Robert J. Krane MD, Department of Urology, Boston Medical Center, 720 Harrison Ave, DOB 606, Boston, MA 02118, USA.
Age-related smooth muscle dysfunction has long been recognized in the respiratory [ 1], gastrointestinal [ 2] and cardiovascular [ 3] systems. Ageing also affects the genitourinary tract and is associated with lower urinary tract symptoms (LUTS) [ 4, 5] and sexual dysfunction [ 6–8] in both men and women. How ageing leads to smooth muscle dysfunction is at best poorly understood. One possible mechanism is chronic ischaemia caused by atherosclerotic occlusive disease. Atherosclerosis-induced arterial insufficiency is a common clinical problem in the elderly and remains the leading cause of death in the adult population [ 9, 10]. The abdominal aorta and its branches, especially the bifurcation of the iliac arteries, is involved earliest and most severely by atherosclerotic lesions [ 10]. Atherosclerosis of the aorto-iliac arterial bed can potentially compromise the blood supply of lower genitourinary tract. For example, atherosclerotic disease of the pudendal and cavernosal arteries has been shown to be a major cause of erectile dysfunction in the elderly patient [ 11]. Major risk factors for atherosclerosis, such as hypertension, hypercholesterolaemia, smoking and diabetes [ 9, 10], have also been found to be associated with smooth muscle dysfunction, e.g. impaired relaxation of vascular smooth muscle [ 12, 13]. In animal models, atherosclerosis-induced pelvic ischaemia can produce functional and structural alterations in detrusor [ 14] and corporal cavernosal smooth muscle [ 15] which parallel the age-related changes in bladder and cavernosal smooth muscle in humans. Therefore, there has been increasing interest in the possible role of atherosclerosis-induced ischaemia in LUTS and erectile dysfunction of the elderly, which will be the subject of this review. The relationship between ischaemia and LUTS is evolving and currently constitutes a hypothesis, while the relationship between ischaemia and erectile dysfunction is more widely accepted.
Ageing and the epidemiology of erectile dysfunction
An association between ageing and male erectile dysfunction was shown in several epidemiological studies. In 1948, Kinsey et al. first showed that there is a decline in sexual activity and erectile function with ageing [ 6]. This study, which involved 15 871 participants aged 10–80 years, remains the largest population-based study of normative data on male sexual behaviour. In that study, the incidence of erectile dysfunction was <1% before age 30 years, <3% before age 45, 6.7% in men aged 45–55 years, 25% in men 65 years old and 75% in men aged 80 years. Likewise, Diokno et al. [ 7] studied the relationship between ageing and sexual behaviour in 296 men and 448 women older than 60 years; they found that the incidence of erectile dysfunction among married men increased from 28.7% when aged 60–64, to 64.3% in men >80 years old. The Massachusetts Male Aging study, a community-based, random-sample observational survey which was performed on 1290 non-institutionalized men in the Boston area, provided very important information about the epidemiology of erectile dysfunction [ 8]. According to this study, it has been estimated that the percentage of potent men will decrease from 60% to 33% between 40 and 70 years of age. However, this study also showed that erectile dysfunction is not necessarily an inevitable consequence of ageing, as it is associated with potentially reversible concomitant pathologies. For example, it has been shown that the age-adjusted probability of complete erectile dysfunction is 39% in men with heart disease, 15% with hypertension and 28% with diabetes, compared with 9.6% in the whole sample population [ 8].
Indeed, atherosclerotic disease of the penile arteries has been known to be the major cause of erectile dysfunction in ageing men [ 11]. For example, in a study of 121 men aged 60–80 years with erectile dysfunction, vascular disease alone or combined with neurogenic problems was the causative factor in 51% of patients [ 16]. Supporting these clinical findings, postmortem studies have revealed that ageing is associated with increasing degrees of atherosclerotic vascular disease in the arterial bed of the penis [ 17]. Epidemiological studies have shown that vascular risk factors such as hypertension, hypercholesterolaemia, smoking and diabetes are associated with a higher probability of erectile dysfunction. For example, it has been reported that ≈10% of untreated hypertensive patients were impotent at the initial diagnosis of hypertension [ 18]. In another study, the incidence of erectile dysfunction among 131 men aged 31–86 years hospitalized for an acute myocardial infarction was 64% [ 19]. Likewise, the incidence of erectile dysfunction among patients who underwent coronary artery surgery was 57% [ 20]. Diokno et al. [ 7] have shown that a history of heart attack is associated with an increase in the incidence of erectile dysfunction from 35% to 59% in men older than 60 years. In the Massachusetts Male Aging study, a decrease in high-density lipoprotein levels from 900 to 300 mg/L was associated with an increase in complete erectile dysfunction from zero to 16% in men aged 56–70 years. Likewise, cigarette smoking was shown to be a risk factor for erectile dysfunction in men with heart disease, with an age-adjusted 56% probability of complete erectile dysfunction in smokers against a probability of 21% in non-smokers [ 8].
As atherosclerosis is a generalized disease of the arteries, it has been suggested that arteriogenic erectile dysfunction can be predictive of a higher risk of cardiovascular disease [ 21, 22]. For example, it has been shown that an abnormal penile brachial pressure index may be predictive for an elevated risk of a myocardial infarction and an abnormal cardiovascular stress test, even in men who have no cardiac symptoms [ 21, 22].
Ischaemia and the pathophysiology of erectile dysfunction
Penile erection is a neurovascular event which requires profound corporal and helicine artery smooth muscle relaxation to allow blood flow to increase by 3–4 times its normal level [ 11]. This allows the engorgement of the lacunar spaces, which causes passive compression of the subtunical veins, thus producing penile rigidity. Neurological innervation for this event is via the sacral parasympathetic nerves, with the final neurotransmitter being nitric oxide (NO) [ 23]. An adequate arterial blood supply is necessary for erectile function to fill the lacunar spaces, increase the cavernosal pressure and activate the veno-occlusive mechanism. Another important function of adequate arterial flow is to increase the intracavernosal oxygen tension from venous levels in the flaccid phase to arterial levels during erection. It has been shown that NO production is regulated by oxygen tension in the corpus cavernosum, where low oxygen tension inhibits NO synthesis [ 24].
The effect of atherosclerosis-induced ischaemia on erectile function has been studied in various animal models [ 15, 25, 26]. In New Zealand white rabbits, Baumgartner’s aorto-iliac balloon de-endothelialization technique [ 27], together with a cholesterol-rich diet, resulted in a diffuse aorto-iliac atherosclerotic occlusive disease also affecting the pudendal and cavernosal arteries, and leading to erectile dysfunction [ 25]. Studies in this animal model have revealed a significant correlation between the severity of arterial occlusive disease, iliac artery blood flow, peak intracavernosal blood flow, equilibrium intracavernosal pressure and cavernosal pressure decay [ 15, 25, 26]. The severity of arterial occlusion in this animal model was significantly correlated with decreased smooth muscle in the corpus cavernosum [ 28] ( Fig. 1). For example, computer-assisted histomorphometric analysis of the cavernosal cross-sections from atherosclerotic animals showed a mean smooth muscle proportion of 33.9%, compared with 45.4% in control animals [ 28]. The decreased proportion of smooth muscle in the corpus cavernosum impairs cavernosal expandability and causes veno-occlusive dysfunction [ 28]. These animal studies support the conclusion that atherosclerosis-induced ischaemia of the corpus cavernosum leads to erectile dysfunction, which is correlated with cavernosal fibrosis. Therefore, a normal cavernosal smooth muscle content appears to be essential for veno-occlusive function.
The important role of cavernosal smooth muscle content in veno-occlusive function has also been shown in human studies. In a study of 24 patients with erectile dysfunction who had a penile prosthesis inserted, the proportion of cavernosal smooth muscle was significantly correlated with cavernosometric variables such as flow to maintain erectile rigidity, venous outflow resistance and intracavernosal pressure decay values [ 29]. As in the animal studies, the corporal smooth muscle content in patients with low, intermediate and severe veno-occlusive dysfunction was in the ranges 39–43%, 30–37% and 13–29%, respectively.
The role of oxygen tension and transforming growth factor β1 (TGF-β1)
The exact pathophysiological mechanism of ischaemia-induced fibrosis of the corpus cavernosum is not clearly understood. However, it is likely that hypoxia-induced over-expression of TGF-β1 may play a role in this process. TGF-β1 is a pleotrophic cytokine that has been shown to be an essential mediator of tissue fibrosis [ 30]. Overproduction of TGF-β1 decreases the smooth muscle to connective tissue ratio by inducing the expression of collagen, fibronectin and proteoglycans, while inhibiting growth of smooth muscle cells and the activity of collagenase [ 30, 31]. In cell culture studies, TGF-β1 has been shown to increase collagen synthesis in corpus cavernosum smooth muscle cells by 2.5–4.5 times the baseline levels [ 32]. In the same study, hypoxia was shown to increase collagen synthesis by inducing TGF-β1 mRNA expression in cavernosal smooth muscle cells. Under continuous ischaemic conditions, TGF-β1 was shown to auto-induce its own mRNA, leading to a further increase in TGF-β1 synthesis that reinforces the development of severe fibrosis [ 30]. Therefore, low oxygen tension in corpus cavernosum appears to be a risk factor for tissue fibrosis.
In another study in patients with various causes of erectile dysfunction, Sattar et al. [ 33] correlated the proportion of cavernosal smooth muscle with cavernosal oxygen tension before and after injection with PGE1. Oxygen tension levels in patients with arteriogenic, venogenic and nonorganic (psychiatric) erectile dysfunction were similar in the flaccid phase. After intracavernosal injection of PGE1, the increase in cavernosal oxygen tension was significantly less in patients with arteriogenic and venogenic erectile dysfunction than in patients with nonorganic erectile dysfunction. The mean percentages of smooth muscle fibres in corpus cavernosum in patients with arteriogenic and venogenic erectile dysfunction was 29.3% and 34.1%, respectively, compared with 43.3% in patients with non-organic erectile dysfunction [ 33]. This decrease in the percentage of cavernosal smooth muscle is similar to the atherosclerosis-induced decrease in the proportion of smooth muscle of the rabbit corpus cavernosum [ 28].
Another important role of cavernosal oxygen tension appears to be the regulation of prostanoid production in the corpus cavernosum [ 34]. It has been shown that low oxygen tension decreases basal and acetylcholine-stimulated production of PGI2, thromboxane A2, PGF2α and PGE2 by inhibiting the activity of prostaglandin H synthase. The role of low oxygen tension in prostanoid production may also play a role in the mechanism of ischaemia-induced cavernosal fibrosis, as decreased levels of PGE1 correlated with increased expression of TGF-β1 mRNA in human corpus cavernosum smooth muscle cells [ 32].
The significant role of cavernosal oxygen tension in maintaining a normal smooth muscle to connective tissue ratio suggests a possible role for nocturnal penile tumescence (NPTs) in oxygenation of the corpus cavernosum. Ageing has been shown to decrease the frequency, duration and degree of these erections [ 35, 36]. The exact mechanism of the age-related impairment of NPT and its role in erectile dysfunction is not clearly understood. It has been widely accepted that impairment of NPT is caused by erectile dysfunction. However, this may not be the case, as NPT has also been found to decrease in potent men as a function of ageing [ 36]. It is an interesting hypothesis that NPT may serve to periodically oxygenate the corpus cavernosum, and that an age-related decrease in the quality and number of NPTs can indirectly affect erectile function by not exposing the penis to sufficient oxygen. Rather than erectile dysfunction leading to impaired NPT, a decreased frequency of NPT may also compromise erectile function. Further studies are needed to clarify the exact role of NPT in the oxygenation of the penile corpus cavernosum.
In summary, oxygen tension appears to be an important regulator of NO, prostanoid and TGF-β1 synthesis in the corpus cavernosum. Decreased oxygen tension can impair the normal balance between smooth muscle cells and connective tissue and lead to veno-occlusive dysfunction. Therefore, it is not unreasonable to hypothesize that the erectile dysfunction of ageing is the result of atherosclerosis-induced cavernosal ischaemia leading to cavernosal fibrosis and veno-occlusive dysfunction. As the erectile dysfunction of ageing appears to be a slowly progressive disorder, it appears wise for the patient to seek medical intervention earlier rather than later, to minimize the development of veno-occlusive dysfunction. If the newer oral agents can alter the natural history of this process, earlier intervention may significantly delay or even prevent the development of corporal fibrosis.
Ageing and the epidemiology of LUTS
Bladder dysfunction is a common problem in the ageing population; traditionally, LUTS in elderly men have been referred to as prostatism and related to bladder outlet obstruction (BOO) caused by benign prostatic hyperplasia (BPH). This view did not consider the ageing bladder as a possible cause of LUTS. However, pressure-flow studies have shown that in about a third of cases, LUTS in ageing men were not associated with BOO [ 37, 38]. This ratio increased to 60% when the study population was older than 80 years [ 5]. Another study has shown that there is no significant difference in voiding pressures in men older than 60 years with and without bladder-emptying symptoms [ 4]. Furthermore, 47% of asymptomatic men had pressure-flow studies indicative of obstruction [ 4].
A discrepancy between LUTS and documented BOO has also been shown by standardized symptom scores, which were introduced as an index of obstruction but have not been found to be specific for BOO [ 39]. These findings explain why treatments aimed at relieving BOO in elderly men fail to completely improve LUTS. For example, 25% of men undergoing TURP will not have a favourable symptomatic outcome when patients are selected on the basis of symptoms [ 40, 41].
LUTS are bothersome in ageing women with the same frequency as in ageing men. In an epidemiological study on non-institutionalized residents older than 60 years, the prevalence of urinary incontinence, difficult bladder emptying and irritative bladder symptoms in men was 19%, 22% and 12%, respectively, and in women 38%, 11% and 17%, respectively [ 42]. Lepor and Machi reported that AUA symptom scores are similar in aged-matched female and male populations [ 43]. Likewise, Chancellor and Rivas showed similar AUA symptom scores in women to those reported in men with BPH [ 44].
Furthermore, histological and urodynamic studies have shown that ageing causes structural and functional changes in the bladder independent of obstruction and gender. For example, histomorphometric and semiquantitative histological studies of the human bladder have shown that the smooth muscle to connective tissue ratio decreases with ageing in both sexes, independent of BOO [ 45, 46]. An ultrastructural study in patients aged 23–92 years showed an age-dependent increase in connective tissue density in conjunction with arteriosclerosis of small vessels, suggesting that ageing may affect the haemodynamics of the bladder [ 47].
Similarly, urodynamic studies have shown that there are age-specific alterations in bladder function affecting both sexes. In a study on 253 men and 183 women aged >40 years, Madersbacher et al. reported a decrease in maximum flow rate, voided volume and bladder capacity, and an increase in postvoided residual volume, with progressing age [ 48]. Bladder compliance also decreases significantly with increasing age after 80 years old [ 49]. The age-related urodynamic changes have also been noted in asymptomatic individuals. For example, in a uroflowmetric study on asymptomatic men aged 50–92 years, the maximum flow rate and voided volume decreased significantly with increasing age [ 50].
The conclusion from these epidemiological, urodynamic and histological studies is that LUTS in the elderly are not necessarily related to BOO or specific to men. It is likely that LUTS of the elderly have a multifactorial aetiology and that age-related changes in bladder function may play a role.
LUTS and vascular disorders
There is a lack of published information about the relation of vascular risk factors and age-related voiding dysfunction. Among very few epidemiological studies, Diokno et al. [ 51] were the first to show that a history of cardiovascular disease was associated with a higher risk of urinary incontinence. Another important finding in their studies was the positive correlation between LUTS and erectile dysfunction [ 7]. They found a doubling of the incidence of erectile dysfunction in men with interrupted urinary stream (61%) when compared with men who did not report urination difficulty (34%). Surveying 225 men of different age groups, Mulligan et al. [ 52] reported an age-adjusted correlation between erectile dysfunction and urinary incontinence. Among vascular risk factors, smoking has been reported to be associated with LUTS in men; however, the mechanism remains unknown [ 53]. Recently, in a study of 440 men older than 50 years, it was reported that sexual dysfunction is positively correlated with LUTS [ 54]. Concomitant hypertension in these patients increased this correlation significantly. Although only a few studies exist in this area, the epidemiological correlation between erectile dysfunction, LUTS and vascular disease is interesting, and suggests that atherosclerotic vascular disease may play a causative role in both erectile dysfunction and LUTS of the elderly.
LUTS and the pathophysiology of ischaemia
Finkbeiner and Lapides were among the earliest to show that distension of the bladder impairs bladder blood flow in dogs [ 55]. Based on these findings, Lapides [ 56] postulated that bladder ischaemia due to high intravesical pressure may be a cause of bladder infection and this led to the introduction of CISC.
Several experimental models have been used to study bladder ischaemia. Models using bilateral ligation of the vesical arteries resulted in bladder necrosis which resembled gangrene rather than atherosclerosis-induced ischaemia [ 57, 58]. Models using unilateral vesical artery ligation produced bladder non-compliance and hyperactivity, with impaired contractile responses to betanechol and electrical field stimulation up to 2 weeks after ligation. However, studies at 4 weeks after unilateral arterial occlusion revealed in vivo and in vitro findings similar to those in sham-operated animals [ 57, 58]. The recovery mechanism of the rabbit bladder after unilateral vesical artery ligation may involve regeneration of collateral circulation.
Because of the difficulty of quantifying bladder blood flow, understanding of bladder haemodynamics remained sparse until the last decade. The development of the Doppler laser flowmeter enabled real-time monitoring of bladder perfusion and study of blood flow changes with filling and emptying [ 59–61]. Using the Doppler laser flowmeter, it has been shown that high-pressure filling and bladder contraction against a closed outlet exposes the bladder wall to significant ischaemia, where intravesical pressure seems to determine the bladder wall blood flow [ 60–62]. These findings support the possibility that bladder ischaemia may develop secondary to BOO. This hypothesis is further supported by another study showing that unilateral ischaemia and BOO induce similar genetic and cellular responses in the rabbit bladder [ 63]. In the latter study, the authors hypothesized that obstruction-induced changes in the bladder may be, at least in part, related to bladder ischaemia.
To investigate the possible role of atherosclerosis-induced ischaemia on bladder function, we used the rabbit model of aorto-iliac atherosclerosis [ 14, 64]. Studies performed 16 weeks after producing atherosclerotic occlusion of the aorto-iliac-hypogastric arteries showed a significant decrease in bladder blood flow [ 14, 64]. The atherosclerotic process has also been found to affect the smaller arteries in the bladder, thus limiting the development of collateral circulation. In vivo studies have shown that bladder ischaemia causes bladder hyperactivity and non-compliance [ 14, 64]. Histomorphometric analysis revealed a mean bladder smooth muscle proportion of 37% in the ischaemic animals, compared with 66% in the control animals [ 14]. As in the case of cavernosal tissue, the degree of bladder fibrosis correlated with the severity of atherosclerotic occlusion and degree of bladder ischaemia [ 14, 64] ( Fig. 1). Immunohistochemical studies revealed that bladder fibrosis in the ischaemic animals correlated with increased expression of TGF-β1 [ 14]. These studies have shown that atherosclerosis-induced arterial insufficiency of the bladder can significantly affect bladder function and structure, producing hyperactivity, non-compliance or decreased smooth muscle percentage. As these changes also occur in the ageing bladder, it can be postulated that atherosclerosis-induced bladder ischaemia may be involved in the pathophysiology of age-related voiding dysfunction.
The presence of a concomitant atherosclerotic disease may also play an important role in the pathophysiology of BPH-induced voiding dysfunction. High intravesical filling and voiding pressures in patients with BOO and arterial insufficiency of the bladder may lead to a more severe bladder ischaemia ( Fig. 2). The degree of bladder ischaemia may dictate the bladder’s compensatory response to obstruction. Patients with more significant ischaemia may be more prone to develop impaired bladder contractility.
At the same time, decreasing outlet resistance, in obstructed or unobstructed patients, with atherosclerotic vascular disease and LUTS may improve symptoms by decreasing voiding pressures and thereby improving bladder blood flow. Indeed, it is well known that TURP improves symptoms in a significant number of unobstructed men. For example, it has been shown that TURP has a favourable symptomatic outcome in 55% of unobstructed men with LUTS [ 65]. Again, it may be hypothesized that improved bladder perfusion may play a role in this outcome, but further studies are needed to test the ischaemia hypothesis in age- as well as BPH-related voiding dysfunction.
We have attempted to present a unifying concept, namely pelvic arterial insufficiency, as underlying the common clinical problems of LUTS and erectile dysfunction. We realize that each disorder has a complicated multifactorial aetiology. However, we are increasingly convinced that strong evidence is accumulating to support ischaemia as an important aetiological factor. Certainly, atherosclerosis can explain the marked association of these conditions with increased age and their association with vascular risk factors. In addition, that the corporal and vesical smooth muscle alterations seen clinically can be reproduced in the ischaemic animal model ( Fig. 1) lends further credence to our unifying hypothesis. As stated in our introduction, the concept of atherosclerosis-induced ischaemia as a cause of erectile dysfunction has been widely accepted. Further research will be necessary to assess the relationship between LUTS and bladder ischaemia. It should also be remembered that ischaemia may lead to functional alterations of other pelvic organs such as the prostate, vagina and clitoris. The age-related changes and disorders involving these organs may also have an ischaemia-induced aetiology to some degree. Further research may lead to the development of therapeutic approaches that can delay or even prevent age-related voiding and erectile dysfunction.