MINIREVIEW
Mechanisms of obesity and related pathologies: Androgen
deficiency and endothelial dysfunction may be the link
between obesity and erectile dysfunction
Abdulmaged M. Traish
1
, Robert J. Feeley
1
and Andre Guay
2
1 Department of Biochemistry and Urology, Boston University School of Medicine, MA, USA
2 Department of Endocrinology, Center for Sexual Function, Lahey Clinic, Peabody, MA, USA
Introduction
Erectile function is a neurovascular process that
depends on the health of the central and peripheral
nervous systems, the vascular health of the erectile
tissue, and on the endocrine milieu [1]. Sexual stimula-
tion activates the non-adrenergic, noncholinergic nerve
and activates the neural nitric oxide (NO) syn-
thase ⁄cGMP pathway. The release of NO facilitates
the relaxation of penile cavernosal arteries and
Keywords
diabetes; dyslipidemia; endothelial
dysfunction; erectile dysfunction;
hypogonadism; inflammatory responses;
metabolic syndrome; testosterone
deficiency; vascular disease; visceral obesity
Correspondence
A. M. Traish, Laboratories for Sexual
Medicine, Institute for Sexual Medicine,
Boston University School of Medicine,
Center for Advanced Biomedical Research,
700 Albany Street, W607, Boston, MA
02118, USA
Fax: +1 617 638 5412
Tel: +1 617 638 4578
E-mail: atraish@bu.edu
(Received 27 February 2009, revised 31 July
2009, accepted 3 August 2009)
doi:10.1111/j.1742-4658.2009.07305.x
Obesity is associated with a high prevalence of erectile dysfunction; how-
ever, the pathophysiological link between obesity and erectile dysfunction
remains poorly understood. In this minireview, we have attempted to eval-
uate the existing literature pertaining to obesity and erectile dysfunction to
determine whether a common pathophysiological link exists. Visceral obes-
ity is associated with increased inflammatory responses, which contribute
to endothelial dysfunction. Furthermore, obesity is also associated with
reduced plasma testosterone levels, which contributes to hypogonadism
and increases the risk of vascular pathology. Endothelial dysfunction and
androgen deficiency have previously been linked to the pathophysiological
mechanisms of erectile dysfunction. The underlying pathophysiological
mechanisms of endothelial dysfunction and testosterone deficiency include
penile vascular insufficiency as a result of the loss of nitric oxide synthase
expression and activity and the loss of tissue compliance, resulting in
reduced hemodynamic properties. Recent progress in the field of sexual
medicine has recognized the impact of vascular disease and hypogonadism
on the management of patients with erectile dysfunction. We suggest that
visceral obesity, a component of the metabolic syndrome, adversely affects
endothelial function and testosterone levels, contributing to hypogandism
and erectile dysfunction. Thus, clinical screening for the risk of erectile
dysfunction in obese patients should include the assessment of waist
circumference, testosterone levels, body mass index and physical inactivity.
Abbreviations
BAT, bioavailable testosterone; BMI, body mass index; CVD, cardiovascular disease; E
2
, estradiol; ED, erectile dysfunction; FT, free
testosterone; IIEF, International Index of Erectile Function; IL, interleukin; IR, insulin resistance; LH, leutinizing hormone; MCP-1, monocyte
chemoattractant protein-1; M-CSF, macrophage-colony stimulating factor; MetS, metabolic syndrome; NO, nitric oxide; PAI-1, plasminogen
activator inhibitor-1; PVD, peripheral vascular disease; SHBG, sex hormone binding globulin; T2DM, type 2 diabetes mellitus; TNFa, tumor
necrosis factor-a; TT, total testosterone; VD, vascular disease; WC, waist circumference; WHR, waist-to-hip ratio.
FEBS Journal 276 (2009) 5755–5767 ª2009 The Authors Journal compilation ª2009 FEBS 5755
resistance arterioles, which causes vasodilation, and
increases blood flow to the corpus cavernosum. The
increased blood flow stimulates the endothelium lining
the lacunar spaces of the corpus cavernosum to release
endothelial NO from the endothelium NO synthase.
These biochemical and physiological processes result in
trabecular smooth muscle relaxation and expansion of
the sinusoids within the corpora cavernosa, leading to
penile engorgement. This expansion of the corpora
cavernosa against the tunica albuginea results in
veno-occlusion and trapping of blood under pressure.
This process is referred to as the ‘veno-occlusive’
mechanism. Neural and endothelial NO synthases are
regulated by androgens. In addition, the tissue
histo-architecture is dependent on androgens. Thus,
any perturbations or alterations in the neural, vascular
or erectile tissue fibroelastic properties will contribute
to erectile dysfunction (ED), by altering the veno-
occlusion mechanism. In this minireview, we discuss
the relationships between obesity, endothelial dysfunc-
tion and reduced plasma androgen levels (hypogona-
dism), and link these pathological states to ED.
Obesity is a major risk factor for the metabolic
syndrome (MetS), vascular disease (VD), diabetes,
hypertension, endothelial dysfunction and androgen
deficiency, all of which contribute to the pathophysiol-
ogy of ED. Approximately 25% of obese individuals
exhibit MetS [2] and obesity, as manifested by an
increased body mass index (BMI), waist circumference
(WC) and waist-to-hip ratio (WHR), which contributes
to the increased prevalence of ED [3–6].
It is universally accepted that fat is stored when the
amount of calories consumed exceeds the amount of
calories expended. However, whether the increased
obesity is attributed solely to increased food intake,
reduced energy expenditure, or both, remains contro-
versial [7]. Recently, Swinburn et al. [8] postulated that
increased obesity in the USA is attributed directly to
increased food intake (i.e. increased calorie consump-
tion). In addition to increased food intake, defective
biochemical pathways involved in energy expenditure
may contribute significantly to obesity [9].
Obesity is a heterogeneous condition and not all
obese patients are characterized by co-morbidities;
however, the accumulation of visceral fat correlates
with a cluster of diabetogenic, atherogenic, prothrom-
botic and proinflammatory metabolic abnormalities
known as the MetS [10,11]. Despres et al. [10,11]
suggested that the most prevalent form of MetS is
associated with abdominal (visceral) obesity and that
this type of obesity contributes to insulin resistance
(IR) and the increased release of cytokines, as well as
the impairment, clearance and storage of triglycerides
in subcutaneous fat. The World Health Organization
defines obesity based on the BMI; however, BMI val-
ues do not always permit stratification of patients
based on their risk for developing subsequent cardio-
vascular disease (CVD) or type 2 diabetes mellitus
(T2DM). When identifying patients who are at high
risk for developing VD, it is important to consider that
there may be subsets comprising overweight patients,
with virtually identical BMI values, who may or may
not suffer from dyslipidemia and IR [12]. In a recent
study, Barter et al. [12] suggested that dyslipidemic
subjects (n= 715, mean BMI = 28.7) are at a higher
risk for developing subsequent CVD and diabetes com-
pared to normolipidemic controls (n= 1073, mean
BMI = 28.2) with similar BMI values. One proposed
hypothesis is that the dyslipidemic patients displayed a
marked reduction in circulating androgen levels, and
such a decrease in androgens, rather than BMI alone,
is a stronger predictor for the onset of CVD and dia-
betes in obese patients. Interestingly, Stefan et al. [13]
suggested that ‘metabolically benign obesity; may exist,
in which obese subjects do not exhibit IR or early ath-
erosclerosis and that ectopic fat in the liver, and not
visceral fat, may be responsible for this phenotype.
Androgen measurements in these metabolically benign
obese patients may provide further clues regarding this
pathophysiological state.
Central (abdominal) obesity, as assessed by several
characteristics including BMI, WC and WHR, is con-
sidered as a hallmark of the MetS and is associated
with reduced plasma androgen levels [14–19]. In this
minireview, we discuss the link between obesity and
ED by focusing on the potential pathophysiological
links including obesity-related androgen deficiency and
obesity-induced endothelial dysfunction.
Obesity promotes inflammatory
responses and contributes to
endothelial dysfunction
The role of endothelial function in erectile physiology
is well established. Any factor that contributes to
endothelial dysfunction will certainly contribute signifi-
cantly to ED. This is supported by the fact that penile
vascular hemodynamics depends on the integrity of
the vascular bed. Furthermore, obesity results in a
hypogonadal state that disrupts the endocrine milieu,
which is critical for maintaining erectile function.
Obesity is a chronic disease and is associated with
an increased risk of T2DM, hypertension and CVD;
however, the exact underlying pathophysiological
mechanisms by which obesity increases the risk of
VD remains the subject of intensive investigation [20].
Obesity and erectile dysfunction A. M. Traish et al.
5756 FEBS Journal 276 (2009) 5755–5767 ª2009 The Authors Journal compilation ª2009 FEBS
Visceral adipose tissue is a dynamic endocrine organ
and this specific adipose component is considered to
directly relate to cardiac risk. Visceral adipose tissue
secretes a host of biochemical modulators and proin-
flammatory factors contributing to systemic and
peripheral vascular inflammation. These include inter-
leukin (IL)-6, IL-1b, plasminogen activator inhibitor-1
(PAI-1), tumor necrosis factor-a(TNFa), angiotensi-
nogen, angiotensin-converting enzyme, vascular endo-
thelial growth factor, and serum amyloid A [21–28]
(Fig. 1). Adipokines are considered to facilitate mono-
cyte adhesion and migration into the vascular wall and
the conversion of monoctyes to macrophages.
Increased levels of TNFacause the enhanced expres-
sion of adhesion molecules in both the endothelium
and in vascular smooth muscle cells, and IL-6 stimu-
lates liver production of C-reactive protein, a nonspe-
cific marker of vascular inflammation [29–32]. Yudkin
[33] suggested a mechanism whereby TNFareleased
from fat stores surrounding a vessel may contribute to
the dysregulation of insulin modulation of endothelin-1
mediated vasoconstriction and NO-mediated vasodila-
tation favoring vasoconstriction.
A decrease in adiponectin levels observed in obese
patients is also thought to contribute to IR and coro-
nary heart disease (Fig. 1) [10,21,34]. Indeed, Pietilainen
et al. [35,36] suggested that only obese subjects who also
have reduced adiponectin levels exhibited endothelial
dysfunction. Adiponectin may play a larger role than
previously thought in both the amelioration and
pathogenesis of endothelial dysfunction. Sowers [34]
reported that: ‘Adiponectin reduces endothelial cell
apoptosis and may reduce the risk for atherosclerosis by
reducing vascular expression of adhesion molecules and
foam cell formation and vascular smooth muscle cell
proliferation …Weight loss in obese and overweight
persons has been found to increase adiponectin levels’.
Couillard et al. [37] noted that oxidized low density
lipoprotein levels were increased in obese men and that
obesity increases IR, which activates the endothelium
to produce adhesion molecules, facilitating macro-
phage migration with concomitant endothelial dysfunc-
tion. Gustafson et al. [25] also noted the up-regulation
of additional adhesion molecules, such as vascular cell
adhesion molecule-1, intracellular adhesion molecule-1
and chemokine monocyte chemoattractant protein-1
(MCP-1). These serum factors may be secondary to
the elevated levels of angiotensinogen and angiotensin-
converting enzymes observed in obese patients.
Increased MCP-1 and macrophage-colony stimulating
factor (M-CSF) expression in obese subjects are linked
to visceral adiposity with an increased risk of T2DM
and coronary heart disease [38].
Kim et al. [39] investigated the effects of exercise in
160 Korean adults separated into an exercise and non-
exercise groups. The authors suggested that central
obesity with high visceral fat is strongly associated
with blood levels of C-reactive protein, IR and endo-
thelial dysfunction-related factors. However, they did
not observe a significant difference in endothelial dys-
function between exercise and non-exercise partici-
pants. They also note that the relative importance of
central adiposity (as opposed to total body adiposity)
in endothelial dysfunction is still unclear. Perhaps, the
use of flow-mediated dilation may have been a better
parameter to capture endothelial dysfunction. Employ-
ing the latter approach together with ultrasonography
would have been a better clinical marker and may
have provided more reliable information.
In summary, increased visceral adiposity is associ-
ated with increased proinflammatory factors such as
increased PAI-1, TNFa, leptin, IL-6 and angiotensino-
gen, as well as reduced adiponectin levels. This imbal-
ance in such cytokines and adipokines results in
increased nuclear factor kB. It also results in a
decrease in NO synthase and NO activity, an increase
Fig. 1. Visceral adipose tissue as a potential endocrine organ.
Visceral fat is considered to comprise an active endocrine organ,
which increases the production of inflammatory cytokines (e.g.
IL-6, TNFa) as well as lipoprotein lipase, angiotensinogen, free fatty
acids, resistin, leptin, lactate, PAI-1, insulin and adipsin, and
reduces the levels of adiponectin. When coupled with visceral fat
hypoxia, the tissue also produces other adipokines, resulting in
inflammation, contributing to the onset of hypertension, dyslipide-
mia, diabetes, thrombosis and atherosclerosis. In addition,
increased conversion of cortisone to cortisol and increased aromati-
zation of androgens to estrogens results in reduced testosterone
levels. Thus, visceral fat contributes to PVD, including the vascular
bed of the penis, with adverse effects on endothelial function and
reduced circulating androgen levels. These processes result in
reduced penile tissue compliance and diminished penile hemody-
namics, and hence a physiological response, leading to ED.
Adapted from Lyon et al. [18]. Trayhum et al. [24] and Eckel et al.
[25].
A. M. Traish et al. Obesity and erectile dysfunction
FEBS Journal 276 (2009) 5755–5767 ª2009 The Authors Journal compilation ª2009 FEBS 5757
in adhesion molecules, and an increase in MCP-1 and
M-CSF, causing endothelial injury and dysfunction
[40] (Fig. 2). This leads to a pathophysiological state
involving reduced hemodynamics to the peripheral
tissues, including the penis. Thus, inflammatory cyto-
kines produced by visceral obesity can have a pro-
found and damaging effect on the endothelium not
only in the systemic vascular circulation, but also in
the peripheral vascular bed, such as the penis, and can
contribute to endothelial dysfunction leading to ED.
Obesity and its comorbidites are
associated with reduced plasma
testosterone levels
Metabolic alterations associated with obesity include
increased insulin, glucose and C-peptide levels, as well
as reduced plasma testosterone levels [14–18,41–45]. A
negative correlation between obesity, total testosterone
(TT), free testosterone (FT) and bioavailable testoster-
one (BAT) levels was inferred from evidence derived
from a host of epidemiological studies [46–48]. Svart-
berg et al. [15,16] investigated the relationship between
WC and TT, FT, and sex hormone binding globulin
(SHBG) levels in 1548 community dwelling men (aged
25–84 years). The authors showed WC to be inversely
related to FT and SHBG levels. A significant inverse
correlation between TT and obesity was reported in
other studies [42,43,45]. It should be noted that, in
massively obese men, serum TT and SHBG were
markedly reduced [49]. Visceral obesity may serve as
an endocrine disrupter and influence the endocrine
milieu by reducing the levels of luetinizing hormone
(LH) and testosterone resulting in hypogondoropic-
hypogonadism. The increased aromatase activity of the
visceral and peripheral fat further diminishes the levels
of testosterone and increases the levels of estradiol
(E
2
), which further contributes to preferential deposi-
tion of more visceral fat and exacerbates the vicious
cycle of reduced LH and testosterone with progressive
hypogonadism and increased obesity.
Corona et al. [50–52] reported a higher prevalence
of hypogonadism in MetS patients and these men had
an increased WC and hyperglycemia. The authors sug-
gested that WC and hypogonadism may predict the
condition of MetS. These observations are supported
by studies conducted by Tsao et al. [53] who showed
that WC was correlated with sexual dysfunction in
young males. In addition, Corona et al. [54] reported
that obese patients with ED had reduced SHBG-bound
and unbound testosterone levels even after adjusting
for obesity-related comorbidities. The authors sug-
gested that obesity itself may be the underlying cause
for a hypoandrogenic state.
Kaplan et al. [55] examined testosterone levels in
864 subjects with and without MetS and found that
obese men with MetS had significantly decreased TT
compared to non-obese men with MetS. Kapoor et al.
[56] investigated the relationship between ED versus
TT, BAT and FT levels in 198 men with T2DM. The
authors also assessed the degree to which cardiovascu-
lar risk factors such as hypertension, visceral adiposity,
smoking and alcohol consumption contributed to ED
in men with T2DM. The authors suggested that ED is
associated with low BAT and FT, but not TT, and
that ED is more frequently observed in men with
higher WC, who smoke, and who have hypertension.
They also found that 55–58% of their diabetic patients
with ED had low TT levels. In another study, Hofstra
et al. [57] assessed the prevalence of isolated hypogo-
nadotrophic hypogonadism in 160 obese men and
found that TT and FT were inversely related to BMI.
These men had decreased libido and ED compared to
eugonadal men. The author concluded that reduced
TT levels, well into the hypogonadal range, are com-
mon in male obesity. In summary, considerable evi-
dence exists linking obesity to reduced TT levels
[18,41,58–67]; however, detailed pathophysiological
mechanisms of obesity-induced androgen deficiency
and hypogonadism-induced obesity remain the subject
of intense investigation.
Several hypotheses have discussed the relationship of
androgen deficiency and obesity. That obesity is
commonly associated with low TT levels is attributed,
in part, to decreased SHBG levels. FT and BAT levels
are less affected in obese subjects early in the process.
Fig. 2. Visceral adiposity-induced endothelial dysfunction. Visceral
obesity induced endothelial dysfunction is mediated, in part, by lep-
tin, IL-6, free fatty acids (FFA), adiponectin, angiotensin II and
TNFa. These factors have a downstream effect on the activity of
NO synthase (NOS) and NO synthesis, adhesion molecules, MCP-1
and M-CSF. Adapted from Chudek and Wiecek [37].
Obesity and erectile dysfunction A. M. Traish et al.
5758 FEBS Journal 276 (2009) 5755–5767 ª2009 The Authors Journal compilation ª2009 FEBS
This is in accordance with findings obtained by Mohr
et al. [68] who suggested that an over-reliance on TT
levels in the diagnosis of androgen deficiency may
result in substantial misclassification. A plausible
mechanism that may account for a decrease in FT
involves elevated serum leptin levels in individuals with
large fat reserves. In obese individuals, elevated leptin
levels may interfere with LH ⁄human chorionic gona-
dotropin-stimulated androgen production, suppressing
androgen biosynthesis [18,69]. Another postulated
hypothesis is that high E
2
levels in obese men may be
attributed to an increased peripheral aromatization of
testosterone and the increased E
2
levels may alter LH
levels in obesity, thus modulating the pituitary gonadal
axis [70,71]. Another link between androgen deficiency
and obesity may involve IR, which has been shown to
contribute towards a reduction in circulating androgen
levels [72]. Alternatively, excess cortisol secretion is
considered to be associated with increased BMI, WC
and WHR, potentially suppressing testosterone pro-
duction via the hypothalamic–pituitary axis [73].
Decreased SHBG and increased aromatization of tes-
tosterone to E
2
in fat cells or cytokine-mediated inhibi-
tion of testicular steroid production in obesity may
represent a pathophysiological mechanism in obesity
[45]. Therefore, men with visceral obesity are in a
vicious cycle because testosterone deficiency leads to
reduced lipolysis, a reduced metabolic rate, visceral fat
deposition and IR.
Obesity and its comorbidities increase
the risk of ED
Based on a large number of studies, obesity may
represent a significant independent risk factor for ED
[74–83]. Corona et al. [50,51] demonstrated that 96.5%
of men with MetS, who also were obese, had ED.
Similar findings were reported in men with organic
ED, in which approximately 43% of men with MetS
had ED and the severity of ED increased with the
components of the MetS [84]. Interestingly, the severity
of ED was positively associated with both MetS and
IR. In a recent study, Corona et al. [54] found that, in
2435 male patients with ED, impaired penile blood
flow had a closer association with obesity-related com-
orbidites than with obesity itself. Clearly, ED repre-
sents a risk factor and may be a warning signal for
MetS and IR, with both being risk factors for CVD.
Interestingly, the largest increase in the incidence of
MetS occurred between men with moderate ED and
men with severe ED (21.7–70%). Corona et al. [54]
also demonstrate that the prevalence of fasting blood
sugar > 110 mgÆdL
)1
, a component of MetS, increases
with the severity of ED. Interestingly, these observa-
tions were not corroborated by Paick et al. [85,86],
who did not find a significant relationship between
ED severity and MetS parameters, except for hyperten-
sion. They hypothesized that the relationship between
MetS and ED severity is not clear-cut and that it may
be selective for specific MetS components.
Wang et al. [87] showed that MetS correlated with
peripheral vascular disease (PVD), suggesting a link
between ED and PVD. The prevalence of ED among
men with MetS increases with the number of MetS com-
ponents [88,89], suggesting that MetS is an independent
risk factor for ED [90] and WC may represent an inde-
pendent predictor of ED [91]. That obesity-related com-
oribidies, such as hypertension, contribute to impaired
penile blood flow is in line with the emerging data sup-
porting a strong link between MetS and ED. Although
there is a tight link between hypogonadism and obesity,
not all hypogonadal patients suffer from ED.
Kupelian et al. [92–94] further suggested that ED
may be a warning sign for the onset of MetS in men
with a BMI < 25 and that early intervention should
be initiated to prevent VD and ED. Although ED
might be considered as a warning sign for MetS, as
noted earlier, the relationship between these two condi-
tions is complex because components of MetS, such as
hypertension, clearly play a role in the etiology of ED.
Therefore, it is likely that each condition potentiates
the other in a vicious cycle.
A study by Zohdy et al. [95] related MetS and
androgen deficiency with ED in 158 obese men and
showed that, with increasing BMI, the frequency of
hypogonadism and ED was increased, whereas TT
showed a strong negative correlation. The authors
found that, for a BMI < 25, three out of 13 men
(23.1%) had vasculogenic ED compared to 32 out
of 54 men (59.3%) with a BMI ‡25. By contrast,
Kupelian et al. [92,93] suggested that ED is a better
predictor of MetS in men with a BMI < 25, although
Kupelian et al. [92–94] did not consider other etiolo-
gies of ED. These observations are not congruent with
other reports [95]. Corona et al. [54] reported that
obesity is associated with low levels of androgens in
men with ED, even after adjustment for comorbidities,
and that obesity-associated comorbidities, especially
hypertension, are among the most important determi-
nants of arteriogenic obesity-associated ED [54].
Although it is clear from the many aforementioned
studies that obesity and ED are tightly linked, Riedner
et al. [75] found that different assessment parameters
of central obesity have different degrees for predicting
the odds of developing ED. The authors presented
data to suggest that a WC > 102 cm has an odds
A. M. Traish et al. Obesity and erectile dysfunction
FEBS Journal 276 (2009) 5755–5767 ª2009 The Authors Journal compilation ª2009 FEBS 5759