Abstract
Objective
This systematic review evaluated the effect of testosterone replacement therapy (TRT) in men with obesity having low testosterone levels (LTLs).
Design and methods
Search strategies were performed in MEDLINE, Embase, LILACS, and CENTRAL databases. Two reviewers selected the studies, assessed the risk of bias, and extracted data from the included studies. A random-effects model was used to pool results across studies, and the Grading of Recommendations Assessment, Development, and Evaluation was used to evaluate the certainty of evidence.
Results
A total of 16 randomized controlled trials were included. With moderate certainty of the evidence, no difference was found between TRT and placebo regarding total adverse events, TRT led to a 2-kg lean body mass gain and slightly improved low-density lipoprotein (LDL), without effects on the blood pressure. Due to imprecision/heterogeneity, effects in cardiovascular events (relative risk: 0.52, 95% CI: 0.26 to 1.05, 7 trials, 583 participants), high-density lipoprotein, hematocrit, prostate-specific antigen, HbA1c, and quality of life were unclear. TRT was effective for waist circumference and BMI; however, large between-study heterogeneity was found, with 95% prediction intervals crossing the null effect line. Meta-regression revealed that the average age of participants was a significant modifier for both outcomes.
Conclusion
TRT slightly improved the lean body mass and LDL in men with obesity having LTLs but did not affect the blood pressure. The effects of TRT on cardiovascular events, HbA1c, and quality of life are unclear. The mean age of participants significantly modified the effect of TRT on weight loss.
Introduction
Obesity, especially central obesity, is an independent risk factor for mortality and cardiovascular diseases (1). An inverse relationship was found between waist circumference (WC) and serum testosterone concentrations in obesity and testosterone in men (2). Some authors pointed out that obesity is more important than age and other chronic diseases with regards to testosterone levels (1, 2). The pathogenic mechanisms are not yet fully understood; however, it was associated with a decline in the levels of sex hormone-binding globulin (3, 4), suppression of the gonadal axis by adipokines, and proinflammatory mediators originating in the adipose tissue (5, 6). In addition, excess leptin present in obesity exerts a negative effect on luteinizing hormone pulsatility (4) and increases aromatase activity in adipocytes (7).
As optimal androgen status in men is associated with healthy body composition and consequently, metabolic health (8), testosterone replacement therapy (TRT) is an attractive treatment for men with obesity having low testosterone levels (LTLs). Testosterone increases lean body mass (9) and regulates carbohydrate, protein, and fat metabolism (10, 11). TRT in men with testosterone deficiency results in glucose utilization normalization and increases lipid oxidation (12). In addition, TRT improves erectile function, increases vigor, and reduces fatigue (9), promoting a better disposition to perform exercises.
However, the benefits of TRT in men with obesity were controversial in the literature. Uncontrolled studies revealed that TRT resulted in substantial and sustained body weight, WC, and BMI reductions (13), but controlled studies showed no significant changes in these anthropometric parameters (14, ). Considering this conflicting evidence and the limited number of trials showing beneficial effects of TRT in men with obesity, a recent narrative review did not recommend TRT as an obesity treatment, but the focus should maintain on lifestyle changes (8).
As no systematic reviews were published emphasizing the effect of TRT in obesity, this systematic review aimed to evaluate the effect of TRT in promoting weight loss, improving quality of life, controlling obesity complications, and preventing cardiovascular events and deaths in men with obesity having LTLs.
Methods
This systematic review was conducted according to the Cochrane Handbook (15) and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (16). Its protocol was registered in the International Prospective Register of Systematic Reviews under CRD42017065598 and previously published (17) with the complementary description shown below.
Eligibility criteria
Randomized controlled trials (RCT) that meet the ‘PICO’ structure described below were included.
Participants (P)
Men with obesity (BMI of ≥30 kg/m2, as well as a BMI of <30 kg/m2, but with central obesity), aged >18 years, with a total testosterone level of ≤350 ng/dL (12.1 nmol/L) were considered. Values higher than this (total or free testosterone) were considered if indicated by the authors as the lower limit of the normal range for healthy young men in their laboratory.
Types of interventions (I)
The intervention group (TRT) comprised of patients who received one of the following testosterones: i.m. injections, buccal mucosa, and transdermally using patches, gels, or liquids.
Comparison (C)
The comparison group was placebo or without TRT.
Outcomes (O)
The primary outcomes were weight loss (assessed by changes in weight, WC, or BMI), safety (frequency of serious adverse events (SAE), total adverse events (TAE), and changes of hematocrit and prostate-specific antigen (PSA) levels from baseline), quality of life (assessed using the aging male symptoms (AMS)), control of obesity complications (improvement of type 2 diabetes mellitus (T2DM), hypertension, dyslipidemia, obstructive sleep apnea, and depression), frequency of cardiovascular events, and deaths.
Secondary outcomes were increased lean body mass, improved symptoms of hypogonadism (using the International Index of Erectile Function, version 5 (IIEF-5)), sustained weight loss, normalized testosterone level, and increased PSA levels.
Identification of studies
General research strategies were created for the following electronic health databases: Embase, MEDLINE, LILACS, and Registry of Controlled Clinical Studies of the Cochrane Collaboration (CENTRAL–Cochrane), Supplementary data (see section on supplementary materials given at the end of this article). Databases were searched on October 10, 2017, and updated on November 23, 2020.
Eligible studies were also surveyed in the Trip Medical Database, SCOPUS, Web of Science, and Cumulative Index to Nursing and Allied Health Literature. Unpublished studies among dissertations and theses (ProQuest Dissertation & Theses Global, WorldCat Dissertations, The Digital Library of Theses and Dissertations of the University of São Paulo, Catalog of Theses & Dissertations-CAPES), ClinicalTrials.gov website, and Brazilian Registry of Clinical Trials (ReBec) were also searched.
Study selection and data extraction
Two reviewers (A S M and L A R B) independently selected the titles and abstracts identified during the literature search. Both used a standardized extraction form so that all information regarding each study is computed.
Assessment of risk of bias
The risk of bias was assessed according to the revised Cochrane risk-of-bias tool for RCTs (RoB 2 tool) (18). Studies without placebo were considered as a high risk of bias, as well as studies with non-balanced losses between groups and those with losses of >10%.
Synthesis and data analysis
The unit of analysis was the data published in the included studies. Similar outcomes were plotted in the meta-analysis using Stata Statistical Software 16 (Stata Statistical Software: Release 16. College Station, TX, StataCorp LLC, USA). The random-effects model was chosen as the analytic model in the meta-analysis.
Dichotomous data were expressed as relative risk (RR) with 95% CI. The mean change from baseline and standard deviations (s.d.) was transformed into mean differences (MD) between the groups plus 95% CI for continuous data. The mean adjusted differences (MADs) were preferred when available. The mean change from baseline was obtained by subtracting the postintervention mean from the baseline mean in studies that only reported postintervention values, and the s.d. was estimated using the following formula: s.d.2change = s.d.2baseline + s.d.2final − (2 × Corr × s.d.baseline ×s.d.final) (15). The estimated mean of the sample and s.d. was obtained from Hozo et al. in studies that reported median and interquartile range (19).
Subgroup analyses
Subgroup analyses were performed according to the time of follow-up, presence of hypogonadal symptoms, average baseline testosterone levels (<8 nmol/L vs levels between 8 and 12 nmol/L), and diagnosis (metabolic syndrome (MS) and/or T2DM and only obesity) for weight loss and HbA1c.
Assessment of statistical heterogeneity
Inconsistencies in the meta-analyses were ascertained by I2 statistic, in which I2 of >50% indicated a moderate probability of heterogeneity, and by chi-squared test (Chi2), where P < 0.10 indicated heterogeneity. Meta-regression was used to explore the causes of inconsistencies. Covariates weeks of treatment, type of intervention, risks of bias, MAD, mean age, and mean BMI of participants were used. The Knapp–Hartung correction was used to calculate the significance of the meta-regression coefficients. In the case of I2 of >30% and more than five studies, the prediction interval (PI) from the random-effects meta-analyses were used since PI predicts the potential underlying effect in a new study, which is different from the average effect from the meta-analyses (20, 21).
Sensitivity analysis
The sensitivity analyses were performed by comparing studies according to the risk of bias, published vs imputed data, and change from baseline vs postintervention values.
Quality of the evidence
The quality of evidence was generated following the Grading of Recommendations Assessment, Development, and Evaluation Working Group (22).
Results
After removing duplicates, search strategies yielded 2535 studies. We selected 25 studies with a high probability of meeting our inclusion criteria (Fig. 1). After a complete references examination, 16 RCTs were included in this review (14, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40). Nine studies were excluded for the following reasons (references on Supplementary data): three had inclusion criteria of testosterone levels above those established in this review; one had most participants with either primary or secondary hypogonadism; two had obesity as an eligibility criterion but not LTLs; and three evaluated outcomes were different from those proposed in this review.
Flow diagram of selected studies.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Flow diagram of selected studies.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Flow diagram of selected studies.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Study characteristics
In addition to LTLs, three studies only required obesity as an eligibility criterion, seven required T2DM diagnoses, and six required T2DM and/or MS. Ten studies used i.m. TRT, and six studies used gel TRT. The control group in four studies received no TRT, whereas the others received a placebo. The mean time of follow-up was 39 weeks (range, 24–56 weeks). The data regarding the main characteristics of each study are presented in Table 1.
Main characteristics of the included studies. Most studies reported that testosterone levels were measured on 2 different days.
| Reference |
Country |
Participants |
HS |
Mean age |
Mean BMI |
Mean total testosterone |
Intervention (no. of participants) |
Control (no. of participants) |
Time of intervention |
Mean HbA1c (%) |
Difference at baseline |
Difference in cointerventions |
Insulin |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 23 | Italy | Men aged between 50 and 65 years, with MS and/or T2DM. TT < 11 nmol/L or FT below 250 pmol/L. Two symptoms of hypogonadism | Yes | 56 | 32 | – | 1000 mg i.m. testosterone undecanoate (32) | i.m. placebo (10) | 52 weeks | 6 | No | No | No |
| 24 | Italy | Men aged between 45 and 65 years, with SM and/or T2DM, TT < 11 nmol/or FT <250 pmol/L. Two symptoms of hypogonadism | Yes | 57.5 | 30.5 | 8.7 nmol/L (ECLIA) | 1000 mg i.m. testosterone undecanoate (40) | i.m. placebo (10) | 24 weeks | 6.5 | No | No | No |
| 25 | US | Veterans, older (age ≥ 65 years), obese (BMI ≥ 30 kg/m2), TT < 10.4 nmol/L, sedentary, stable medications for 6 months | No | 73 | 37 | 7.45 nmol/L (LC-MS/MS) | Androgel 1.62% (42) | Placebo gel (41) | 26 weeks | No | No | No | |
| 26 | US | Men aged between 30 and 65 years, with T2DM with HbA1c ≤ 8%, stable treatment for 3 months. FT serum level < 6.5 ng/dL | No | 53 | 39.8 | 8.7 nmol/L (LC-MS/MS) | 250 mg i.m. testosterone cypionate (20) | i.m. placebo (14) | 24 weeks | 7 | No | Yes | |
| 27 | Denmark | Men aged between 60 and 78 years, with BT < 7.3 nmol, and WC > 94 cm. The cutoff level of BT was defined from observations in a thoroughly characterized population of men aged 20–30 years | No | 67.5 | 30 | 12.6 nmol/L (LC-MS/MS) | 50 mg testosterone gel/day for 3 weeks and increased to 100 mg if testosterone <7.3 nmol (20) | Placebo gel (18) | 26 weeks | – | No | No | No |
| 1 | Australia | Men aged between 18 and 70 years, BMI ≥ 30 kg/m2. TT <12 nmol/L on two different days | No | 53.5 | 37.5 | 6.9 nmol/L (ECLIA) | 1000 mg i.m. testosterone undecanoate (49) | i.m. placebo (51) | 56 weeks | 6 | No | No | No |
| 29 | Australia | Men aged between 35 and 70 years. T2DM. TT ≤ 12 nmol/L on two different days | No | 62 | 32.5 | 10.8 nmol/L (LC-MS/MS) | 1000 mg i.m. testosterone undecanoate (46) | i.m. placebo (43) | 30 weeks | 7 | No | No | No |
| 31 | Slovenia | Men aged > 35 years, BMI ≥ 30 kg/m2, T2DM treated with oral antidiabetic medications. Confirmed untreated late-onset hypogonadism | Yes | 60 | 33.5 | 7.59 nmol/L | 1000 mg i.m. testosterone undecanoate (28) | i.m. placebo (27) | 52 weeks | 8 | No | No | No |
| 33 | UK | Men aged between 18 and 80 years. Symptoms of hypogonadism. TT ≤ 12.0 nmol/L or FT ≤ 0.25 nmol/L. T2DM | Yes | 61.5 | 33 | 9.1 (LC-MS/MS) | 1000 mg i.m. testosterone undecanoate (92) | i.m. placebo (98) | 30 weeks | 7.6 | No | No | Yes |
| 34 | Germany | Men with MS and newly diagnosed with T2DM2. TT < 12 nmol/L on two separate days | No | 56.6 | 32 | 10.5 nmol/L | 50 mg daily transdermal testogel (16) | No TRT (16) | 52 weeks | 7.5 | No | No | No |
| 35 | Belgium, France, Germany, Italy, Netherlands, Spain Sweden, UK | Men ≥ 40 years old, T2DM2 and/or Ms TT ≤ 11 nmol/L or FT ≤255 pmol/L. Two symptoms of hypogonadism | Yes | 59.9 | 32 | 9.4 (ECLIA) | 60 mg testosterone gel (108) | Placebo transdermal gel (102) | 52 weeks | 7.5 | No | No | No |
| 36 | Russia | Hypogonadal men aged between 35 and 70 years suffering from the MS. Total testosterone serum level < 12 nmol/L or free testosterone < 225 pmol/L | No | 52 | 35 | 7.1 nmol/L | 1000 mg i.m. testosterone undecanoate (113) | i.m. placebo (71) | 30 weeks | – | No | No | Yes |
| 38 | Russia | Men newly diagnosed with T2DM, TT < 12.1 nmol/L or FT two times < 243 pmol/L, at least two symptoms of hypogonadism | Yes | 53.5 | 34 | 9.75 nmol/L (ELISA) | 50 mg testosterone gel (38) | No TRT (38) | 39 weeks | 7.8 | No | No | No |
| 14 | Denmark | Men, aged between 50 and 70 years, BT < 7.3 nmol/L, T2DM between 3 months and 10 years, metformin use for more than 3 months. The cutoff of BT was defined from men aged 20–29 years | Yes | 60 | 30.6 | 8.7 nmol/L (LC-MS/MS) | 50 mg testosterone gel for 3 weeks and if BF < 7.3 nmol/L, increased to 100 mg testosterone gel daily (20) | Placebo gel (19) | 24 weeks | 6.5 | No | No | No |
| 39 | Japan | Men with MS and FT ≤ 11.8 pg/mL (Japanese criterium for hypogonadism) | No | 68 | – | 6.9 pg/mL* (RIA) | 250 mg i.m. testosterone enanthate every 4 weeks (32) | No TRT (33) | 52 weeks | 6.4 | Yes** | No | No |
| 40 | Japan | Men with T2DM and free testosterone level ≤ 11.8 pg/mL (Japanese criterium for hypogonadism) | No | 67 | 23.5 | 7.5 pg/mL* (RIA) | 250 mg i.m. testosterone enanthate (47) | No TRT (39) | 52 weeks | 6.6 | No | No | Yes |
*Free testosterone; **Systolic blood pressure was significantly higher in the control group compared to that of the TRT group, and the TRT group had a significantly higher fasting glucose level compared to the control group.
–, no information provided; BT, bioavailable testosterone; ECLIA, electrochemiluminescence; ELISA, DRG Elisa test system (Germany) FT, free testosterone; HS, hypogonadism symptoms; LC-MS/MS, liquid chromatography-tandem mass spectrometry; MS, mass spectrometry; MS, metabolic syndrome; T2DM, type 2 diabetes mellitus; TRT, testosterone replacement therapy; TT, total serum testosterone; WC, waist circumference.
Risk of bias
Figure 2 shows the risk of bias corresponding to the included studies for weight loss outcomes (BMI, WC, and weight) and lean body mass. This assessment had the same classification for other primary outcomes. In Jone’s study (35), a high risk for HbA1c and blood pressure due to missing data were considered.
Risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis
Weight loss
The meta-analysis of WC obtained an effect of −2.42 cm from the TRT (95% CI −3.7 to −1.15 cm, 13 trials, 1186 participants, I2 = 84.2%, Fig. 3). However, a large between-study heterogeneity was found, with 95% PI of −7 to 2.2 cm. Using meta-regression, only the average age of participants was a significant modifier (P = 0.033, Adj R-squared = 41.6%, Fig. 4), which means that a study with an average age close to 50 years has an expected effect between −6 and −2 cm. Contrarily, a study with a mean age of participants older than 60 years would have an expected effect near zero. The meta-analysis of BMI obtained an effect of −0.40, but without a statistical difference (95% CI: −0.99 to 0.18, 13 trials, 1165 participants, I2 = 92.9%, Fig. 5). Only the average age of participants was a significant modifier of the effect size (P = 0.001, Adj R-squared = 70.45%, Supplementary data). For both outcomes, the subgroup and sensitivity analyses did not change the TRT effectiveness (Supplementary data). For weight, the meta-analysis did not show a statistical difference between the groups (Supplementary data), but with heterogeneity, and meta-regression was not performed due to the number of studies included (<10).
Meta-analysis of waist circumference; n, number of participants in the study; RoB2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis of waist circumference; n, number of participants in the study; RoB2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis of waist circumference; n, number of participants in the study; RoB2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
(A) Meta-regression graph of waist circumference and covariate age (bubble plot). A study with an average age of participants close to 50 years would have an expected effect between −8 and −2 cm. By contrast, a study with a mean age of participants older than 60 years would have an expected effect near 0. (B) The percentage of residual variation that is attributable to between-study heterogeneity (I2res) is 78.12%, with the other 21.88% attributable to within-study sampling variability. The proportion of between-study variance explained by the covariate age is 41.6%.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
(A) Meta-regression graph of waist circumference and covariate age (bubble plot). A study with an average age of participants close to 50 years would have an expected effect between −8 and −2 cm. By contrast, a study with a mean age of participants older than 60 years would have an expected effect near 0. (B) The percentage of residual variation that is attributable to between-study heterogeneity (I2res) is 78.12%, with the other 21.88% attributable to within-study sampling variability. The proportion of between-study variance explained by the covariate age is 41.6%.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
(A) Meta-regression graph of waist circumference and covariate age (bubble plot). A study with an average age of participants close to 50 years would have an expected effect between −8 and −2 cm. By contrast, a study with a mean age of participants older than 60 years would have an expected effect near 0. (B) The percentage of residual variation that is attributable to between-study heterogeneity (I2res) is 78.12%, with the other 21.88% attributable to within-study sampling variability. The proportion of between-study variance explained by the covariate age is 41.6%.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis of BMI; n, number of participants in the study; RoB_2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis of BMI; n, number of participants in the study; RoB_2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis of BMI; n, number of participants in the study; RoB_2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Control of obesity complications
For HbA1c, the meta-analysis showed a statistically significant difference in favor of TRT (MD: −0.32%, 95% CI: −0.58 to −0.06, I2 = 83.8%, 13 trials, 909 participants, Fig. 6A). However, due to heterogeneity, the PI was calculated as −1.25 and 0.62%. The meta-regression revealed a P -value of >0.05 in the joint and individual tests for categorical and no categorical covariates. Subgroup analyses did not change the effectiveness (Supplementary data). However, the sensitivity analysis revealed that the statistical difference in favor of the TRT was not maintained in studies classified with a low risk of bias (Fig. 6B).
Meta-analysis of HbA1c. (A) Meta-analysis with prediction interval. (B) Sensitivity analysis according to the risk of bias. SC, some concerns
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis of HbA1c. (A) Meta-analysis with prediction interval. (B) Sensitivity analysis according to the risk of bias. SC, some concerns
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis of HbA1c. (A) Meta-analysis with prediction interval. (B) Sensitivity analysis according to the risk of bias. SC, some concerns
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
The meta-analyses of both systolic and diastolic blood pressure did not show a statistical difference between groups (MD: 1.6 mmHg, 95% CI: −0.58 to 3.79 mmHg; MD: 1.29, 95% CI: −0.19 to 2.76, respectively, I2 = 0, 8 trials, 698 participants, moderate certainty of the evidence, Supplementary data).
The meta-analysis of HDL did not show a statistical difference between the groups (MD: −0.02 mmol/L, 95% CI: −0.08 to 0.03, 12 trials 1097 participants, I2 = 74.8%, Supplementary data). In the meta-regression, the average age of the participants was a significant modifier of the effectiveness (P = 0.018, Adj R-squared = 58.82%, Supplementary data). In seven trials with a low risk of bias, the effect favored placebo (MD: −0.06 mmol/L, 95% CI: −0.10 to −0.02 mmol/L, 826 participants, I2 = 16.4%, Supplementary data).
The meta-analysis of low-density lipoprotein (LDL) showed a small difference in favor of TRT (MD: −0.10 mmol/L, 95% CI: −0.2 to −0.003 mmol/L, 10 trials, 948 participants, Supplementary data). The positive effect was consistent with sensitivity analyses according to the risk of bias.
Cardiovascular events
Six studies reported cardiovascular events; however, none of them had this endpoint as the primary outcome. Thus, the optimal information size was not achieved in the meta-analysis, and evidence was not reported to support a difference between the groups (RR 0.52, 95% CI 0.26 to 1.05, 583 participants, very low-certainty evidence, Supplementary data).
Safety
No difference was found between the groups in the frequency of TAE and SAE (RR: 0.88, 95% CI: 0.7 to 1.12, 7 RCTs, and RR: 0.98, 95% CI: 0.51 to 1.88, 9 RCTs, Supplementary data). However, the certainty of the evidence is low for SAE due to wide CI (Supplementary data).
The meta-analyses of the hematocrit and PSA revealed a small increase in the TRT, 0.03 and 0.15 μg/L, respectively (Supplementary data). However, a large between-study heterogeneity was found (I2 = 94.2 and 65.8%), and the PIs overlap zero. The meta-regression of the joint and individual tests gave a P -value of >0.05, indicating no evidence for the association of these covariates with the TRT effectiveness.
Quality of life
Six studies used the AMS scale, and the summary MD estimate was −3.32 (95% CI: −6.14 to −0.5, 748 participants, I2 = 79.1%, Supplementary data), but the PIs crosses zero. A meta-regression was not performed due to the number of studies included.
Secondary outcomes
Seven studies evaluated the lean body mass, and the meta-analysis showed a statistically significant difference in favor of the TRT (MD: 2 kg, 95% CI 1.46 to 2.63 kg, I2 = 34.7%, 432 participants, moderate certainty of the evidence, Fig. 7). Due to I2 of 35%, the PI was calculated 0.7 to 3.39 kg, which revealed that TRT is beneficial when applied in at least 95% of the individual study settings. Only four studies evaluated the hypogonadism symptom improvements (using the IIEF-5), and the effect was unclear due to imprecision and heterogeneity between the groups (Supplementary data).
Meta-analysis with a prediction interval of the lean body; n, number of participants in the study; Rob2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis with a prediction interval of the lean body; n, number of participants in the study; Rob2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Meta-analysis with a prediction interval of the lean body; n, number of participants in the study; Rob2, risk of bias assessed according to the revised Cochrane risk-of-bias tool for randomized controlled trials; SC, some concerns.
Citation: European Journal of Endocrinology 186, 1; 10.1530/EJE-21-0473
Obstructive sleep apnea, sustained weight loss, depression, and death were not reported as primary or secondary outcomes in the included studies.
Publication bias
Publication bias was investigated for HbA1c, WC, and BMI outcomes. In the funnel plot, the presence of asymmetries was not observed in any of them. Thus, the Egger test was performed, with a P -value of >0.05 for all outcomes (Supplementary data).
Discussion
This systematic review included 16 RCTs that evaluated the effect of TRT in men with obesity having LTLs. Owing to imprecision or large between-study heterogeneity, a clear effect of TRT on cardiovascular events, quality of life, and diabetes control was not found. On average, the TRT is effective for weight loss; however, the PI crosses the null effect line, which means that the true effect is different from the average effect generated from the meta-analysis when applied within an individual study setting. The meta-regression indicated that the studies with an average age of participants close to 50 years had a positive effect of TRT on weight loss and HDL increase but not for studies with an average age higher than 60 years. A small difference was found in favor of TRT for LDL and lean body mass. No difference was found between groups on the frequency of TAE and SAE.
This is the first systematic review to evaluate the effect of TRT in men with obesity having LTLs. The literature revealed two existing systematic reviews of RCTs that evaluated the metabolic effects of TRT on men with hypogonadism with T2DM (41, 42). The review by Cai et al. (41) included five trials, and the meta-analysis of the postintervention values showed that TRT reduced HbA1c. However, the I2 was 36%, and causes of heterogeneity were not explored. Li et al. (42) enrolled 18 studies in their review, and their results showed that TRT reduced HbA1c, BMI, and WC. However, a large between-study heterogeneity was found (I2 = 89%, I2 = 95%, and I2 = 92%), and neither PIs nor meta-regression was provided. Additionally, non-RCTs (43, 44) and studies with men without obesity were included (45, 46, 47).
Ponce et al. conducted a systematic review and meta-analysis of RCTs to determine the effects of TRT in men with hypogonadism. They found four RCTs (including 1779 patients) at low risk of bias, and TRT was associated with a small but significant increase in sexual desire or libido, erectile function, and sexual satisfaction compared to that of placebo (48). However, their eligibility criteria differ from ours; they included RCTs in adult men with morning total serum testosterone (TT) levels of <300 ng/dL, but with one or more symptoms or signs of hypogonadism, with obesity as not mandatory. Consequently, they did not assess weight loss and obesity-related complications.
The random-effect analyses of the hematocrit and PSA levels showed a small increase in the TRT, but with large between-study heterogeneity. The increase in the hematocrit and PSA levels during TRT was higher in older men than in young men (49, 50); however, the meta-regression did not show evidence of an association between the mean age of the participants and the TRT effectiveness. However, low statistical power is not ruled out.
An observational analysis by Hackett et al. studied 857 men who were screened for their RCT (study included in our systematic review) regarding mortality (32, 51). These men were stratified first by testosterone levels (>12 nmol/L vs ≤12 nmol/L), then those with LTLs were further stratified into those on TRT and those not on TRT. Men with TT of >12 nmol/L and men on TRT who did not discontinue the replacement have lower mortality compared with individuals who were not on TRT. However, some confounders in this association were found: compared with those not on therapy replacements, men on TRT were younger (58.3% vs 65.5%, P < 0.001), more with controlled diabetes diet (19.4% vs 10.5%, P = 0.005), fewer with insulin treatment (9.7% vs 20.4%, P = 0.002), and more with phosphodiesterase 5 inhibitors (37.14% vs 14.64%).
The only robust positive effect of TRT found in our review was lean body mass gain. Body fat was associated with a higher risk of major adverse cardiovascular events, and lean body mass was associated with a lower risk of these events (52). However, a post hoc analysis of data from the Action to Control Cardiovascular Risk in Diabetes study that investigated the relationship between the predicted lean body mass or fat mass and major adverse cardiovascular events revealed that increasing lean body mass was not associated with major adverse cardiovascular events in T2DM (hazard ratio 0.92, 95% CI 0.72–1.17; 10 251 participants) (53).
The main limitation of our review is that none of the included studies evaluated cardiovascular events and/or death as primary outcomes. In addition, a variable that could explain the between-study heterogeneity in the effect of TRT on HbA1c was not found. Regarding our published protocol, a few modifications were performed as follows: the units of analysis used were not individuals but data published in the studies, and the effect measure change from baseline was used instead of postintervention values for continuous outcomes. However, the sensitivity analyses using the final values led to the same direction of change score results. Four RCTs without placebo usage were included; however, sensitivity analysis according to the risk of bias was performed, which did not change the direction and TRT effectiveness except for HbA1c and HDL.
Conclusion
Implications for practice
TRT slightly improved the lean body mass and LDL concentrations in men with obesity having LTLs; however, it did not increase the frequency of TAE and did not affect the blood pressure. Due to imprecision and/or heterogeneity, the effect of TRT on cardiovascular events, HbA1c, quality of life, and hypogonadism symptoms is unclear. On average, the TRT is effective for weight loss; however, it is not always beneficial in an individual setting, with the mean age of participants being a significant modifier of the magnitude of the effectiveness.
Implications for research
Further large and well-designed RCTs are needed to identify other causes of heterogeneity between studies, in particular, the groups of participants that benefit most.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/EJE-21-0473
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
This work was supported by the São Paulo Research Foundation (FAPESP) (Grant numbers: 2018/11836-6 and 2018/14411-6).
Acknowledgements
The authors thank Beatriz Ventura Franco e Rebeca Soares Nogueira, who helped us in the gray literature. The authors thank the Brazilian National Research Council (CNPq) for providing us two high school scholarships. The authors would like to thank Enago for the English language review.
References
- 1↑
Fui MNT, Prendergast LA, Dupuis P, Raval M, Strauss BJ, Zajac JD, Grossmann M. Effects of testosterone treatment on body fat and lean mass in obese men on a hypocaloric diet: a randomised controlled trial. BMC Medicine 2016 14 153. (https://doi.org/10.1186/s12916-016-0700-9)
- 2↑
Garvey WT, Mechanick JI, Brett EM, Garber AJ, Hurley DL, Jastreboff AM, Nadolsky K, Pessah-Pollack R, Plodkowski R & reviewers of the AACEOCPG. American Association of Clinical Endocrinologists and American College of Endocrinology Comprehensive Clinical Practice guidelines for medical care of patients with obesity. Endocrine Practice 2016 22 (Supplement 3) 1–203. (https://doi.org/10.4158/EP161365.GL)
- 3↑
Fui MNT, Hoermann R, Prendergast LA, Zajac JD, Grossmann D. Symptomatic response to testosterone treatment in dieting obese men with low testosterone levels in a randomized, placebo-controlled clinical trial. International Journal of Obesity 2017 41 420–426.
- 4↑
Isidori AM, Caprio M, Strollo F, Moretti C, Frajese G, Isidori A, Fabbri A. Leptin and androgens in male obesity: evidence for leptin contribution to reduced androgen levels. Journal of Clinical Endocrinology and Metabolism 1999 84 3673–3680. (https://doi.org/10.1210/jcem.84.10.6082)
- 5↑
Veldhuis J, Yang R, Roelfsema F, Takahashi P. Proinflammatory cytokine infusion attenuates LH’s feedforward on testosterone secretion: modulation by age. Journal of Clinical Endocrinology and Metabolism 2016 101 539–549. (https://doi.org/10.1210/jc.2015-3611)
- 6↑
Kalyani RR, Dobs AS. Androgen deficiency, diabetes, and the metabolic syndrome in men. Current Opinion in Endocrinology, Diabetes, and Obesity 2007 14 226–234. (https://doi.org/10.1097/MED.0b013e32814db856)
- 7↑
Saboor Aftab SA, Kumar S, Barber TM. The role of obesity and type 2 diabetes mellitus in the development of male obesity-associated secondary hypogonadism. Clinical Endocrinology 2013 78 330–337. (https://doi.org/10.1111/cen.12092)
- 8↑
Lapauw B, Kaufman JM. MANAGEMENT OF ENDOCRINE DISEASE: Rationale and current evidence for testosterone therapy in the management of obesity and its complications. European Journal of Endocrinology 2020 183 R167–R183. (https://doi.org/10.1530/EJE-20-0394)
- 9↑
Traish AM Testosterone and weight loss: the evidence. Current Opinion in Endocrinology, Diabetes, and Obesity 2014 21 313–322. (https://doi.org/10.1097/MED.0000000000000086)
- 10↑
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)
- 11↑
Bhasin S, Taylor WE, Singh R, Artaza J, Sinha-Hikim I, Jasuja R, Choi H, Gonzalez-Cadavid NF. The mechanisms of androgen effects on body composition: mesenchymal pluripotent cell as the target of androgen action. Journals of Gerontology: Series A, Biological Sciences and Medical Sciences 2003 58 M1103–M1110. (https://doi.org/10.1093/gerona/58.12.m1103)
- 12↑
Host C, Gormsen LC, Christensen B, Jessen N, Hougaard DM, Christiansen JS, Pedersen SB, Jensen MD, Nielsen S, Gravholt CH. Independent effects of testosterone on lipid oxidation and VLDL-TG production: a randomized, double-blind, placebo-controlled, crossover study. Diabetes 2013 62 1409–1416. (https://doi.org/10.2337/db12-0440)
- 13↑
Saad F, Yassin A, Doros G, Haider A. Effects of long-term treatment with testosterone on weight and waist size in 411 hypogonadal men with obesity classes I–III: observational data from two registry studies. International Journal of Obesity 2016 40 162–170. (https://doi.org/10.1038/ijo.2015.139)
- 14↑
Magnussen LV, Glintborg D, Hermann P, Hougaard DM, Hojlund K, Andersen M. Effect of testosterone on insulin sensitivity, oxidative metabolism and body composition in aging men with type 2 diabetes on metformin monotherapy. Diabetes, Obesity and Metabolism 2016 18 980–989. (https://doi.org/10.1111/dom.12701)
- 15↑
Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (Eds). Cochrane Handbook for Systematic Reviews of Interventions. Oxford: The Cochrane Collaboration and John Wiley & Sons Ltd., 2019.
- 16↑
Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Annals of Internal Medicine 2009 151 W65–W94. (https://doi.org/10.7326/0003-4819-151-4-200908180-00136)
- 17↑
Mangolim AS, Brito LAR, Nunes-Nogueira VS. Effectiveness of testosterone therapy in obese men with low testosterone levels, for losing weight, controlling obesity complications, and preventing cardiovascular events: protocol of a systematic review of randomized controlled trials. Medicine 2018 97 e0482. (https://doi.org/10.1097/MD.0000000000010482)
- 18↑
Higgins JPT, Sterne JAC, Savović J, Page MJ, Hróbjartsson A, Boutron I, Reeves B, Eldridge S. A revised tool for assessing risk of bias in randomized trials. In: Cochrane Methods. Eds Chandler J, McKenzie J, Boutron I, Welch V. Cochrane Database of Systematic Reviews 2016. (https://doi.org/10.1002/14651858.CD201601)
- 19↑
Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Medical Research Methodology 2005 5 13. (https://doi.org/10.1186/1471-2288-5-13)
- 20↑
Cumpston M, Li T, Page MJ, Chandler J, Welch VA, Higgins JP, Thomas J. Updated guidance for trusted systematic reviews: a new edition of the Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Database of Systematic Reviews 2019 10 ED000142. (https://doi.org/10.1002/14651858.ED000142)
- 21↑
Riley RD, Higgins JP, Deeks JJ. Interpretation of random effects meta-analyses. BMJ 2011 342 d549. (https://doi.org/10.1136/bmj.d549)
- 22↑
Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, Norris S, Falck-Ytter Y, Glasziou P & DeBeer H et al.GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables. Journal of Clinical Epidemiology 2011 64 383–394. (https://doi.org/10.1016/j.jclinepi.2010.04.026)
- 23↑
Aversa A, Bruzziches R, Francomano D, Spera G, Lenzi A. Efficacy and safety of two different testosterone undecanoate formulations in hypogonadal men with metabolic syndrome. Journal of Endocrinological Investigation 2010 33 776–783. (https://doi.org/10.1007/BF03350341)
- 24↑
Aversa A, Bruzziches R, Francomano D, Rosano G, Isidori AM, Lenzi A, Spera G. Effects of testosterone undecanoate on cardiovascular risk factors and atherosclerosis in middle-aged men with late-onset hypogonadism and metabolic syndrome: results from a 24-month, randomized, double-blind, placebo-controlled study. Journal of Urology 2011 185 634.
- 25↑
Barnouin Y, Armamento-Villareal R, Celli A, Jiang B, Paudyal A, Nambi V, Bryant MS, Marcelli M, Garcia JM & Qualls C et al.Testosterone replacement therapy added to intensive lifestyle intervention in older men with obesity and hypogonadism. Journal of Clinical Endocrinology and Metabolism 2021 106 e1096–e1110. (https://doi.org/10.1210/clinem/dgaa917)
- 26↑
Dhindsa S, Ghanim H, Batra M, Kuhadiya ND, Abuaysheh S, Sandhu S, Green K, Makdissi A, Hejna J & Chaudhuri A et al.Insulin resistance and inflammation in hypogonadotropic hypogonadism and their reduction after testosterone replacement in men with type 2 diabetes. Diabetes Care 2016 39 82–91. (https://doi.org/10.2337/dc15-1518)
- 27↑
Frederiksen L, Hojlund K, Hougaard DM, Brixen K, Andersen M. Testosterone therapy increased muscle mass and lipid oxidation in aging men. Age 2012 34 145–156. (https://doi.org/10.1007/s11357-011-9213-9)
- 28↑
Ng Tang Fui M, Prendergast LA, Dupuis P, Raval M, Strauss BJ, Zajac JD, Grossmann M. Effects of testosterone treatment on body fat and lean mass in obese men on a hypocaloric diet: a randomised controlled trial. BMC Medicine 2016 14 153. (https://doi.org/10.1186/s12916-016-0700-9)
- 29↑
Gianatti EJ, Dupuis P, Hoermann R, Strauss BJ, Wentworth JM, Zajac JD, Grossmann M. Effect of testosterone treatment on glucose metabolism in men with type 2 diabetes: a randomized controlled trial. Diabetes Care 2014 37 2098–2107. (https://doi.org/10.2337/dc13-2845)
- 30↑
Gianatti EJ, Dupuis P, Hoermann R, Zajac JD, Grossmann M. Effect of testosterone treatment on constitutional and sexual symptoms in men with type 2 diabetes in a randomized, placebo-controlled clinical trial. Journal of Clinical Endocrinology and Metabolism 2014 99 3821–3828. (https://doi.org/10.1210/jc.2014-1872)
- 31↑
Groti K, Zuran I, Antonic B, Forsnaric L, Pfeifer M. The impact of testosterone replacement therapy on glycemic control, vascular function, and components of the metabolic syndrome in obese hypogonadal men with type 2 diabetes. Aging Male 2018 21 158–169. (https://doi.org/10.1080/13685538.2018.1468429)
- 32↑
Hackett G, Cole N, Bhartia M, Kennedy D, Raju J, Wilkinson P & BLAST Study Group. Testosterone replacement therapy improves metabolic parameters in hypogonadal men with type 2 diabetes but not in men with coexisting depression: the BLAST study. Journal of Sexual Medicine 2014 11 840–856. (https://doi.org/10.1111/jsm.12404)
- 33↑
Hackett G, Cole N, Bhartia M, Kennedy D, Raju J, Wilkinson P. Testosterone replacement therapy with long-acting testosterone undecanoate improves sexual function and quality-of-life parameters vs. placebo in a population of men with type 2 diabetes. Journal of Sexual Medicine 2013 10 1612–1627. (https://doi.org/10.1111/jsm.12146)
- 34↑
Heufelder AE, Saad F, Bunck MC, Gooren L. Fifty-two-week treatment with diet and exercise plus transdermal testosterone reverses the metabolic syndrome and improves glycemic control in men with newly diagnosed type 2 diabetes and subnormal plasma testosterone. Journal of Andrology 2009 30 726–733. (https://doi.org/10.2164/jandrol.108.007005)
- 35↑
Jones TH, Arver S, Behre HM, Buvat J, Meuleman E, Moncada I, Morales AM, Volterrani M, Yellowlees A & Howell JD et al.Testosterone replacement in hypogonadal men with type 2 diabetes and/or metabolic syndrome (the TIMES2 study). Diabetes Care 2011 34 828–837. (https://doi.org/10.2337/dc10-1233)
- 36↑
Kalinchenko SY, Tishova YA, Mskhalaya GJ, Gooren LJ, Giltay EJ, Saad F. Effects of testosterone supplementation on markers of the metabolic syndrome and inflammation in hypogonadal men with the metabolic syndrome: the double-blinded placebo-controlled Moscow study. Clinical Endocrinology 2010 73 602–612. (https://doi.org/10.1111/j.1365-2265.2010.03845.x)
- 37↑
Giltay EJ, Tishova YA, Mskhalaya GJ, Gooren LJ, Saad F, Kalinchenko SY. Effects of testosterone supplementation on depressive symptoms and sexual dysfunction in hypogonadal men with the metabolic syndrome. Journal of Sexual Medicine 2010 7 2572–2582. (https://doi.org/10.1111/j.1743-6109.2010.01859.x)
- 38↑
Khripun I, Vorobyev S, Belousov I, Kogan M, Zitzmann M. Influence of testosterone substitution on glycemic control and endothelial markers in men with newly diagnosed functional hypogonadism and type 2 diabetes mellitus: a randomized controlled trial. Aging Male 2019 22 241–249. (https://doi.org/10.1080/13685538.2018.1506918)
- 39↑
Shigehara K, Konaka H, Nohara T, Izumi K, Kitagawa Y, Kadono Y, Iwamoto T, Koh E, Mizokami A, Namiki M. Effects of testosterone replacement therapy on metabolic syndrome among Japanese hypogonadal men: a subanalysis of a prospective randomised controlled trial (EARTH study ). Andrologia 2018 50 e12815. (https://doi.org/10.1111/and.12815)
- 40↑
Shigehara K, Konaka H, Kato Y, Iijima M, Nakashima K, Kawaguchi S, Nohara T, Izumi K, Namiki M, Mizokami A. Effect of testosterone replacement therapy on sexual function and glycemic control among hypogonadal men with type 2 diabetes mellitus. International Journal of Impotence Research 2019 31 25–30. (https://doi.org/10.1038/s41443-018-0065-z)
- 41↑
Cai X, Tian Y, Wu T, Cao CX, Li H, Wang KJ. Metabolic effects of testosterone replacement therapy on hypogonadal men with type 2 diabetes mellitus: a systematic review and meta-analysis of randomized controlled trials. Asian Journal of Andrology 2014 16 146–152. (https://doi.org/10.4103/1008-682X.122346)
- 42↑
Li SY, Zhao YL, Yang YF, Wang X, Nie M, Wu XY, Mao JF. Metabolic effects of testosterone replacement therapy in patients with type 2 diabetes mellitus or metabolic syndrome: a meta-analysis. International Journal of Endocrinology 2020 2020 4732021. (https://doi.org/10.1155/2020/4732021)
- 43↑
Francomano D, Lenzi A, Aversa A. Effects of five-year treatment with testosterone undecanoate on metabolic and hormonal parameters in ageing men with metabolic syndrome. International Journal of Endocrinology 2014 2014 527470. (https://doi.org/10.1155/2014/527470)
- 44↑
Yan W, Wei W, Al E, Sue S, Liu D, Chunyan D. Effect of androgen replacement therapy on late onset hypogonadism in men with diabetes mellitus. Chinese Journal of Gerontology 2015 13 06.
- 45↑
Kapoor D, Goodwin E, Channer KS, Jones TH. Testosterone replacement therapy improves insulin resistance, glycaemic control, visceral adiposity and hypercholesterolaemia in hypogonadal men with type 2 diabetes. European Journal of Endocrinology 2006 154 899–906. (https://doi.org/10.1530/eje.1.02166)
- 46↑
Boyanov MA, Boneva Z, Christov VG. Testosterone supplementation in men with type 2 diabetes, visceral obesity and partial androgen deficiency. Aging Male 2003 6 1–7. (https://doi.org/10.1080/tam.6.1.1.7)
- 47↑
Gopal RA, Bothra N, Acharya SV, Ganesh HK, Bandgar TR, Menon PS, Shah NS. Treatment of hypogonadism with testosterone in patients with type 2 diabetes mellitus. Endocrine Practice 2010 16 570–576. (https://doi.org/10.4158/EP09355.OR)
- 48↑
Ponce OJ, Spencer-Bonilla G, Alvarez-Villalobos N, Serrano V, Singh-Ospina N, Rodriguez-Gutierrez R, Salcido-Montenegro A, Benkhadra R, Prokop LJ & Bhasin S et al.The efficacy and adverse events of testosterone replacement therapy in hypogonadal men: a systematic review and meta-analysis of randomized, placebo-controlled trials. Journal of Clinical Endocrinology and Metabolism 2018 103 1745 - 1754. (https://doi.org/10.1210/jc.2018-00404)
- 49↑
Coviello AD, Kaplan B, Lakshman KM, Chen T, Singh AB, Bhasin S. Effects of graded doses of testosterone on erythropoiesis in healthy young and older men. Journal of Clinical Endocrinology and Metabolism 2008 93 914–919. (https://doi.org/10.1210/jc.2007-1692)
- 50↑
Bhasin S, Singh AB, Mac RP, Carter B, Lee MI, Cunningham GR. Managing the risks of prostate disease during testosterone replacement therapy in older men: recommendations for a standardized monitoring plan. Journal of Andrology 2003 24 299–311. (https://doi.org/10.1002/j.1939-4640.2003.tb02676.x)
- 51↑
Hackett G, Cole N, Mulay A, Strange RC, Ramachandran S. Long-term testosterone therapy in type 2 diabetes is associated with reduced mortality without improvement in conventional cardiovascular risk factors. BJU International 2019 123 519–529. (https://doi.org/10.1111/bju.14536)
- 52↑
Medina-Inojosa JR, Somers VK, Thomas RJ, Jean N, Jenkins SM, Gomez-Ibarra MA, Supervia M, Lopez-Jimenez F. Association between adiposity and lean mass with long-term cardiovascular events in patients with coronary artery disease: no paradox. Journal of the American Heart Association 2018 7 e007505. (https://doi.org/10.1161/JAHA.117.007505)
- 53↑
Xing Z, Tang L, Chen J, Pei J, Chen P, Fang Z, Zhou S, Hu X. Association of predicted lean body mass and fat mass with cardiovascular events in patients with type 2 diabetes mellitus. CMAJ 2019 191 E1042–E1048. (https://doi.org/10.1503/cmaj.190124)
