Neuronal NOS activity by nonadrenergic noncholinergic

neurons is known to be decreased in both

in vitro

and

in vivo

models of smoking (Xie

et al.

, 1997). Compo-

nents of burned, but not unburned, tobacco are in part

responsible for the loss of neuronal NO through enzy-

matic blockade (Demady

et al.

, 2003). It has been well

established

the

vascular

literature

that

cigarette

smoke damages the endothelium and impairs eNOS-

mediated vasodilation (Fig. 1) (Celermajer

et al.

, 1993;

Butler

et al.

, 2001).

Furthermore, in addition to cigarettes’ effect on enzy-

matic synthesis, superoxide anions produced by the

metabolites from smoke directly decrease the level of free

NO in the corpora cavernosa. This is partially mediated

via activation of the NADH oxidase enzyme family

(Orosz

et al.

, 2007). Superoxide anions are increased in

smokers, and their presence shunts NO into a peroxyni-

trite pathway that lessens the vasoactive availability of

NO (Peluffo

et al.

, 2009).

The Rho-associated kinase (ROK), which regulates sen-

sitivity to calcium contractility in the smooth muscle cell,

is known to maintain the ±accid state, and, as such, ROK

inhibitors can be theorised to induce erection (Mills

et al.

, 2001). Intracellular NO functions to inhibit ROK

allowing for vasodilation. Similarly, decreases in NO lev-

els secondary to smoking disinhibit ROK and act to fur-

ther worsen ED (Chitaley

et al.

, 2001). Furthermore,

smokers have decreased ROK activity in peripheral leuco-

cytes correlating to poor nitroglycerin vasodilation further

hinting at a connection between ROK signalling and

smoking (Hidaka

et al.

, 2010).

Smoking also causes intrinsic damage to vessels pre-

venting elastic dilation despite strong paracrine signals.

Smoking alters the elastin of the extracellular matrix and

induces calci²cation of medial elastic ²bres producing

arterial stiffness (Guo

et al.

, 2006).

Epidemiological associations between ED and

smoking

There have been numerous cross-sectional studies that

have established a correlation between cigarette smoking

and ED (Austoni

et al.

, 2005; Kupelian

et al.

, 2007; Chew

et al.

, 2009; Ghalayini

et al.

, 2010; Wu

et al.

, 2012). The

studies have included populations from China, the Mid-

dle East, Europe and the Americas. Each study exhibited

a variable baseline smoking prevalence. The odds ratio of

smokers with ED has ranged between 1.4 and 3.1 with

statistically signi²cant con²dence intervals in the vast

majority of these studies. Typically, populations were

selected to minimise other known causes of ED such as

psychotropic medications and prostate cancer due to

treatment effects. In a speci²c cohort of young men

40 years of age, smoking was a signi²cant risk factor for

ED. In these men, the multivariate analysis did not show

signi²cance in other vascular risk factors strongly indicat-

ing a role for smoking in the pathogenesis of ED in

younger men (Elbendary

et al.

, 2009). While the majority

of these studies accounted for other vascular risk factors

(i.e. age, hypertension, obesity and diabetes), it was dif²-

cult to determine signi²cance of any isolated risk factors

as many existed together and were impossible to separate.

To address the inherent bias in cross-sectional studies,

a series of long-term cohorts were created in the 1990s to

determine possible links between smoking and ED. In the

Male Health Professionals Study, of the 22 086 men with-

out baseline ED, the relative risk that smokers developed

ED over a follow-up of 14 years was 1.4 (95% CI

1.3

–

1.6) (Bacon

et al.

, 2006). Likewise in Minnesota

cohort, the odds ratio of smokers to develop ED was 1.42

after adjusting for age (95% CI 1.00

–

2.02). The strength

of the association was greatest in men under the age of

70 years, and this association decreased with progressively

older age groups. The investigators felt this may have

been due to survivorship bias or the exclusion of men

from the study who had undergone prostate surgery or

who had prostate cancer (Gades

et al.

, 2005). The Massa-

chusetts Male Aging Study followed 513 middle-aged men

with good erectile function and excluded diabetics and

patients with baseline cardiac disease. Cigarette smokers

Fig. 1

Interplay of neuronal and endothelial effects on vascular relax-

ation. The main mediator of penile arterial relaxation is nitric oxide

(NO). NO is released directly from neurons and indirectly via endothe-

lial production. Cigarette smoke has been shown to directly inhibit

both neuronal and endothelial isoforms of nitric oxide synthase (NOS).

In addition, superoxide anions from cigarette smoke directly degrade

NO. Rho-kinase (ROK) is upregulated in men who smoke, thus activat-

ing myosin light-chain (MLC) phosphatase that prevents NO-induced

relaxation.

1088

2014 Blackwell Verlag GmbH

Andrologia

2015,

, 1087–1092

Smoking and male erectile function

J. R. Kovac

et al.