Comparison of the Effect of Endurance, Strength, and Endurance-Strength Training on Inflammatory Markers and Adipokines Levels in Overweight and Obese Adults: Systematic Review and Meta-Analysis of Randomised Trials

The aim of this meta-analysis was to compare the effects of endurance, strength, and combined training on inflammatory markers and adipokine concentrations in overweight and obese adults. We performed a literature search of the Cochrane Library, PubMed, Scopus, and Web of Science databases and identified 24 randomised control trials published prior to June 2021. Our findings indicate that endurance training was significantly more beneficial than strength training in reducing C-reactive protein (CRP) (standard mean difference (SMD): −1.317, 95% confidence intervals (CI): −2.565, −0.070, p = 0.0385), interleukin 6 (IL-6) (SMD: −0.363, 95% CI: −0.648, −0.078, p = 0.0126), and visfatin (SMD: −0.618, 95% CI: −1.015, −0.222, p = 0.0023) concentrations. Moreover, combined training was more beneficial than strength training alone in lowering tumour necrosis factor-alpha (TNF-α) levels (SMD: 0.890, 95% CI: −0.301, 1.478, p = 0.0030). There were no differences between the effects of different types of training programmes on adiponectin and leptin concentrations. In conclusion, compared with strength training, endurance training is more effective in lowering CRP, IL-6, and visfatin concentrations, while combined training is more beneficial in reducing TNF-α levels in overweight and obese adults. Further studies are needed to determine which type of training has a better effect on adiponectin and leptin concentrations in this population.


Introduction
The World Health Organisation (WHO) defines being overweight or obese as having abnormal or excessive fat accumulation that presents a health risk, and, since 1997, has classified these as global epidemics [1]. These conditions are the major public health problems of modern times. Recent data show that almost 2 billion adults worldwide are overweight, of which more than 670 million are obese [2]. Obesity is associated with a high risk of morbidity and mortality, as well as reduced life expectancy [3]. Being overweight or obese increases the risk of developing multiple diseases, such as type 2 diabetes mellitus (DM2), cardiovascular disease, several types of cancers, non-alcoholic fatty liver disease, an array of musculoskeletal disorders, and poor mental health. Obesity is a chronic metabolic disease characterised by energy intake exceeding energy expenditure [1,3,4].
A large number of studies have confirmed that exercise intervention is one of the effective means to prevent and treat obesity and reduce the risk of developing concomitant diseases [3,5,6]. Physical training is known to be associated with body weight and fat mass Moreover, manual searches of the reference lists of included papers were performed to identify further relevant studies and potential studies not captured in the electronic database searches. The research was conducted from database inception to June 2021.

Eligibility Criteria
Original studies were included in this systematic review if they met the following inclusion criteria: 1.
Language: articles published in English; 3.
Population: free-living adult (≥18 years old) overweight and obese subjects (overweightness, obesity, or one of the following criteria should be listed in the inclusion criteria: body mass index (BMI) ≥25 kg/m 2 [30] (≥23 kg/m 2 for Asian populations [31]), waist circumference (WC) ≥80 cm for women and ≥94 cm for men [32], and percentage of fat mass (%FM) >32% for women and 25% for men [33], or equivalent) of either gender and without restrictions based on the ethnicity of study participants, location of study, or sample size; 4.
Types of interventions: studies that compare the effects of ET vs. ST training, or/and ET vs. CT, or/and ST vs. CT on inflammatory markers or adipokine levels without any dietary consultation or intervention (study populations should be instructed not to change dietary habits and should not take any dietary supplements), with a duration for the intervention of at least two weeks; 5.
The exclusion criteria were as follows: 1.
Types of studies: non-randomised trials, uncontrolled trials, observational studies, cohort studies, cross-sectional studies, case-control, case-series, case-report studies, editorial letters, systematic reviews, meta-analyses, conference reports, studies available only as abstracts, and studies with animal models; 2.
Language: articles published in any language other than English; 3.
Population: children, adolescents, pregnant and breastfeeding women, subjects with rare comorbidities, and subjects living in non-public (closed-type) houses where subjects cannot freely decide on their eating habits or where all residents received the same diet.

Study Selection
Two investigators independently evaluated each database (the Cochrane Library: MK and AM, PubMed: MJ and NK, Scopus: MK and AŚ, and Web of Science: NK and AM). All articles were assessed in three main stages of the assessment process (see Figure 1). First, the reviewers screened article titles, then abstracts, and finally full texts for eligibility based on the inclusion and exclusion criteria. Disagreements were resolved by discussion between the reviewers until a consensus was reached. All reviewers agreed on the final decision of studies to be included. With regard to missing data, primary authors were contacted for more information.
3. Population: children, adolescents, pregnant and breastfeeding women, subjects with rare comorbidities, and subjects living in non-public (closed-type) houses where subjects cannot freely decide on their eating habits or where all residents received the same diet.

Study Selection
Two investigators independently evaluated each database (the Cochrane Library: MK and AM, PubMed: MJ and NK, Scopus: MK and AŚ, and Web of Science: NK and AM). All articles were assessed in three main stages of the assessment process (see Figure  1). First, the reviewers screened article titles, then abstracts, and finally full texts for eligibility based on the inclusion and exclusion criteria. Disagreements were resolved by discussion between the reviewers until a consensus was reached. All reviewers agreed on the final decision of studies to be included. With regard to missing data, primary authors were contacted for more information.

Data Collection Process
A data extraction sheet was developed. Then, the sheet was tested and refined. Two investigators (JG and MJ) independently extracted the data from the included studies and the third investigator (MK) checked the extracted data. The fourth investigator (AM) converted each outcome to the same units, to facilitate the interpretation of the data. Disagreements were resolved by discussion between the investigators. In the case of missing or unclear information, corresponding authors were contacted by e-mail.

Data Item
The following information was extracted from each included trial:

Data Collection Process
A data extraction sheet was developed. Then, the sheet was tested and refined. Two investigators (J.G. and M.J.) independently extracted the data from the included studies and the third investigator (M.K.) checked the extracted data. The fourth investigator (A.M.) converted each outcome to the same units, to facilitate the interpretation of the data. Disagreements were resolved by discussion between the investigators. In the case of missing or unclear information, corresponding authors were contacted by e-mail.

Data Item
The following information was extracted from each included trial: 1.
General information: title of the articles, author list, journal name, publication year, country, and continent; 2.
Characteristics of the study: study name and design (parallel or cross-over randomised trial) and inclusion and exclusion criteria; Type of outcomes measured: pre-intervention and post-intervention values of each outcome, changes (∆) for each outcome (post-intervention minus pre-intervention values).

Risk of Bias of Individual Studies
The risk of bias was assessed by two independent investigators (J.G. and M.J.) using the Cochrane risk of bias tool for randomised trials, where the following domains are included: bias due to randomisation, bias due to deviations from intended intervention, bias due to missing data, bias due to outcome measurement, and bias due to selection of reported results [34]. Criteria for low risk, some concerns, and high risk of bias as per the Cochrane handbook for systematic reviews of interventions were used [28].

Statistical Analysis
The meta-analysis was performed using the Comprehensive Meta-Analysis software, version 3.0 (Biostat, Inc., Englewood, CO, USA). A p < 0.05 was considered to be statistically significant. Post-intervention means and standard deviations (SD) were used to perform the meta-analysis. When a standard error was reported, SD was calculated from the standard error of a mean by multiplying by the square root of the sample size. If a 95% confidence interval (CI) was available and the sample size in each group was large (n ≥ 100), the SD for each group was obtained by dividing the width of the CI by 3.92, and then multiplying by the square root of the sample size in that group. If the sample size was smaller than 100, the CI was calculated using a value from a t distribution [28]. When the publications revealed median and range, the mean was calculated by the method of Hozo et al. [35]. If logarithmic values were presented, data were transformed back to the raw scale. If a study included at least two groups of the same type of training, but with different intensities, the groups were combined into a single group according to the formula provided in the Cochrane Handbook [28]. The GetData Graph Digitizer software was used to extract data from figures [36]. For studies that reported changes in outcomes >2 time points, the last measurement was used in the meta-analysis. A meta-analysis was carried out on each outcome that was assessed in at least two studies. The original values presented in the publications were used to perform the meta-analysis, while the tables show the values after unifying the units for easier data interpretation. We performed analyses to compare the effect of the following: 1. ET vs. ST; 2. ET vs. CT; and 3. ST vs. CT. Data synthesis was undertaken including a calculation of effect sizes with 95% CI, using fixed-effects models (if no heterogeneity is present), and random-effects models (to analyse outcomes with moderate and high heterogeneity) with inverse variance weighting. Standard mean differences (SMD) of post-intervention values were used as a summary statistic, to allow the comparison of effect sizes across studies. The SMD measures the absolute difference between the mean value in two groups of a trial. Forest plots were generated to illustrate the study-specific effect sizes, along with 95% CI. To determine the presence of publication bias, funnel plots were generated. Additionally, Begg's and Egger's tests were performed. Subgroup analyses were not performed. Heterogeneity between studies was evaluated using Cochran Q statistics with p < 0.1 indicating significant heterogeneity. The I 2 test was used to evaluate consistency between studies in which a value <30% indicates a low risk of heterogeneity, 30% to 75% indicates a moderate risk of heterogeneity, and >75% indicates a high risk of heterogeneity, which were consistent with the interpretation thresholds for the I 2 statistics, according to the Cochrane handbook for systematic reviews (<40% indicates a low risk of heterogeneity, 30% to 60% indicates a moderate risk of heterogeneity, 50 to 90% indicates a significant risk of heterogeneity, and >75% indicates a significant risk of heterogeneity) [28]. Template data collection forms, data extracted from included studies, data used for analysis, analytic code, and any other materials used in the review are available on reasonable request from the corresponding author.

Effect of Physical Training on CRP Concentrations
The effects of the training interventions on inflammatory markers are presented in Table 3. Ten studies compared the effect of different training programmes on CRP levels [14][15][16][21][22][23]25,40,43,48]: three studies evaluated the effect of ET and ST [21,22,40], three assessed the effect of ET and CT [14,23,48], and four evaluated the effect of ET, ST, and CT [15,16,25,43]. However, only one study reported significant differences between the effect of ET and ST, as well as between CT and ST [15].
The results of this meta-analysis showed that a CT programme is significantly more beneficial in reducing TNF-α levels than an ST programme (ST vs. CT: fixed-effects model, SMD: 0.890, 95% CI: −0.301, 1.478, p = 0.0030, Figure 4A

Effect of Physical Training on Leptin Levels
The effect of different training programmes on levels of adipokines is reported in Table 4. Four studies evaluated the effect of exercise on leptin concentrations [19,40,45,47]. Two studies compared the effect of ET and CT training, and two evaluated the effect of ET and ST [19,40]. One study found significant differences between ET and ST [45].
However, the results of the meta-analysis did not show any significant differences between the effect of ET and ST or between ET and CT programmes on leptin levels (ET vs.
Nevertheless, this meta-analysis did not report any significant differences between the effects of ET and

Effect of Physical Training on Visfatin Levels
Four studies evaluated the effect of exercise programmes on visfatin levels [15,18,25,37]. All studies evaluated the effect of ET, ST, and CT. Two studies reported significant differences between the effect of ST and CT [15,37], one paper found significant differences between ET and CT training [18], and one between ET and ST [15].
Funnel plots of standard error by standard differences in means of adipokine levels are included in the Supplementary Materials (see Figures S4 and S5).       6 Least square means (means adjusted for baseline) with (95% confidence intervals); 7 Value after Bonferroni; 8 correction Mean and 95% confidence intervals; 9 High-intensity interval endurance training; 10 Moderate-intensity continuous endurance training; 11 Mean ± standard deviation; 12 Data shown as log; 13 Adjusted mean ± standard error; 14
Funnel plots of standard error by standard differences in means of adipokine levels are included in the Supplementary Materials (see Figures S4 and S5).

Discussion
Our study incorporated 24 trials which included data from 1145 overweight and obese adults. The effects of ET, ST, and CT were compared by assessing their influence on the levels of inflammatory markers (CRP, IL-6, and TNF-α) and adipokines (leptin, adiponectin, and visfatin). The results of our meta-analysis clearly show a more beneficial effect of ET training in reducing CRP, IL-6, and visfatin levels, compared with ST. Moreover, our study indicates that CT is more effective in reducing TNF-α levels compared with ST alone. However, we did not identify any differences between the effects of different training programmes on adiponectin and leptin concentrations.
Previously, it has been shown that lifestyle interventions aiming to reduce weight in overweight or obese adult populations reduce mortality, regardless of their success in achieving weight loss [49]. It is now also believed that adipose tissue is an active endocrine organ that secretes various adipokines and pro-inflammatory cytokines, which, in obesity, can lead to a low level of systemic inflammation [50]. It is also associated with changes in levels of CRP, which is produced mainly by the liver in a response to pro-inflammatory cytokines, such as IL-6 and TNF-α, but has also been shown to be produced in adipose tissue and atherosclerotic plaques [51]. A recent meta-analysis comparing the independent effects of ET, ST, and CT on subcutaneous abdominal adipose tissue (SAT) in adults has shown that all these types of training lead to SAT reduction, while endurance exercise was shown to produce the greatest efficacy in decreasing SAT [52]. Several systematic reviews and meta-analyses have provided evidence of improvement in some inflammatory markers after different training sessions in various populations [26,[53][54][55][56][57].
Focusing on meta-analyses, Zheng et al. [26] have shown that ET significantly decreased CRP, TNF-α, and IL-6 without reducing IL-4 levels in healthy middle-aged and elderly people when compared to the control group. Meanwhile, Hayashino et al. [53] assessed the effects of any type of supervised exercise (endurance, strength, and combined) or physical exercise advice on inflammatory markers and adipokine levels in adults with T2D, and observed that training, overall, resulted in improved IL-6 and CRP comparing to the inactive control group in this population. Moreover, this exercise was more effective in lowering IL-6 levels where programmes had longer durations and a greater number of sessions. A meta-analysis by Monteiro-Junior et al. [54] has also shown a significant reduction in IL-6 and CRP concentrations, but not TNF-α levels, after chronic overall exercise intervention in older adults. However, Meneses-Echávez et al. [55] conducted a meta-analysis evaluating the influence of overall exercise training on mediators of inflammation in breast cancer survivors, and only observed improvement in the concentrations of IL-6, TNF-α, IL-8, and IL-2, without any differences in the concentrations of CRP when compared to a control group who received no intervention. Finally, Khalafi et al. [56], in their meta-analysis, compared the effect of exercise alone versus caloric re-

Discussion
Our study incorporated 24 trials which included data from 1145 overweight and obese adults. The effects of ET, ST, and CT were compared by assessing their influence on the levels of inflammatory markers (CRP, IL-6, and TNF-α) and adipokines (leptin, adiponectin, and visfatin). The results of our meta-analysis clearly show a more beneficial effect of ET training in reducing CRP, IL-6, and visfatin levels, compared with ST. Moreover, our study indicates that CT is more effective in reducing TNF-α levels compared with ST alone. However, we did not identify any differences between the effects of different training programmes on adiponectin and leptin concentrations.
Previously, it has been shown that lifestyle interventions aiming to reduce weight in overweight or obese adult populations reduce mortality, regardless of their success in achieving weight loss [49]. It is now also believed that adipose tissue is an active endocrine organ that secretes various adipokines and pro-inflammatory cytokines, which, in obesity, can lead to a low level of systemic inflammation [50]. It is also associated with changes in levels of CRP, which is produced mainly by the liver in a response to pro-inflammatory cytokines, such as IL-6 and TNF-α, but has also been shown to be produced in adipose tissue and atherosclerotic plaques [51]. A recent meta-analysis comparing the independent effects of ET, ST, and CT on subcutaneous abdominal adipose tissue (SAT) in adults has shown that all these types of training lead to SAT reduction, while endurance exercise was shown to produce the greatest efficacy in decreasing SAT [52]. Several systematic reviews and meta-analyses have provided evidence of improvement in some inflammatory markers after different training sessions in various populations [26,[53][54][55][56][57].
Focusing on meta-analyses, Zheng et al. [26] have shown that ET significantly decreased CRP, TNF-α, and IL-6 without reducing IL-4 levels in healthy middle-aged and elderly people when compared to the control group. Meanwhile, Hayashino et al. [53] assessed the effects of any type of supervised exercise (endurance, strength, and combined) or physical exercise advice on inflammatory markers and adipokine levels in adults with T2D, and observed that training, overall, resulted in improved IL-6 and CRP comparing to the inactive control group in this population. Moreover, this exercise was more effective in lowering IL-6 levels where programmes had longer durations and a greater number of sessions. A meta-analysis by Monteiro-Junior et al. [54] has also shown a significant reduction in IL-6 and CRP concentrations, but not TNF-α levels, after chronic overall exercise intervention in older adults. However, Meneses-Echávez et al. [55] conducted a meta-analysis evaluating the influence of overall exercise training on mediators of inflammation in breast cancer survivors, and only observed improvement in the concentrations of IL-6, TNF-α, IL-8, and IL-2, without any differences in the concentrations of CRP when compared to a control group who received no intervention. Finally, Khalafi et al. [56], in their meta-analysis, compared the effect of exercise alone versus caloric restriction alone, as well as exercise combined with caloric restriction versus caloric restriction alone, on inflammatory parameters in overweight and obese subjects, and showed that a combination of exercise with caloric restriction may be more effective than caloric restriction alone, causing a larger decrease in IL-6 and TNF-α, and tending to decrease CRP in this population. The most recent meta-analysis by Khalafi et al. [57] indicated that not only overall exercise but also ET, ST, and CT alone significantly reduced IL-6, TNF-α, and CRP concentrations in postmenopausal women when compared to an inactive control group. On the other hand, other systematic reviews and meta-analyses have not indicated an effective reduction in IL-6 and TNF-α concentrations after chronic overall exercise [54] or resistance training [58] in older adults compared to a control group who received no exercise intervention. Previous studies only assessed the overall effect of exercise, and the authors did not focus on comparing the effects of different types of training on the levels of inflammatory parameters. Only some of the works presented additional subgroup analysis to check whether a given type of training influences the inflammatory parameters, but even these did not compare the individual types of training. In our meta-analysis, we focused on the comparison of ET, ST, and CT, and indicated that ET has a more beneficial effect in reducing CRP and IL-6 levels in this population compared with ST alone. Moreover, we observed that CT is more beneficial than ST in reducing TNF-α levels.
The differences between the effects of particular types of training on inflammatory parameters may be explained by the promotion of other specific cardiovascular and neuromuscular adaptations [59]. ET causes adaptations of the musculoskeletal and cardiovascular systems that support an increase in performance and exercise capacity [60], while ST promotes neuromuscular adaptations that lead to power development and muscle strength improvement [61]. On the other hand, CT, as a combination of ET and ST, is a promising way to increase performance by training both cardiorespiratory fitness and muscle strength [62]. In general, our findings align with other studies, indicating positive effects of exercise on inflammatory parameters, and reinforcing the appropriateness of exercise prescription for different populations. Moreover, they indicate that, in order to obtain better therapy effects, it is important to select the appropriate type of training.
Being overweight or obese contributes to increased leptin and visfatin levels and decreased adiponectin concentrations [63,64]. A lot of studies have shown the positive effect of exercise on levels of adipokines in adults [53,57,[65][66][67], and in the paediatric population [68,69]. However, the evidence from these studies has not been conclusive. In a systematic review and meta-analysis investigating possible beneficial effects of exercise on adiponectin and leptin levels in overweight and obese subjects, Yu et al. [65] revealed that exercise, particularly endurance training, significantly increased adiponectin and reduced serum leptin concentrations compared to a control. Another meta-analysis assessing the influence of exercise on adipokine levels in adults with T2D has shown that overall exercise did not alter adiponectin or leptin concentrations, and that only an ET program was associated with a significant change in leptin levels [53]. However, overall exercise has been shown to be effective in increasing adiponectin compared to a control in postmenopausal women [57], while, in a meta-analysis of the adult population, adipokine levels also increased after ST [66]. Moreover, in the meta-analysis by Rostás et al. [67], ST appeared to be more efficient in reducing leptin concentrations than ET alone in middle-aged or older overweight or obese subjects. On the other hand, in a meta-analysis assessing the influence of exercise on adipokine levels in the obese paediatric population, CT resulted in greater increases in adiponectin levels than ET alone [68], while leptin concentrations decreased significantly after both ET and CT [69]. Previous studies show that different types of training can positively affect leptin and adiponectin levels, but there are no clear conclusions as to which type of training most effectively improves the concentrations of these adipokines. In our meta-analysis, we did not find any differences between the effect of ET, ST, and CT on leptin and adiponectin concentrations in overweight and obese adults. A possible explanation for our results is that the changes in these parameters are not related to the type of exercise, but to the change in body weight [70,71] or the duration [72,73] and intensity [74] of the intervention. In a three-year weight loss study, Madsen et al. [70] indicated that weight loss greater than 10% can improve the levels of circulating adiponectin, as well as inflammatory parameters in obese subjects. A greater than 10% increase in adiponectin levels after weight loss was also confirmed in older obese adults with and without periodontal disease, 3-18 months post enrolment [71]. It is accepted that chronic exercise resulting in weight reduction corresponds to an increase in adiponectin concentrations [75] and a decline in leptin levels [76]. Studies on the effects of acute exercise and the corresponding changes or lack thereof in leptin levels are less conclusive. In the study by Weltman et al. [72], 30 min of exercise at various intensities and caloric expenditures did not appear to be sufficient to affect leptin concentrations during exercise, or 3.5 h after training in healthy young men. On the other hand, Nindl et al. [73] indicated that, after acute strength exercises with an energy expenditure of 855 ± 114 kcals, leptin concentrations were lower compared to the control nine hours following the exercise. Moreover, in an RTC on overweight inactive elderly subjects, Fatouros et al. [74] indicated that, after six months of ST followed by six months of detraining, leptin concentrations decreased after low, moderate, and intensive ST, whereas adiponectin levels increased only after intensive ST. Furthermore, the small number of studies included in our analysis may also reinforce the need for more trials to confirm which type of training has a better effect on adipokines.
Our results do indicate that ET was more effective at lowering visfatin levels than ST. This may be related to the fact that ET leads to greater adipocyte tissue loss than ST, which in turn leads to a greater decrease in visfatin concentrations [52]. There is a lack of meta-analysis assessing the influence of exercise on visfatin concentrations in the adult population; however, a review of paediatric obesity indicated that overall exercise has an impact on the release of visfatin in this population [77]. Some intervention studies have evaluated the influence of different training on visfatin levels. Twelve-week CT intervention has been found to be effective in reducing visfatin levels in middle-aged obese women [78], while the same duration of ET did not significantly change visfatin levels in these obese women [79]. On the contrary, another researcher reported that twelve weeks of ET reduced the levels of visfatin in obese young subjects, with T2D or normal glucose tolerance [80]. Our results assessing the effect of different training programmes on visfatin levels should be interpreted with caution, due to the small number of studies included in the analysis.
To our knowledge, this is one of the first meta-analyses comparing the effect of ET, ST, and CT on inflammatory markers and levels of adipokines in overweight and obese adults. The other strengths of this meta-analysis include the detailed characteristics of the studies included and study populations, as well as specific inclusion and exclusion criteria. In addition, excluding studies with any dietary consultation or intervention allowed us to evaluate the actual impact of various training programmes on inflammatory markers and adipokine levels in overweight and obese adults. Moreover, this meta-analysis was written based on a search of PubMed, Web of Science, Scopus, and Cochrane, which are the largest and the most available databases. Furthermore, double counting of subjects from overlapping publications was prevented during the meta-analysis. However, heterogeneity was still significant owing to differences in the lengths, types, and durations of the exercise interventions, and the participants involved in the studies. Significant heterogeneity may also be related to different sampling and preparation methods, as well as the time elapsed between previous exercise sessions and sample measurements, which may affect inflammatory markers and adipokine levels. Another limitation of our publication is that the availability of outcome data that could be used for meta-analysis was limited, and the information needed was not always provided by the authors after contact. Moreover, for comparisons of ST versus CT, there was a lack of studies showing leptin concentrations suitable for meta-analysis. Other limitations include the fact that only six studies had a low risk of bias, while seven studies were assessed as having some concerns about bias, and a high risk of bias was found in nine studies. This could have impacted our analysis of the actual effects of the interventions. However, due to the specificity of the exercise interventions, it was not possible to conduct double-blind trials; therefore, a performance bias may be unavoidable in studies of this nature. In addition, several studies evaluated in the meta-analysis included subjects with comorbidities, such as DM2, prediabetes, or metabolic syndrome. However, these are common diseases in overweight and obese people.
To reduce heterogeneity, we disqualified studies involving subjects with rare comorbidities unrelated to obesity, such as cancers, lung, musculoskeletal, or gastrointestinal diseases. Another limitation of our study is the lack of subgroup analysis in terms of the duration of the intervention, body weight, or training intensity, which was not possible due to the small number of studies included in the meta-analysis. Moreover, we did not perform a sensitivity or meta-regression analysis to remove the sources of heterogeneity or variance in the studies included in the meta-analysis.
Our meta-analysis indicates the effectiveness of exercise therapy for reducing inflammatory markers and adipokine levels in overweight and obese adults. However, considering the limited number of included studies and the fact that we did not identify any differences between the effects of particular training programmes on adiponectin and leptin concentrations, we see a need for randomised control trials with larger sample sizes to determine the most suitable method to reduce levels of these adipokines. Moreover, the potential anti-inflammatory effects of the compared training programmes should be considered in future meta-analyses to clarify the influence of individual training programmes on inflammatory markers in subjects of different ages and with specific diseases.

Conclusions
In summary, this systematic review and meta-analysis of randomised control trials provides evidence that an endurance training programme is more beneficial in reducing CRP, IL-6, and visfatin concentrations in overweight and obese adults compared with a strength training programme. Additionally, a combined training programme appeared to be significantly more beneficial in lowering TNF-α levels compared with a strength training programme. Therefore, in the obese and overweight adult population, our findings suggest that training programmes including only strength exercise are the least appropriate for reducing inflammatory parameters and adipokine levels. However, we found no difference between the effects of different types of training on adiponectin and leptin concentrations, which may be related to the small number of studies included in the meta-analysis. Further randomised control trials need to be conducted to determine which type of training has a greater effect on the levels of these adipokines in the obese or overweight adult population.  Data Availability Statement: Template data collection forms, data extracted from included studies, data used for analysis, analytic code, and any other materials used in the review are available on reasonable request from the corresponding author (J.W.).