Men with obesity, the metabolic syndrome, and type 2 diabetes have low total and free
testosterone and low sex hormone–binding globulin (SHBG). Conversely, the presence
of low testosterone and/or SHBG predicts the development of metabolic syndrome and
type 2 diabetes. Visceral adiposity present in men with low testosterone, the metabolic
syndrome, and/or type 2 diabetes acts through proinflammatory factors. These inflammatory
markers contribute to vascular endothelial dysfunction with adverse sequelae such
as increased cardiovascular disease (CVD) risk and erectile dysfunction. This review
focuses on the multidirectional impact of low testosterone associated with obesity
and the metabolic syndrome and its effects on erectile dysfunction and CVD risk in
men with type 2 diabetes. Whenever possible in this review, we will cite recent reports
(after 2005) and meta-analyses.
Epidemiological studies of low testosterone, obesity, metabolic status, and erectile
dysfunction
Epidemiological studies support a bidirectional relationship between serum testosterone
and obesity as well as between testosterone and the metabolic syndrome. Low serum
total testosterone predicts the development of central obesity and accumulation of
intra-abdominal fat (1–3). Also, low total and free testosterone and SHBG levels are
associated with an increased risk of developing the metabolic syndrome, independent
of age and obesity (1–3). Lowering serum T levels in older men with prostate cancer
treated with androgen deprivation therapy increases body fat mass (4). Conversely,
high BMI, central adiposity, and the metabolic syndrome are associated with and predict
low serum total and to a lesser extent free testosterone and SHBG levels (1–3,5).
Because obesity suppresses SHBG and as a result total testosterone concentrations,
alterations in SHBG confound the relationship between testosterone and obesity.
Low total testosterone or SHBG levels are associated with type 2 diabetes, independent
of age, race, obesity, and criteria for diagnosis of diabetes (6,7). In longitudinal
studies, low serum total and free testosterone and SHBG levels were independent predictors
of type 2 diabetes (6,8). In these studies, SHBG levels were stronger predictors of
diabetes than total or free testosterone. Because type 2 diabetes is often associated
with obesity, which suppresses SHBG and in turn total testosterone levels, both obesity
and SHBG levels represent important confounding factors in the relationship between
testosterone and type 2 diabetes. The prevalence of low free testosterone levels is
higher in diabetic men compared with nondiabetic men (6). However, a recent longitudinal
study found that free testosterone did not predict the development of type 2 diabetes.
In this study, the association of total testosterone and of SHBG with diabetes was
not significant after adjusting for waist circumference or central obesity (9). Also,
low SHBG was found to be a strong independent predictor of type 2 diabetes (10,11).
Finally, in prospective studies, androgen deprivation therapy either using bilateral
orchidectomy or gonadotropin-releasing hormone agonist in older men with prostate
cancer is associated with an increased risk of diabetes and CVD (12).
A number of epidemiological studies support associations of obesity (13,14), the metabolic
syndrome (15,16), type 2 diabetes (17), and low serum testosterone (18) with sexual
dysfunction including erectile dysfunction (ED) (19). These studies highlight the
complex often multidirectional relationships among obesity, metabolic status, low
testosterone, and ED in men.
Pathobiology of low testosterone in type 2 diabetes and its impact on sexual dysfunction
and CVD risk
Obesity is a proinflammatory state resulting in increased release and secretion of
proinflammatory cytokines and adipokines, free fatty acids, and estrogens from adipose
tissue. These increases are important risk factors that may contribute to the development
of metabolic syndrome and type 2 diabetes as well as androgen deficiency (20). Visceral
fat is an active secretory tissue producing inflammatory cytokines, adipokines, biochemical
modulators, and other proinflammatory factors including interleukin (IL)-6, IL-1β,
plasminogen activator inhibitor-1, tumor necrosis factor (TNF)-α, angiotensinogen,
vascular endothelial growth factor, and serum amyloid A (Fig. 1). These factors contribute
to systemic and peripheral vascular inflammation and dysfunction (21). As shown in
Fig. 1, one potential mechanism of how visceral adiposity and inflammatory response
modulate insulin sensitivity involves the release of free fatty acids. Free fatty
acids activate nuclear factor-κB pathways resulting in increased synthesis of TNF-α.
TNF-α further activates lipolysis as well as increased synthesis of IL-6 and macrophage
chemoattractant protein-1, which increases recruitment of more macrophages and modulates
insulin sensitivity. Increased production of TNF-α also enhances expression of adhesion
molecules in both endothelium and vascular smooth muscle cells. IL-6 stimulates hepatic
synthesis of C-reactive protein, a nonspecific marker of vascular inflammation. In
addition, TNF-α contributes to the dysregulation of insulin modulation of endothelin-1–mediated
vasoconstriction and nitric oxide–mediated vasodilation, hence promoting vasoconstriction.
Release of adipokines facilitates monocyte adhesion and migration into the vessel
wall as well as the conversion of monoctyes to macrophages.
Figure 1
Complex multidirectional interactions between testosterone and obesity, metabolic
syndrome, and type 2 diabetes mediated by cytokines and adipokines leading to comorbidities
such as ED and increased CVD risk. FFA, free fatty acids; GnRH, gonadotropin-releasing
hormone; LH, luteinizing hormone; PAI-1, plasminogen activator inhibitor-1.
Aromatase, the enzyme that converts testosterone to estradiol, is mainly located in
adipose tissue. Obesity is associated with elevated estrogen in men activating hypothalamic
estrogen receptors triggering inhibition of the hypothalamic-pituitary-gonadal axis.
Treatment with aromatase inhibitors reverses the hypogonadotropic hypogonadism associated
with obesity (22). Men with obesity and insulin resistance showed attenuated Leydig
cell responsiveness to exogenous gonadotropin stimulation (23).
There are data supportive of a direct role of testosterone in insulin sensitivity.
Acute withdrawal of testosterone in hypogonadal men for 2 weeks reduced insulin sensitivity
without apparent changes in body composition, suggesting that sex steroids directly
modulate insulin sensitivity (24). Others have reported that normalizing testosterone
levels in older men had favorable effects on body composition, which could improve
insulin sensitivity but not effects on postprandial triglyceride metabolism (25).
Recently, it was demonstrated, using glucose clamp studies, that the interplay between
insulin sensitivity, triglycerides, and sex steroids are almost immediate and not
facilitated by changes in body composition. Concomitantly, increasing testosterone
and decreasing estradiol levels for 1 week in young men improved postprandial triglyceride
handling, postprandial glucose-dependent insulinotropic polypeptide (GIP) release,
and insulin sensitivity (26). These studies indicate that interactions between low
testosterone and visceral adiposity acting through proinflammatory agents (Fig. 1)
result in insulin resistance and vascular endothelial dysfunction, which are potential
causal factors for increased CVD and ED (20).
Sexual dysfunction and low testosterone in type 2 diabetes
A national survey of sexual activity in the U.S. found that over 60% of people with
partners were sexually active, including individuals with diabetes (27). Similarly,
68.7% of 383 men with diabetes in the Look Ahead Study were sexually active (28).
The clinical observation that ED occurs at an earlier age and with greater frequency
in men with diabetes compared with nondiabetic men is supported by multiple population-based
epidemiological studies (27) and by surveys of clinical practices (29). In the Look
Ahead Study (28), 49.8% of men with diabetes reported mild or moderate ED. ED was
associated with age (odds ratio 1.05, 95% CI 1.01–1.10), baseline hemoglobin A1c (1.31,
1.05–1.63), hypertension (2.41, 1.34–4.36), and the metabolic syndrome (3.05, 1.31–7.11).
There are few studies evaluating the prevalence of reduced libido in men with diabetes.
Decreased sexual desire is primarily affected by the presence of ED and by depression.
An observational study of 253 men with type 2 diabetes in Sri Lanka found that after
excluding men with ED (33%), the prevalence of reduced libido was 25% (30). In a population-based
survey, premature ejaculation occurred in 36.3% (95% CI 26–48) of diabetic men and
22.9% (18–28.6) of nondiabetic men (27). Inability to climax was reported in 26% of
diabetic men versus 15.9% of nondiabetic men. Premature ejaculation was reported in
40% of the patients from Sri Lanka who did not have severe or complete ED (30).
In the European Male Aging Study database of 3,369 men between the ages of 40 and
79 years, three sexual symptoms (poor morning erections, low sexual desire, and ED)
had a syndromic relationship with decreased testosterone levels (18). Moreover, in
the European Male Aging Study, low serum testosterone was more frequent in men with
comorbidities such as obesity, metabolic syndrome, and type 2 diabetes. In studies
from diabetes clinics, total, bioavailable, and free testosterone levels were low
in men with type 2 diabetes (31). When comparing testosterone levels in men with and
without ED and type 2 diabetes, these investigators found significantly lower serum
bioavailable testosterone (P < 0.006) and free testosterone (P < 0.027) in men with
ED, but there was no significant difference in total testosterone levels. The lower
the serum testosterone, the greater the severity of ED (32). Corona et al. (33) evaluated
1,200 men with ED and reported that 16% had type 2 diabetes. Serum total testosterone
levels were below the reference range (<300 ng/dL or <10.4 nmol/L) in 24.5% of men
with diabetes versus 12.6% of nondiabetic subjects (P < 0.0001) after adjustment for
age and BMI. In addition, hypogonadism in men with type 2 diabetes was associated
with decreased sexual desire, more symptoms of depression, and lower luteinizing hormone
levels.
ED in the past was ascribed to autonomic neuropathy or obliterative vascular disease;
more recent studies identify endothelial dysfunction as an early abnormality that
is potentially more amenable to therapy (20). Animal studies have demonstrated testosterone
effects on nerve structure and function, nitric oxide synthase activity, and smooth
muscle growth and differentiation, which mediate penile erections (34). Obesity and
androgen deficiency are associated with increased proinflammatory cytokines, which
also results in vascular endothelial dysfunction (20).
Men with type 2 diabetes can have other causes of ED. In a study of 8,373 men with
type 2 diabetes (35), ED was associated with poor metabolic control, smoking, alcohol,
antidepressants, antihypertensives, CVD medications, and histamine 2 receptor antagonists.
There are multiple causes for low libido in the general population and in men with
type 2 diabetes in addition to testosterone deficiency, including medications (e.g.,
serotonin reuptake inhibitors, antiandrogens), alcoholism, recreational drugs, fatigue,
systemic illness, depression, relationship problems, other sexual dysfunction (fear
of humiliation), hypoactive sexual disorder, and sexual aversion disorder.
The Look Ahead study reported that weight loss and increased physical activity were
mildly beneficial in maintaining erections or improving ED in men with type 2 diabetes
(36). Although improvement in glucose control is associated with some improvement
in erectile function in some studies, most clinicians have not found this to be a
reliable and effective treatment for ED. The Testosterone Replacement in Older Men
with either Metabolic Syndrome or Type 2 Diabetes (TIMES 2) trial recruited hypogonadal
men with total testosterone <318 ng/dL (11 nmol/L) or free testosterone <6.5 ng/dL
(225 pmol/L) and either metabolic syndrome or type 2 diabetes. Testosterone treatment
improved libido (37). Two meta-analyses of many clinical trials analyzed the effects
of testosterone on different domains of sexual function (38,39). Testosterone treatment
moderately improved the number of nocturnal erections, sexual thoughts and motivation,
number of successful intercourse sessions, scores of erectile function, and overall
sexual satisfaction in men with baseline serum testosterone <346 ng/dL (<12 nmol/L).
The effects of testosterone on libido were more consistent than on erectile function.
Testosterone replacement can restore nocturnal erections in hypogonadal men, but the
effects are greater when testosterone and a phosphodiesterase (PDE)-5 inhibitor are
administered together.
ED in many men with diabetes is improved by one of the PDE-5 inhibitors when used
on demand. A recently published randomized double-blind placebo-controlled multicenter
study evaluated the effectiveness of daily oral dosing of tadalafil in 298 men with
diabetes and ED. Daily dosing of tadalafil showed significant improvement in vaginal
penetration, completion of intercourse, and overall treatment satisfaction (40). Testosterone
replacement therapy has been reported to improve erections in men who did not respond
satisfactorily to a PDE-5 inhibitor alone (41). Larger trials using testosterone in
addition to a PDE-5 inhibitor in hypogonadal men with ED who have testosterone levels
<300 ng/dL (10.4 nmol/L) are needed.
Low testosterone, CVD risks, and type 2 diabetes
There is increasing evidence from multiple studies after adjustment of confounding
variables that low serum testosterone is associated with an increase in all-cause
mortality that is independent of the metabolic syndrome and diabetes (42–45) (Table
1). Low testosterone predicted the increased risk of CVD independent of age, obesity,
hyperlipidemia, and lifestyle in men with or without type 2 diabetes (43–45). In patients
with CVD, excess mortality was noted in the testosterone-deficient men compared with
men with normal serum testosterone concentrations (46) (Table 1). Testosterone deficiency
and CVD are both associated with visceral fat accumulation, metabolic syndrome, type
2 diabetes, increased inflammatory cytokines, hyperlipidemia, and abnormalities of
coagulation (47).
Table 1
Low testosterone is associated with increased mortality in older men
Study design
n
Follow-up (years)
Mortality
Hazard ratio (95% CI)
Recent studies
Retrospective
858
8
All-cause
1.88 (1.34–2.63)*
Shores et al. (42)
Prospective
794
20
All-cause and CVD
1.40 (1.14–1.71)*
1.38 (1.02–1.85)*
Laughlin et al. (45)
Prospective
2,314
10
All-cause and CVD
2.29 (1.60–3.26)*
Khaw et al. (43)
Prospective
1,954
7.2
All-cause and CVD
2.32 (1.38–3.89)*
Haring et al. (44)
Prospective
930
6.9
All-cause and CVD in men with CVD
2.27 (1.45–3.60)*
Malkin et al. (46)
*On the basis of recent publications in which the number of subject is >500 and age
of the subjects is >60 years.
In intervention studies on a small number of subjects, administration of testosterone
caused coronary artery dilation, decreased myocardial ischemia, and improved angina
during stress tests (48,49). Others suggest that testosterone may improve chronic
heart failure (50,51). In a recent study, Testosterone in Older Men with Mobility
Limitations (TOM), the older men had multiple comorbidities including obesity, hypertension,
diabetes, and hyperlipidemia, the application of relatively high doses of transdermal
testosterone gel was associated with significantly higher CVD event rates than in
patients treated with placebo gel (52). An increase in CVD event rate was not observed
in another study of frail elderly men treated with lower doses of transdermal testosterone
gel compared with the placebo-treated men (53). The adverse CVD events in the TOM
trial suggest that monitoring for cardiovascular adverse events is essential in a
testosterone intervention study of older men with or without type 2 diabetes and we
need a better understanding of testosterone effects on coagulation.
Clinical studies of testosterone replacement in men with obesity, metabolic syndrome,
type 2 diabetes, and low testosterone concentrations
The major goal of testosterone replacement therapy is to increase serum testosterone
concentrations to physiological concentrations with the purpose of resolving symptoms
and biological sequelae of hypogonadism. The advent of new modes of testosterone delivery
such as transdermal testosterone gels and depot intramuscular testosterone undecanoate
injections have made physiological replacement possible. Testosterone replacement
in the hypogonadal man with type 2 diabetes and/or metabolic syndrome should aim to
have beneficial effects on multiple outcomes including sexual health; general well-being;
body composition; and reducing CVD risk factors, including central adiposity, glycemic
control, and atherogenic lipid profile.
Evidence from several studies (Table 2) demonstrated that testosterone promotes insulin
sensitivity in hypogonadal men with and without type 2 diabetes. Mårin et al. (54)
were the first to report that testosterone improved insulin sensitivity assessed by
euglycemic clamp studies in obese men while reducing central adiposity. More recently,
a randomized double-blind crossover trial demonstrated a significant reduction in
insulin resistance in hypogonadal men with type 2 diabetes (55). This finding was
confirmed in three further studies in men with metabolic syndrome and/or type 2 diabetes
(37,56,57). Hypogonadism with either of these conditions was the major inclusion criterion.
Subjects were not selected for poor diabetic control (Table 2). Even so, three studies
reported decreases in HbA1c levels in the men with diabetes (37,55,56).
Table 2
Randomized trials of testosterone replacement in hypogonadal men with metabolic syndrome
or type 2 diabetes
Study
Kapoor et al. (55)
Heufelder et al. (56)
Kalinchenko et al. (57)
Jones et al.* (37)
Subjects
Type 2 diabetes
New type 2 diabetes/metabolic syndrome
Type 2 diabetes/metabolic syndrome
Type 2 diabetes/metabolic syndrome
Study design
RCT-c
NRCT
RCT-p
RCT-p
n
24
32
184
220
Duration (months)
3
12
6
6/12*
Medications for diabetes
Diet, oral, insulin
Naive
Diet, oral
Diet, oral
Baseline serum testosterone (nmol/L)
≤8.6
≤10.5
≤6.7
≤10.2
Testosterone formulation
TES injections (200 mg/2 weeks)
Testosterone gel (50 mg/day)
TU depot injections
Testosterone gel (40–80 mg/day)
Treatment effects (changes)
HOMA-IR
−1.7
−0.9
−1.49
−0.54
Fasting glucose (nmol/L)
−1.6
−0.3 (AS)
↔
−0.42 (AS)
Fasting insulin (mIU/mL)
↔
↓
↔
↓(AS)
HbA1c
−0.37
−0.80
ND
↔ [−0.45]†
Total cholesterol (nmol/L)
−0.4
ND
↔
↔ [−0.13]
LDL cholesterol (nmol/L)
↔
ND
↔
↔
‡
HDL cholesterol (nmol/L)
↔
↑§
↔
−0.049‡
Triglycerides
↔
↓
↔
↔
Lipoprotein a
ND
ND
ND
↓
BMI
↔
↔
↔
↔
Waist circumference
↓
↓
↓
↔
% Body fat
ND
ND
ND
↔
Blood pressure
↔
↓‖
ND
↔
↔, No significant change; ↑, significant increase; ↓, significant decrease; AS, approaching
significance (P = 0.05–0.07); ND, not done; NRCT, randomized open label, not placebo-controlled
parallel trial; RCT-c, randomized placebo-controlled crossover; RCT-p, randomized
placebo-controlled parallel; TES, mixed testosterone esters; TU, testosterone undecanoate
depot injections after the first injection followed by another injection at 6 weeks
and then injections every 12 weeks. Testosterone gel was dose-adjusted to give total
testosterone level >17 nmol/L.
*The study by Jones et al. (TIMES2) had no medication changes in the first 6 months
unless overriding clinical needs, but medication changes were allowed in the second
6 months for ethical reasons (intention-to-treat group, modified per protocol group
where no changes in medications occurred; data not shown).
†Significant difference compared with placebo observed after 9 months, but result
may be confounded by allowed medication changes.
‡Metabolic syndrome subgroup showed significant changes in total cholesterol (−0.34
mmol/L), LDL cholesterol (−0.21 mmol/L), and HDL cholesterol (−0.058 mmol/L).
§No figure quoted.
‖Diastolic blood pressure only.
Insulin resistance commonly occurs in chronic heart failure, and it has been shown
to improve with testosterone replacement therapy (58). As discussed above, the mechanisms
by which testosterone improves insulin sensitivity is multifactorial and likely to
be due to a combination of testosterone effects on liver, muscle, and adipose tissues
and by reducing the production of inflammatory cytokines (e.g., TNF-α, IL-1β, and
IL-6), which cause insulin resistance (Fig. 1) (59).
It is well known that testosterone replacement reduces body fat mass and waist circumference
in hypogonadal men with and without obesity (54,59). In men with the metabolic syndrome
and/or type 2 diabetes, a decrease in central adiposity was observed in all but one
study with testosterone treatment (55–57). BMI improved in only one trial (56) and
body fat decreased in another in those men who did not have changes in medications
that affect body weight (37). Leptin levels correlate with body fat content and have
been shown to decrease with testosterone replacement in type 2 diabetes and the metabolic
syndrome (57,60). The effect of testosterone on lipid profile was investigated in
several studies including those on coronary heart disease, metabolic syndrome, and
diabetes (59). The majority of studies have found that testosterone therapy results
in a small but significant fall in total cholesterol and in some LDL cholesterol (37,55,59)
(Table 2). HDL cholesterol may fall, rise, or remain unchanged (59). There is some
evidence that after an initial decrease, HDL cholesterol levels then return to baseline
(37). Most reports found no change in triglycerides. Lipoprotein a, which has the
strongest positive correlation with premature coronary heart disease than any other
component of the lipid profile, was found to fall significantly after testosterone
treatment of men with the metabolic syndrome and/or type 2 diabetes (37).
Current evidence, albeit from mainly small-scale studies, does demonstrate some beneficial
effects of testosterone on important CVD risk factors, which include insulin resistance,
glycemic control, lipid profile, central adiposity, body composition, and inflammatory
state in hypogonadal men with type 2 diabetes, as well as sexual health. None of these
clinical trials reported any adverse effects on blood pressure, cardiovascular events,
or mortality.
Conclusions
The multidirectional interrelationships between serum testosterone and SHBG with obesity,
metabolic syndrome, and type 2 diabetes are complex. Obesity is accompanied by increased
adipokines, cytokines, and other proinflammatory factor production from adipocytes
and macrophages mainly in visceral fat. These factors may alter insulin responsiveness
in fat, liver, muscle, and endothelial function resulting in metabolic syndrome, type
2 diabetes, ED, and CVD. Many men with type 2 diabetes, especially those who are obese,
have low serum total testosterone and SHBG levels. Small-scale studies of testosterone
treatment in men with metabolic syndrome or type 2 diabetes and borderline low or
normal testosterone levels showed small improvement in glycemic control. Many of these
studies in men with type 2 diabetes are associated with confounders, including changes
in medications for diabetes. More randomized placebo-controlled interventional trials
of testosterone treatment in testosterone-deficient men with the metabolic syndrome
and poorly controlled type 2 diabetes are needed to evaluate the putative role of
testosterone in the interruption of the vicious cycle contributed by metabolic imbalances.
At present, it is important for the clinician to recognize that low testosterone and
sexual dysfunction are commonly found in patients with obesity, metabolic syndrome,
and type 2 diabetes. Testosterone replacement, in addition to diet, exercise, glycemic
control, and PDE-5 inhibitors, should be considered in symptomatic hypogonadal men
with type 2 diabetes and serum testosterone below the reference range. During testosterone
treatment, monitoring should include assessment of improvement of symptoms, glycemic
control, lipid levels, hematocrit, and potential adverse effects including CVD and
prostate diseases in older men.