Oral contraceptives and endothelial function

Oral contraceptives (OC) have been widely used for some decades. As early as in the 1980s, epidemiological studies suggested a higher cardiovascular risk in women taking OCs containing high-dose oestrogen– progestagen [1]. Third generation OCs, with lower doses of ethinylestradiol and synthetic progesterones, appear to have a lower cardiovascular risk potential but their use is still associated with a trend to increased atherothrombotic diseases [2]. Moreover, the role of hormonal supplementation has been questioned due to recent observations in postmenopausal women, for example, in the ERA [3], HERS [4] and WHI [5] studies. In this issue of the journal, Virdis et al. [6] investigated the influence of a third generation OC (30 μg ethinylestradiol and 75 μg gestodene daily) on endothelial function in healthy women. Endothelial dysfunction is recognized as an initial step in the development of atherosclerosis [7,8], and the authors speculate that an increased cardiovascular risk in OC users might be at least partly explained by an impairment of endothelial function. The endothelium is a complex endocrine/paracrine organ modulating vascular tone, with nitric oxide (NO) being the most important relaxing factor. NO derives from the l-arginine pathway promoted by NO synthase (NOS). The healthy endothelium releases NO continuously and upon stimulation by certain agonists, as well as by shear stress [9]. Furthermore, the endothelium exerts anti-atherosclerotic and anti-thrombotic effects by inhibiting platelet aggregation, adhesion of monocytes and proliferation of smooth muscle cells [9]. Endothelial function can be assessed by biochemical and, in particular, functional parameters. The most widely used technique to assess NO-dependent vasodilatation is the measurement of forearm blood flow by strain-gauge venous plethysmographic changes, as performed by Virdis et al. [6]. Endothelial dysfunction is usually represented by a decline of vasodilatation due to decreased NO activity. This may be a consequence of either reduced release or augmented degradation of NO [9]. Impaired endothelium-dependent vasodilatation has been recognized in essential hypertension [10,11] and hypercholesterolemia [12]. In both conditions, oxidative stress appears to be the major mechanism reducing NO availability. Besides other factors, age and gender are important in affecting endothelial function [13–15]. The age-associated decline of endothelial function is initially caused by reduction of endothelial NOS (eNOS) [13], followed by rise of endothelium-derived constriction factors [11]. Oestrogens and endothelial function Endothelial function fluctuates during a menstrual cycle in parallel with changes in endogenous oestrogen release and premenopausal women appear to be protected from age-related decline of endothelial function [15]. Thus, endothelium-dependent vasodilatation is significantly better in premenopausal women than in age-matched men. This appears to be due to increased NO availability as well as reduced oxidative stress [16,17]. To further elucidate the influence of oestrogen on endothelium-dependent vasodilatation, it is helpful to take a closer look at differences between pre- and postmenopausal women. Postmenopausal women, with similar or even lower oestrogen levels than age-matched men, have a significant lower increase of forearm blood flow after reactive hyperemia than premenopausal women [12]. How do oestrogens influence endothelium-dependent vasodilatation? Endothelial cells possess two separate oestrogen receptors (ERα and ERβ) [18–20]. Their activation increases NO availability by two different mechanisms. First, activation of these receptors stimulates the expression of a gene or genes for eNOS and, second, direct eNOS production [18,19,21], Furthermore, oestrogen serves as an endogenous antioxidant [20]. Accordingly, improvement of endothelial function by oestrogen supplementation has been demonstrated in postmenopausal women [12,16,22–26]. Oral contraceptives and endothelial function Whether OCs improve or alter endothelial function is an important issue. In this context, Virdis et al. [6] studied the effect of a third-generation OC on endothelial function in healthy young women. Endothelial function remained unchanged after 6 months of OC use (30 μg ethinyl-oestradiol + 75 μg gestodene daily) despite unfavourable changes of the lipid profile that occurred during the study. The authors speculate that adverse effects of increased lipid levels are counterbalanced by the assumed beneficial influences of the oestrogen component. Their assumption implies a neutral effect of the gestodene component. However, data on the influence of progesterones on endothelial function are still inconsistent. In some studies, the favourable effect of oestrogen on endothelial function was unchanged by progesterones [22,27], whereas other studies found that progesterones reduce or offset oestrogen-mediated effects on endothelial function [28]. Moreover, the single premenopausal use of depot medroxyprogesterone acetate impairs hyperemia induced flow-mediated vasodilatation [29]. By contrast, in postmenopausal women with already established oestrogen deficiency, neither micronized progesterone nor synthetic progestin adversely affected vascular function [30,31]. Taking all the available data together, a direct effect of progesterone on endothelial function appears unlikely. Oral contraceptives, lipids, and endothelial function Virdis et al. [6] found elevated levels of low density lipoprotein (LDL)-cholesterol in women using OCs. The assumption that increased levels of LDL- cholesterol impaired endothelial function is crucial for their interpretation. Indeed, it is well established that hyperlipidemia, especially an elevated level of LDL-cholesterol, adversely affects endothelial function [32,33]. Therefore, it is conceivable that ethinylestradiol counteracted the adverse lipid action, thus contributing to the overall neutral effect on the endothelium, which is the main finding in the study by Virdis et al. [6]. In line with their conclusion, John et al. [34] reported an enhanced basal production and release of NO in women using OCs. However, the results of Virdis et al. [6] raise another question. Apparently, the third generation OC induced a rise of LDL-cholesterol during 6 months of use. This is surprising because multiple studies of hormone replacement therapy (HRT) clearly demonstrated favourable effects of oestrogens on the lipid profile [3–5,35]. Any unfavourable effects of progesterones on the lipid profile are worthy of mention. Progesterones have been shown to attenuate beneficial effects of oestrogens on the lipid profile in many but not all studies. In particular, the increase in HDL-cholesterol appears to be blunted, whereas the reduction of LDL-cholesterol was not influenced in most studies [35]. In the ERA study, the lowering of LDL-cholesterol was even higher (16.5 versus 9.4%) in women using oestrogen combined with progesterone than it was in women using oestrogen alone [3]. By contrast to HRT and the study by Virdis et al. [6], a recent study using the same third generation OC in premenopausal women failed to demonstrate significant changes in the lipid profile [36]. Considering the available data, the observed remarkable increase of LDL-cholesterol, and a slight decrease of HDL-cholesterol by Virdis et al. [6], is unlikely to be caused by the progesterone component. Thus, the reason for the changes in lipid profile remain unclear and need further investigation. Oxidative stress is the most important mechanism by which hyperlipidemia affects endothelial function. This is supported by the finding of Virdis et al. [6] indicating that, in hypercholesterolemic women, vasodilatation to acetylcholine was restored to values of healthy women by intra-arterial infusion of the antioxidant vitamin C. It is likely that one mechanism by which hyperlipidemia induces oxidative stress is LDL-cholesterol accumulation and LDL-oxidation in the vascular wall. In addition to this functional impairment of endothelial function, long-term hypercholesterolemia promotes atherosclerosis. Thus, endothelial dysfunction in hypercholesterolemic patients may partly be explained by structural damage. Nevertheless, the short-term effects of lipids on endothelial function are able to affect endothelium-dependent vasodilatation [37]. Despite the newly developed hypercholesterolemia, Virdis et al. [6], who observed an overall neutral effect of OC on acetylcholine-induced increase of forearm blood flow, draw an indirect conclusion. They suggest that the neutral effect could be the balance between a deleterious effect of hypercholesterolemia and a protective effect of OC on endothelial function. This conclusion remains open to discussion. Oral contraceptives and cardiovascular risk The results of Virdis et al. [6] are also interesting because they shed light on the problem of an increased cardiovascular risk, even in women receiving OC in third generation formulations. Considering that alteration of endothelium-dependent vasodilatation, which is the earliest detectable step towards atherosclerotic disease, remained unchanged in the study by Virdis et al. [6], this does not appear to be the mechanism by which OC increases cardiovascular risk. Nevertheless, Virdis et al. [6] studied NO-dependent vasodilatation, which is an important, but not the sole, component of endothelial function. The intact endothelium also exerts anti-thrombotic and anti-inflammatory effects, which are not explored or detected by the assessment of NO-dependent vasodilatation. Thus, cardiovascular benefits of exogenous oestrogen may be offset by previously underestimated pro-thrombotic or pro- inflammatory effects, both in postmenopausal women [38,39] and in OC users. In postmenopausal women, HRT both with unopposed oestrogen and oestrogen/progestin, has been shown to increase certain thrombotic markers and C-reactive protein (CRP) [39,40]. These alterations have been proposed to explain the unexpected increase of cardiovascular risk by HRT [3–5]. Thus, even in the light of findings of Virdis et al. [6] and others with respect to beneficial effects of oestrogens on NO-dependent vasodilatation, caution is required when attempting to extrapolate such observations to the protection of endothelial function in general and to associated cardiovascular risk reduction.