Effects Of Moderate-Frequency Resistance Training On Cardiometabolic Risk Factors In Adults With Overweight And Obesity: A Systematic Review And Meta-Analysis

1. Introduction

Over recent decades, significant environmental and lifestyle changes have contributed to the increasing prevalence of overweight and obesity [1,2]. Research strongly links these conditions to cardiometabolic health issues including, including hypertension, hyperglycemia, insulin resistance, high cholesterol, and elevated triglyceride levels [3]. These factors collectively heighten the risk of cardiovascular disease (CVD) and mortality, potentially shortening lifespan [4].

The term “cardiometabolic risk factors (CRFs)” encompasses all risk factors associated with diabetes and CVD, including being overweight and obese, elevated blood glucose and blood pressure, insulin resistance, and abnormal lipid metabolism [5]. Addressing all CRFs is essential due to their interplay, which exacerbates CVD vulnerability and contributes to its pathogenesis [5]. The American Heart Association has issued a statement advocating physical activity as a strategy to ameliorate CRFs [6].

Resistance training (RT) involves utilizing equipment such as resistance machines, weights, elastic bands, or one’s own body weight to improve muscle strength and endurance. This form of exercise primarily enhances skeletal muscle strength, muscle cross-sectional area, muscle fiber number, and cardiac pumping capacity, while also facilitating adaptive vasodilation [7,8]. Conversely, a reduction in skeletal muscle mass coupled with an increase in body fat (e.g., sarcopenic obesity) is detrimental to the well-being of adults with obesity. This scenario can lead to decreased physical performance, increased cardiometabolic risk, and increased likelihood of adverse clinical outcomes [9,10]. Notably, RT is the only non-pharmacological intervention that consistently counteracts age-related declines in skeletal muscle mass, strength, and power [11,12].

Aerobic exercise is known to enhance CRFs in populations with obesity. However, it usually requires prolonged exercise duration and a conductive environment, which may reduce exercise compliance among individuals with overweight and obesity [13]. Additionally, previous meta-analyses have shown that the effect of aerobic exercise on blood glucose depends on the appropriateness of the training load, which may be compromised if individuals are unable or unwilling to complete the prescribed training [14]. Conversely, RT does not encounter the constraints of space, equipment, or time commitments, redenring it a convenient option for individuals with overweight and obesity, whether at home or another location [15]. Indeed, RT aligns with global physical activity guidelines that advocate for 2-3 RT sessions per week for adults [16,17].

It is worth noting that in recent years, concurrent training (a combination of both aerobic training and RT) has exhibited efficacy in improving CRFs in adults [18,19]. However, concurrent training is time-demanding, and the interference effects between aerobic training and RT warrant careful consideration due to the possible compromise of skeletal muscle mass and function [20]. Recent studies have reported the positive effects of high-intensity interval training(HIIT) on CRFs in the general population [21], but these effects have not been validated in populations with overweight and obesity. Considering that overweight and obesity are often associated with various metabolic and other diseases, performing HIIT without professional supervision may increase the risk of injury or impair the function of certain organs in this population [22,23]. On the contrary, RT can be adjusted to individual perceptions and, when performed with an optimal dose interval, can mitigate barriers to participation and positively influence the intervention outcome [24]. Based on the advantages of RT, it becomes imperative to elucidate the effects of moderate-frequency RT (two or three times weekly) on cardiometabolic health outcomes, to support further research endeavors and the practical application of RT.

Currently, although some studies have reported CRFs as a secondary outcome, there is limited research employing it as the primary outcome for RT, especially in populations with overweight and obesity. Moreover, interpreting the impact of RT on CRFs is limited by methodological diversity, such as RT frequency and dietary regulation. Previous meta-analyses have explored the effects of short-, medium- and long-term RT on CRFs in adults, while ignoring the assessment of improved cardiometabolic health at the prevailing and recommended frequency of RT endorsed by the global physical activity guidelines [25]. A high-quality systematic review and meta-analysis can help overcome these challenges by providing more precise effect sizes while accounting for bias and heterogeneity. Therefore, the objective of this systematic review is to comprehensively assess the effects of moderate-frequency RT on CRF outcomes in adults with overweight and obesity, with a subgroup analysis examining the role of dietary control.

2. Methods

2.1. Protocol

This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist [26] and was registered with PROSPERO(ID: CRD42022343167).

2.2. Eligibility Criteria

To determine the inclusion criteria for the literature, we followed the PICOS model (see Table 1).

2.3. Information Sources and Search Strategy

A systematic search was conducted on PubMed, Cochrane Library, and Web of Science databases from inception until April 2023. Medical subject heading (MeSH) terms and text words related to overweight, obesity, resistance training, glucose metabolism, lipid metabolism, and randomized controlled trials were used to identify studies that could be potentially relevant. We also checked the reference lists of all trials that met the screening criteria, although no additional eligible trials were found.

The search strategy employed a combination of subject headings and free text words, adjusted for each database's characteristics. The search was conducted using four main categories of terms: (ⅰ) population; (ⅱ) intervention; (ⅲ) study type; and (ⅳ) outcome. For example, the specific search strategy used in PubMed is shown in Table 2. The reference lists of the included studies were also searched to supplement the obtained information.

2.4. Study Selection and Data Extraction

To identify relevant studies, we applied the following inclusion criteria: (i) reports that included clear participant information for body mass index (BMI) and age, adhering to the WHO definitions of overweight and obesity used (BMI ≥ 25 for overweight and BMI ≥ 30 for obesity); (ii) studies that included a control group, i.e., no additional physical exercise; and (iii) studies that employed a randomized controlled trial design. Exclusions were applied to studies that met the following criteria: (i) studies involving pregnant participants; (ii) studies involving exercise forms other than resistance training, such as a combination of aerobic exercise and resistance training; (iii) studies that used an acute resistance training intervention [25]; and (iv) studies that were published in a language other than English. After identifying studies that met the inclusion criteria, we first eliminated duplicate content. Subsequently, articles were subjected to title and abstract screening to eliminate those that did not meet the eligibility criteria. Finally, we carefully reviewed the full texts of these articles, excluding those that did not meet the inclusion criteria.

The data extraction process from the selected reference articles was performed using Microsoft Excel spreadsheets (Microsoft Corporation, Redmond, WA, USA). To ensure accuracy, another author scrutinized the data extraction, and any discrepancies were resolved through discussion. In situations where a consensus could not be reached between the two authors, a third person was consulted for resolution. The extracted data included information on study source (authors, publication year, country, and region), participant demographics (age, gender, and BMI), resistance training intervention details (total time, frequency, and intensity), and outcome measurements. If a study had multiple intervention arms (such as different resistance training doses, regimens, or participant populations), and at least one of these arms met the inclusion criteria, those individual arms were treated as distinct studies, termed trials. In addition, one author independently reviewed the titles and abstracts of all referenced articles, screening them against the inclusion criteria. Following the removal of duplicate articles, the other author further reviewed the full texts of the remaining articles. Any discrepancies between the two authors were resolved through discussion to reach a consensus.

2.5. Risk of Bias

We used the Cochrane Risk of Bias tool to assess the following aspects of bias risk: (1) random sequence generation (selection bias); (2) allocation concealment (selection bias); (3) blinding of participants and personnel (performance bias); (4) blinding of outcome assessment (detection bias); (5) incomplete outcome data (attrition bias); (6) selective reporting (reporting bias); and (7) other biases [27]. In parallel, we used funnel plots to assess potential publication bias [28].

2.6. Data Synthesis and Analysis

Meta-analysis was conducted using Review Manager (RevMan V.5.4) when more than two studies reported on the same outcome. Effect sizes were expressed as mean differences with 95% confidence intervals. To assess heterogeneity, both the I² statistic and p-value were used, with I²>50% and P<0.10 indicating significant heterogeneity. In instances of significant heterogeneity, a random-effects model was employed for analysis. Conversely, if heterogeneity was not significant, a fixed-effects model was applied.

Sensitivity analysis was conducted to examine the impact of individual studies on the overall effect size by systematically excluding each study from the analysis. If significant heterogeneity was detected, but no clinical heterogeneity was absent, a random-effects model was used for analysis. If significant heterogeneity was observed but its source could not be identified, descriptive analysis was performed. Furthermore, funnel plots were utilized to assess publication bias, with a significance level of α=0.05. The Z-test was used to determine statistical significance, with p<0.05 indicating a significant result.

3. Results

3.1. Search results

Initially, an extensive search across multiple databases yielded a total of 7011 references. After removing 573 duplicates and 2478 references using the Endnote automatic screening tool, 3925 references were excluded for other reasons such as animal experiments, systematic reviews, studies irrelevant to the research topics, and lack of full text availability. This left 35 references for a detailed screening of titles and abstracts. Subseqently, 12 articles were excluded due to the absence of resistance training or had non-pure resistance training interventions. One reference involved pregnant subjects, seven were written in languages other than English, and two lacked sufficient data for meaningful analysis. Ultimately, 13 articles, involving 651 adults with overweight and obesity, were included in the final analysis, with two trials featuring two resistance training groups (Refer to Figure 1 for a visual representation).

3.2. Characteristics of the included studies

Table 2 presents the basic characteristics of the included studies. These 15 randomized controlled trials encompassed 691 participants, with 343 in the control group and 348 in the resistance training group. The trials were conducted in 9 countries at various settings, such as hospitals, specialized research centers, universities, and fitness centers. The number of participants in the resistance training group varied from 7 to 62 individuals per study, with 5 studies only including female participants and 1 study only including male participants. All participants in the studies were adults with overweight and obesity, ranging from 21 to 71 years old.

The duration of resistance training interventions exhibited a range from 7 weeks to 24 weeks, with resistance training sessions performed two to three times per week. All resistance training involved multiple muscle groups in different body regions. The majority of resistance training entailed a progressive intensity approach ranging from 40% 1RM to 85% 1RM. The regimen encompassed 1 to 8 sets per session, each comprising 8 to 20 repetitions per set, with a rest period between each set.

Table 3. Summary of descriptive characteristics of included articles.

Author	Year	Samples(F/M)	Age(years)	Duration	Intervention	Primary outcomes
Dejan Reljic[29]	May-21	23/7	51.7 ± 11.7	12 weeks, twice a week	Weeks 1-4:50-60%1RM; Weeks 5-8:60-75%1RM; Weeks 9-12:70-80%1RM, one exercise per set	A, B, C, D, E, F, I, J, K
Dejan Reljic [29]	May-21	23/7	51.7 ± 11.7	12 weeks, three times a week	Weeks 1-4:50-60%1RM; Weeks 5-8:60-75%1RM; Weeks 9-12:70-80%1RM, three exercises per set	A, B, C, D, E, F, I, J, K
Javier Ibáñez [30]	Aug-09	12/0	51.4 ± 5.5	16 weeks, twice a week	During the first 8 weeks of the training period, subjects were loaded at 50 to 70% of the individual 1-RM, and during the last 8 weeks of the training period, the load was 70 to 80% of the maximum load. In addition, from week 8 to week 16, subjects performed a partial (20%) calf extension and bench press set-up with loads ranging from 30% to 50% of the maximum load.	A, F, G, H, I, J, K, L
Itamar Levinger [31]	Oct-07	5/9	51.9 ± 5.8	10 weeks, three times a week	With a 48-h recovery between sessions, the initial training intensity was two sets of 15-20 reps at 40-50% 1RM each. Starting in weeks 2-10, subjects performed three sets of each exercise at 50-85% 1RM for 8-20 repetitions.	A, B, C, F, I, J
Itamar Levinger [31]	Oct-07	6/4	48.9±7.4	10 weeks, three times a week	The total intervention duration was 10 weeks, with 48-h of recovery between sessions, and the initial training intensity was two sets of 15-20 reps at 40-50% 1RM. Starting in weeks 2-10, subjects performed three sets of each exercise at 50-85% 1RM for 8-20 repetitions.	A, B, C, F, I, J
Steven K [32]	Sep-13	7/0	20.9 ± 1.59	7 weeks, three times a week	At 60% 1-RM for 3 days/week and approximately 60 minutes/repetition, subjects performed three sets of 8-12 repetitions with 90-120 seconds of rest between sets.	A, F, L
Robert G. Memelink [33]	Dec-20	24/38	65.8 ± 6.4	13 weeks, three times a week	Total intervention length 13 weeks, three times a week, one hour each time	A, E, F, G, L
Ramin Shabani [34]	Apr-15	10/0	51.3 ± 6.63	12 weeks, three times a week	In the initial 1-3 weeks, the intensity is 40-50% 1-RM, 4-8 weeks, 50-65% 1-RM, with a total of 8 sets of movements, each set repeated 8-12 times, with 3 minutes rest between sets	F
D.W. Dunstan [35]	Feb-98	5/5	51.1 ± 2.2	8 weeks, three times a week	The total duration of the intervention was 8 weeks, three times a week, with a weight of 50-55% 1RM for each exercise.	B, C. F. L
Cíntia E. Botton [36]	Oct-18	8/14	68.6 ± 7.06	12 weeks, three times a week	The initial training load is determined during the familiarization training and increases with the load until a maximum of 15 repetitions is reached.	E, F, H, I, J, K
Sophie Drapeau [37]	Apr-11	22/0	58.5±4.6	24 weeks, three times a week	Phase 1: Start training (3 weeks, 15 repetitions, 2-3 sets each, 90-120 seconds between sets); Phase 2 (5 weeks, 12 repetitions, 2-3 sets each, 90 seconds between sets); Phase 3 (9 weeks, 8-10 repetitions, 2-4 sets each, 120-180 seconds between sets); Phase 4 (8 weeks, 10-12 repetitions, 3-4 sets each, 60-90 seconds between sets))	F, G, H, I, J, K, L
RC Plotnikoff [38]	Jun-10	13/8	54±12	16 weeks, three times a week	Total intervention duration of 16 weeks, 3 training sessions per week	A, B, C, E, F, I, J, K, L
Mika Venojärvi [39]	Aug-12	0/40	54 ±7.2	12 weeks, three times a week	Increase exercise intensity and load every 4 weeks	A, B, C, E, F, G, H, I, J, K, L
Carmen Castaneda [40]	Sep-02	19/12	66 ± 5.57	16 weeks, three times a week	For 45 minutes each time, subjects performed three sets of eight repetitions on each machine at a time.	A, B, C, F, H, J, K
Marisol García-Unciti [41]	Dec-12	12/0	51.4±5.5	16 weeks, twice a week	During the first 8 weeks of the training period, subjects were loaded at 50-70% of the individual 1-RM, and during the last 8 weeks of the training period, the load was 70-80% of the maximum load.	A, F, H, I, J, K, L

Note: Values are presented as means ± SD. F: female; M: male; A: Waist circumference (cm); B: Systolic blood pressure(mmHg); C: Diastolic blood pressure(mmHg); D: Mean arterial pressure(mmHg); E: Glycated hemoglobin (%); F: Fasting glucose (mmol·L^-1); G: HOMA-IR(Homeostasis model assessment for insulin resistance); H: Serum total cholesterol (mmol·L^-1); I: Triglyceride (mmol·L^-1); J: High-density lipoprotein (mmol·L^-1); K: Low-density lipoprotein (mmol·L^-1); L: Fasting insulin (pmol·L^-1).

Table 3. Summary of descriptive characteristics of included articles.

Author	Year	Samples(F/M)	Age(years)	Duration	Intervention	Primary outcomes
Dejan Reljic[29]	May-21	23/7	51.7 ± 11.7	12 weeks, twice a week	Weeks 1-4:50-60%1RM; Weeks 5-8:60-75%1RM; Weeks 9-12:70-80%1RM, one exercise per set	A, B, C, D, E, F, I, J, K
Dejan Reljic [29]	May-21	23/7	51.7 ± 11.7	12 weeks, three times a week	Weeks 1-4:50-60%1RM; Weeks 5-8:60-75%1RM; Weeks 9-12:70-80%1RM, three exercises per set	A, B, C, D, E, F, I, J, K
Javier Ibáñez [30]	Aug-09	12/0	51.4 ± 5.5	16 weeks, twice a week	During the first 8 weeks of the training period, subjects were loaded at 50 to 70% of the individual 1-RM, and during the last 8 weeks of the training period, the load was 70 to 80% of the maximum load. In addition, from week 8 to week 16, subjects performed a partial (20%) calf extension and bench press set-up with loads ranging from 30% to 50% of the maximum load.	A, F, G, H, I, J, K, L
Itamar Levinger [31]	Oct-07	5/9	51.9 ± 5.8	10 weeks, three times a week	With a 48-h recovery between sessions, the initial training intensity was two sets of 15-20 reps at 40-50% 1RM each. Starting in weeks 2-10, subjects performed three sets of each exercise at 50-85% 1RM for 8-20 repetitions.	A, B, C, F, I, J
Itamar Levinger [31]	Oct-07	6/4	48.9±7.4	10 weeks, three times a week	The total intervention duration was 10 weeks, with 48-h of recovery between sessions, and the initial training intensity was two sets of 15-20 reps at 40-50% 1RM. Starting in weeks 2-10, subjects performed three sets of each exercise at 50-85% 1RM for 8-20 repetitions.	A, B, C, F, I, J
Steven K [32]	Sep-13	7/0	20.9 ± 1.59	7 weeks, three times a week	At 60% 1-RM for 3 days/week and approximately 60 minutes/repetition, subjects performed three sets of 8-12 repetitions with 90-120 seconds of rest between sets.	A, F, L
Robert G. Memelink [33]	Dec-20	24/38	65.8 ± 6.4	13 weeks, three times a week	Total intervention length 13 weeks, three times a week, one hour each time	A, E, F, G, L
Ramin Shabani [34]	Apr-15	10/0	51.3 ± 6.63	12 weeks, three times a week	In the initial 1-3 weeks, the intensity is 40-50% 1-RM, 4-8 weeks, 50-65% 1-RM, with a total of 8 sets of movements, each set repeated 8-12 times, with 3 minutes rest between sets	F
D.W. Dunstan [35]	Feb-98	5/5	51.1 ± 2.2	8 weeks, three times a week	The total duration of the intervention was 8 weeks, three times a week, with a weight of 50-55% 1RM for each exercise.	B, C. F. L
Cíntia E. Botton [36]	Oct-18	8/14	68.6 ± 7.06	12 weeks, three times a week	The initial training load is determined during the familiarization training and increases with the load until a maximum of 15 repetitions is reached.	E, F, H, I, J, K
Sophie Drapeau [37]	Apr-11	22/0	58.5±4.6	24 weeks, three times a week	Phase 1: Start training (3 weeks, 15 repetitions, 2-3 sets each, 90-120 seconds between sets); Phase 2 (5 weeks, 12 repetitions, 2-3 sets each, 90 seconds between sets); Phase 3 (9 weeks, 8-10 repetitions, 2-4 sets each, 120-180 seconds between sets); Phase 4 (8 weeks, 10-12 repetitions, 3-4 sets each, 60-90 seconds between sets))	F, G, H, I, J, K, L
RC Plotnikoff [38]	Jun-10	13/8	54±12	16 weeks, three times a week	Total intervention duration of 16 weeks, 3 training sessions per week	A, B, C, E, F, I, J, K, L
Mika Venojärvi [39]	Aug-12	0/40	54 ±7.2	12 weeks, three times a week	Increase exercise intensity and load every 4 weeks	A, B, C, E, F, G, H, I, J, K, L
Carmen Castaneda [40]	Sep-02	19/12	66 ± 5.57	16 weeks, three times a week	For 45 minutes each time, subjects performed three sets of eight repetitions on each machine at a time.	A, B, C, F, H, J, K
Marisol García-Unciti [41]	Dec-12	12/0	51.4±5.5	16 weeks, twice a week	During the first 8 weeks of the training period, subjects were loaded at 50-70% of the individual 1-RM, and during the last 8 weeks of the training period, the load was 70-80% of the maximum load.	A, F, H, I, J, K, L

Note: Values are presented as means ± SD. F: female; M: male; A: Waist circumference (cm); B: Systolic blood pressure(mmHg); C: Diastolic blood pressure(mmHg); D: Mean arterial pressure(mmHg); E: Glycated hemoglobin (%); F: Fasting glucose (mmol·L^-1); G: HOMA-IR(Homeostasis model assessment for insulin resistance); H: Serum total cholesterol (mmol·L^-1); I: Triglyceride (mmol·L^-1); J: High-density lipoprotein (mmol·L^-1); K: Low-density lipoprotein (mmol·L^-1); L: Fasting insulin (pmol·L^-1).

3.3. Risk of bias

3.3.1. Selection bias

All studies employed random allocation for group assignment, with 7 randomized controlled trials using a describable method for sequence generation, and the remaining 6 studies having unclear risk due to insufficient information on randomization methods. The majority of studies did not report allocation concealment and were considered to have unclear risk.

3.3.2. Performance and detection bias

All studies exhibited a notable susceptibility to performance bias (i.e., lack of blinding among intervention and outcome assessors). This absence of blinding among researchers could have introduced potential biases into the measurement of blood pressure and the extent of improvement in some indicators such as glucose and lipid metabolism among participants. Most studies were judged as unclear for detection bias (i.e., blinding of outcome assessors) due to insufficient information provided in the studies. Two studies stood out with a pronounced risk of detection bias.

3.3.3. Attrition bias

The majority of studies were categorized as having a low risk of attrition bias. However, three trials received a high-risk rating due to either high dropout rates or the exclusion of specific participants from the analysis.

3.3.4. Reporting bias

All studies were rated as low risk for selective reporting bias.

3.3.5. Publication bias

All measurements were plotted in funnel plots, which were asymmetrical, indicating potential publication bias.

Figure 2. Results of the risk of bias assessment of the 15 studies included in this systematic review. Green: low risk of bias. Yellow: unclear risk of bias. Red: high risk of bias.

Figure 2. Results of the risk of bias assessment of the 15 studies included in this systematic review. Green: low risk of bias. Yellow: unclear risk of bias. Red: high risk of bias.

Figure 3. Results of the risk of bias assessment of the 15 studies included in this systematic review. Green: low risk of bias. Yellow: unclear risk of bias. Red: high risk of bias.

Figure 3. Results of the risk of bias assessment of the 15 studies included in this systematic review. Green: low risk of bias. Yellow: unclear risk of bias. Red: high risk of bias.

3.4. Study outcomes

Waist

Figure 4 illustrates the waist circumference data encompassing 550 participants from 11 trials. Evaluating the effect of resistance training on waist circumference in adults with overweight and obesity, no heterogeneity was observed among the studies (I² =0%, p=0.5). Employing a fixed-effects model for analysis yielded a combined effect size of MD = 0.27, with 95% CI spanning from = 0.10 to 0.44 (p=0.002). Subgroup analyses showed no between-study heterogeneity in the dietary control group, with a total effect size of MD = 0.30, and 95%CI of = 0.13 to 0.47 (p＜0.001); there was no heterogeneity in the no-dietary-control group, with a total effect size of MD = of -1.33, and 95%CI from = -2.62 to -0.03 (p=0.04). These findings suggest that resistance training does not manifest a favorable impact on the enhancement of waist circumference in adults with overweight and obesity.

Systolic blood pressure

Figure 5 presents the systolic blood pressure data encompassing 372 participants from 8 trials. Exploring the effect of resistance training on systolic blood pressure in adults with overweight and obesity, heterogeneity was observed among the studies (I²=60%, p=0.01). Employing a random-effects model for the analysis, a combined effect size of MD = -4.66 is derived, accompanied by a 95% CI of = -9.34 to 0.02 (p=0.05). These findings suggest that resistance training has a positive effect on improving systolic blood pressure in adults with overweight and obesity.

Diastolic blood pressure

Figure 6 presents the diastolic blood pressure data encompassing 372 participants from 8 trials. Delving into the effect of resistance training on diastolic blood pressure in adults with overweight and obesity, no heterogeneity was observed among the studies (I²=0%, p=0.8). The analysis adopted a fixed-effects model, yielding a combined effect size of MD = -1.32, with 95% CI of = -3.03 to 0.38 (p=0.13). These findings indicate that resistance training could potentially yield a favorable influence on improving diastolic blood pressure in adults with overweight and obesity.

Mean arterial pressure

Figure 7 illustrates the mean arterial pressure data involving 116 participants from 2 trials. Exploring the potential effect of resistance training on mean arterial pressure in adults with overweight and obesity, no heterogeneity was observed among the studies (I²=0%, p=0.81). Employing a fixed-effects model for analysis, a combined effect size of MD = -6.48 is derived, with 95% CI of = -10.63 to -2.33 (p=0.002). These findings indicate that resistance training exerts a significant positive effect on the enhancement of mean arterial pressure in adults with overweight and obesity.

Glycated hemoglobin (HbA_1c ^(%))

Figure 8 illustrates the glycated hemoglobin (HbA_1c) data derived from 418 participants across 6 trials. In examining the effect of resistance training on HbA_1c levels in adults with overweight and obesity, a substantial degree of heterogeneity was observed among the studies (I²=89%, p＜0.001). Employing a random-effects model, the resulting combined effect size is MD = 0.2, with a 95% CI spanning from = -0.33 to 0.73 (p=0.46). Subgroup analyses revealed a total effect size of MD = 1.30, with 95%CI from = -2.40 to 5.00 (p=0.49) in the dietary control group; notable heterogeneity (I²=91%, p＜0.001) is observed in the no-dietary-control group, with a total effect size of MD = 0.18,and 95%CI spanning from = -0.36 to 0.71 (p=0.52). These findings indicate that resistance training does not yield a favorable impact on the enhancement of HbA_1c levels in adults with overweight and obesity.

Fasting glucose

Figure 9 illustrates the fasting glucose data from 681 participants across 15 trials. In examining the effect of resistance training on fasting glucose levels in adults with overweight and obesity, no heterogeneity was observed among the studies (I²=0%, p=0.6). Utilizing a fixed-effects model for analysis, the resultant combined effect size stands at MD=-0.01, with a 95% CI spanning from = -0.08 to 0.05 (p=0.75). Subgroup analyses revealed a notable level of between-study heterogeneity (I²=67%, p=0.05) in the dietary control group, with a total effect size of MD = -0.50, along with a 95%CI spanning from = -0.24 to 0.14 (p=0.6); there was no heterogeneity within the no-dietary-control group, with a total effect size of MD = -0.01, and a 95%CI ranging from = -0.07 to 0.06 (p=0.88). These findings indicate that resistance training does not yield a positive influence on the enhancement of fasting glucose levels in adults with overweight and obesity.

Homeostasis model assessment for insulin resistance

Figure 10 illustrates the homeostasis model assessment for insulin resistance (HOMA-IR) data encompassing 277 participants from 4 trials. In examining the effect of resistance training on HOMA-IR in adults with overweight and obesity, a moderate level of heterogeneity was observed among the studies (I²=48%, p=0.12). Employing a fixed-effects model for analysis, the resulting combined effect size is MD = -1.49, with a 95% CI spanning from = -1.63 to -1.35 (p＜0.001). Subgroup analyses revealed no between-study heterogeneity in the dietary control group, with a total effect size of MD = -1.52, accompanied by a 95%CI ranging from = -1.66 to -1.38 (p＜0.001); within the no-dietary group a total effect size of MD = -0.6 is observed, with a 95%CI spanning from = -1.41 to 0.21 (p=0.15). These findings indicate that resistance training exerts a significant and positive effect on the enhancement of HOMA-IR in adults with overweight and obesity.

Total cholesterol

Figure 11 illustrates the total cholesterol data derived from 278 participants spanning 6 trials. In examining the effect of resistance training on total cholesterol in adults with overweight and obesity, a substantial degree of heterogeneity was observed among the studies (I²=63%, p=0.02). Utilizing a random-effects model for analysis, the resulting combined effect size is MD=-0.53, with a 95% CI ranging from = -1.14 to 0.08 (p=0.09). Subgroup analyses revealed no between-study heterogeneity in the dietary control group, with a total effect size of MD = -1.94, alongside a 95%CI ranging from = -3.07 to -0.82 (p＜0.001); within the no-dietary-control group low level of heterogeneity (I²=10%, p=0.33) is observed, with a total effect size of MD = 0.00, alongside a 95%CI ranging from = -0.26 to 0.27 (p=0.98). These findings indicate that resistance training could potentially yield a favorable impact on the enhancement of total cholesterol in adults with overweight and obesity.

Total triglycerides

Figure 12 illustrates total triglyceride data from 419 participants across 10 trials. In examining the effect of resistance training on total triglyceride levels in adults with overweight and obesity, no heterogeneity was observed among the studies (I²=0%, p=0.83). Utilizing a fixed-effects model for analysis, a combined effect size of MD = -0.03 is observed, with a 95% CI ranging from = -0.20 to 0.14 (p=0.72). Subgroup analyses revealed no between-study heterogeneity in the dietary control group, with a total effect size of MD = 0.30, alongside a 95%CI ranging from = -0.96 to 1.56 (p=0.64); no heterogeneity is observed within the no-dietary-control group, with a total effect size of MD = -0.04, alongside a 95%CI ranging from = -0.21 to 0.13 (p=0.67). These findings indicate that resistance training does not yield a favorable impact on the enhancement of total triglyceride levels in adults with overweight and obesity.

High-density lipoprotein

Figure 13 illustrates high-density lipoprotein data from 491 participants across 11 trials. In examining the effect of resistance training on high-density lipoprotein levels in adults with overweight and obesity, a low degree of heterogeneity was observed among the studies (I²=27%, p=0.19). Subgroup analyses yielded no between-study heterogeneity in the dietary control group, with a total effect size of MD = -0.49, alongside a 95%CI ranging from = -0.82 to -0.17 (p=0.002); no heterogeneity was observed within the no-dietary-control group, with a total effect size of MD = 0.05, alongside a 95%CI ranging from = -0.04 to 0.15 (p=0.25). Utilizing a fixed-effects model for analysis, a combined effect size of MD = 0.01 was observed, with a 95% CI ranging from = -0.08 to 0.10 (p=0.79). These findings indicate that resistance training does not yield a favorable impact on the enhancement of high-density lipoprotein levels in adults with overweight and obesity.

Low-density lipoprotein

Figure 14 illustrates low-density lipoprotein data from 442 participants across 9 trials. In examining the effect of resistance training on low-density lipoprotein levels in adults with overweight and obesity, a substantial degree of heterogeneity was observed among the studies (I²=69%, p=0.001). Utilizing a random-effects model for analysis, a combined effect size of MD = -0.06 was observed, with a 95% CI ranging from = -0.49 to 0.37 (p=0.79). Subgroup analyses yielded high between-study heterogeneity (I²=88%, p＜0.001) in the dietary control group, with a total effect size of MD = -0.29, alongside a 95%CI ranging from = -3.07 to 2.48 (p=0.84); low degree of heterogeneity (I²=42%, p=0.12) was observed within the no-dietary-control group, with a total effect size of MD = -0.09, alongside a 95%CI ranging from = -0.39 to 0.21 (p=0.57). These findings indicate that resistance training does not yield a favorable impact on the enhancement of low-density lipoprotein in adults with overweight and obesity.

Fasting insulin

Figure 15 illustrates the fasting insulin data from 390 participants across 8 trials. In examining the effect of resistance training on fasting insulin levels in adults with overweight and obesity, a substantial degree heterogeneity was observed among the studies (I²=98%, p＜0.001). Utilizing a random-effects model for analysis, a combined effect size of MD = -12.52 was observed, with a 95% CI ranging from = -24.17 to -0.88 (p=0.04). Subgroup analyses yielded high between-study heterogeneity (I²=99%, p＜0.001) in the dietary control group, with a total effect size of MD = -8.56, alongside a 95%CI ranging from = -26.37 to 9.25 (p=0.35); a moderate degree of heterogeneity (I²=51%, p=0.11) was observed within the no-dietary-control group, with a total effect size of MD = -20.51, alongside a 95%CI ranging from = -40.86 to -0.17 (p=0.05). These findings indicate that resistance training has a significant and positive effect on the enhancement of fasting insulin levels in adults with overweight and obesity.