Abstract
Objective
There is no consensus of opinion if exercise beneficially affects sex hormones if added to weight-loss diets. The purpose of this study was to perform a systematic review and meta-analysis of controlled clinical trials to evaluate the effect of adding exercise to a hypo-caloric diet during a weight-loss program, on serum testosterone, estradiol, and sex hormone-binding globulin (SHBG) in adults with overweight/obesity.
Design
Systematic review and meta-analysis of the literature.
Methods
Online databases including PubMed/MEDLINE, EMBASE, Scopus, ISI Web of Science, and Google Scholar were searched up to April 2021. A random-effects model was applied to compare mean changes in sex hormones and SHBG between participants undergoing a hypo-caloric diet with or without exercise.
Results
In total, 9 eligible clinical trials with 462 participants were included. Out of these, seven, three, and four studies illustrated changes in testosterone, estradiol, and SHBG, respectively. The meta-analysis revealed that exercise had no significant effect on circulating testosterone (WMD = −0.03 nmol/L, 95% CI: −0.11, 0.06, P = 0.51), estradiol (WMD = −0.46 pg/mL, 95% CI: −1.57, 0.65, P = 0.42), and SHBG (WMD = 0.54 nmol/L, 95% CI: −2.63, 3.71, P = 0.74) when added to low-calorie diets.
Conclusion
The addition of exercise to a hypo-caloric diet provided no additional improvement in sex hormone profiles. Further, well-designed randomized controlled trials with longer follow-up periods in both sexes are recommended to confirm and expand the current results.
Introduction
The prevalence of obesity is increasing dramatically worldwide (1), causing considerable health, psychological, and socio-economic challenges (2, 3). Excess body fat is a risk factor for numerous cardio-metabolic diseases such as cardiovascular diseases (CVD), type 2 diabetes mellitus, dyslipidemia, and several non-metabolic comorbidities such as musculoskeletal disorders, obstructive sleep apnea, and certain cancers.
In addition to these well-known consequences, obesity and the distribution of fat in the body can be related to the levels of certain sex hormones such as testosterone and estradiol, and SHBG in men and women (4, 5). In men, obesity has been shown to decrease serum testosterone and SHBG concentrations and increase estradiol concentrations (2, 6), while in obese women a decline in serum SHBG concentration and elevation in testosterone and estradiol concentrations have been observed (5, 7). The decline in testosterone and SHBG are known to be risk factors for CVD, metabolic syndrome, and diabetes in men (8, 9), whereas higher serum estrogen and testosterone and lower SHBG considerably increase the risk of breast cancer and CVD in women (10, 11, 12).
Lifestyle modifications are strongly encouraged in overweight and obese individuals with sex hormones disorders (13, 14). A sedentary lifestyle and poor dietary intake are two modifiable risk factors related to obesity (15). Weight loss (through a hypo-caloric diet with and without exercise) may play an important role in improving the levels of sex hormones (2, 16, 17). However, individual studies examining whether adding exercise to a hypo-caloric diet leads to greater improvement in profiles of sex hormones, compared to a hypo-caloric diet alone, are mixed. For example, in women, Thomson et al. have shown that adding exercise to a diet results in a more and significant reduction in testosterone compared to diet alone (16), while according to the results of Nybaka et al., a hypo-caloric diet alone significantly reduced testosterone (18). In other studies, however, neither hypo-caloric diet alone nor adding exercise to the diet caused significant changes in testosterone (5, 19). Also, only in one study, the addition of exercise to the diet resulted in a significant increase in SHBG (16). Other studies have demonstrated no significant effects (5, 6, 18).
To the best of our knowledge, we are not aware of any systematic review and/or meta-analysis on this topic. Therefore, the present study aimed to summarize the results of trials that examined the influence of hypo-caloric diets plus exercise compared to hypo-caloric diets alone on serum testosterone, estradiol, and SHBG in adults with overweight/obesity.
Methods
The reporting methods used in the present study were in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement (20). The study protocol was registered in the international prospective register of systematic reviews (PROSPERO) database (with registration code CRD42020173434) in April 2020. The detailed protocol of the current systematic review and meta-analysis has been provided elsewhere (21).
Search strategy
Relevant articles were identified from the earliest available online indexing year up to April 2021 through systematic searches in the following databases/search engines: PubMed/MEDLINE, ISI (Web of sciences), EMBASE, Scopus, and Google Scholar, without language or any other restriction. The search strategy included Medical subject headings (MESH) and also non-MeSH keywords that might appear in the title and/or abstract of the related publications using the following keywords: ‘exercise’ OR ‘physical activity’ OR ‘training’ OR ‘physical fitness’ AND ‘nutrition’ OR ‘nutrition therapy’ OR ‘caloric restriction’ OR ‘weight-reducing’ OR ‘diet’ OR ‘hypocaloric’ OR ‘hypo-caloric’ AND ‘Intervention Studies’ OR ‘intervention’ OR ‘controlled trial’ OR ‘randomized’ OR ‘randomized’ OR ‘random’ OR ‘randomly’ OR ‘placebo’ OR ‘assignment’ OR ‘clinical trial’ OR ‘trial’. Search strategies used for searching each database are provided elsewhere (21). Unpublished studies were recognized by searching pre-print indexing websites such as research square (https://www.researchsquare.com/) and WHO-approved registered clinical trials. The full-text of the articles that seemed to contain the information needed for the current study were evaluated. The corresponding author was contacted at least two times (1 week apart) if we could not access the full text. The bibliographies of the related studies were also checked manually to identify additional relevant articles.
Eligibility criteria
Titles/abstracts derived from database search were divided between six reviewers (Z Y, S S, S B, S H R, S M T, and T Z) for screening and cross-checking, to independently identify the relevant or potentially relevant studies for full-text screening. In the absence of a consensus between reviewers, disagreements were resolved by discussion and consensus with the senior author (A S A). This review included all controlled clinical trials with either a parallel, cross-over, or factorial design, which have been performed on overweight or obese adults (≥18 years of age and BMI ≥25 kg/m2), and examining the effect of hypo-caloric diets plus exercise in comparison with hypo-caloric diets alone on serum testosterone, estradiol, and/or SHBG. It should be noted that diet and energy restriction should be iso-caloric in both arms. Studies were excluded if (1) they investigated the effects on pregnant or lactating women and (2) the follow-up period was less than 2 weeks. In the case of multiple publications from one study, we selected only the earliest or most informative study.
Data extraction
The following data were extracted from the eligible papers, by two independent investigators (S M T, S H R): first authors’ family name, publication year, geographic research location, number of study arms, age and gender of participants, number of participants in the intervention groups, the health condition of participants, the study design (parallel/cross-over/others), duration of intervention, amount of calorie restriction considered for weight-loss diets, type of diets, type of exercise programs and intensity, frequency and delivery of studies, compliance to the diet and exercise programs, funding source(s), mean and s.d. for outcome measures in pre-and post-intervention, and their change, if possible, and P-values for within-group and between-group comparison.
Risk of bias assessment
The quality of the eligible studies was assessed using Cochrane collaboration’s tool (22), based on the following sources of bias: (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 outcome reporting (reporting bias), and (7) compliance to the diet as another possible source of bias. According to the recommendations, each domain was judged as ‘low risk of bias’, ‘high risk of bias’, or ‘unclear risk of bias’. Since blinding is not possible for trials examining the effect of diet and exercise, the blinding of participants and reviewers was not considered as a key component for assessing the risk of bias. The overall quality of the included studies was considered as ‘low risk’ (if all domains were categorized as ‘low risk’), “high risk’ (if one or more domains were categorized as ‘high risk’), and ‘unclear risk’ (if one or more domains were categorized as ‘unclear risk’).
For non-randomized trials, the risk of bias was evaluated based on ROBIN-I, through seven domains: confounding factors, participant selection, interventions classification, deviations from intended interventions, missing data, outcomes measurement, and selective reporting. General judgment about the overall risk of bias was reported as low risk of bias (if all domains were categorized as ‘low risk’), moderate risk of bias (if all domains were categorized as ‘low risk’ or ‘moderate risk’), serious risk of bias (if at least one domain was categorized as ‘serious risk’ but not at ‘critical risk’), critical risk of bias (if at least one domain was categorized as ‘critical risk’), or no information (if one or more domains were categorized as ‘no information’) (23).
To assess the quality of the evidence of this systematic review and meta-analysis on the outcome level for this meta-analysis, the GRADE (grading of recommendations assessment, development, and evaluation) approach was incorporated (24), by using GRADEpro, V.3.6 software. The quality of evidence may be ‘downgraded’ due to limitations in the quality of studies, indirectness, imprecision, inconsistency, and risk of publication bias. Each domain was judged as not serious, serious, or very serious. The four levels used to classify the resulting quality assessment of the evidence made by the GRADE tool are high, moderate, low, or very low.
Statistical analysis
Mean changes and s.d. for each outcome variable within the hypo-caloric diet group alone and hypo-caloric diet plus exercise group were derived to calculate the mean difference and its s.e., which was used to determine the effect sizes for the meta-analyses. Since the change values were not reported, the change values s.d. was calculated by selecting 0.5 as the correlation coefficient between baseline and follow-up values. To ensure that the meta-analyses were not sensitive to the selected correlation coefficient, all analyses were replicated with a correlation coefficient of 0.2 and 0.8. The weighted mean difference (WMD) and its corresponding 95% CI were calculated using the random-effects model which takes the between-study heterogeneity into account (25). The between-study heterogeneity was checked using Cochran’s Q test and I-squared (I2) statistic (26). The Cochran’s Q test P-value < 0.05 was considered as significant heterogeneity. I2 values between ‘0 and 25%’ and ‘75 and 100%’ were indicated low and considerable heterogeneity, respectively. Subgroup analyses were also completed for sex (male/female) and exercise type (aerobic/resistance) to detect potential sources of heterogeneity. Sensitivity analyses were performed by removing the studies from the meta-analysis one by one to examine the robustness of the overall results (27). Since the number of studies in each meta-analysis was less than ten, potential publication bias was determined by visual inspection of Begg’s funnel plots (28). STATA, version 11.2 (Stata Corp, College Station, TX) was used to perform the statistical analyses and a P-value ≤ 0.05 was considered statistically significant.
Results
Study selection
A total of 42,189 potentially relevant papers were identified through electronic searching of databases. Following the removal of duplicate records (5607 duplicates), 36,582 papers remained for screening the titles or abstracts. Studies that were not trials or did not have any group with weight-loss diet alone and weight-loss diet plus exercise groups were excluded (n = 36,113) and 469 full-text articles were assessed for eligibility, accurately. Of these studies, 459 were excluded for the following reasons (repeated publications on the same study (n = 119), did not impose equal energy deficit in the diet alone and diet plus exercise groups (n = 31), no weight-loss program (n = 76), did not assess the outcome variables (n = 178), and other reasons (letter, pregnant and lactating women, <18 years old) (n =55)). In total, ten studies were included in the systematic review. One study was omitted due to lack of sample size reporting (29) and therefore nine clinical trials were eligible to be included in the meta-analysis (5, 6, 16, 18, 19, 30, 31, 32, 33). Out of these, seven articles had reported data on testosterone (5, 6, 16, 18, 19, 30, 31), three on estradiol (5, 32, 33), and four on SHBG (5, 6, 16, 31). A flow diagram outlining the study selection process is provided in Fig. 1.
The study selection process.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
The study selection process.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
The study selection process.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Studies’ characteristics
The characteristics of the ten eligible papers are listed in Table 1. In general, 519 participants were included in these trials with a mean age range of 29.9–70 years. All studies used a randomized controlled parallel design, except one that was non-randomized (31). The duration of the intervention period for each study ranged between 3 and 48 weeks. Two of the studies conducted interventions for 24 weeks (6, 33), two for 16 weeks (18, 32), and two for 12 weeks (19, 30). The intervention period was 3 (31), 8 (29), 20 (16), and 48 weeks (5) in four studies. Six trials were exclusively performed on women (5, 16, 18, 19, 31, 32), three studies were done on men (6, 29, 30), and one remaining trial included both sexes (33). In the majority of trials, a moderate caloric restriction was applied for energy intake. In three studies, aerobic plus resistance exercises were prescribed for one group and aerobic exercises for another group of participants (16, 19, 30). The recommendations to perform only aerobic exercise, resistance exercise, and aerobic plus resistance exercise were accomplished in two (5, 31), one (32), and four articles (6, 16, 18, 33), respectively. Also, in one trial, two groups of participants were provided with either aerobic or resistance exercise (29).
Characteristics of individual studies evaluating the effects of weight-loss programs (including a hypo-caloric diet with or without exercise) on serum sex hormones.
| Reference | Country | Sex/ population characteristics | Age | Number of participants in each group | Design of study | Duration (week) | Type of diet | Calorie restriction | Type of exercise | Duration of exercise | Outcome | Results |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (6) | USA | Male/ obese with sedentary lifestyle | >65 | D: 9; D + E: 12 | Parallel | 24 | 1 g of high-quality protein per kilogram of body weight per day | 500–750 kcal/day from their daily energy requirement; ~10 % weight loss at 6 months | Aerobic exercises, resistance training, and exercises to improve flexibility and balance | Three group exercise-training sessions per week. Each session ~ 90 min | Testosterone, estradiol, SHBG | Decrease in estradiol, and increase in testosterone and SHBG in both groups |
| (29) | USA | Female and male/ sedentary, nonsmoking with good health, BMI: 25–52 kg/m2 | 19–49 | D: 17; D + E: 24 for aerobic exercise and 16 for resistance exercise | Parallel | 8 | Liquid formula; CHO: 24%; fat: 24; pro: 52%; | Formula packets were combined with variable amounts of 1%-fat milk to provide 70% of RMR. | Aerobic training (leg cycle ergometer, upper-body ergometer, arm cycling direction)/resistance weight training. | Aerobic: 3 times/ week, each session ~30 min; resistance: 3 times/ week, each session ~60 min | Testosterone | Testosterone was measured only in men and the number of men in each group was not specified. |
| (18) | Sweden | Female/ PCOS, BMI >27 kg/m2 | 18–40 | D: 14; D + E: 14 | Parallel | 16 | CHO: 55–60%; fat: 25–30%; SFA: 10%; prot: 10–15%; energy intake | Reduced by at least 600 kcal/day in comparison with pre-intervention intake | Aerobics (walking, jogging, swimming) and muscle strength training, with a moderate to vigorous exertion level | 2–3 times/ week, each time 45–60 min | Testosterone, SHBG | Decrease in testosterone in the D group |
| (5) | USA | Female/ postmenopausal, BMI ≥25 kg/m2 | 50–76 | D: 115; D + E: 116 | Parallel | 48 | Fat ≤30% | Total daily energy intake of 1200–2000 kcal/day based on baseline weight; a 10% reduction in body weight by 6 months with maintenance to 12 months | Moderate-to vigorous-intensity aerobic exercise, such as brisk walking | 5 days per week, each time 45 min (225 min/week) | Testosterone, estradiol, SHBG | Testosterone decreased, estradiol decreased, SHBG increased in both groups |
| (33) | USA | Female and male/ obese, sedentary lifestyle | >65 | D: 26 (17 female/ 9 male); D + E: 28 (16 female/ 12 male) | Parallel | 24 | 500–750 kcal/day, achieved a 10% weight loss at 6 months | Flexibility exercises, aerobic exercise, progressive resistance training, and balance exercises | 3 days/ week, each time 90 min | Estradiol | Estradiol decreased in both groups | |
| (32) | Spain | Female/ nonsmoking, obese, BMI: 30–40 kg/m2 | 40–60 | D: 12; D + E:13 | Parallel | 16 | CHO: 55%, prot: 15% and the rest as fat | 500 kcal/day | Strength training program (bilateral leg press and bilateral knee extension exercises, the bench-press and exercises for the main muscle groups of the body) | 2 times/ week, each time 45–60 min | Estradiol | Estradiol decreased in the D group |
| (16) | Australia | Female/ overweight and obese with PCOS, BMI: 25–55 kg/m2 | 18–41 | D: 14; D + E: 18 (for aerobic group) and 20 (for aerobic + resistance group) | Parallel | 20 | High-protein diet (fat: 30%; SFA: 8%; CHO: 40%; pro: 30%) | 1195–1434 kcal/day (weight loss of 8–12 kg over the study period) | Aerobic resistance exercise (walking/jogging + bench press, lat pulldown, leg press, knee extension, and sit-ups) | Aerobic: 3 times/ week (from 25 to30 min during the first week to 45 min by study end) + resistance: 2 times/ week | Testosterone, SHBG | Testosterone decreased and SHBG increased in both groups |
| (30) | USA | Male/ healthy overweight (≥120% of desirable weight) | Mean age D: 40 mean age D + E: 37.8 (for aerobic group) and 39.8 (for aerobic + resistance group) | D: 8; D + E: 11 (for aerobic group) and 10 (for aerobic + resistance group) | Parallel | 12 | Prot ≥ RDA and ≥1 g/kg IBW | Moderate caloric restriction (designed to create a 6–9 kg weight loss) | Aerobic (treadmill walking/jogging, stationary cycling, seated rowing, and stationary stair climbing) + resistance (squat exercise performed and additional Nautilus machine) consisting of military press, bench press, lat pull down, seated rows, sit-ups, lower back, leg press, hamstring curls, calf raises, and arm curls) | Aerobic: 3 times per week (30 min first week and then 50 min) resistance: load between heavy day (5–7 RM) and moderate day (8–10 RM) loads | Testosterone | No significant changes were observed. |
| (19) | USA | Female/ healthy premenopausal, overweight (either ≥120% of desirable weight or a BMI ≥27 kg/m2) | Mean age D: 34.6; mean age D + E: 35.6 | D: 8 D + E: 9 for aerobic exercise and 8 for aerobic + resistance exercise |
Parallel | 12 | Whole body aerobic endurance exercise (treadmill walking/jogging, stationary cycling, seated rowing, and stationary stair climbing) and heavy-resistance training program consisted of a squat exercise | First week, each session lasted 30 min and gradually increased to 50 min over the subsequent weeks. resistance training took place 3 times per week | Testosterone | No significant changes were observed. | ||
| (31) | Sweden | Female/ obese | Middle-aged | D: 9 D + E: 9 |
Parallel (non-randomized trial) | 3 | Liquid diet (Cambridge diet) consisting of 47.8 g protein, 67 g carbohydrates, 5.0 g fat per day (500 kcal total per day) | 500 kcal/ day in the form of a liquid diet (Cambridge diet) | Training program was standardized for each subject in relation to maximal work capacity | 3 times/ week, each time 55 min | Testosterone, SHBG | No significant changes was observed in testosterone, but SHBG was decreased in both group. |
IBW, ideal body weight; RMR, resting metabolic rate; RDA, recommended daily allowance.
Assessment of the risk of bias
Using the Cochrane risk of bias tool for RCTs, six studies reported the details of random sequence generation and had a low risk of bias for this domain (5, 6, 16, 18, 29, 33); however, three studies did not state any method for randomizing (19, 30, 32). None of the studies explained the methods used for allocation concealment. The blinding of outcomes assessors was asserted in four articles (5, 6, 29, 33). The majority of the studies didn’t have incomplete outcome data (5, 6, 16, 18, 30, 32, 33). None of the studies had selective reporting and dietary compliance had been evaluated in six articles (5, 6, 19, 30, 32, 33). By evaluating the quality of clinical trials according to the domains suggested by Cochranes collaboration’s tool, the quality of eight studies was categorized as ‘unclear risk’ (5, 6, 16, 18, 19, 30, 32, 33). Quality assessment details of studies included in this systematic review are illustrated in Table 2.
Risk of bias assessment for included studies according to the Cochrane collaboration’s tool.
| Reference | Random sequence generation (selection bias) | Allocation concealment (selection bias) | Blinding of participants and personnel (performance bias) | Blinding of outcome assessment (detection bias) | Incomplete outcome data (attrition bias) | Selective reporting (reporting bias) | Other sources of bias (evaluating the dietary compliance) | Overall quality |
|---|---|---|---|---|---|---|---|---|
| (6) | L | U | NA | L | L | L | L | Unclear risk |
| (29) | L | U | NA | L | H | L | U | High risk |
| (5) | L | U | NA | L | L | L | L | Unclear risk |
| (32) | U | U | NA | U | L | L | L | Unclear risk |
| (30) | U | U | NA | U | L | L | L | Unclear risk |
| (19) | U | U | NA | U | U | L | L | Unclear risk |
| (18) | L | U | NA | U | L | L | U | Unclear risk |
| (33) | L | U | NA | L | L | L | L | Unclear risk |
| (16) | L | U | NA | U | L | L | U | Unclear risk |
H, high risk of bias; L, low risk of bias; NA, not assessed because the blinding was not possible in this type of clinical trials; U, unclear risk of bias.
The design of one study was a non-randomized trial (31). According to the ROBINS-I tool, risks of bias of this study were low in four domains (’selection of participants’, ‘classification of interventions’, ‘deviations from intended interventions’, and ‘missing data’) and medium in the relevant confounding domain. Also, due to insufficient data in ‘outcomes measurement’ and ‘selection of reported results’ domains, the overall risk of bias for this study was categorized as ‘No information’ (Table 3).
Risk of bias assessment for the non-randomized study (31) according to the ROBINS-I tool.
| Risk of bias | |
|---|---|
| Confounding | Moderate |
| Selection of participants | Low |
| Classification of interventions | Low |
| Deviations from intended interventions | Low |
| Missing data | Low |
| Outcomes measurement | No information |
| Selection of reported results | No information |
| Overall quality | No information |
Assessment of the quality of evidence
Using the GRADE approach, testosterone and SHBG outcomes were downgraded by one level due to serious risk of bias and imprecision, respectively, so their evidence qualities were considered as moderate. However, evidence was low for estradiol (downgraded by two levels due to indirectness, imprecision.) (Table 4).
Assessment of the quality of evidence for included studies according to the GRADE approach.
| Intervention | Participants (No. of studies) | Risk of bias1 | Inconsistency2 | Indirectness3 | Imprecision4 | Publication bias5 | Certainty of evidence6 |
|---|---|---|---|---|---|---|---|
| Testosterone | 404 (7) | Serious | Not serious (I2 = 0) | Not serious | Not serious | Not assessed | Moderate |
| Estradiol | 310 (3) | Not serious | Not serious (I2 = 0) | Serious | Serious | Not assessed | Low |
| SHBG | 332 (4) | Not serious | Not serious (I2 = 0) | Not serious | Serious | Not assessed | Moderate |
1Downgraded one level (serious) if 50–75% of RCTs were at high risk of bias and two levels (very serious) if more than 75% of RCTs were at high risk of bias. 2Downgraded one level for inconsistency when I2 was ≥50%. Where I2 >50% and predefined subgroups (calorie restriction, behavioral, physical activity) did not explain the source of heterogeneity. The I2 ≤50% inconsistency was considered as a not serious limitation. 3Downgraded one level for indirectness if more than 75% included trials have been conducted in the same geographical location. 4Downgraded one level for imprecision if the number of participants was less than 400. 5Assessed for potential publication bias when ≥10 trials were available. 6Data from RCTs begin with a grade of ‘HIGH’.
Downgraded for high risk of bias, inconsistency, indirectness, imprecision, and publication bias.
Meta-analysis
Testosterone
Seven trials including 404 participants provided data on changes in testosterone following a hypo-caloric diet or hypo-caloric diet combined with exercise intervention (5, 6, 16, 18, 19, 30, 31). It is noteworthy that one study measured serum testosterone in men but was not included in the meta-analysis due to a lack of data on sample size in each group (29). As shown in Table 3, the results indicated that adding exercise to a hypo-caloric diet had no significant effect on circulating testosterone (WMD = −0.03 nmol/L, 95% CI: −0.11, 0.06, P = 0.51, Fig. 2). There was no heterogeneity between studies (Q statistic = 5.61, P = 0.47; I2 = 0%). Subgroup analysis was performed based on the sex and exercise type (aerobic training or aerobic + resistance training) and no significant changes were observed in these subgroups (Table 5).
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum testosterone.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum testosterone.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum testosterone.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
The effects of weight-loss programs (including a hypo-caloric diet with or without exercise) on serum sex hormones, based on different subgroups as well as the overall analysis (all analyses were conducted using random effects model).
| Number of studies | Meta-analysis | Heterogeneity | |||||
|---|---|---|---|---|---|---|---|
| WMD (95% CI) | P effect | Q statistic | P within group | I2 (%) | P between group | ||
| Testosterone | 7 | −0.03 (−0.11, 0.06) | 0.51 | 5.61 | 0.47 | 0.0 | |
| Gender | |||||||
| Female | 5 | −0.02 (−0.17, 0.12) | 0.76 | 5.60 | 0.23 | 28.6 | 0.91 |
| Male | 2 | 0.04 (−1.20, 1.29) | 0.95 | 0.00 | 0.99 | 0.0 | |
| Exercise type | |||||||
| aerobic | 5 | −0.04 (−0.18, 0.10) | 0.60 | 4.95 | 0.29 | 19.3 | 0.87 |
| Aerobic + resistance | 5 | −0.06 (−0.35, 0.23) | 0.69 | 2.31 | 0.68 | 0.0 | |
| Estradiol | 3 | −0.46 (−1.57, 0.65) | 0.42 | 1.70 | 0.43 | 0.0 | |
| SHBG | 4 | 0.54 (−2.63, 3.71) | 0.74 | 2.86 | 0.41 | 0.0 | |
Estradiol
The meta-analysis of three relevant studies (included 310 subjects) (5, 32, 33) indicated that hypo-caloric diets plus exercise compared with hypo-caloric diets alone had no significant effect on estradiol (WMD = −0.46 pg/mL, 95% CI: −1.57, 0.65, P = 0.42), and no heterogeneity was observed between studies (Q statistic = 1.70, P = 0.43; I2 = 0%) (Fig. 3).
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum estradiol.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum estradiol.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum estradiol.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
SHBG
Four trials with 332 participants assessed changes in SHBG following weight-loss diets with and without exercise (5, 6, 16, 18). They showed that when exercise was added to a low-calorie diet, there was no significant change in SHBG compared to a low-calorie diet alone (WMD = 0.54 nmol/L, 95% CI: −2.63, 3.71, P = 0.74). No heterogeneity existed between study results (Q statistic = 2.86, P = 0.41; I2 = 0%) (Fig. 4).
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum SHBG.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum SHBG.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Forest plot illustrating the effect of adding exercise to a hypo-caloric diet on serum SHBG.
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Sensitivity analysis and publication bias
The sensitivity analysis indicated that the effect of a hypo-caloric diet with or without exercise on circulating sex hormones and SHBG was not substantially changed by excluding a single study. According to observations based on the funnel plots, there was no evidence of publication bias for meta-analysis of studies assessing the effect of adding exercise to a hypo-caloric on serum testosterone, estradiol, and SHBG (Fig. 5).
Begg’s funnel plots (with pseudo 95% CIs) depicting the effect sizes (difference in means) vs their s.e. for studies that assessed effects of adding exercise to a hypo-caloric diet on serum testosterone (A), estradiol (B), and SHBG (C).
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Begg’s funnel plots (with pseudo 95% CIs) depicting the effect sizes (difference in means) vs their s.e. for studies that assessed effects of adding exercise to a hypo-caloric diet on serum testosterone (A), estradiol (B), and SHBG (C).
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Begg’s funnel plots (with pseudo 95% CIs) depicting the effect sizes (difference in means) vs their s.e. for studies that assessed effects of adding exercise to a hypo-caloric diet on serum testosterone (A), estradiol (B), and SHBG (C).
Citation: European Journal of Endocrinology 186, 2; 10.1530/EJE-21-0675
Discussion
To the best of our knowledge, this is the first systematic review and meta-analysis of randomized trials to assess the effects of adding exercise to hypo-caloric diets on the serum profiles of testosterone, estradiol, and SHBG in overweight/obese adults. The current evidence revealed that the effects of a weight loss diet on serum testosterone, estradiol, and SHBG were not significantly altered with the addition of exercise. Subgroups analysis for sex and type of exercise revealed no effects in these results.
A previous meta-analysis in men indicated a significant increase in total testosterone, SHBG, and gonadotropins, and a decrease in estradiol through hypo-caloric diet and bariatric surgery. Greater beneficial effects of weight loss were observed in non-diabetic and younger participants who were more obese; although, this could be due to a higher bodyweight reduction (2). Another meta-analysis in PCOS women indicated that a lifestyle intervention (dietary treatment in combination with exercise) and/or exercise alone increased SHBG and decreased total testosterone and androstenedione (14). Moreover, another meta-analysis suggested that independent from menopausal status, exercise reduces circulating total and free estradiol, free testosterone, and androstenedione, and increased circulating SHBG, significantly. Although, these effects were relatively modest and probably not clinically significant. The decrease in free estradiol and free testosterone was greater than total estradiol and total testosterone, which was due to an increase in SHBG (17). However, some clinical trials aimed at weight loss had indicated that the addition of exercise training to the hypo-caloric diet did not help to lose more weight or alter hormonal status (16, 30). Also, it has been shown that low-fat diets without weight loss did not change estrogen, considerably (34, 35). These findings reveal that the main factor linking diet or exercise changes to sex hormone variation is weight loss (5) and apparently makes no difference as to whether weight loss is through hypo-caloric diet alone or through hypo-caloric diet with exercise. It has been suggested that more than a 10% weight loss may be required to improve hormonal profiles (6).
Several mechanisms might explain the effect of obesity on sex hormone abnormality and demonstrate that reductions in adipose tissue can be the most appropriate approach. The aromatase enzyme (the enzyme that is responsible for converting androgen precursors to estrogens (including estradiol)) is highly expressed by fat tissue; therefore, in participants with overweight and/or obesity, increased estradiol levels are observed (2). On the other hand, in men, high levels of estrogen resulting from increased aromatase activity, first suppress gonadotropin production from the pituitary gland and then testosterone production from the testis (6). Adiposity is associated with insulin resistance, resulting in increased pancreatic insulin secretion (hyperinsulinemia). However, insulin has an inhibitory effect on hepatic SHBG production. Therefore, the reduction in SHBG is a potential consequence of being overweight or obese and can affect the bioavailability of androgens and estrogens (36). Moreover, hyperinsulinemia has a hyperandrogenic effect through increasing the activity of the key enzyme in androgen synthesis (P450C17) on the ovaries. Hence, testosterone levels might increase in obese women (37).
The present study had several strengths that need to be considered. First, the duration of intervention in most studies was long enough to observe the effects of the interventions on hormonal profiles. Exercise training had been supervised in all studies, except one (16) and compliance to the diet had been assessed in the majority of studies (5, 6, 19, 30, 31, 32, 33). Also, there was no heterogeneity between the studies included in meta-analyses. Some limitations should be taken into account while interpreting our results: (1) the number of studies conducted in this area is limited, particularly for estradiol and SHBG. (2) Also, the sample size for 8 of 9 included studies was not large enough (the number of participants in each group ranged from 8 to 20). Therefore, more high-quality clinical trials with more sample size should be performed to achieve a more definitive conclusion about the effect of adding exercise to a hypo-caloric diet on sex hormone levels. (3) Based on the risk of bias assessment, the methodological quality of eight included studies was ‘unclear’ and had no reports on ‘selection bias’, ‘attrition data’, didn’t mention to blinding of outcomes, or didn’t evaluate the dietary compliance (5, 6, 16, 18, 19, 30, 32, 33). In trials, lack of adherence to interventions can be a potential problem and can vitiate the benefits of randomization. The GRADE approach revealed that there was moderate confidence in the non-significant effect of exercise + weight-loss diets compared with weight-loss diets alone on serum testosterone and SHBG. Also, the quality of evidence was low for estradiol (3). The positive effect of weight loss on hormone levels is short-lived and depends on how long participants adhere to the recommended diet and exercise to maintain their weight loss (6). (4) Another limitation of the present study is that in the study completed by Campbell et al., due to the lack of information on hormonal changes at the end of 24 weeks of intervention, data at the end of 48 weeks (24 weeks weight-loss intervention + 24 weeks weight-maintenance intervention) were used for the meta-analysis. It should be noted that some weight regain may have occurred during the weight maintenance phase, which might dilute the results.
In conclusion, the present systematic review and meta-analysis provide evidence that adding exercise to a hypo-caloric diet offers no additional improvement in sex hormones and SHBG profiles. Well-designed randomized controlled trials exploring the effects of different types of exercise along with a weight-loss diet on sex hormone levels with longer follow-up periods in both sexes are still required to confirm the current results.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of this study.
Funding
The present study was supported by the Research Council of the Nutrition and Food Security Research Center, Shahid Sadoughi University of Medical Sciences, Yazd, Iran.
Author contribution statement
A S A designed the study and will be the guarantor of the review. A S A, Z Y, and S S contributed to defining the search strategy; Z Y, S S, S B, S H R, S M T, and T Z performed the systematic search and study selection; S M T and S H R extracted the data; S B analyzed the data; S M T wrote the first draft of the manuscript, A S A, M K, J S B, and S F edited the manuscript and all authors read and approved the final version of the manuscript.
Acknowledgements
The authors would like to appreciate the research council of Nutrition and Food Security Research Center for their scientific support.
References
- 1↑
Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, Mullany EC, Biryukov S, Abbafati C & Abera SF et al.Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 2014 384 766–781. (https://doi.org/10.1016/S0140-6736(1460460-8)
- 2↑
Corona G, Rastrelli G, Monami M, Saad F, Luconi M, Lucchese M, Facchiano E, Sforza A, Forti G & Mannucci E et al.Body weight loss reverts obesity-associated hypogonadotropic hypogonadism: a systematic review and meta-analysis. European Journal of Endocrinology 2013 168 829–843. (https://doi.org/10.1530/EJE-12-0955)
- 3↑
Marques A, Peralta M, Naia A, Loureiro N, de Matos MG. Prevalence of adult overweight and obesity in 20 European countries, 2014. European Journal of Public Health 2018 28 295–300. (https://doi.org/10.1093/eurpub/ckx143)
- 4↑
Haffner SM, Valdez RA, Stern MP, Katz MS. Obesity, body fat distribution and sex hormones in men. International Journal of Obesity and Related Metabolic Disorders 1993 17 643–649.
- 5↑
Campbell KL, Foster-Schubert KE, Alfano CM, Wang CC, Wang CY, Duggan CR, Mason C, Imayama I, Kong A & Xiao L et al.Reduced-calorie dietary weight loss, exercise, and sex hormones in postmenopausal women: randomized controlled trial. Journal of Clinical Oncology 2012 30 2314–2326. (https://doi.org/10.1200/JCO.2011.37.9792)
- 6↑
Armamento-Villareal R, Aguirre LE, Qualls C, Villareal DT. Effect of lifestyle intervention on the hormonal profile of frail, obese older men. Journal of Nutrition, Health and Aging 2016 20 334–340. (https://doi.org/10.1007/s12603-016-0698-x)
- 7↑
De Pergola G, Maldera S, Tartagni M, Pannacciulli N, Loverro G, Giorgino R. Inhibitory effect of obesity on gonadotropin, estradiol, and inhibin B levels in fertile women. Obesity 2006 14 1954–1960. (https://doi.org/10.1038/oby.2006.228)
- 8↑
Laaksonen DE, Niskanen L, Punnonen K, Nyyssönen K, Tuomainen TP, Valkonen VP, Salonen R, Salonen JT. Testosterone and sex hormone-binding globulin predict the metabolic syndrome and diabetes in middle-aged men. Diabetes Care 2004 27 1036–1041. (https://doi.org/10.2337/diacare.27.5.1036)
- 9↑
Tambo A, Roshan MH, Pace NP. Testosterone and cardiovascular disease. Open Cardiovascular Medicine Journal 2016 10 1–10. (https://doi.org/10.2174/1874192401610010001)
- 10↑
Brand JS, Van Der Schouw YT. Testosterone, SHBG and cardiovascular health in postmenopausal women. International Journal of Impotence Research 2010 22 91–104. (https://doi.org/10.1038/ijir.2009.64)
- 11↑
Woolcott CG, Shvetsov YB, Stanczyk FZ, Wilkens LR, White KK, Caberto C, Henderson BE, Le Marchand L, Kolonel LN, Goodman MT. Plasma sex hormone concentrations and breast cancer risk in an ethnically diverse population of postmenopausal women: the Multiethnic Cohort Study. Endocrine-Related Cancer 2010 17 125–134. (https://doi.org/10.1677/ERC-09-0211)
- 12↑
Farhat GN, Cummings SR, Chlebowski RT, Parimi N, Cauley JA, Rohan TE, Huang AJ, Vitolins M, Hubbell FA & Manson JE et al.Sex hormone levels and risks of estrogen receptor–negative and estrogen receptor–positive breast cancers. Journal of the National Cancer Institute 2011 103 562–570. (https://doi.org/10.1093/jnci/djr031)
- 13↑
Buvat J, Maggi M, Gooren L, Guay AT, Kaufman J, Kaufman A, Kaufman C, Tan HM, Torres LO, Yassin A. Endocrine aspects of male sexual dysfunctions. Journal of Sexual Medicine 2010 7 1627–1656. (https://doi.org/10.1111/j.1743-6109.2010.01780.x)
- 14↑
Haqq L, McFarlane J, Dieberg G, Smart N. Effect of lifestyle intervention on the reproductive endocrine profile in women with polycystic ovarian syndrome: a systematic review and meta-analysis. Endocrine Connections 2014 3 36–46. (https://doi.org/10.1530/EC-14-0010)
- 15↑
van Gemert WA, Schuit AJ, van der Palen J, May AM, Iestra JA, Wittink H, Peeters PH, Monninkhof EM. Effect of weight loss, with or without exercise, on body composition and sex hormones in postmenopausal women: the SHAPE-2 trial. Breast Cancer Research 2015 17 120. (https://doi.org/10.1186/s13058-015-0633-9)
- 16↑
Thomson RL, Buckley JD, Noakes M, Clifton PM, Norman RJ, Brinkworth GD. The effect of a hypocaloric diet with and without exercise training on body composition, cardiometabolic risk profile, and reproductive function in overweight and obese women with polycystic ovary syndrome. Journal of Clinical Endocrinology and Metabolism 2008 93 3373–3380. (https://doi.org/10.1210/jc.2008-0751)
- 17↑
Ennour-Idrissi K, Maunsell E, Diorio C. Effect of physical activity on sex hormones in women: a systematic review and meta-analysis of randomized controlled trials. Breast Cancer Research 2015 17 139. (https://doi.org/10.1186/s13058-015-0647-3)
- 18↑
Nybacka Å, Carlström K, Fabri F, Hellström PM, Hirschberg AL. Serum antiMüllerian hormone in response to dietary management and/or physical exercise in overweight/obese women with polycystic ovary syndrome: secondary analysis of a randomized controlled trial. Fertility and Sterility 2013 100 1096–1102. (https://doi.org/10.1016/j.fertnstert.2013.06.030)
- 19↑
Kraemer WJ, Volek JS, Clark KL, Gordon SE, Incledon T, Puhl SM, Triplett-McBride NT, McBride JM, Putukian M, Sebastianelli WJ. Physiological adaptations to a weight-loss dietary regimen and exercise programs in women. Journal of Applied Physiology 1997 83 270–279. (https://doi.org/10.1152/jappl.1997.83.1.270)
- 20↑
Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart LA & PRISMA-P Group. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Systematic Reviews 2015 4 1. (https://doi.org/10.1186/2046-4053-4-1)
- 21↑
Beigrezaei S, Yazdanpanah Z, Soltani S, Rajaie SH, Mohseni-Takalloo S, Zohrabi T, Kaviani M, Forbes SC, Baker JS, Salehi-Abargouei A. The effects of exercise and low-calorie diets compared with low-calorie diets alone on health: a protocol for systematic reviews and meta-analyses of controlled clinical trials. Systematic Reviews 2021 10 120. (https://doi.org/10.1186/s13643-021-01669-7)
- 22↑
Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2021. (available at: https://handbook-5-1.cochrane.org). Accessed on 18 April 2021.
- 23↑
Sterne JA, Hernán MA, Reeves BC, Savović J, Berkman ND, Viswanathan M, Henry D, Altman DG, Ansari MT & Boutron I et al.Robins-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ 2016 355 i4919. (https://doi.org/10.1136/bmj.i4919)
- 24↑
Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, Schünemann HJ & GRADE Working Group. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008 336 924–926. (https://doi.org/10.1136/bmj.39489.470347.AD)
- 25↑
DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clinical Trials 1986 7 177–188. (https://doi.org/10.1016/0197-2456(8690046-2)
- 26↑
Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in Medicine 2002 21 1539–1558. (https://doi.org/10.1002/sim.1186)
- 27↑
Egger M, Davey-Smith G, Altman D. Systematic Reviews in Health Care: Meta-Analysis in Context. John Wiley & Sons, 2008. (https://doi.org/10.1002/9780470693926)
- 28↑
Higgins J, Green S. Cochrane Handbook for Systematic Reviews of Interventions, 672 p. John Wiley & Sons, 2011. (https://doi.org/10.1002/9780470712184)
- 29↑
Geliebter A, Ochner CN, Dambkowski CL, Hashim SA. Obesity-related hormones and metabolic risk factors: a randomized trial of diet plus either strength or aerobic training versus diet alone in overweight participants. Journal of Diabetes and Obesity 2014 1 1–7.
- 30↑
Kraemer WJ, Volek JS, Clark KL, Gordon SE, Puhl SM, Koziris LP, McBride JM, Triplett-McBride NT, Putukian M & Newton RU et al.Influence of exercise training on physiological and performance changes with weight loss in men. Medicine and Science in Sports and Exercise 1999 31 1320–1329. (https://doi.org/10.1097/00005768-199909000-00014)
- 31↑
Krotkiewski M, Toss L, Björntorp P, Holm G. The effect of a very-low-calorie diet with and without chronic exercise on thyroid and sex hormones, plasma proteins, oxygen uptake, insulin and c peptide concentrations in obese women. International Journal of Obesity 1981 5 287–293.
- 32↑
Ibáñez J, Izquierdo M, Martínez-Labari C, Ortega F, Grijalba A, Forga L, Idoate F, García-Unciti M, Fernández-Real JM, Gorostiaga EM. Resistance training improves cardiovascular risk factors in obese women despite a significative decrease in serum adiponectin levels. Obesity 2010 18 535–541. (https://doi.org/10.1038/oby.2009.277)
- 33↑
Shah K, Armamento-Villareal R, Parimi N, Chode S, Sinacore DR, Hilton TN, Napoli N, Qualls C, Villareal DT. Exercise training in obese older adults prevents increase in bone turnover and attenuates decrease in hip bone mineral density induced by weight loss despite decline in bone-active hormones. Journal of Bone and Mineral Research 2011 26 2851–2859. (https://doi.org/10.1002/jbmr.475)
- 34↑
Berrino F, Bellati C, Secreto G, Camerini E, Pala V, Panico S, Allegro G, Kaaks R. Reducing bioavailable sex hormones through a comprehensive change in diet: the diet and androgens (DIANA) randomized trial. Cancer Epidemiology, Biomarkers and Prevention 2001 10 25–33.
- 35↑
Prentice RL, Caan B, Chlebowski RT, Patterson R, Kuller LH, Ockene JK, Margolis KL, Limacher MC, Manson JE & Parker LM et al.Low-fat dietary pattern and risk of invasive breast cancer: the Women’s Health Initiative Randomized Controlled Dietary Modification Trial. JAMA 2006 295 629–642. (https://doi.org/10.1001/jama.295.6.629)
- 36↑
Lim SS, Norman RJ, Davies MJ, Moran LJ. The effect of obesity on polycystic ovary syndrome: a systematic review and meta-analysis. Obesity Reviews 2013 14 95–109. (https://doi.org/10.1111/j.1467-789X.2012.01053.x)
- 37↑
Barber TM, McCarthy MI, Wass JA, Franks S. Obesity and polycystic ovary syndrome. Clinical Endocrinology 2006 65 137–145. (https://doi.org/10.1111/j.1365-2265.2006.02587.x)
