Journal of Endocrinological Investigation

, Volume 39, Issue 9, pp 967–981 | Cite as

Testosterone supplementation and body composition: results from a meta-analysis of observational studies

  • G. Corona
  • V. A. Giagulli
  • E. Maseroli
  • L. Vignozzi
  • A. Aversa
  • M. Zitzmann
  • F. Saad
  • E. Mannucci
  • M. Maggi
Review

Abstract

Purpose

The concept of testosterone (T) supplementation (TS) as a new anti-obesity medication in men with testosterone deficiency syndrome (TDS) is emerging. Data from placebo-controlled trials are more conflicting. The aim of this study is to systematically review and meta-analyze available observational and register studies reporting data on body composition in studies on TS in TDS.

Methods

An extensive MEDLINE, Embase, and Cochrane search was performed including the following words: “testosterone” and “body composition.” All observational studies comparing the effect of TS on body weight and other body composition and metabolic endpoints were considered.

Results

Out of 824 retrieved articles, 32 were included in the study enrolling 4513 patients (mean age 51.7 ± 6.1 years). TS was associated with a time-dependent reduction in body weight and waist circumference (WC). The estimated weight loss and WC reduction at 24 months were −3.50 [−5.21; −1.80] kg and −6.23 [−7.94; −4.76] cm, respectively. TS was also associated with a significant reduction in fat and with an increase in lean mass as well as with a reduction in fasting glycemia and insulin resistance. In addition, an improvement of lipid profile (reduction in total cholesterol as well as of triglyceride levels and an improvement in HDL cholesterol levels) and in both systolic and diastolic blood pressure was observed.

Conclusions

Present data support the view of a positive effect of TS on body composition and on glucose and lipid metabolism. In addition, a significant effect on body weight loss was observed, which should be confirmed by a specifically designed RCT.

Keywords

Testosterone Weight loss Body composition Obesity diabetes Hypogonadism 

Introduction

Overweight and obesity have become an epidemic problem, on the rise not only in Western societies, but also in developing countries. Body mass index (BMI) is universally used to categorize overweight and obesity, along with its severity [1, 2]. Although the classification of obesity is based on BMI, the crucial abnormality that characterizes this condition is an excess of fat mass, rather than body weight. In addition, BMI does not take into account the accumulation of visceral fat, which characterizes the most morbigenous form of obesity: visceral (central) obesity, i.e., abdominal fat depots around visceral organs. Waist circumference (WC) is an alternative measure that reflects visceral adiposity, and it is superior to BMI in predicting cardiovascular (CV) disease [3]. This is largely explained by the association of increased visceral adipose tissue (VAT) with a range of metabolic and CV abnormalities included in the construct of metabolic syndrome (MetS) [4, 5].

MetS and visceral obesity are nowadays considered the main determinants of male hypogonadism (HG), in particular when these conditions become manifest during adulthood (testosterone deficiency syndrome, TDS). In fact, both cross-sectional and longitudinal studies indicate that obesity increases the risk of testosterone deficiency ([6, 7, 8, 9, 10], see in Refs. [11, 12, 13, 14] for review).

Testosterone (T) is an anabolic hormone that plays a relevant role in decreasing VAT accumulation and increasing muscle mass. Preclinical studies show that, in differentiating adipocytes, an androgen-dependent increase in Wnt signaling promotes differentiation of resident mesenchymal cells into myocytes and osteocytes, while suppressing their commitment toward the adipocyte lineage and terminal differentiation, by inhibiting the expression of peroxisome proliferator-activated receptor gamma (PPARγ) and CAAT/enhancer-binding protein alpha (CEBPα), the central regulators of adipogenesis [4, 15]. Accordingly, several observational and register studies indicate that T supplementation (TS) in hypogonadal men is able to dramatically reduce waist circumference, body weight and BMI in a time-dependent fashion [16, 17, 18, 19, 20, 21, 22, 23, 24]. In longer-term studies assessing anthropometric parameters, results with testosterone undecanoate (TU) in TDS-associated obesity seem to be superior even to other lifestyle or medical interventions and comparable to those obtained with the most invasive therapies, such as bariatric surgery [25, 26].

Randomized controlled trials (RCTs), designed for a specific purpose, are generally considered the gold standard for investigating the efficacy of a particular health intervention, including medical treatment. In fact, RCTs are at the top of the evidence-based hierarchy of the best available data to develop clinical guidelines aimed at standardizing procedures and therapies. However, we are unaware of studies specifically designed to obtain evidence for evaluating the effect of TS on body composition in HG, at least as a primary endpoint. Information derived from RCTs designed for other purposes is available, although they are more fragmentary and often contradictory. In this case, meta-analyses may aid in clinical reasoning and establish a more solid foundation for causal inferences, because they are able to overcome between-studies heterogeneity and to address questions across many different research domains for which data sources are conflicting or lack adequate statistical power. A recent meta-analysis on all available RCTs comparing the effect of TS on different metabolic endpoints showed that TS determined a significant reduction of fat and an increase in lean mass as well as a reduction in fasting glycemia and insulin resistance [27]. Similar results were observed for total cholesterol and triglyceride levels when only RCTs enrolling hypogonadal (total T < 12 mol/L) subjects were considered [27]. Interestingly, the effect of TS on body composition was even better when subjects with metabolic diseases at enrollment were considered [27]. This finding is not surprising since previous meta-analysis of available evidence documented that TS was associated with an improvement of metabolic profile either in patients with MetS or in those with type 2 diabetes mellitus (T2DM) [4]. However, in the aforementioned meta-analysis on body composition, TS does not appear to induce weight loss. The dual, opposite effects of TS on body composition, i.e., an increase in lean mass and a reduction in fat mass, might justify the overall null efficacy of TS on body weight or BMI. The peculiar study design of these RCTs might also justify the lack of efficacy of TS on weight parameters. In fact, RCTs are performed under idealized and rigorously controlled conditions, which are different from everyday clinical practice. Hence, results of RCTs offer an indication of the efficacy of an intervention rather than its effectiveness in everyday practice. For example, patients enrolled in RCTs are not representative of the whole population of patients in the real world, because of the exclusion of certain categories of subjects (e.g., elderly subjects, those with comorbidities, individuals with low compliance to treatments). In addition, patients who participate may receive better care, regardless of which arm of the trial they are in. In contrast, observational and register studies maintain the integrity of the context in which medical care is provided [28, 29]. As a result, whereas RCTs provide an indication of the minimal effect of an intervention, observational studies offer an estimate of the maximal one. In addition, observational studies are more likely to provide an indication of daily medical practice and to measure the effectiveness of an intervention in ‘real world’ scenarios [28, 29].

The aim of this study is to systematically review and meta-analyze available observational and register studies reporting data on body composition in studies on TS in HG. A comparison with results obtained in RCTs [27] was also performed.

Methods

This meta-analysis was performed according to the MOOSE Guidelines for Meta-Analyses and Systematic Reviews of Observational Studies [30].

An extensive MEDLINE, Embase, and Cochrane search was performed including the following words: (“testosterone”[MeSH Terms] OR “testosterone”[All Fields]) AND (“body composition”[MeSH Terms] OR (“body”[All Fields] AND “composition”[All Fields]) OR “body composition”[All Fields] AND English[lang] AND “male”[MeSH Terms]). The search accrued data from January 1, 1969, up to August 31, 2014. In addition, completed but still unpublished trials evaluating the effects of TS on different outcomes were identified through a search on the http://www.clinicaltrials.gov website. The identification of relevant studies was performed independently by two of the authors (V.G., G.C.), and conflicts resolved by the third investigator (A.A.). We did not employ search software. We hand-searched bibliographies of retrieved papers for additional references. The principal source of information was derived from published articles; if data were missing from the publication, an attempt at retrieval was made through the clinicaltrial.gov website.

Study selection

We included all observational studies, comparing the effect of TS on different endpoints ([16, 17, 18, 19, 20, 21, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56]; see also Fig. 1; Tables 1, 2, 3). Studies not specifically evaluating body composition or glycometabolic outcomes were excluded from the analysis (not shown [57, 58, 59, 60, 61, 62; see also Table 4). Studies using androgens other than TS as well as studies with simultaneous treatment with other hormones, and drugs were also excluded, unless there was a clearly defined treatment arm that received only T treatment. To compare results in observational studies to those reported in RCTs [27], we restricted the analysis to the same time frame as before (August 2014).
Fig. 1

Trial flow diagram. Unpublished studies were identified through a search of http://www.clinicaltrials.gov website. N number, HIV human immunodeficiency virus; GnRH gonadotropin-releasing hormone; Gn gonadotropins; T testosterone

Table 1

Characteristics of the clinical studies included in the meta-analysis

References

# Patients (T/placebo or controls)

Trial duration (months)

Age (years)

Type of population

Baseline total T (nmol/l)

T levels

Dose

Type of study

Valdermasson et al. [31]

10/0

9

LOH

1.8

<8 nM

TE 250 mg/3–4 weeks

BA

Rebuffé-Scrive et al. [32]

11/0

1.5

42

Overweight/obesity

13.8

Mixed

TU 120–160 mg/day

BA

Forbes et al. [33]

7/0

4

Healthy

Eugonadism

TE 42 mg/kg/week

BA

Marin and Krotkievski et al. [34]

11/0

1.5

42

Obesity

13.8

Mixed

TU oral 160 mg/day

CBA

Marin and Krotkievski et al. [34]

8/0

3

42

Obesity

14.1

Mixed

T gel 250 mg/day

CBA

Brodsky et al. [35]

5/0

6

LOH

3.7

<8 nM

TC 3 mg/kg/2 weeks

BA

Katznelson et al. [36]

29/0

13

57

LOH

6.4

<10.4 nM

TE or TC 100 mg/week

BA

Wang et al. [37]

67/0

6

LOH

4.1

<8.7

T sublingual 15 mg/day

BA

Zgliczynski et al. [38]

22/0

12

58.5

Elderly men

4.3

<12

TE 200 mg/2 weeks

BA

Bhasin et al. [39]

7/0

2.5

34.7

LOH

2.5

<8

TE 100 mg/week

BA

Tan et al. [40]

11/0

4

33.3

LOH

5.5

<8

TE 250 mg/4 weeks

CBA

Brill et al. [41]

10/0

1

68.1

Elderly men

15

Mixed

T patch 5 mg/day

BA

Minnemann et al. [42]

25/0

57

LOH

14.3

Mixed

TU 1000 mg/12 weeks from week 6

BA

Naharci et al. [43]

24/0

6

20.7

HH

5.7

<8

Mixed ester 250 mg/3 weeks

CBA

Saad et al. [16]

28/0

13

 

LOH with ED

7.5

<10.4

TU 1000 mg/12 weeks from week 6

BA

Saad et al. [17]

27/0

9

60

LOH with ED

7.5

<12

TU 1000 mg/12 weeks from week 6

BA

Saad et al. [17]

27/0

9

60

LOH with ED

7.5

<12

T gel 50 mg/day

BA

La Vignera et al. [44]

7/0

3

58

LOH with MetS

<8

T gel 50 mg/day

CBA

Moon et al. [45]

133/0

6

54

LOH

8.6

<12

TU 1000 mg/12 weeks from week 6

BA

Permpongkosol et al. [46]

161/0

13.5

60.4

LOH consulting urological center

9.4

<10,4

TU 1000 mg/12 weeks from week 6

BA

Garcia et al. [47]

29/0

25.5

55.5

LOH and diabetes

<10,4

TU 1000 mg/12 weeks from week 6

BA

Schwarz &Willix [48]

56/0

18

52.3

Overweight or obesity

15

Mixed

TC 80–200 mg/week + diet + training

BA

Arafa et al. [49]

56/0

12

55.5

T2DM

<10.4

TU 1000 mg/12 weeks from week 6

CBA

Schroeder et al. [50]

29/0

4

71

Elderly men

13.1

Mixed

T patch 5 or 10 mg/day

CBA

Jo et al. [51]

18/0

26.8

35.9

Klinefelter syndrome

3.1

<8

TU 1000 mg/12 weeks from week 6

BA

Ko et al. [52]

246/0

14.7

58.5

LUTS

<12

TU 1000 mg/12 weeks from week 6

BA

Rodriguez-Tolrà et al. [53]

50/0

12

59.1

LOH

10.2

Mixed

T gel 25–100 mg/day

BA

Saad et al. [18]

255/0

60

58

Mixed urological population

10.0

Mixed

TU 1000 mg/12 weeks from week 6

BA

Tirabassi et al. [54]

15/0

18.5

55.7

LOH

5.2

<8

TU 1000 mg/12 weeks from week 6

BA

Zitzmann et al. [55]

1438/0

10.5

49.2

LOH

9.6

Mixed

TU 1000 mg/12 weeks from week 6

BA

Francomano et al. [19]

20/0

60

57.5

MetS

8.3

<11

TU 1000 mg/12 weeks from week 6

CBA

Pexman-Fieth et al. [56]

669/0

6

53

LOH

Mixed

T gel 50, 75 or 100 mg/day

BA

Yassin et al. [20]

261/0

54

59.5

LOH

7.7

<12

TU 1000 mg/12 weeks from week 6

BA

Zitzmann et al. [21]

381/0

60

42.6

LOH

5.2

<8

TU 1000 mg/12 weeks from week 6

BA

T Testosterone, TE testosterone enanthate, TU testosterone undecanoate, TC testosterone cypionate, LOH late onset hypogonadism, T2DM type 2 diabetes mellitus, MetS metabolic syndrome, BA controlled cohort before-and-after comparisons in the same group of patients, CBA controlled before-and-after study between two or more groups of participants receiving different interventions

Table 2

Outcomes of the clinical studies included in the meta-analysis

References

Body weight

Body mass index

WC

Fat mass

Lean mass

Fasting glycemia

HOMA

Total cholesterol

HDL cholesterol

TG

Diastolic blood pressure

Systolic blood pressure

Valdemarsson et al. [31]

X

      

x

x

x

  

Rebuffé-Scrive et al. [32]

X

      

x

 

x

x

x

Forbes et al. [33]

X

  

X

X

       

Marin and Krotkievski et al. [34]

X

  

X

X

x

      

Marin and Krotkievski et al. [34]*

X

  

X

X

x

      

Brodsky et al. [35]

X

  

X

X

  

x

x

x

  

Katznelson et al. [36]

X

  

X

X

  

x

x

x

  

Wang et al. [37]

X

x

X

X

X

   

x

 

x

 

Zgliczynski et al. [38]

 

x

     

x

x

x

  

Bhasin et al. [39]

X

  

X

X

       

Tan et al. [40]

       

x

x

x

  

Brill et al. [41]

 

x

 

X

X

       

Minnemann et al. [42]

       

x

x

x

  

Naharci et al. [43]

X

x

 

X

X

x

X

     

Saad et al. [16]

X

 

X

  

x

 

x

x

x

x

x

Saad et al. [17]

X

 

X

  

x

 

x

x

x

x

x

Saad et al. [17]*

X

 

X

  

x

 

x

x

x

x

x

La Vignera et al. [44]

 

x

X

     

x

x

  

Moon et al. [45]

 

x

   

x

 

x

x

x

  

Permpongkosol et al. [46]

 

x

X

X

   

x

    

Garcia et al. [47]

  

X

  

x

 

x

 

x

x

x

Schwarz & Willix [48]

X

x

 

X

 

x

 

x

x

x

  

Arafa et al. [49]

 

x

   

x

 

x

x

x

  

Schroeder et al. [50]

    

X

      

x

Jo et al. [51]

X

x

          

Ko et al. [52]

 

x

          

Rodriguez-Tolrà et al. [53]

   

X

X

       

Saad et al. [18]

X

x

X

  

x

 

x

x

x

x

 

Tirabassi et al. [54]

X

 

X

  

x

X

x

x

x

x

x

Zitzmann et al. [55]

X

 

X

    

x

x

x

x

x

Francomano et al. [19]

X

x

X

   

X

  

x

x

x

Pexman-Fieth et al. [56]

 

x

X

         

Yassin et al. [20]

X

x

X

  

x

 

x

x

x

x

x

Zitzmann et al. [21]

X

x

X

  

x

 

x

x

x

x

x

WC waist circumference, HDL high-density lipoprotein, TG triglycerides

Data on *Testosterone gel

Table 3

Characteristics and outcomes of the randomized clinical trials and observational studies included in the meta-analysis

References

Design

Blinding

Dropout

Intention-to-treat

Eligibility criteria listed

Valdemarsson et al. [31]

NAP

NAP

NA

No

Yes

Rebuffé-Scrive et al. [32]

NAP

NAP

NA

No

No

Forbes et al. [33]

NAP

NAP

NA

No

No

Marin and Krotkievski et al. [34]

NAP

NAP

NA

No

No

Marin and Krotkievski et al. [34]

NAP

NAP

NA

No

No

Brodsky et al. [35]

NAP

NAP

NA

No

Yes

Katznelson et al. [36]

NAP

NAP

A

Yes

Yes

Wang et al. [37]

NAP

NAP

A

No

Yes

Zgliczynski et al. [38]

NAP

NAP

NA

Yes

Yes

Bhasin et al. [39]

NAP

NAP

A

Yes

Yes

Tan et al. [40]

NAP

NAP

NA

No

Yes

Brill et al. [41]

NAP

NAP

NA

No

Yes

Minnemann et al. [42]

NAP

NAP

A

Yes

Yes

Naharci et al. [43]

NAP

NAP

NA

Yes

Yes

Saad et al. [16]

NAP

NAP

NA

No

Yes

Saad et al. [17]

NAP

NAP

NA

No

Yes

Saad et al. [17]*

NAP

NAP

NA

No

Yes

La Vignera et al. [44]

NAP

NAP

A

Yes

Yes

Moon et al. [45]

NAP

NAP

A

Yes

Yes

Permpongkosol et al. [46]

NAP

NAP

NA

No

No

Garcia et al. [47]

NAP

NAP

NA

Yes

Yes

Schwarz and Willix [48]

NAP

NAP

A

Yes

Yes

Arafa et al. [49]

NAP

NAP

A

Yes

Yes

Schroeder et al. [50]

NAP

NAP

NA

Yes

Yes

Jo et al. [51]

NAP

NAP

NA

Yes

Yes

Ko et al. [52]

NAP

NAP

NA

Yes

Yes

Rodriguez-Tolrà et al. [53]

NAP

NAP

A

Yes

Yes

Saad et al. [18]

NAP

NAP

A

Yes

No

Tirabassi et al. [54]

NAP

NAP

NAP

NAP

Yes

Zitzmann et al. [55]

NAP

NAP

A

A

Yes

Francomano et al. [19]

NAP

NAP

NA

No

Yes

Pexman-Fieth et al. [56]

NAP

NAP

A

A

Yes

Yassin et al. [20]

NAP

NAP

A

Yes

Yes

Zitzmann et al. [21]

NAP

NAP

Yes

No

Yes

A Adequately described, NA not adequately described, NAP not applicable

Data on *Testosterone gel

Table 4

Studies that met inclusion criteria but did not provide data for meta-analysis

First Authors

Years

Short summery of the study and main conclusions

Conway et al. [57]

1988

With the aims of comparing the pharmacokinetics and pharmacodynamics of three different formulations of T, 15 hypogonadal men were studied. However, those results were not included in our meta-analysis since the group of participants consisted of not all unaware patients

Sorva et al. [58]

1988

The study consisted in evaluating the TRT effects on serum lipids and lipoproteins in 13 hypogonadal men. This work was ruled out given that the patient group was made up of prepubertal and postpubertal men

Ozata et al. [59]

1996

The effects of gonadotropin and T therapies on lipids and lipoprotein(a) were studied in 31 hypogonadal males. The data were not considered for our study since the patients were prepubertal men

Tripathy et al. [60]

1998

This study aims to look at the T role in developing insulin resistance and other cardiovascular risks factors in men. The data were not included in the meta-analysis as they were obtained from a group of subjects affected by delayed puberty

Andrade et al. [61]

2009

Observational cohort before-and-after study between 2 groups of participants receiving T (T cyprionate 200 mg/3 weeks) or not. There are no data about circulating T both at baseline and at the end of the study. In addition, the data are not complete

Wu et al. [62]

2009

Observational cohort before-and-after study comparison in the same group of prepubertal hypogonadic men. TRT (T undercyclate 40 mg o.s.) increased weight and BMI. The data were not considered for our study since the patients were prepubertal boys

Outcome

The principal outcome of this analysis was the effect of TS, as compared to baseline, on body composition modification including fat and lean mass (Table 2). Secondary outcomes included several other glycometabolic parameters (Table 2).

Quality assessment

The quality of trials was assessed using the Cochrane criteria ([63]; see also Table 3). In particular, we evaluated the following criteria: the weaknesses of the designs that have been used (such as noting their potential to ascertain causality), the execution of the studies through a careful assessment of their risk of bias, especially the potential for selection bias and confounding to which all observational studies are susceptible, and the potential for reporting biases, including selective reporting of outcomes. For each study, we also assessed how the population was selected, the duration and route of TS, and the adequacy of study follow-up [64].

Statistical analysis

Heterogeneity was assessed by using I2 statistics. Even when low heterogeneity was detected, a random-effects model was applied, because the validity of tests of heterogeneity can be limited with a small number of component studies. To estimate possible publication or disclosure bias, we used funnel plots and the Begg adjusted rank correlation test [65, 66]. However, because these tests have low statistical power when the number of trials is small, undetected bias may still be present. In addition, since in some trials the significance of between group comparisons (p) was not reported, the analysis was performed evaluating endpoint values of each parameter in different treatment groups, in a non-paired fashion (non-paired analysis). Considering that most of the studies, which did not describe p values, reported nonsignificant differences across groups, the mean (paired) analysis, which excludes those data, is likely to overestimate the effect of treatments. On the other hand, the non-paired analysis is a very conservative approach, which could underestimate treatment effect. Since body fat mass and lean mass were evaluated through different approaches and expressed in different ways, the mean difference for each study was divided by the pooled estimate of the SD, in order to express the effect size for each study in a common metric, namely the standardized mean difference (SMD). According to Cohen [66], a small treatment-effect size is considered to be about 0.2, a medium effect size to be about 0.5, and a large effect size to be about 0.8. All other data were expressed as weight mean differences. A meta-regression analysis was performed to test the effect of different parameters on TS reduction in weight and waist circumference. All analyses were performed using Comprehensive Meta-analysis version 2, Biostat (Englewood, NJ, USA). Multivariate analyses were performed on SPSS (Statistical Package for the Social Sciences; Chicago, USA) for Windows 22.0.

Results

Out of 824 retrieved articles, 32 were included in the study (Fig. 1). The characteristics of the retrieved trials (including parameters on trial quality) and type of outcomes considered are reported in Tables 1, 2, and 3. Retrieved studies included 4513 subjects, all receiving TS. TS was administered in different doses, formulations, and cohorts (Table 1).

When compared to a previous meta-analysis [27] of RCT only, patients included in observational trials were younger and had lower T levels, higher prevalence of diabetes mellitus, and longer duration of follow-up (see Table 5). Conversely, no difference between groups in BMI levels at enrollment was observed (Table 5).
Table 5

Comparison of the most important clinical characteristics at enrollment between observational (included in the present meta-analysis) and randomized controlled trials [27]

Parameter

Observation trials

Randomized controlled trials

p

Age (years)

51.7 ± 6.1

62.0 ± 8.5

<0.0001

Testosterone levels (nM)

7.2 ± 3.6

11.6 ± 2.7

<0.0001

Trial duration (months)

18.7 ± 19.0

9.1 ± 8.1

<0.0001

Diabetes mellitus (%)

20.4 ± 13.6

18.3 ± 33.3

0.001

Body mass index (kg/m2)

29.3 ± 2.9

28.6 ± 2.6

NS

Body composition

Among studies reporting several outcomes, 19 included information on weight (Table 2). I2 was 92.44, p < 0.0001. Funnel plot and Begg adjusted rank correlation test suggested no major publication bias (Kendall’s τ: −0.14; p = 0.38, see also Fig. 2). TS was associated with a reduction in body weight (Fig. 3a), BMI (Fig. 3b), and WC (Fig. 3c). In addition, meta-regression analysis showed that the reduction in both weight and WC at endpoint was more evident in longer studies, in a time-dependent manner (Fig. 4a, b, respectively). The estimated weight loss and WC reduction at 12 and 24 months were −0.62 [−1.97; 0.72] kg; −3.47 [−4.82; −2.24] cm and −3.50 [−5.21; −1.80] kg; −6.23 [−7.94; −4.76] cm, respectively. Meta-regression also indicates that results with TS on weight loss and WC are more evident in younger (Fig. 4c, d, respectively) and more hypogonadal subjects (Fig. 4e, f, respectively) at enrollment. Similarly, the effect of TS on weight loss and WC was higher according to the prevalence of diabetes mellitus at baseline (S = −0.309 [−0.356; 0.006]; p < 0.0001 and I = −0.243 [−0.391; −0.096]; p < 0.0001 and S = −0.653 [−0.752; −0.555]; p < 0.0001 and I = 8.798 [6.092; 11.505]; p < 0.0001 for weight loss and waist circumference, respectively).
Fig. 2

Funnel plot of observational trials examining the effects of testosterone therapy on body weight. *Testosterone gel-treated men

Fig. 3

Weighted mean differences (with 95 % CI) of body weight (a), waist circumference (b), and body mass index (c) at endpoint in comparison with enrollment data

Fig. 4

Influence of trial duration (a, b), age (c, d) and testosterone levels at enrollment (e, f) on weighted mean differences (with 95 % CI) of body weight (a, c, e) and waist circumference (b, d, f) at endpoint after testosterone supplementation. The size of the circles reflects the sample dimension

Information on fat mass and lean mass was available from 11 and 10 studies, respectively (Table 2). TS was associated with a significant reduction in fat and with an increase in lean mass (Table 6; see also Supplementary Figure 1, panels A–B).
Table 6

Mean difference or mean standardized differences in several clinical parameters after testosterone substitution as derived from meta-analysis of the included studies

Parameter

Outcome

p value

Body composition

Fat mass (standardized mean)

−0.62 [−1.06; −0.18]

<0.01

Lean mass (standardized mean)

0.62 [0.18; 1.05]

<0.01

Glucose profile

Fasting glycemia (mM)

−0.47 [−0.62; −0.32]

<0.0001

HOMA index

−2.85 [−3.13; −2.58]

<0.0001

Lipid profile

Total cholesterol (mM)

−0.83 [−1.15; −0.52]

<0.0001

Triglycerides (mM)

−0.47 [−0.61; −0.33]

<0.0001

HDL cholesterol (mM)

0.12 [0.02; 0.22]

<0.01

Blood pressure

Systolic blood pressure (mmHg)

−6.62 [−12.19; −1.05]

<0.0001

Diastolic blood pressure (mmHg)

−6.37 [−10.19; −2.56]

<0.0001

Glyco-metabolic profile and blood pressure

TS was associated with a reduction in fasting glycemia and insulin resistance (IR), as detected by HOMA-IR index (Table 6; see also Supplementary Figure 1, panels C, D). Meta-regression analysis showed that the effect of TS on fasting glycemia was higher in those studies enrolling a higher prevalence of diabetic subjects at baseline (S = −0.010 [−0.016; −0.262]; p < 0.0001 and I = −0.595 [−1.452; 0.263]; p < 0.0001). In addition, when lipid profile was analyzed, reductions in total cholesterol as well as triglyceride levels were observed (Table 6; see also Supplementary Figure 1, panels E–F). Finally, an improvement in HDL cholesterol levels as well as in both systolic and diastolic blood pressure was detected (Table 6; see also Supplementary Figure 1, panels G–I).

Discussion

The present meta-analysis of observational studies on TS in HG shows that TS is associated with weight loss. On average, TS is associated with a pooled 3-kg decrease in weight, one-point decrease in BMI, and a 6-cm loss in WC. This result is at variance with that of RCTs, as previously reported [27]. In general, RCTs are considered more reliable for the assessment of treatment effects, because they allow a comparison with a randomly allocated control group, thus eliminating spontaneous variations of the parameters investigated, together with the effect of being included in a study [28, 29]. On the other hand, RCTs are performed on a selected population of patients, who may differ from that of subjects actually receiving the treatment in clinical practice [28, 29]. In fact, the observational studies summarized here enrolled patients who were younger and showed a lower baseline T level and higher prevalence of diabetes mellitus than those included in RCTs. Meta-regression analysis indicated that a higher prevalence of diabetes mellitus, lower age, and a lower baseline T are predictors of a greater weight loss with TS. It is conceivable that these differences might explain, at least partially, the different weight outcomes observed between observational studies and RCTs. Another, even more relevant, difference between observational studies and RCTs is the duration of follow-up (only 9 months in RCTs, on average, vs. more than 18 months in observational studies). Weight loss, like weight gain, is a slow process, which requires sufficient time. It is not surprising that, if a treatment produces weight loss, this effect becomes evident only in studies of sufficient duration. In particular, the positive effect of TS on weight loss in observational studies, involving unselected cohorts of HG men, was apparent only after 24- but not 12-month follow-up. It is interesting to note that only few RCTs extended their follow-up past this limit. Theoretically, according to equations derived from the present meta-regression analysis, a 12-kg weight loss and 14.5-cm reduction in WC can be extrapolated in 5 years. Recently, real-life data from two independent observational registries involving 411 obese, hypogonadal men receiving TS for 8 years were published [24]. A total weight loss of 17.4 ± 0.5, 25.3 ± 0.5, and 30.5 ± 0.7 kg, in obesity classes I, II and III, respectively, was reported. These figures even extended our meta-analytic broadcast, most probably because in the Saad study [24] only obese subjects were enrolled. Accordingly, more pronounced positive effects of TS on weight were evident also in other cohorts considering obese subjects [22, 23, 67]. The study cited above [24] clearly shows that weight loss with T, as well as with any weight-reducing treatment, is greater in patients with higher baseline body weight. In the evaluation of the results of TS on weight loss, we should consider that both RCTs and observational studies enrolled a relevant proportion of lean and mildly overweight subjects, in whom a larger weight loss is unlikely. It should be noted that none of the studies analyzed here were designed as weight loss studies. The profound weight loss achieved in the long-term studies was rather an unexpected and accidental finding. In contrast, weight loss studies exclusively enroll obese subjects who have the intention to lose weight.

In addition, TS is known to increase muscle mass [27], possibly attenuating the reduction in total body weight. The present meta-analysis confirmed the increase in lean mass, which had already been reported in RCTs [27]. In fact, the reduction of fat mass, which had been observed also in RCTs, was confirmed here for observational studies. Unfortunately, this parameter was available only for a fraction of studies, simply because data obtained in “real-life” settings would not include measurement of body composition at regular intervals.

Considering the T anabolic properties, the role of TS in improvement frailty and in particular muscle strength and physical performance has attracted considerable attention [14, 68, 69]. However, the currently available evidence indicates that the T alone is relatively ineffective in improving the aging-associated impaired elderly physical performance [4]. Accordingly, a recent RCT evaluating the effect of T gel or placebo in 790 men aged 65 years or older on several outcome showed no benefit with respect to vitality or walking distance [70].

In line with its positive effects on body composition, we also report here a significant T-induced improvement in lipid profile, which was only marginally apparent in RCTs [27]. A significant, beneficial effect of TS was also noted for both systolic and diastolic blood pressure, most probably reflecting the improvement in body composition. Finally, we here confirm [27, 71, 72, 73] that TS improves glucose metabolism, by decreasing glycemia and increasing insulin sensitivity (HOMA index).

The improvement of several CV risk factors observed in this study might suggest a possible reduction in CV risk. However, CV safety remains one of the most controversial issues related to TS in aging men. So far, six systematic meta-analyses on the effect of TS as derived from placebo-controlled RCTs are currently available [74, 75, 76, 77, 78, 79]. The only meta-analysis reporting an increased CV risk associated with TS was published 3 years ago in BMC Medicine by Xu et al., [79]. The authors used as the primary outcome “composite CV events,” defined as anything reported as such by the study’s authors [79]. The construct “composite CV-related events” included all investigator-reported adverse events assigned to the CV system. The Xu et al. [79] meta-analysis was in contrast with all the previous ones [75, 76, 77], and in fact, all the other meta-analyses showed no significant difference between the T and placebo groups for all incident CV events, or for each type of event (CV death, fatal and non-fatal myocardial, revascularization procedures, arrhythmia, cerebrovascular events), except the increase in hematocrit over 50 %, which was significantly more prevalent in the T group [75]. In line with the latter evidence, we performed the last, updated, systematic review and meta-analysis of RCTs on TS, using the incidence of major adverse cardiac events (MACE) as a primary endpoint [74]. MACE are easier to detect and less controversial in diagnosis and are the events requested by regulatory authorities to verify the safety of newly registered drugs (http://www.fda.gov/Drugs/) [74]. By meta-analyzing the largest number of studies collected so far, any increase in CV risk associated with TRT was not observed when either composite or single CV endpoints were considered [74]. Hence, available data suggest that when hypogonadism is properly diagnosed and replacement therapy correctly performed CV risk is not an issue [14]. Accordingly, the Coordination Group for Mutual Recognition and Decentralised Procedures—Human (CMDh), a regulatory body representing European Member States—after a review by the European Medicines Association (EMA)’s Pharmacovigilance Risk Assessment Committee (PRAC)—has agreed by consensus that there is no consistent evidence of an increased risk of heart problems with T medicines in HG men (http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/referrals/Testosterone-containing_medicines/human_referral_prac_000037.jsp&mid=WC0b01ac05805c516f).

The interpretation of the results of the present meta-analysis should be cautious because of the relevant potential biases. Although specific analyses seem to exclude a publication bias in this case, significant results are more likely to be published, thus providing a potential distortion in favor of weight loss. In addition, in observational studies the completeness of follow-up and the management of missing data are crucial; it can be speculated that patients experiencing a relevant weight loss have a higher chance of reporting body weight to investigators, thus producing an overestimate of weight loss. Long-term RCTs involving young HG subjects with obesity having weight loss as a primary endpoint are needed to confirm the surprising results obtained in the observational studies scrutinized here.

It is also important to note, however, that diagnosis of HG allows only for a short period of placebo-controlled design, because the condition is associated with an increased risk of osteoporosis or other metabolic and sexual side effects. Hence, long-term, placebo-controlled studies are scarcely feasible. Although the evidence strength is usually lower than that of RCTs, observational studies have the advantage of reflecting the actual prescribing of T in health care and the advantages of including broader patient populations also from non-teaching hospitals, all resulting in higher generalizability.

In conclusion, the lesson derived from everyday clinical practice, summarized here, supports the view of a positive effect of TS on body composition and on glucose and lipid metabolism, as already suggested by the meta-analysis of RCTs [27]. Observational studies also suggest a significant effect on body weight loss, which should be confirmed by a specifically designed RCT.

Notes

Compliance with Ethical Standards

Conflict of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the review.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Informed consent

Since this is a meta-analysis informed consent is not applicable.

Supplementary material

40618_2016_480_MOESM1_ESM.docx (13 kb)
Supplementary material 1 (DOCX 12 kb)
40618_2016_480_MOESM2_ESM.pptx (274 kb)
Supplementary material 2 (PPTX 273 kb)

References

  1. 1.
    Dietz WH, Baur LA, Hall K, Puhl RM, Taveras EM, Uauy R et al (2015) Management of obesity: improvement of health-care training and systems for prevention and care. Lancet 385:2521–2533CrossRefPubMedGoogle Scholar
  2. 2.
    Corona G, Rastrelli G, Filippi S, Vignozzi L, Mannucci E, Maggi M (2014) Erectile dysfunction and central obesity: an Italian perspective. Asian J Androl 16:581–591CrossRefPubMedPubMedCentralGoogle Scholar
  3. 3.
    Yusuf S, Hawken S, Ounpuu S, Bautista L, Franzosi MG, Commerford P, INTERHEART Study Investigators et al (2005) Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case–control study. Lancet 366:1640–1649CrossRefPubMedGoogle Scholar
  4. 4.
    Corona G, Rastrelli G, Morelli A, Vignozzi L, Mannucci E, Maggi M (2011) Hypogonadism and metabolic syndrome. J Endocrinol Investig 34:557–567CrossRefGoogle Scholar
  5. 5.
    Corona G, Vignozzi L, Sforza A, Mannucci E, Maggi M (2015) Obesity and late-onset hypogonadism. Mol Cell Endocrinol 418:120–133CrossRefPubMedGoogle Scholar
  6. 6.
    Brand JS, Rovers MM, Yeap BB, Schneider HJ, Tuomainen TP, Haring R et al (2014) Testosterone, sex hormone-binding globulin and the metabolic syndrome in men: an individual participant data meta-analysis of observational studies. PLoS ONE 9:e100409CrossRefPubMedPubMedCentralGoogle Scholar
  7. 7.
    Rastrelli G, Corona G, Mannucci E, Maggi M (2015) Vascular and chronological age in subjects with erectile dysfunction: a cross-sectional study. J Sex Med 12:2303–2312CrossRefPubMedGoogle Scholar
  8. 8.
    Iglesias P, Prado F, Macías MC, Guerrero MT, Muñoz A, Ridruejo E et al (2014) Hypogonadism in aged hospitalized male patients: prevalence and clinical outcome. J Endocrinol Invest 37:135–141CrossRefPubMedGoogle Scholar
  9. 9.
    Calderón B, Gómez-Martín JM, Vega-Piñero B, Martín-Hidalgo A, Galindo J, Luque-Ramírez M et al (2016) Prevalence of male secondary hypogonadism in moderate to severe obesity and its relationship with insulin resistance and excess body weight. Andrology 4:62–67CrossRefPubMedGoogle Scholar
  10. 10.
    Rastrelli G, Corona G, Mannucci E, Maggi M (2016) Vascular and chronological age in men with erectile dysfunction: a longitudinal study. J Sex Med 13(2):200–208CrossRefPubMedGoogle Scholar
  11. 11.
    Corona G, Rastrelli G, Vignozzi L, Mannucci E, Maggi M (2012) How to recognize late-onset hypogonadism in men with sexual dysfunction. Asian J Androl 14:251–259CrossRefPubMedPubMedCentralGoogle Scholar
  12. 12.
    Saad F, Aversa A, Isidori AM, Gooren LJ (2012) Testosterone as potential effective therapy in treatment of obesity in men with testosterone deficiency: a review. Curr Diabetes Rev 8:131–143CrossRefPubMedPubMedCentralGoogle Scholar
  13. 13.
    Zarotsky V, Huang MY, Carman W, Morgentaler A, Singhal PK, Coffin D et al (2014) Systematic literature review of the risk factors, comorbidities, and consequences of hypogonadism in men. Andrology 2:819–834CrossRefPubMedGoogle Scholar
  14. 14.
    Isidori AM, Balercia G, Calogero AE, Corona G, Ferlin A, Francavilla S et al (2015) Outcomes of androgen replacement therapy in adult male hypogonadism: recommendations from the Italian society of endocrinology. J Endocrinol Investig 38:103–112CrossRefGoogle Scholar
  15. 15.
    Maneschi E, Morelli A, Filippi S, Cellai I, Comeglio P, Mazzanti B et al (2012) Testosterone treatment improves metabolic syndrome-induced adipose tissue derangements. J Endocrinol 215:347–362CrossRefPubMedGoogle Scholar
  16. 16.
    Saad F, Gooren L, Haider A, Yassin A (2007) An exploratory study of the effects of 12 month administration of the novel long-acting testosterone undecanoate on measures of sexual function and the metabolic syndrome. Arch Androl 53:353–357CrossRefPubMedGoogle Scholar
  17. 17.
    Saad F, Gooren LJ, Haider A, Yassin A (2008) A dose-response study of testosterone on sexual dysfunction and features of the metabolic syndrome using testosterone gel and parenteral testosterone undecanoate. J Androl 29:102–105CrossRefPubMedGoogle Scholar
  18. 18.
    Saad F, Haider A, Doros G, Traish A (2013) Long-term treatment of hypogonadal men with testosterone produces substantial and sustained weight loss. Obesity 21:1975–1981 (Silver Spring) CrossRefPubMedGoogle Scholar
  19. 19.
    Francomano D, Lenzi A, Aversa A (2014) Effects of five-year treatment with testosterone undecanoate on metabolic and hormonal parameters in ageing men with metabolic syndrome. Int J Endocrinol 2014:527470CrossRefPubMedPubMedCentralGoogle Scholar
  20. 20.
    Yassin DJ, Doros G, Hammerer PG, Yassin AA (2014) Long-term testosterone treatment in elderly men with hypogonadism and erectile dysfunction reduces obesity parameters and improves metabolic syndrome and health-related quality of life. J Sex Med 11:1567–1576CrossRefPubMedGoogle Scholar
  21. 21.
    Zitzmann M, Saad F, Kliesch S (2014) Long-term treatment with testosterone undecanoate injections leads to sustained weight loss and improvement of metabolic syndrome parameters in 381 hypogonadal men. J Sex Med 11(Suppl. 1):7Google Scholar
  22. 22.
    Haider A, Saad F, Doros G, Gooren L (2014) Hypogonadal obese men with and without diabetes mellitus type 2 lose weight and show improvement in cardiovascular risk factors when treated with testosterone: an observational study. Obes Res Clin Pract 8:e339–e349CrossRefPubMedGoogle Scholar
  23. 23.
    Traish AM, Haider A, Doros G, Saad F (2014) Long-term testosterone therapy in hypogonadal men ameliorates elements of the metabolic syndrome: an observational, long-term registry study. Int J Clin Pract 68:314–329CrossRefPubMedGoogle Scholar
  24. 24.
    Saad F, Yassin A, Doros G, Haider A (2016) 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. Int J Obes 40:162–170CrossRefGoogle Scholar
  25. 25.
    Corona G, Maseroli E, Maggi M (2014) Injectable testosterone undecanoate for the treatment of hypogonadism. Expert Opin Pharmacother 15:1903–1926CrossRefPubMedGoogle Scholar
  26. 26.
    Neff KJ, le Roux CW (2013) Bariatric surgery: a best practice article. J Clin P athol 66:90–98CrossRefGoogle Scholar
  27. 27.
    Corona G, Giagulli VA, Maseroli E, Vignozzi L, Aversa A, Zitzmann M, Saad F, Mannucci E, Maggi M (2016) THERAPY OF ENDOCRINE DISEASE: testosterone supplementation and body composition: results from a meta-analysis study. Eur J Endocrinol 174(3):R99–R116CrossRefPubMedGoogle Scholar
  28. 28.
    Cohen AT, Goto S, Schreiber K, Torp-Pedersen C (2015) Why do we need observational studies of everyday patients in the real-life setting? Eur Heart J Suppl 17(Supplement D):D2–D8CrossRefGoogle Scholar
  29. 29.
    Cox JL, Pieper K (2015) Harnessing the power of real-life data. Eur Heart J Supp. 17(Supplement D):D9–D14CrossRefGoogle Scholar
  30. 30.
    Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group et al (2000) Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 283:2008–2012CrossRefPubMedGoogle Scholar
  31. 31.
    Valdemarsson S, Hedner P, Nilsson-Ehle P (1987) Increase in hepatic lipase activity after testosterone substitution in men with hypogonadism of pituitary origin. Acta Med Scand. 221:363–366CrossRefPubMedGoogle Scholar
  32. 32.
    Rebuffé-Scrive M, Mårin P, Björntorp P (1991) Effect of testosterone on abdominal adipose tissue in men. Int J Obes 15:791–795Google Scholar
  33. 33.
    Forbes GB, Porta CR, Herr BE, Griggs RC (1992) Sequence of changes in body composition induced by testosterone and reversal of changes after drug is stopped. JAMA 267:397–399CrossRefPubMedGoogle Scholar
  34. 34.
    Mårin P, Krotkiewski M, Björntorp P (1992) Androgen treatment of middle-aged, obese men: effects on metabolism, muscle and adipose tissues. Eur J Med 1:329–336PubMedGoogle Scholar
  35. 35.
    Brodsky IG, Balagopal P, Nair KS (1996) Effects of testosterone replacement on muscle mass and muscle protein synthesis in hypogonadal men—a clinical research center study. J Clin Endocrinol Metab 81:3469–3475PubMedGoogle Scholar
  36. 36.
    Katznelson L, Finkelstein JS, Schoenfeld DA, Rosenthal DI, Anderson EJ, Klibanski A (1996) Increase in bone density and lean body mass during testosterone administration in men with acquired hypogonadism. J Clin Endocrinol Metab 81:4358–4365PubMedGoogle Scholar
  37. 37.
    Wang C, Eyre DR, Clark R, Kleinberg D, Newman C, Iranmanesh A et al (1996) Sublingual testosterone replacement improves muscle mass and strength, decreases bone resorption, and increases bone formation markers in hypogonadal men—a clinical research center study. J Clin Endocrinol Metab 81:3654–3662PubMedGoogle Scholar
  38. 38.
    Zgliczynski S, Ossowski M, Slowinska-Srzednicka J, Brzezinska A, Zgliczynski W, Soszynski P et al (1996) Effect of testosterone replacement therapy on lipids and lipoproteins in hypogonadal and elderly men. Atherosclerosis 121:35–43CrossRefPubMedGoogle Scholar
  39. 39.
    Bhasin S, Storer TW, Berman N, Yarasheski KE, Clevenger B, Phillips J et al (1997) Testosterone replacement increases fat-free mass and muscle size in hypogonadal men. J Clin Endocrinol Metab 82:407–413PubMedGoogle Scholar
  40. 40.
    Tan KC, Shiu SW, Pang RW, Kung AW (1998) Effects of testosterone replacement on HDL subfractions and apolipoprotein A-I containing lipoproteins. Clin Endocrinol (Oxf). 48:187–194CrossRefPubMedGoogle Scholar
  41. 41.
    Brill KT, Weltman AL, Gentili A, Patrie JT, Fryburg DA, Hanks JB et al (2002) Single and combined effects of growth hormone and testosterone administration on measures of body composition, physical performance, mood, sexual function, bone turnover, and muscle gene expression in healthy older men. J Clin Endocrinol Metab 87:5649–5657CrossRefPubMedGoogle Scholar
  42. 42.
    Minnemann T, Schubert M, Hübler D, Gouni-Berthold I, Freude S, Schumann C et al (2007) A four-year efficacy and safety study of the long-acting parenteral testosterone undecanoate. Aging Male 10:155–158CrossRefPubMedGoogle Scholar
  43. 43.
    Naharci MI, Pinar M, Bolu E, Olgun A (2007) Effect of testosterone on insulin sensitivity in men with idiopathic hypogonadotropic hypogonadism. Endocr Pract 13:629–635CrossRefPubMedGoogle Scholar
  44. 44.
    La Vignera S, Calogero AE, Condorelli R, Lanzafame F, Giammusso B, Vicari E (2009) Andrological characterization of the patient with diabetes mellitus. Minerva Endocrinol 34:1–9PubMedGoogle Scholar
  45. 45.
    Moon DG, Park MG, Lee SW, Park K, Park JK, Kim SW, Park NC et al (2010) The efficacy and safety of testosterone undecanoate (Nebido®) in testosterone deficiency syndrome in Korean: a multicenter prospective study. J Sex Med 7(6):2253–2260CrossRefGoogle Scholar
  46. 46.
    Permpongkosol S, Tantirangsee N, Ratana-olarn K (2010) Treatment of 161 men with symptomatic late onset hypogonadism with long-acting parenteral testosterone undecanoate: effects on body composition, lipids, and psychosexual complaints. J Sex Med 7:3765–3774CrossRefPubMedGoogle Scholar
  47. 47.
    Garcia JA, Sanchez PE, Fraile C, Escovar P (2011) Testosterone undecanoate improves erectile dysfunction in hypogonadal men with the metabolic syndrome refractory to treatment with phosphodiesterase type 5 inhibitors alone. Andrologia 43:293–296CrossRefPubMedGoogle Scholar
  48. 48.
    Schwarz ER, Willix RD Jr (2011) Impact of a physician-supervised exercise-nutrition program with testosterone substitution in partial androgen-deficient middle-aged obese men. J Geriatr Cardiol 8:201–206PubMedPubMedCentralGoogle Scholar
  49. 49.
    Arafa M, Zohdy W, Aboulsoud S, Shamloul R (2012) Prevalence of late-onset hypogonadism in men with type 2 diabetes mellitus. Andrologia. 44(Suppl 1):756–763CrossRefPubMedGoogle Scholar
  50. 50.
    Schroeder ET, He J, Yarasheski KE, Binder EF, Castaneda-Sceppa C, Bhasin S et al (2012) Value of measuring muscle performance to assess changes in lean mass with testosterone and growth hormone supplementation. Eur J Appl Physiol 112:1123–1131CrossRefGoogle Scholar
  51. 51.
    Jo DG, Lee HS, Joo YM, Seo JT (2013) Effect of testosterone replacement therapy on bone mineral density in patients with Klinefelter syndrome. Yonsei Med J 54:1331–1335CrossRefPubMedPubMedCentralGoogle Scholar
  52. 52.
    Ko YH, Moon DG, Moon KH (2013) Testosterone replacement alone for testosterone deficiency syndrome improves moderate lower urinary tract symptoms: one year follow-up. World J Mens Health 31:47–52CrossRefPubMedPubMedCentralGoogle Scholar
  53. 53.
    Rodriguez-Tolrà J, Torremadé Barreda J, del Rio L, di Gregorio S, Franco Miranda E (2013) Effects of testosterone treatment on body composition in males with testosterone deficiency syndrome. Aging Male. 16:184–190CrossRefPubMedGoogle Scholar
  54. 54.
    Tirabassi G, Delli Muti N, Corona G, Maggi M, Balercia G (2013) Androgen receptor gene CAG repeat polymorphism regulates the metabolic effects of testosterone replacement therapy in male postsurgical hypogonadotropic hypogonadism. Int J Endocrinol. 2013:816740CrossRefPubMedPubMedCentralGoogle Scholar
  55. 55.
    Zitzmann M, Mattern A, Hanisch J, Gooren L, Jones H, Maggi M (2013) IPASS: a study on the tolerability and effectiveness of injectable testosterone undecanoate for the treatment of male hypogonadism in a worldwide sample of 1,438 men. J Sex Med 10:579–588CrossRefPubMedGoogle Scholar
  56. 56.
    Pexman-Fieth C, Behre HM, Morales A, Kan-Dobrosky N, Miller MG (2014) A 6-month observational study of energy, sexual desire, and body proportions in hypogonadal men treated with a testosterone 1% gel. Aging Male 17:1–11CrossRefPubMedGoogle Scholar
  57. 57.
    Conway AJ, Boylan LM, Howe C, Ross G, Handelsman DJ (1988) Randomized clinical trial of testosterone replacement therapy in hypogonadal men. Int J Androl 11:247–264CrossRefPubMedGoogle Scholar
  58. 58.
    Sorva R, Kuusi T, Taskinen MR, Perheentupa J, Nikkilä EA (1988) Testosterone substitution increases the activity of lipoprotein lipase and hepatic lipase in hypogonadal males. Atherosclerosis 69:191–197CrossRefPubMedGoogle Scholar
  59. 59.
    Ozata M, Yildirimkaya M, Bulur M, Yilmaz K, Bolu E, Corakci A et al (1996) Effects of gonadotropin and testosterone treatments on Lipoprotein(a), high density lipoprotein particles, and other lipoprotein levels in male hypogonadism. J Clin Endocrinol Metab 81:3372–3378PubMedGoogle Scholar
  60. 60.
    Tripathy D, Shah P, Lakshmy R, Reddy KS (1998) Effect of testosterone replacement on whole body glucose utilisation and other cardiovascular risk factors in males with idiopathic hypogonadotrophic hypogonadism. Horm Metab Res 30:642–645CrossRefPubMedGoogle Scholar
  61. 61.
    Andrade ES Jr, Clapauch R, Buksman S (2009) Short term testosterone replacement therapy improves libido and body composition. Arq Bras Endocrinol Metabol 53:996–1004CrossRefPubMedGoogle Scholar
  62. 62.
    Wu XY, Mao JF, Lu SY, Zhang Q, Shi YF (2009) Testosterone replacement therapy improves insulin sensitivity and decreases high sensitivity C-reactive protein levels in hypogonadotropic hypogonadal young male patients. Chin Med J (Engl) 122:2846–2850Google Scholar
  63. 63.
    Cochrane Handbook for Systematic Reviews of Interventions. Available at http://www.cochrane-handbook.org. Last accessed 26 Dec 2015
  64. 64.
    Begg CB, Mazumdar M (1994) Operating characteristics of a rank correlation test for publication bias. Biometrics 50:1088–1101CrossRefPubMedGoogle Scholar
  65. 65.
    Egger M, Smith GD, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315:629–634CrossRefPubMedPubMedCentralGoogle Scholar
  66. 66.
    Cohen J (1977) Statistical power analysis for the behavioral sciences. Academic Press, New YorkGoogle Scholar
  67. 67.
    Yassin A, Doros G (2013) Testosterone therapy in hypogonadal men results in sustained and clinically meaningful weight loss. Clin Obes 3:73–83CrossRefPubMedPubMedCentralGoogle Scholar
  68. 68.
    Crewther BT, Kilduff LP, Cook CJ (2014) Trained and untrained males show reliable salivary testosterone responses to a physical stimulus, but not a psychological stimulus. J Endocrinol Investig 37:1065–1072CrossRefGoogle Scholar
  69. 69.
    Sgrò P, Romanelli F, Felici F, Sansone M, Bianchini S, Buzzachera CF, Baldari C, Guidetti L, Pigozzi F, Lenzi A, Di Luigi L (2014) Testosterone responses to standardized short-term sub-maximal and maximal endurance exercises: issues on the dynamic adaptive role of the hypothalamic-pituitary-testicular axis. J Endocrinol Investig 37:13–24CrossRefGoogle Scholar
  70. 70.
    Snyder PJ, Bhasin S, Cunningham GR, Matsumoto AM, Stephens-Shields AJ, Cauley JA et al (2016) Effects of testosterone treatment in older men. N Engl J Med 374:611–624CrossRefPubMedGoogle Scholar
  71. 71.
    Corona G, Rastrelli G, Maggi M (2015) The pharmacotherapy of male hypogonadism besides androgens. Expert Opin Pharmacother 16:369–487. Erratum in: Expert Opin Pharmacother. 2015;16:941. Ratrelli, Giulia [corrected to Rastrelli, Giulia]Google Scholar
  72. 72.
    Corona G, Rastrelli G, Vignozzi L, Maggi M (2012) Emerging medication for the treatment of male hypogonadism. Expert Opin Emerg Drugs 17:239–259CrossRefPubMedGoogle Scholar
  73. 73.
    Rastrelli G, Giovannini L, Calogero AE, Gianfrilli D, Serra E, Pizzocaro A, Giagulli VA, Motta G, Vancieri G, Sperandio A, Andò S, Selice R, Luca G, Cocchiara F, Canale D, Maggi M (2016) Predictors and clinical consequences of starting androgen therapy in men with low testosterone: results from the SIAMO-NOI registry. J Endocrinol Invest 39(6):695–708CrossRefPubMedGoogle Scholar
  74. 74.
    Corona G, Maseroli E, Rastrelli G, Isidori AM, Sforza A, Mannucci E et al (2014) Cardiovascular risk associated with testosterone-boosting medications: a systematic review and meta-analysis. Expert Opin Drug Saf 13:1327–1351CrossRefPubMedGoogle Scholar
  75. 75.
    Calof OM, Singh AB, Lee ML, Kenny AM, Urban RJ, Tenover JL et al (2005) Adverse events associated with testosterone replacement in middle-aged and older men: a meta-analysis of randomized, placebo-controlled trials. J Gerontol A Biol Sci Med Sci 60:1451–1457CrossRefPubMedGoogle Scholar
  76. 76.
    Haddad RM, Kennedy CC, Caples SM, Tracz MJ, Boloña ER, Sideras K et al (2007) Testosterone and cardiovascular risk in men: a systematic review and meta-analysis of randomized placebo-controlled trials. Mayo Clin Proc 82:29–39CrossRefPubMedGoogle Scholar
  77. 77.
    Fernández-Balsells MM, Murad MH, Lane M, Lampropulos JF, Albuquerque F, Mullan RJ et al (2010) Clinical review 1: adverse effects of testosterone therapy in adult men: a systematic review and meta-analysis. J Clin Endocrinol Metab 95:2560–2575CrossRefPubMedGoogle Scholar
  78. 78.
    Borst SE, Shuster JJ, Zou B, Ye F, Jia H, Wokhlu A, Yarrow JF (2014) Cardiovascular risks and elevation of serum DHT vary by route of testosterone administration: a systematic review and meta-analysis. BMC Med 12:211CrossRefPubMedPubMedCentralGoogle Scholar
  79. 79.
    Xu L, Freeman G, Cowling BJ, Schooling CM (2013) Testosterone therapy and cardiovascular events among men: a systematic review and meta-analysis of placebo-controlled randomized trials. BMC Med 11:108CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© Italian Society of Endocrinology (SIE) 2016

Authors and Affiliations

  • G. Corona
    • 1
  • V. A. Giagulli
    • 2
  • E. Maseroli
    • 3
  • L. Vignozzi
    • 3
  • A. Aversa
    • 4
    • 5
  • M. Zitzmann
    • 6
  • F. Saad
    • 7
    • 8
  • E. Mannucci
    • 9
  • M. Maggi
    • 3
  1. 1.Endocrinology Unit, Medical DepartmentAziendaUsl Bologna Maggiore-Bellaria HospitalBolognaItaly
  2. 2.Unit of Metabolic Diseases and EndocrinologyConversanoItaly
  3. 3.Andrology and Sexual Medicine Unit, Department of Experimental and Clinical Biomedical SciencesUniversity of FlorenceFlorenceItaly
  4. 4.Department of Experimental MedicineSapienza University of RomeRomeItaly
  5. 5.Department of Experimental and Clinical MedicineUniversity Magna GraeciaCatanzaroItaly
  6. 6.Centre for Reproductive Medicine and AndrologyMuensterGermany
  7. 7.Bayer Pharma, Global Medical Affairs AndrologyBerlinGermany
  8. 8.Gulf Medical University School of MedicineAjmanUnited Arab Emirates
  9. 9.Diabetes AgencyCareggi HospitalFlorenceItaly

Personalised recommendations