Physical Activity during Pregnancy and Childhood Obesity: Systematic Review and Meta-Analysis

Background and Objectives: The repercussions of childhood overweight and obesity are multifaceted, extending beyond the realm of physiology and giving rise to psychological and emotional disturbances in affected children. The precise effects of gestational physical activity (PA) on parameters related to childhood overweight and obesity remain inadequately understood. The aim of this study (Registration CRD42022372490) was to evaluate the literature regarding the influence of PA during pregnancy on the risk of childhood overweight and obesity. Materials and Methods: Only randomized controlled trials (RCTs) were considered for inclusion. Determinant parameters of childhood obesity were analyzed. A total of 30 studies involving 16,137 pregnant women were examined. Five meta-analyses about the effects of PA during pregnancy on determinants of childhood overweight and obesity were conducted. Results: Although favorable trends were observed, Meta-Analyses showed no statistical differences in the effects of PA on weight at birth (Z = 0.03, p = 0.97), Ponderal Index at birth (Z = 0.64, p = 0.52), Macrosomia and Large for Gestational Age at birth (Z = 0.93, p = 0.35), children’s BMI (Z = 0.78, p = 0.44), weight (Z = 0.50, p = 0.62), and skinfold thicknesses (Z = 0.45, p = 0.65). Conclusions: The engagement in physical activity during pregnancy exhibits a favorable trend in parameters associated with childhood overweight and obesity while presenting no adverse effects on such outcomes.


Introduction
The growing percentage of children who are overweight and obese has emerged as a pervasive global concern, exhibiting a relentless upward trend that poses formidable challenges to mitigation efforts.Globally, in 2022, an estimated 37 million children under 5 years were classified as overweight.What was once perceived as primarily a concern in countries with high incomes has now become more prevalent in low-and middle-income territories.In Africa, the prevalence of overweight children under 5 years of age has surged by nearly 23% since 2000.Notably, nearly half of the children under 5 years of age who were overweight or obese in 2022 resided in Asia [1].
The repercussions of childhood overweight and obesity are multifaceted, extending beyond the realm of physiology and giving rise to psychological and emotional disturbances in affected children.Remarkably, chronic conditions that were historically prevalent only in adults are now increasingly diagnosed among children afflicted by obesity [2,3].Such conditions encompass hypertension [4], type 2 diabetes, asthma [5], sleep apnea [6], joint disorders, and musculoskeletal pain [7], as well as gastrointestinal and hepatic complications [2,8,9].Additionally, obesity can profoundly impact their psychological and emotional well-being, leading to conditions such as depression, diminished self-esteem [10][11][12], and body dissatisfaction [2,12].
Within the realm of potential causative factors, a substantial body of scientific literature grounded in an epigenetic framework delves into the genesis of this issue during the gestational period.It substantiates the notion of an unfavorable intrauterine environment, particularly in the metabolic context, as a paramount factor intricately linked to the burgeoning epidemic of childhood obesity.This assertion finds robust support in a plethora of studies dedicated to this subject matter [13][14][15][16].
This scientific evidence underscores the pivotal role of the gestational period in shaping the health of future populations.In this context, the repercussions of a healthy or imbalanced pregnancy extend far beyond the gestational period, affecting both the mother and the offspring in detectable ways [17].
Recent research indicates that the perinatal environment contributes to childhood obesity, irrespective of genetic predisposition and dietary habits [18,19].Consequently, the pregnancy period emerges as a critical window for early prevention [20,21].
In this context, research suggests that fetal exposure to a metabolically adverse intrauterine environment is linked to subsequent complications in newborns.Indeed, specific studies even posit that genetic alterations in metabolic regulation during fetal life might impact the appetite, energy regulation, and metabolism of the newborn [22,23].
From a metabolic standpoint, a pivotal determinant for maintaining a healthy intrauterine environment is the significant aspect of appropriate maternal weight gain during pregnancy, which is closely associated with a woman's pre-pregnancy BMI.The imbalance of these factors, coupled with their attendant complications, such as gestational diabetes (GD) and hypertension, is strongly correlated with an elevated risk of childhood obesity [24][25][26][27][28][29][30].Current data indicate concerning trends.Approximately 34% of pregnant women exhibit a BMI exceeding 25 kg/m 2 [31], although some studies suggest an even higher percentage nearing 50% [29,32].Notably, 60% of pregnant women with obesity and 40% with normal weight surpass the recommendations established by the Institute of Medicine [33], amplifying the aforementioned risks.
Children born to mothers who are obese or overweight during pregnancy face elevated risks, including an increased likelihood of developing insulin resistance [42], requiring neonatal intensive care, being born preterm, having congenital anomalies, and needing treatment for hypoglycemia [32,36,38,41].
It is imperative to develop and implement strategies to manage maternal weight gain during pregnancy effectively, along with exploring other relevant epigenetic markers, to mitigate the escalating epidemic of childhood overweight and obesity.Achieving this goal necessitates a fundamental shift in maternal behaviors during pregnancy.The objective is to identify practical interventions that can substantially enhance the lifestyle of pregnant individuals, thereby fostering the well-being of future generations in the long term [43].
Engaging in physical activity during pregnancy, including exercises such as walking, swimming, yoga, and low-impact aerobics, offers numerous benefits.It is recommended that pregnant individuals accumulate moderate-intensity aerobic physical activity with a minimum of three days per week, although daily activity (avoiding a sedentary lifestyle) is encouraged to ensure consistent benefits and minimize the risk of injury [44].
In this context, over the last three decades, the practice of physical activity through pregnancy has emerged as a preventive factor against pre-, peri-, and postnatal complications, a fact substantiated by substantial scientific evidence.The recent recommendation by the World Health Organization [45], advocating for at least 150 min of moderate-intensity physical activity weekly for all healthy pregnant women to promote well-being, exemplifies this reality.However, the global data indicate a low prevalence of physical activity among pregnant women [46] which apparently could be directly associated with the increasing prevalence of childhood obesity.
The impact of prenatal physical activity on reducing childhood obesity rates has been well-documented.However, despite significant research supporting this trend, a broad and concrete consensus has not yet been established [15,17].Rigorous review studies of the scientific literature on this topic are necessary to draw solid conclusions [14].In this context, systematic reviews and meta-analyses of experimental studies, particularly those examining randomized controlled trials (RCTs), are considered the most reliable type of study.Given this context, the objective of this study was to systematically analyze the published literature regarding the influence of physical activity during pregnancy on childhood overweight and obesity risk.

Materials and Methods
A Systematic Review and Meta-Analysis (SR + MA) design was carried out based on the PRISMA guidelines [47].
Search strategy, database information, and data extraction were registered in PROS-PERO, Registration No. CRD42022372490.

Eligibility Criteria
The PICOS framework was used to set the search strategy as registered in the protocol [47].

Population
The study population comprised healthy pregnant women without any contraindications to participate in a physical exercise intervention.Absolute contraindications were delineated as conditions to stop physical activity due to imminent risk for the mother or the fetus [44].

Intervention (Exposure)
Information on the characteristics of the intervention has been examined: type of intervention, supervision class, duration of class, and adherence of the participants to the intervention (Table 1).

Comparison
The comparator was defined as no exercise, which constituted the control group in the selected studies.Typically, this involved pregnant women undergoing regular follow-ups during pregnancy at health centers.

Outcome
The main outcome of the study was childhood obesity or directly related parameters and determinants of childhood obesity such as birth weight, newborn body mass index, macrosomia, Ponderal Index, or Large for Gestational Age (LGA).Relevant secondary variable outcomes were other parameters of the children (cardiovascular, motor, mental).Regarding birth weight, only studies that examined this parameter in relation to other childhood parameters were included.

Study Design
Only randomized controlled trials (RCTs) were considered for inclusion.Other types of study design were excluded from analysis in this review.

Data Sources
The initial search was performed by three different reviewers (

Study Selection and Data Extraction
In addition to RCTs, we conducted a complementary search for previously published systematic reviews in the same field to facilitate a comparative analysis of our results.RCTs published from 2010 to March 2024, in both languages, were evaluated, as we tried to focus on the latest scientific publications to update current knowledge on this topic.
Three researchers conducted an abstract/title search to identify exercise interventions during gestation.The main objective was to extract pertinent details to ensure that the resulting articles met the eligibility criteria, thus enabling the full-text reviews of these valid articles to be considered going forward.
After that, two researchers (GD and PL) screened the full-text files to evaluate the validity of the selected articles and to filter results of interest for data extraction.In cases where a study had multiple publications, we prioritized the most recent as the primary source.
In situations of one recommended article exclusion, both people endeavored to reach a consensus to have a final decision regarding its inclusion.In cases of persistent disagreement, a third person (RM) provided the opinion on whether the study should be included or excluded.
Data extraction was performed through an Excel spreadsheet.One researcher (MS-P) performed the extraction of the data, which was subsequently verified by a content expert to facilitate further analysis.A manual search of reference lists and relevant journals was performed to ensure a comprehensive review.

Statistical Analysis
Version 5.4 of the RevMan statistical package was used.Two distinct meta-analyses were conducted: one involving continuous variables, where the overall confidence interval (CI) was computed using the standardized mean difference, and another involving categorical variables (yes/no), where the number of events in each study group and its relative risk (RR) was documented.The total RR was calculated using a fixed effects model, accounting for the weight assigned to each study based on its sample size, thus establishing a compensated average [48].
In both analyses, the I 2 statistic was employed to quantify the heterogeneity resulting from different interventions and study designs in each article.This statistic measures the variability in the effect of each intervention and is non-random.Heterogeneity was categorized as follows: low (25%), moderate (50%), and high (75%) [49].

Quality of Evidence and Risk of Bias Assessments
To assess the quality of analyzed evidence through selected articles, two reviewers (DZ and CS) utilized the GRADE framework in its website tool (GRADEpro) version [50].Results of GRADE Assesment were included as Supplementary Materials.This tool, by inputting the results of performed analyses and the risk of bias of selected articles, allows researchers to determine the certainty of evidence and the importance of each outcome.
For evaluating the risk of bias (RoB) of analyzed articles, the RoB version 2 of the Cochrane Handbook was used, through RevMan software version 5.4.Potential sources of bias assessed included selection bias (including the generation of the sequence of randomization evaluation and how was the participant allocation), performance bias (compliance with the intervention for RCTs, i.e., blinding involving personnel or/and participants), detection bias (flawed outcome measurement, if exists outcome blinding), attrition bias (incomplete outcome data or a high rate of lost to follow-up), and reporting bias (selective or incomplete outcome reporting) [51].
To evaluate publication bias in each analysis, the Egger regression test was utilized due to its sensitivity in detecting publication bias in all included articles.Typically, significant publication bias is indicated when p < 0.1 [52].LGA, BMI. 4 months: BMI, Weight-for-length 6 months: BMI, Weight-for-length.12 months: BMI, Weight-for-length.

Study Selection
A total of 317 references were initially evaluated during the first search, thus removing 208 before screening: duplicate records (n = 172) and other reasons (n = 36).A total of 59 articles were excluded due to not meeting inclusion criteria (n = 57) and reports not retrieved (n = 2).Subsequently, 20 articles were not included due to the following reasons: being a review article (n = 3) or observational study (n = 5), no variable of interest (n = 6), or inadequate population (n = 6).The detailed study selection process is illustrated in Figure 1.

Risk of Bias Assessment
The collective certainty of the evidence was established as "high" and the importance of each outcome was established as "important".Regarding RoB analysis, it was observed that blinding of participants to each group is typically unfeasible due to the characteristics of the physical activity intervention, resulting in unclear or high risk of bias depending on how it was reported.In certain instances, other sources of bias included the inability to access the published article protocol (for comparing planned and measured outcomes) and insufficient transparency in reporting the randomization.Overall, the studies exhibited a low risk of bias across the types assessed.The risk of bias analysis is depicted individually in each of the following meta-analyses.Ultimately, 30 studies were identified that met the inclusion criteria, with 16,137 women from 11 countries and three continents.All studies were RCTs, and the intervention used was supervised physical exercise [53,56,68,69,74] or advice and recommendations [54,55,[57][58][59][60][61][62][63][64][65][66][67][70][71][72][73][75][76][77][78][79][80][81][82] (Table 1).

Risk of Bias Assessment
The collective certainty of the evidence was established as "high" and the importance of each outcome was established as "important".Regarding RoB analysis, it was observed that blinding of participants to each group is typically unfeasible due to the characteristics of the physical activity intervention, resulting in unclear or high risk of bias depending on how it was reported.In certain instances, other sources of bias included the inability to access the published article protocol (for comparing planned and measured outcomes) and insufficient transparency in reporting the randomization.Overall, the studies exhibited a low risk of bias across the types assessed.The risk of bias analysis is depicted individually in each of the following meta-analyses.

PA Effect during Pregnancy on Ponderal Index at Birth
A total of six studies were included in this analysis [56,59,64,67,71].The findings indicate that engaging in regular exercise during pregnancy does not exhibit a significant

PA Effect during Pregnancy on Ponderal Index at Birth
A total of six studies were included in this analysis [56,59,64,67,71].The findings indicate that engaging in regular exercise during pregnancy does not exhibit a significant association with Ponderal Index at birth (Z = 0.64; p = 0.52) (Standardized Mean Difference, Random Effects Model, 95% Confidence Interval = −0.05(−0.22,0.11), I 2 = 52%, P heterogeneity = 0.06).Figure 3 shows the corresponding graphic of this Meta-Analysis.The evaluation of publication bias across the reviewed articles indicated minimal evidence of bias, with a p-value of 0.48.

Effect of PA during Pregnancy on Macrosomia and Large for Gestational Age at Birth
This analysis included a total of 12 studies [57,58,62,67,[70][71][72][73][74]76,78,80].There was no significant association (Z = 0.93; p = 0.35) observed between exercise during gestation and the likelihood of Macrosomia and Large for Gestational Age at birth (Relative Risk = 0.95, 95% Confidence Interval = 0.86, 1.06,I 2 = 55%, Pheterogeneity = 0.003).Figure 4 illustrates the corresponding graphic for the current meta-analysis.The assessment of bias in publication across the analyzed articles revealed negligible evidence of bias, as reflected by a p-value of 0.09.

Effect of PA during Pregnancy on Macrosomia and Large for Gestational Age at Birth
This analysis included a total of 12 studies [57,58,62,67,[70][71][72][73][74]76,78,80].There was no significant association (Z = 0.93; p = 0.35) observed between exercise during gestation and the likelihood of Macrosomia and Large for Gestational Age at birth (Relative Risk = 0.95, 95% Confidence Interval = 0.86, 1.06,I 2 = 55%, Pheterogeneity = 0.003).Figure 4 illustrates the corresponding graphic for the current meta-analysis.The assessment of bias in publication across the analyzed articles revealed negligible evidence of bias, as reflected by a p-value of 0.09.

Effect of PA during Pregnancy on Children's BMI (1 Month-7 Years)
A total of 11 studies were included in this analysis [53,54,60,61,65,68,69,71,74,78,80].The findings indicate that engaging in a physical activity program across pregnancy does not exhibit a significant association with children´s BMI (Z = 0.78; p = 0.44) (Standardized Mean Difference, Random Effects Model, 95% Confidence Interval = 0.04 (−0.06, 0.15), I 2 = 63%, P heterogeneity = 0.003).Figure 5 illustrates the corresponding plot of the Meta-Analysis.Analysis of the publication bias risk in the surveyed articles showed no significant evidence of bias, evidenced by a p-value of 0.78.

Effect of PA during Pregnancy on Children's BMI (1 Month-7 Years)
A total of 11 studies were included in this analysis [53,54,60,61,65,68,69,71,74,78,80].The findings indicate that engaging in a physical activity program across pregnancy does not exhibit a significant association with children s BMI (Z = 0.78; p = 0.44) (Standardized Mean Difference, Random Effects Model, 95% Confidence Interval = 0.04 (−0.06, 0.15), I 2 = 63%, Pheterogeneity = 0.003).Figure 5 illustrates the corresponding plot of the Meta-Analysis.Analysis of the publication bias risk in the surveyed articles showed no significant evidence of bias, evidenced by a p-value of 0.78.

Discussion
Our study aimed to discern the impact of physical activity during pregnancy on childhood obesity indicators, such as birthweight, BMI, skinfold thickness, and the incidence of macrosomia and LGA outcomes.To the best of our knowledge, this represents the first SR + MA conducted on this particular scientific subject.
The absence of notable differences in birthweight and BMI among children born to exercising mothers during pregnancy invites contemplation of underlying biological and physiological mechanisms.Exercise may modulate maternal energy balance without disrupting fetal growth parameters, possibly through adaptive responses that preserve fetal nourishment.This equilibrium potentially reflects evolved maternal capacities to support fetal development despite varied energy expenditures.Further exploration into how maternal exercise alters gestational resource allocation may illuminate the subtleties of this relationship.
The present meta-analysis indicates that maternal physical activity during pregnancy does not result in a significant difference in fetal birthweight compared to non-exercising controls, further supporting the growing consensus that maternal exercise is a safe practice for fetal development.
This lack of significant impact on fetal weight may be attributed to the adaptive capabilities of the placenta to regulate fetal nutrition and growth, as hypothesized by various researchers [83].It is plausible that exercise-induced enhancements in placental function may offset potential changes in fetal growth parameters [84].Moreover, the homeostatic mechanisms of the mother's body during pregnancy are known to be highly efficient in maintaining energy balance.Even vigorous exercise in active women showed reassuring fetal Doppler indices, heart tracings, and biophysical profiles post-exercise, supporting the safety of continuing or initiating moderate to vigorous exercise during pregnancy [85].
Nevertheless, the substantial heterogeneity observed (I 2 = 60%) suggests variability among the included studies, which could be due to differences in the type, frequency, duration, and intensity of the exercise regimes prescribed.Additionally, the potential impact of maternal nutrition and pre-pregnancy BMI on birthweight outcomes cannot be overlooked as these factors have been shown to have a significant influence on fetal growth [86].
Our findings suggest that engagement in physical activity during pregnancy does not significantly affect the likelihood of macrosomia.The pooled odds ratio (OR) for macrosomia across studies was close to 1, suggesting no substantial difference between the intervention (exercise) and control groups.This aligns with previous research suggesting that moderate exercise during pregnancy is not associated with an increased risk of delivering a macrosomic infant [59,67].Similarly, the incidence of LGA infants did not significantly differ between the two groups.These results are particularly reassuring given the concerns regarding the potential for maternal exercise to alter fetal growth patterns.
In examining the longer-term effects of prenatal exercise, our analysis did not reveal significant differences in BMI during childhood (ages 5-7 years) or in BMI at birth.These findings contribute to the growing body of evidence suggesting that prenatal physical activity does not adversely affect child growth trajectories as measured by BMI.
Moreover, the standard mean differences (SMD) for birth weight and BMI indices in infancy and early childhood were negligible, further underscoring the safety profile of prenatal exercise with regard to these outcomes.The heterogeneity observed in some of the analyses was moderate, indicating that, while there may be variations across individual studies, the overall trend suggests no adverse impact.
The results of this meta-analysis also support current recommendations encouraging pregnant women to engage in regular, moderate-intensity physical activity, as it does not seem to compromise neonatal weight or childhood development, which are important predictors of health in later life.This is consistent with guidelines from health authorities that advocate for the benefits of physical activity during pregnancy for both maternal and fetal health [44,47].
Previous studies have offered varying results, some suggesting that exercise during pregnancy may reduce the risk of obesity in offspring, while others observed no effect.Our results align with the latter, adding to the narrative that exercise is a nuanced factor in the gestational period.Discrepancies across studies could be attributed to variability in exercise regimes, intensity, and duration, as well as methodological differences, such as the timing of outcome measurement and the specific populations studied.
The burgeoning prevalence of obesity in pregnant women and its ramifications not just for maternal health but for the lifelong health of offspring demand multifaceted intervention strategies.
The synthesis of these findings spotlights a critical intervention point: the promotion of exercise during pregnancy.Exercise has been touted for its broad spectrum of health benefits, including weight management, improved cardiovascular health, and enhanced mood and energy levels.By extrapolation, engaging in regular physical activity during pregnancy could hypothetically counteract the factors contributing to the development of childhood obesity.For instance, if maternal obesity can induce alterations in gut microbiota that predispose children to obesity, as outlined by Isolauri et al. (2016) [16], then the potential of exercise to improve gut health and reduce inflammation could be a key to breaking this cycle.
Exercise during pregnancy might also mitigate the risk factors outlined by Fraser et al. (2010) [28], who report that maternal weight gain is associated with obesity and adverse metabolic traits in childhood.By potentially moderating gestational weight gain, exercise could reduce the risks of delivering LGA infants, a concern raised in the consensus by Poston et al. (2011) [30], which also notes the lifelong health impact on the child due to maternal obesity.
Furthermore, considering Athukorala et al.'s (2010) findings [29], which connect adverse pregnancy outcomes with maternal overweight and obesity, exercise might serve as a preventive measure against such outcomes.Additionally, it could alleviate some of the gestational diabetes risks that Gillman et al. (2003) [26] indicate as precursors to adolescent obesity.
Therefore, promoting physical activity during pregnancy could be a strategic public health recommendation.Such an approach aligns with the view that preventive strategies focusing on maternal health may offer a viable route to curb the transgenerational perpetuation of obesity.While further research is required to establish a definitive causal link, the collective evidence points toward exercise during pregnancy as a promising intervention for improving both maternal and child health outcomes.
In our study, we leveraged a robust meta-analytic methodology which allowed us to integrate data across multiple studies enhancing the statistical power and reliability of our findings due to the large, combined sample size.The comprehensive nature of our meta-analysis, encompassing a broad spectrum of exercise types, intensities, and durations during pregnancy, contributes to a more generalized understanding of the effects of exercise on neonatal outcomes.
However, there are several limitations that must be acknowledged.The heterogeneity observed among the included studies is noteworthy.Variability in exercise regimens-ranging from type and intensity to duration-can significantly affect the comparability of the outcomes.This heterogeneity might limit the specificity of our conclusions regarding which exercise parameters are most beneficial during pregnancy.Furthermore, the quality of data available from the included studies is a concern, especially when considering that many relied on self-reported exercise, which is susceptible to recall bias.This method of data collection may not accurately capture the true exercise behaviors of the participants and could skew the associations observed.
These limitations should be carefully considered when interpreting the results.While our analysis suggests certain trends and associations, the impact of the aforementioned factors on our conclusions cannot be overlooked.Future research would benefit from more standardized protocols for exercise interventions during pregnancy and more objective measures of physical activity to reduce the risk of bias and improve the accuracy of findings.
Future research should aim to establish standardized exercise protocols during pregnancy to better quantify the relationship between maternal exercise and neonatal outcomes.Moreover, long-term follow-up studies that extend beyond early childhood could provide further insights into the potential impact of maternal exercise on offspring health.
Our findings indicate that physical activity during pregnancy did not significantly alter the birthweight or BMI during the early years of life, which is consistent with prior research suggesting that gestational exercise is beneficial for maternal health without adversely affecting neonatal outcomes.These results may be explained by the hypothesis that physical activity optimizes maternal energy balance and fosters an environment conducive to healthy fetal development.Exercise might trigger adaptive mechanisms that safeguard the fetus, maintaining nutrient delivery even amidst increased maternal energy expenditure.Finally, while physical activity offers numerous benefits during pregnancy, it is important to consider potential risks, such as musculoskeletal injuries and falls, which could adversely affect the pregnancy.Appropriate exercise modifications and precautions can help mitigate these risks.Additionally, maternal age plays a significant role in pregnancy outcomes, with older mothers potentially facing higher risks of complications.Therefore, tailored exercise recommendations based on maternal age and fitness levels are crucial for optimizing both maternal and fetal health.

Conclusions
The present study found no statistical differences in the effects of physical activity on birth weight, Ponderal Index at birth, Macrosomia, LGA, children's BMI, weight, and skinfold thicknesses.In conclusion, our comprehensive analysis adds to the evidence base indicating that engagement in physical activity during pregnancy demonstrates a favorable trend in parameters associated with childhood overweight and obesity while posing no discernible adverse effects.These findings should reassure health professionals and pregnant women about the safety of maintaining an active lifestyle during pregnancy.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm13133726/s1,Results of GRADE Assesment and Data Base Search Strategy.
Author Contributions: D.Z. played a pivotal role in conceptualizing the research study and actively collecting data.R.M. drafted the initial version and meticulously revised it for intellectual depth and coherence.The statistical expertise of P.L. and G.D.R. was instrumental in conducting rigorous analyses and interpreting the complex results.The fieldwork conducted by M.S.-P., including the gathering of primary data, was essential in establishing the empirical basis of the manuscript.Theoretical insights and expertise provided by C.S.-J.greatly enriched the manuscript's conceptual framework.Author R.B. demonstrated exceptional organizational and leadership skills by conducting rigorous analyses, interpreting the complex results, and managing various administrative aspects of the project, facilitating its successful execution.All authors have read and agreed to the published version of the manuscript.
being a review article (n = 3) or observational study (n = 5), no variable of interest (n = 6), or inadequate population (n = 6).The detailed study selection process is illustrated in Figure 1.

Figure 1 .
Figure 1.Flow diagram of the search.

Figure 1 .
Figure 1.Flow diagram of the search.

Table 1 .
Characteristics of the analyzed references.
Author: study reference.Year: publication year.N: total number of participants.EG: number of participants in the intervention group.CG: number of participants in the control group.Fq.: frequency in a week of exercise sessions.Intens.: intensity.Dur. of prog.: program duration in weeks.Type of intervent.: type of intervention: exercise program (EP: aerobic, muscle strengthening, etc.) or physical activity and nutritional advice.Superv.Class: whether or not intervention was supervised.Dur. of class: minutes of each session.Adh.: adherence (compliance) to the intervention (%).