IL-6 and IL-10 are available in Supplementary Fig. 2.
Standardized mean difference meta-analysis for resistance exercise compared to control groups on tumor necrosis factor alpha. n, sample size; T2DM (yr), years since the first diagnosis of type 2 diabetes mellitus; HbA1c, glycosylated hemoglobin; ES, effect size; CI, confidence interval; NR, not reported.
Standardized mean difference meta-analysis for resistance exercise compared to control groups on interleukin-6 levels. n, sample size; T2DM (yr), years since the first diagnosis of type 2 diabetes mellitus; HbA1c, glycosylated hemoglobin; ES, effect size; CI, confidence interval; NR, not reported.
Sensitivity analysis
The negative SMD on CRP values was not modified when removing each included study one by one. Further details are displayed in Supplementary Table 5.
Subgroup analyses and meta-regression
The subgroup analysis considering potential comorbidities (i.e., overweight/obesity, hypertension) showed no differences in CRP values between participants with reported comorbidities (SMD, –0.12; 95% CI, –0.34 to 0.10; I2=0%) and those without them (SMD, –0.31; 95% CI, –0.63 to 0.01; I2=23%) (Supplementary Table 6). Regarding the coefficients and P values of the meta-regression models conducted for mean age (P=0.62), sample size (P=0.33), length of the intervention (P=0.22), T2DM duration (years) (P=0.78), total body fat percentage (P=0.66), HbA1c (%) (P=0.68), blood glucose levels (mg/dL) (P=0.76), and percentage of women (P=0.50), none of them significantly influenced the effects of resistance exercise on CRP levels (Supplementary Table 7). Meta-regression models conducted for the raw MD in CRP values showed a significant effect of baseline HbA1c (%) for the resistance exercise group (coef=0.78; P=0.01), while for the CG there was no significant effect of baseline HbA1c (%) (coef=–0.32; P=0.39) (Supplementary Table 8 and Supplementary Fig. 3). Finally, baseline HbA1c levels (%) influenced the response of resistance exercise when compared to CG, showing a significant reduction of –0.71 mg/dL (95% CI, –0.81 to –0.61; I2=0%) in CRP values in those patients with mean baseline HbA1c <7.5% (Supplementary Fig. 4).
Publication bias
Funnel plot visual assessment and Egger’s test confirmed no significant publication bias regarding CRP levels among the RCTs included in the meta-analysis (P=0.388) (Supplementary Fig. 5).
DISCUSSION
This study aimed to systematically synthesize the evidence regarding resistance exercise effectiveness on inflammatory markers in adults diagnosed with T2DM. Our findings suggest that resistance exercise may reduce CRP values in patients with T2DM. When exploring potential effect modifiers such as comorbidities, age, sample size, length of the intervention, time since T2DM diagnosis, total body fat percentage, blood glucose levels, and percentage of women, none of them significantly influenced the effect size estimates for CRP levels. As an exception, baseline HbA1c (%) had a significant effect on the MD in CRP levels for resistance exercise groups, showing that when patients had higher levels of HbA1c, the anti-inflammatory effects of resistance exercise would be reduced. However, evidence regarding TNF-α, IL-6, and IL-10 levels is scarce, and the corresponding results are far from consistent.
Circulating CRP levels were the most reported outcome. Most studies, except one [29], did not show a significant reduction in CRP values [25-34,36] compared to the control condition. Otherwise, when clinical differences were considered with the raw MD estimates, our results showed a significant reduction of CRP for resistance exercise while CG increased their values of CRP in mg/dL. Accordingly, previous evidence has demonstrated agreement with these results for patients with T2DM [37] and healthy participants [38]. However, these increased CRP levels after control conditions observed in seven out 10 studies were only significant in one study [30]. Of note, in this specific study, changes in CRP levels of the participants could be associated with their baseline clinical characteristics because they were obese (BMI, 32 kg/m2) with high fat mass percentage (34%), triglyceride levels (195 mg/dL), and insulin values (10.24 mU/L) [30]. Moreover, the overall increased CRP in the CG was markedly heterogeneous and may be related to the different recommendations given to the CG (i.e., no intervention, usual care, physical activity or healthy diet advice, low-impact activities, etc.). Despite this, a clinical message could be extrapolated: the resistance exercise seems to be a better option than the different control conditions to manage systemic inflammation for patients with T2DM. It would be worth noting that there are some potential factors not fully described among the studies that should be considered as the baseline diet quality [39], diets based on anti-inflammatory foods [40] or the physical activity levels sustained throughout life that may attenuate the progression on inflammation caused by the disease and the aging process [41].
Despite this modest benefit on circulating CRP levels, other key factors associated with high oxidative stress in patients with T2DM could influence resistance exercise effectiveness, such as smoking status, physical inactivity, increased adiposity levels, diet [42], and pharmacological treatments [43], which could mask the effect of exercise in the T2DM population. In this sense, when we explored the meta-regression models for raw MD on circulating CRP levels, we found that the baseline HbA1c (%) may modify significantly the effect in the resistance exercise groups but no in CG. Of note that the pooled baseline HbA1c values for participants in the resistance exercise groups were slightly lower (7.5%) than for CG (7.7%), and only two studies reported good metabolic control (<7%) for the resistance exercise groups [26,33]. Because of the scarcity of trials making it difficult to clearly interpret this, it would be interesting that further studies explore the potential moderating role of HbA1c in the effectiveness of resistance exercise in patients with T2DM.
Most studies have reported lower circulating TNF-α levels after resistance exercise [26,35,36]. Our meta-analysis showed a slight nonsignificant effect of resistance exercise for reducing TNF-α levels when compared to CG, which is similar to the results reported in a previous systematic review exploring the effects of aerobic exercise on inflammatory markers in T2DM patients [37]. However, two studies reported increased TNF-α levels after resistance exercise [24,29]. Specifically, in the study of Gordon et al. [24], the authors reported an increase in TNF-α expression in the trained muscle, but suggested that the poorly controlled T2DM and the elevated number of years with T2DM (10.5 years) could affect their results. Moreover, in Jorge et al. [29], the increased TNF-α levels could be explained because of the baseline differences in the HbA1c levels (%) for the resistance exercise group (8.27%) and the CG (6.99%); in addition, the authors also suggest that this surprising negative effect could be attributed by the poor metabolic control of participants. Even if the impact of resistance exercise on TNF-α levels are still unclear and needs to be further explored, it should be considered that increased muscle mass (hypertrophy) may be associated with decreased inflammation through the improvement in blood glucose levels and insulin resistance [25]. Moreover, when resistance exercise is prescribed to patients with T2DM, we should consider some factors that negatively impact muscle strength and progression of the disease such as age, diabetes duration, or fat percentage [44].
Despite the potential role of IL-6 and IL-10 have in the inflammatory response, few studies have reported their circulating levels. In this sense, our data showed a slight nonsignificant effect of resistance exercise for reducing IL-6 values when compared to CG, with some studies showing a reduction in IL-6 [28,35,36], while others showed a nonsignificant increase after resistance exercise [26,29]. We must be cautious when interpreting these results as the increase in IL-6 reported in Jorge et al. [29] could be affected due to the high HbA1c levels (%) shown in the intervention group (8.27%) compared to the CG (6.99%); thus, this fact should be further explored in future trials. Moreover, in Rech et al. [26] the age of the participants may affect as older patients could need longer interventions to account for positive effects on inflammatory biomarkers or glycemic response; moreover, their life habits exceed the accuracy of the questionnaire (mainly in the CG), which may also have influenced the results.
When considering the anti-inflammatory cytokine IL-10, the response after resistance exercise was not homogenous, showing nonsignificant reduced [35] or increased [26] circulating levels of IL-10. Although the scarcity of studies reporting this cytokine makes it difficult to better clarify the potential role of resistance exercise in the IL-10 response, it is recommended that future RCTs include IL-6 and IL-10 accounting for factors (i.e., physical activity status, diet, characteristics of the resistance exercise program, etc.) that could be potentially mediate the effect of resistance exercise on the response of these biomarkers.
Some mechanisms should be stated when considering the potential benefits of resistance exercise on inflammation. First, acute exercise stimulates the release of IL-6 in muscular tissue, which may act at the systemic level, inhibiting proinflammatory cytokines such as TNF-α and increasing anti-inflammatory cytokines such as IL-10 [45]. Second, resistance exercise is associated with increased muscle mass, which may improve insulin sensitivity due to the key role of skeletal muscle in glucose uptake [25]. Third, resistance exercise could act on adipose tissue, diminishing adiposity, thus improving insulin sensitivity [46] and increasing vasodilatation, angiogenesis and blood flow, which may cause a reduction in hypoxia, macrophage infiltration and chronic inflammation in adipose tissue [45]. Moreover, mechanisms contributing to sarcopenia are also crucial to metabolic disorder pathogenesis (bidirectional relationship), with inflammation being a typical process involved in T2DM and skeletal muscle structure [47]. Finally, it is worth noting the influence of regular resistance exercise on the immune system, such as an increased T-cell count, which is related to IL-10 release and reduced expression of receptors associated with the production of inflammatory cytokines [45].
Otherwise, the inconclusive results shown in some studies might be due to some related factors, such as weight gain, physical inactivity, and genetic predisposition, that may cause adipose tissue dysfunction and increased secretion of CRP, TNF-α, and IL-6 and decreased adiponectin levels (anti-inflammatory protein) [4]. In addition, age is a crucial factor in the progression of T2DM, the aging process is associated with increased levels of some inflammatory cytokines [48]. Of note, weight gain produced by some usual medications in T2DM, such as insulin and sulfonylureas, may also impact the results.
Adding to the abovementioned confounding factors that may impact the results from primary studies, our review has some limitations that should be noted. First, the included exercise studies were not blinded because it would be impossible to truly blind participants to treatment allocation. Second, the heterogeneity among pharmacological treatments in patients with T2DM may introduce some kind of bias in our estimates. Third, the small total sample size (n=568), added to the heterogeneity among resistance exercise protocols and the nonreported exercise intensity in some studies, should be considered. Fourth, the overall metabolic control of participants included could have affected our estimates. Finally, studies evaluating the resistance exercise effect on the inflammatory response in patients with T2DM in the long term are lacking.
In summary, our analyses suggest a significant reduction in CRP values after resistance exercise in patients with T2DM. Moreover, potential effect modifiers such as comorbidities, age, sample size, length of the intervention, time since T2DM diagnosis, total body fat percentage, blood glucose levels, and percentage of women did not significantly influence our effect size estimates, except for baseline HbA1c (%). Regarding TNF-α and IL-6 levels, resistance exercise showed a nonsignificant reduction. However, future trials would help to elucidate the controversial and heterogenous results for TNF-α, IL-6, and IL-10 levels. Thus, further studies should consider confounding factors that may have a direct or indirect impact on inflammatory biomarker levels (i.e., HbA1c levels, diabetes duration, diet quality, physical activity status, body composition, pharmacological treatments). Additionally, assessments of resistance exercise benefits in inflammatory biomarkers in patients with T2DM in the long term are strongly required. Current evidence indicates that prescribing resistance exercise to patients with T2DM can reduce inflammatory marker levels, specifically CRP in addition to the known benefits on body composition and metabolic parameters.
SUPPLEMENTARY MATERIALS
Supplementary materials related to this article can be found online at https://doi.org/10.4093/dmj.2022.0007.
Search strategy for each database
Studies excluded after full text reading with reasons
Covariates of interest among included studies
GRADE assessment for C-reactive protein
Sensitivity analyses
Subgroup analyses models for SMD meta-analysis
Meta-regressions models for standardized mean difference meta-analysis
Meta-regression models for raw mean differences meta-analysis
Quality assessment of randomized controlled trials according to RoB2. (A) Per protocol. (B) Intention-to-treat. CRP, C-reactive protein; IL-6, interleukin-6; TNF-α, tumor necrosis factor alpha; IL-10, interleukin-10.
Summary of the standardized mean differences (SMDs) of resistance exercise versus control group for interleukin-6 (IL-6) and IL-10 outcomes. CI, confidence interval.
Meta-regressions models for glycosylated hemoglobin (HbA1c, %) in the raw mean difference meta-analysis. (A) HbA1c total for strength training groups. (B) HbA1c total for control groups. CRP, C-reactive protein.
Mean difference (MD) in C-reactive protein between groups considering the baseline mean values of glycosylated hemoglobin (HbA1c) of intervention and control groups (<7.5% vs. >7.5%). CI, confidence interval; NR, not reported.
Funnel plot and Egger’s test for C-reactive protein (CRP) meta-analysis. SE, standard error; SMD, standardized mean difference.
Notes
CONFLICTS OF INTEREST
No potential conflict of interest relevant to this article was reported.
AUTHOR CONTRIBUTIONS
Conception or design: R.F.R., S.M.C., B.B.P.
Acquisition, analysis, or interpretation of data: R.F.R., S.M.C., M.G.M., A.E.M., V.M.V.
Drafting the work or revising: R.F.R., S.M.C.
Final approval of the manuscript: R.F.R., S.M.C., B.B.P., M. G.M., A.E.M., V.M.V.
FUNDING
This study was funded by the European Regional Development Fund. The funding source had no such involvement or restrictions regarding publication. Bruno Bizzozero-Peroni is supported by a grant from the Universidad de Castilla-La Mancha co-financed by the European Social Fund (2020-PREDUCLM-16746). Arthur Eumann Mesas was supported by a ‘Beatriz Galindo’ contract (BEAGAL18/00093) by the Spanish Ministry of Education, Culture and Sport.
Acknowledgements
None
