Capillaroscopy is conceptually a simple technique, but nevertheless it can provide valuable diagnostic information in the clinical microvascular setting. Indeed, capillaries play a critical role in cardiovascular function being responsible for nutrients and waste products exchange between the tissues and circulation [2]. Understanding their structure and physiology in health and disease is therefore very important.

The physiologic pattern of finger capillaries consists of homogeneous distribution of hairpin-like, parallel capillary loops with a mean length of 200–300 μm and a density of 9–14 capillaries/mm (average 10) [5]. However, capillaroscopic pattern in healthy subjects is characterized by a great inter- and intra-variability of findings [5, 6]. Irregularity in their capillary morphology and lower loop density than in adults have been observed in children, whereas old adults can progressively develop mild, nonspecific morphological changes including tortuosity and microaneurysms [7].

The main clinical application of capillaroscopy is in the rheumatologic field where it is of outstanding importance in patients with Raynaud’s phenomenon and connective tissue diseases as well as in the early diagnosis and monitoring of systemic sclerosis. Systemic sclerosis is a multi-organ disease characterized by tissue fibrosis and immune/microvascular abnormalities. The most specific finding in this pathological condition is the so-called scleroderma pattern, characterized by the presence of dilated capillaries, hemorrhages, avascular areas, and neoangiogenesis [8]. In particular, the presence of giant capillaries and micro-hemorrhages on nail fold capillaries is sufficient to identify the “early” scleroderma pattern, and an increase in these features along with the progressive loss of capillaries (active pattern) is followed by neoangiogenesis, fibrosis, and “desertification” (late pattern) [5]. These different stages reflect the development of the disease processes and correlate with visceral involvement [9, 10]. In addition, capillaroscopy has been suggested to possess prognostic value in evaluation of the risk of developing systemic sclerosis in patients with Raynaud’s phenomenon or digital ulcers in patients with systemic sclerosis [11].

Significant microangiopathy is also often present in patients with dermatomyositis, Sjogren’s syndrome, systemic lupus erythematosus, and undifferentiated connective tissue disease, albeit the changes observed are often not specific [11].

Apart from rheumatologic diseases, changes in nail fold capillary morphology, including microaneurysms, apical dilatations, branching, and hemorrhagic extravasations, may be detected also in diabetes mellitus. The presence of a dilatation at the apex of loops is quite commonly observed in diabetes, but these changes do not seem to be related to disease duration [11].

Nonspecific nail fold microangiopathy has been also identified in several other conditions such as glaucoma, wound healing, including venous ulcer healing and critical limb ischemia [11].

In essential hypertension, many abnormalities are known to occur in the capillary circulation. These include capillary hypertension, increased looping, increased transcapillary filtration, and reduced capillary density [12–15]. Particularly, decreased capillary density, or rarefaction, is a consistent finding in patients with essential hypertension [15–18].

Vascular rarefaction may be defined either as a functional rarefaction, when the vessels are temporarily not perfused or “recruited,” or an anatomical rarefaction, when vessels are actually missing. Many studies have reported microvascular rarefaction in some, but not all vascular beds in hypertensive animals; data available for humans are relatively scanty. A reduction of arteriolar (vessels smaller than 100 μm) and capillary number in skeletal muscle and other vascular beds of spontaneously hypertensive rats has been observed, together with a rarefaction of small vessels in the cremasteric muscle of renal hypertensive rats [19, 20]. A reduction of arteriolar and capillary density in conjunctival microcirculation of hypertensive patients has been also detected by direct visualization in vivo [14]. Similarly, a 20 % reduction of capillary density in the nail fold capillaries using capillary microscopy was observed [16].

Using the same technique, Antonios and other groups [15, 21–23] have demonstrated the presence of capillary rarefaction in the skin of fingers of patients with essential hypertension or borderline hypertension and also of normotensive offspring of patients with essential hypertension, suggesting that structural rarefaction seems to be due to a primary abnormality that antedates the onset of sustained hypertension, rather than being secondary to the elevation of blood pressure.

From a pathophysiological point of view, microvascular rarefaction increases peripheral vascular resistance, thereby increasing blood pressure and aggravating blood pressure-related target organ damage [24]. Indeed, not only the diameter of individual vessels but also the absolute number of perfused vessels contributes to total vascular resistance. A rarefaction of about 42 % of third order arterioles would increase tissue flow resistance by 21 % [25]. Moreover, a reduction in the microvascular network may decrease tissue perfusion [26] inducing a nonuniform distribution of microvascular flow among exchange vessels [25] and alterations of skeletal muscle perfusion and metabolism (i.e., oxygen uptake and insulin-mediated glucose uptake) [27].

In the last decades, direct intravital videomicroscopy has been widely used to investigate whether capillary rarefaction in essential hypertension is caused by a structural (anatomic) absence of capillaries or by functional non-perfusion secondary to severe vasoconstriction upstream [15]. As the visualization of capillaries, without using specific dyes, depends on the presence of red blood cells inside, standard capillaroscopy cannot directly show capillaries that are not perfused at resting conditions. Indeed, different methods have been assessed, such as venous congestion and postocclusive reactive hyperemia, in order to maximize the number of visible perfused capillaries during intravital capillary microscopy [15, 17, 28, 29].

In venous congestion, a miniature blood pressure cuff is applied to the base of the fourth finger of the nondominant hand, and the cuff is then inflated and maintained at 60 mmHg for 2 min. The consequent increase in venous back pressure allows to passively open up, and therefore to visualize, non-perfused and intermittently perfused capillaries. In reactive hyperemia, arterial blood flow in the forearm and hand is stopped for 3 min by inflating a sphygmomanometer cuff applied to the upper arm at 200 mmHg. The cuff is then deflated abruptly, and subsequent capillaroscopic images are obtained continuously usually for 15 s (up to 2 min). Differently from venous congestion, reactive hyperemia induces a vasodilator response possibly mediated by both myogenic and/or local chemical factors [30, 31]. Therefore, capillary density during venous congestion depends mainly on the anatomic number of capillaries [15, 29], whereas postocclusive reactive hyperemia may detect functional recruitment of initially non-perfused capillaries (microvascular reactivity) [28, 29, 32], thus reflecting both functional and structural factors [33]. In particular, it has been suggested [15, 23] that venous congestion allows to visualize a greater number of capillaries compared to postocclusive reactive hyperemia, making this technique probably the best method to maximize capillary number [29].

A reduction of capillary density at rest and after venous congestion [15] has been shown in hypertensive patients suggesting that much of the reduction in capillary density in these patients is due to the structural (anatomical) absence of capillaries, rather than to functional non-perfusion. However, no differences in capillary density at rest between normotensive and hypertensive subjects have been observed in other studies [17, 28]. In addition, hypertensive patients, as compared to normotensive subjects, show a decreased capillary recruitment after arterial occlusion [28] albeit this is not confirmed by all the studies [17]. Discrepancy of findings from different studies may be possibly due by the different definitions of capillary density in the resting condition and during postocclusive reactive hyperemia (capillary recruitment) depending on different periods for acquisition of images used in these studies [33].

Importantly, structural capillary rarefaction in the nail fold may suggest a generalized microvascular abnormality that may play a role in the pathogenesis of cardiovascular diseases. Indeed, patients with anginal chest pain and normal coronary arteriograms have significantly lower skin capillary density both at baseline and after maximization with venous congestion than matched healthy volunteers, independently on the presence of hypertension [34]. Accordingly, Pedrinelli and other groups [35–38] have demonstrated a significant higher minimal forearm vascular resistance during maximal postischemic vasodilation in patients with syndrome X compared to controls, indicating structural vasodilatory abnormalities. However, the pathophysiological importance of capillary rarefaction in these patients remains still unknown since a similar reduction of capillary density is present also in asymptomatic hypertensive patients.

Interestingly, also obese subjects show structural and functional alterations in skin microcirculation that are proportional to the degree of global and central obesity [39]. In addition, in obese subjects with metabolic syndrome, the cutaneous capillaries at rest are already maximally recruited, indicating an absence of functional capillary reserve. This may be related to the insulin resistance observed in obese individuals with metabolic syndrome [39]. Recently, the presence of microvascular rarefaction has been also detected in obese patients with or without concomitant increase of blood pressure values [40].