Human endothelial function and microvascular ageing


  • Phillip E. Gates,

    1. Diabetes and Vascular Medicine, Peninsula NIHR Clinical Research Facility and Institute of Biomedical and Clinical Sciences, Peninsula Medical School, Exeter, Devon, UK
    Search for more papers by this author
  • W. David Strain,

    1. Diabetes and Vascular Medicine, Peninsula NIHR Clinical Research Facility and Institute of Biomedical and Clinical Sciences, Peninsula Medical School, Exeter, Devon, UK
    Search for more papers by this author
  • Angela C. Shore

    1. Diabetes and Vascular Medicine, Peninsula NIHR Clinical Research Facility and Institute of Biomedical and Clinical Sciences, Peninsula Medical School, Exeter, Devon, UK
    Search for more papers by this author

Corresponding author P. E. Gates: Diabetes and Vascular Medicine, Peninsula Medical School (Exeter), Royal Devon and Exeter Hospital (Wonford), Barrack Road, Exeter, Devon EX2 5AX, UK. Email:


Age is a primary risk factor for cardiovascular disease, and this is an increasingly important public health concern because of an increase in the absolute number and proportion of the population at an older age in many countries. A key component of cardiovascular ageing is reduced function of the vascular endothelium, and this probably contributes to the impaired microvessel function observed with ageing in multiple vascular beds. In turn, impaired microvessel function is thought to contribute to the pathophysiology of cardiovascular and metabolic diseases. Here we review evidence that the first signs of altered endothelial and microvessel function can appear in childhood and at all stages of the human lifespan; low-birth-weight babies have reduced endothelial function in skin microvessels at 3 months, and by age 10 years their brachial artery endothelial function is reduced in comparison with normal-birth-weight babies. In overweight/obese adolescent children with clustering of traditional cardiovascular disease risk factors, endothelial function is reduced compared with normal-weight children, and this appears to persist into early adulthood. Adult ageing is associated with impaired microvessel endothelial function and an increase in capillary blood pressure. Biological and lifestyle factors that influence microvessel function include body fat and visceral adiposity, sex hormone status, diet and physical activity. The mechanisms underlying age-associated changes in microvessel function are uncertain but may involve alterations in nitric oxide, prostanoid, endothelium-derived hyperpolarizing factor(s) and endothelin-1 pathways.

Age is a primary risk factor for cardiovascular diseases (Najjar et al. 2005), and with an increase in the absolute number and proportion of the population at an older age in many countries, the mechanisms by which ageing increases cardiovascular risk are an increasingly relevant social and economic concern. A key cardiovascular change observed with advancing age is an alteration in vascular endothelial function, often measured as an impaired endothelium-dependent vasodilatation in response to a physiological or pharmacological stimulus. Endothelial cells play a critical role in maintaining vascular homeostasis (Moncada et al. 1991), and impaired vascular endothelial function is mechanistically involved in the initiation and development of atherosclerosis (Lüscher & Barton, 1997; Vita & Keaney, 2002; Davignon & Ganz, 2004; Deanfield et al. 2007) and is prognostic of future cardiovascular disease (Schachinger et al. 2000; Suwaidi et al. 2000; Halcox et al. 2002). Because of this, impaired vascular endothelial function with advancing age has been widely studied as a putative mechanism by which age acts to increase cardiovascular risk.

Most research on human vascular endothelial function has focused on conduit arteries and resistance vessels, often using a forearm model. This work has revealed a key role of nitric oxide bioavailability (Celermajer et al. 1994; Taddei et al. 1995) and a pro-inflammatory and oxidative stress phenotype (Taddei et al. 2001; Donato et al. 2007) in impaired endothelial function in carefully screened healthy, older people. In the microcirculation, comprising small arterioles, capillaries and venules (Levy et al. 2001; Sernéet al. 2007), the contribution of endothelial function to alterations in microvascular structure and function with advancing age has been the subject of increased interest in recent years. This is partly due to recognition that the microcirculation might be importantly involved in cardiovascular and metabolic diseases and because a large proportion of the endothelium lies within the microcirculation. For example, severity of microalbuminuria, an indirect marker of renal microvascular damage, is associated with the incidence of stroke, myocardial infarction and death due to cardiovascular disease in type II diabetic and non-diabetic men and women (Miettinen et al. 1996). Similarly, retinopathy indicative of microvascular disease is associated with a threefold increase in risk of heart failure and is an independent predictor of heart failure, even in people without pre-existing coronary heart disease, diabetes or hypertension (Wong et al. 2005). Cerebral microvascular disease indicated by the presence of white matter lesions (Pantoni & Garcia, 1997) was associated with increased risk of stroke in the Atherosclerosis Risk In Communities (ARIC) study. This study also showed that if retinopathy was also present, the risk ratio rose from 2.6 to 18.1 (Wong et al. 2002), suggesting that greater risk might be conferred by systemically abnormal microvascular function. As such, surrogate markers of microvascular function might be clinically useful, and the microcirculation of the skin has been widely used for this purpose. Skin microvessel function is correlated with Framingham cardiovascular risk scores (Vuilleumier et al. 2002), is inversely associated with indices of target organ damage (Strain et al. 2005b) and is tightly correlated with microalbuminuria, which is highly predictive of clinical end-points, including stroke, heart failure and renal failure (Strain et al. 2005a). However, the extent to which skin microvessels accurately reflect disturbances in other microvascular regions (e.g. heart, kidney and brain) and the ability of the skin microcirculation to predict cardiovascular end-points remains a point of debate (Holowatz et al. 2008; Minson & Green, 2008).

The functional importance of the microcirculation

The microcirculation provides the interface for the delivery of oxygen and nutrients, the removal of waste products and carbon dioxide, transvascular exchange and tissue fluid economy (Shore, 1996). Organs, tissues and most cells are therefore dependent on adequate perfusion by the microcirculation, and adequate microvessel function is essential for the cardiovascular system to fulfil this fundamental role (Shore, 1996). As such, changes in microvessel function with age may have important consequences for organ function and overall human health. Abnormalities of the microcirculation have been reported in a number of disease states, including hypertension (Tooke et al. 1991; Feihl et al. 2006), insulin resistance (Sernéet al. 2007) and type I and II diabetes (Tooke et al. 1996). The structure and function of the microcirculation have been covered elsewhere in detail (Shore, 1996; Levick, 2003; Sernéet al. 2007).

Assessment of skin microvessel function in humans

The skin is probably the most studied organ in microcirculatory research and represents an easily accessible and non-invasive site for measuring microvessel function. Studies are typically made in an area of skin that has few arterio-venous anastomoses and that is less exposed and less prone to ageing, often the volar aspect of the forearm. An index of skin blood flow is normally obtained using laser Doppler fluximetry. Endothelium-dependent and -independent vasorelaxation can be assessed pharmacologically by iontophoresis of acetylcholine and sodium nitroprusside, respectively. Physiological stimulation of skin blood vessels can be achieved by measuring the maximal skin blood flow response to local heating to 42°C (using a brass disk) or by reactive hyperaemia using a 3 or 4 min cuff occlusion. Capillary function can be assessed by direct cannulation of capillaries in the nail-fold to measure capillary pressure and by capillary count, using video microscopy, at rest and with venous or arterial occlusion to expose non-perfused vessels (capillary recruitment). More recent advances in studying skin microcirculation include microdialysis and wavelet analysis.

Microvessel ageing throughout the human lifespan

Current evidence indicates that a number of factors contribute to alterations in microvascular function from birth to older age. These include genetic factors, developmental factors, age-related biological factors, environmental and lifestyle factors, as well as the interaction of disease processes that probably combine to determine microcirculatory health. A conceptual schematic diagram of these interactions is presented in Fig. 1.

Figure 1.

Simplified diagram showing the interrelationships of some of the factors that influence the structure and function of the microcirculation thoughout the lifespan

Genetic factors Studies of parents and offspring suggest that hypertensive persons who have hypertensive parents inherit attenuated microvessel function. In the offspring of parents with high blood pressure, the maximal skin blood flow response to heating, reactive hyperaemia and capillary recruitment are blunted in those who have high blood pressure compared with those who have low blood pressure (Noon et al. 1997). As such, a genetic predisposition to poor microvessel function may constitute an important part of the pathophysiology of hypertension, and poor microvessel function might be a requisite inherited phenotype for the development of hypertension in these patients. In contrast, offspring of parents with low blood pressure demonstrate no difference in microvessel function between those with high and low blood pressure (Noon et al. 1997).

Ethnic differences in microvascular function also suggest a genetic component to microvascular function. Black men and women of African-Caribbean descent have poorer reactive hyperaemia compared with European counterparts after correcting for conventional risk factors (Strain et al. 2005c). Target organ damage is also greater in African-Caribbeans, and some of this difference is explained by poorer microvessel function (Strain et al. 2005b). The mechanisms underlying inherited differences in microvessel function are uncertain (Lee et al. 2001), but inherent differences in microvascular structure and function may have important implications for appropriate clinical management of disease.

Developmental factors Fetal development may also influence microvessel function. Low-birth-weight babies born at full term have lower maximal hyperaemia at 3 months of age compared with babies of higher birth weight (Goh et al. 2001). This persists into childhood (Martin et al. 2000) and may contribute to increased cardiovascular risk associated with low birth weight (Leon et al. 1998). Impaired endothelial function and increased arterial stiffness (Goodfellow et al. 1998) have also been reported in adults with low birth weight, suggesting that low birth weight is associated with systemically poor micro- and macrovascular function.

Alterations to microvascular function are also observed during childhood development at puberty. A postural challenge induced by lowering the leg from a supine position to below heart level induces vasoconstriction in healthy post- but not prepubertal children. This function is protective of small vessels exposed to a postural challenge and is absent in adult diabetics, resulting in capillary hypertension upon standing (Rayman et al. 1986; Shore et al. 1994). Adaptation during pubertal development may be beneficial because of the greater hydrostatic challenge incurred as a result of the growth spurt and progression towards adult height (Shore et al. 1994). Children with type I diabetes do not fully acquire this microcirculatory adaptation during puberty, even if they are free of diabetic complications (Shore et al. 1994). This has the potential to leave the microcirculation of the feet exposed to high hydrostatic pressures associated with upright posture, perhaps contributing to the pathophysiology of the diabetic foot. Impaired microvessel function in peripubertal children has also been reported in association with cardiovascular risk factor clustering and increased body fat (Khan et al. 2003). Microvessel endothelium-dependent vasodilatation assessed by iontophoresis of acetylcholine was negatively associated with increased body fat and was lowest in children with the highest compared with the lowest quintile of fasting glucose (Khan et al. 2003). Taken together, these findings suggest that microvascular function during development and childhood may have important implications for disease pathophysiology and the manifestation of diseases in adulthood. Importantly, these studies have led to speculation that microvessel endothelial dysfunction precedes some clinically manifest cardiovascular and metabolic diseases and is an early part of the pathophysiology of these diseases.

Adult ageing and the influence of biological and lifestyle factors In a comparison of younger and older normotensive men, the vasodilatory response of skin microvessels to acetylcholine but not to sodium nitroprusside was blunted in older men, indicating impaired endothelium-dependent vasodilatation (James et al. 2006). No differences were observed between older men with and without hypertension (James et al. 2006). Similarly, capillary pressure was higher in older compared with the younger men, but there was no difference between older men with and without hypertension (James et al. 2006). These data suggest that microvessel endothelial function and capillary pressure are subject to ageing independently of changes in brachial artery blood pressure. Excised small arteries from a subgroup of the same cohort showed that normotensive older men had a preserved nitric oxide pathway compared with hypertensive men. This was hypothesized to compensate for diminished prostanoid vasodilator pathways or to offset enhanced vasoconstrictor pathways (James et al. 2006). Microvessel dilatation in response to skin heating is also diminished with age (Kenney et al. 1997). This appears to be mediated at least partly through alterations to nitric oxide signalling, but non-nitric oxide signalling mechanisms were also important (Minson et al. 2002). Capillary rarefaction is commonly reported in younger hypertensives, but in the oldest cohort reported to date, rarefaction was not apparent in older normotensive or hypertensive men (James et al. 2006). This may be because capillary rarefaction is associated with diastolic hypertension, and diastolic blood pressure was similar in all groups in this study. The mechanisms underlying structural alterations and capillary rarefaction are uncertain but may involve altered mechanical/haemodynamic forces and/or impaired angiogenesis, which may in turn be related to impaired vasoregulatory function originating from poor endothelial function (Feihl et al. 2006).

In women, hormone status may impart important influences on microvessel function. Capillary pressure is lower in women compared with men below age 50 years, but there is no difference between the sexes after age 50 years (Shore et al. 1995). However, hormone replacement therapy does not appear to improve microvesel function in healthy postmenopausal women and may be detrimental to the microcirculation in diabetic women (Gooding et al. 2005).

Microvessel function appears to be influenced by a number of biological and lifestyle factors that might interact with the ageing process and modulate ‘normal’ age-associated changes in the microcirculation. Adult ageing is associated with weight gain (Seals & Gates, 2005), and the overweight/obese phenotype is a major challenge for modern healthcare. Capillary recruitment and endothelium-dependent vasodilatation may be impaired in obese compared with lean older women in the basal state and during experimental hyperinsulinaemia (de Jongh et al. 2004b). Blunted vasodilatation by nutritive microvessels in response to insulin may contribute to insulin resistance and glucose disposal and form an important part of the pathophysiology of insulin resistance (de Jongh et al. 2004b) and hypertension (Sernéet al. 2001). Visceral adiposity appears to be particularly detrimental to microvessel function, perhaps mediated by a pro-inflammatory state (de Jongh et al. 2006). Circulating free fatty acid concentrations (de Jongh et al. 2004a) and dietary intake of salt may also importantly influence microvessel function (de Jongh et al. 2007).

Aerobic exercise has been reported to attenuate the age-related decline in resistance vessel endothelial function through a nitric-oxide-dependent mechanism (DeSouza et al. 2000). A similar finding has recently been reported in skin microvessels studied using microdialysis. Microvessel function was worse in older sedentary compared with older active or younger men, and this was attributable to impaired nitric oxide signalling (Black et al. 2008). An exercise training programme in the sedentary older cohort restored microvessel function and improved nitric oxide signalling.

The Framingham risk score combines modifiable and non-modifiable risk factors, including age, to provide a 10 year cardiovascular disease risk probability score. Reactive hyperaemia has been shown to be blunted in women in the highest two quartiles of Framingham risk score (Vuilleumier et al. 2002). Framingham risk score was also associated with impaired skin microvessel endothelium-dependent vasodilatation in healthy men and women (IJzerman et al. 2003). Whether interventions that improve microvessel function or attenuate the age-associated changes in microvessel function reduce cardiovascular risk through improved microcirculatory structure and function has not been determined to date.

Mechanisms of change in microvessel endothelial function throughout the lifespan

The mechanisms underlying endothelial dysfunction have been reviewed comprehensively elsewhere (Feletou & Vanhoutte, 2006). Nitric oxide, prostacyclin and endothelium-derived hyperpolarizing factor(s) are all thought to contribute to microvessel dilatation, and endothelin-1, thromboxane A2 and prostaglandin H2 may contribute to microvessel vasoconstriction (Shore, 1996). A local angiotensin II system on the endothelial cell wall may increase vasoconstriction through endothelin-1, and insulin may stimulate vasodilatation in insulin-sensitive but not insulin-resistant adults (Shore, 1996; Feletou & Vanhoutte, 2006). Superoxide generation with low grade inflammation may also reduce the bioavailability of nitric oxide and impair vasorelaxation (Crimi et al. 2007). The exact contribution and balance between these signalling mechanisms are incompletely defined, and alterations with age and disease remain uncertain.

Capillary rarefaction may be a product of vessel destruction, impaired angiogenesis, impaired vasculogenesis or all of these. Oxidant stress may contribute to capillary rarefaction by inducing endothelial cell apoptosis and/or reducing the nitric oxide needed for vascular budding and stimulation of vascular endothelial growth factor (Feihl et al. 2006).


The microcirculation is the interface between the cardiovascular system and respiring cells, tissues and organs. It appears to contribute independently to target organ damage and is associated or implicated with the pathophysiology of several diseases. Skin microvessel function provides a means of assessing the microcirculation and is altered in many disease states, although interpolation of results from one vascular bed to another must be performed with care. Alterations in both structure and function of the microcirculation occur during ageing, and this may play an important role in the pathophysiology of cardiovascular and metabolic diseases associated with ageing. However, the mechanisms underlying alterations in microvascular function throughout the lifespan are uncertain, and their exact role in the pathophysiology of cardiovascular and metabolic diseases is not fully elucidated.



The authors are grateful to the Wellcome Trust, British Heart Foundation, Diabetes UK and the National Institute for Health Research for Funding this work. The views expressed in this publication are those of the authors and not necessarily those of the funding bodies.