Microcirculation is recognized as a crucial element in many diseases. Nailfold capillaroscopy (NFC), as a non-invasive, highly sensitive, and convenient technique, is designed to assess multiple features of the nailfold capillaries, including blood flow, density, and different kinds of abnormalities. It can provide visual inspection for microcirculation in living body. Since the 1970s, NFC has been used in many systemic disorder researches, like rheumatologic condition (Redisch et al., 1970), diabetes (Chazan et al., 1970), schizophrenia (Pokorny et al., 1970), and so on. Microvascular abnormalities are also recognized as important factors in ocular diseases, like retinal arterial occlusions, retinal vein occlusions glaucoma, non-arteritic anterior ischemic neuropathy, and glaucoma (Flammer et al., 2001). Recently, some studies demonstrated nailfold microcirculation abnormalities were encountered in ocular diseases involved in vascular factors (Pasquale et al., 2015; Chen et al., 2018; Philip et al., 2019). The development of optical coherence tomography angiography (OCTA) has revolutionized the resolution with which the retinal vasculature is visualized, allowing the potential to evaluate retinal microcirculation quantificationally. The previous literatures have showed there is reduction of retinal superficial vascular density measured by OCTA in some ocular diseases, even in systemic diseases (Vadalà et al., 2019). Based on these evidence, we hypothesized that nailfold capillaries, as the terminal capillary networks, were associated with retinal capillaries. To test this hypothesis, using NFC and OCTA, we conducted the first study to examine the relationship among nailfold capillaries, peripapillary retinal nerve fiber layer (RNFL) thickness, and the macular and peripapillary superficial vessel density (VD).

Fifty eyes of 50 healthy subjects aged 18–64 years old were recruited in this study, including 23 males and 27 females (Table 1). Among all the subjects, 26(52%) had nailfold capillary density ≤ 9/mm; 23 (46%) had avascular zones; 17 (34%) had dilated capillaries; and nine (18%) had hemorrhages in the nailfold microcirculation. Compared with subjects whose nailfold density was > 9/mm, subjects with nailfold density ≤ 9/mm had thinner RNFL thickness (p = 0.030), less cpVD of RPCs (p = 0.026), but higher idVD (p = 0.024). Moreover, RNFL thickness and wiVD of RPCs were less in subjects with avascular zones than that in subjects without nailfold avascular zones (p = 0.002, p = 0.014). Finally, superficial pfVD was less in subjects with dilated capillaries than that in subjects without dilated capillaries (p = 0.004) (Table 2 and Figure 4).

Systemic factors, including gender, DBP, and age, were also associated with abnormalities of nailfold microcirculation (Table 3). Compared to subjects without avascular zones, subjects with avascular zones had significantly higher proportion of female (p = 0.042) and lower DBP (p = 0.029). Subjects with nailfold hemorrhages were older than subjects without hemorrhages (p = 0.032).

We controlled for gender, DBP, and age to explore the relationship between nailfold microcirculation and retinal microcirculation. Multivariate regression analyses showed a direct relationship between RNFL thickness and nailfold capillary density (OR = 1.09, 95% CI = 1.001–1.187, p = 0.046). Thinner RNFL thickness and fewer wiVD were correlated with nailfold avascular zones (OR = 0.855, 95% CI = 0.763–0.957, p = 0.007; OR = 0.596, 95% CI = 0.402–0.883, p = 0.010). Fewer superficial pfVD was associated with nailfold dilated capillaries (OR = 0.794, 95% CI = 0.663–0.952, p = 0.012) (Table 4 and Figure 5).

In this study, it was the first time to analyze the difference in retinal microcirculation between subjects with and without one kind of nailfold capillary abnormality. The subjects with less nailfold density had thinner RNFL thickness and fewer cpVD, but possessed higher idVD. We supposed the RPCs in inside disc might compensate for less VD in circumpapillary. The subjects with nailfold avascular zones and dilated capillaries had thinner RNFL thickness and reduced retinal VD. The results are in line with previous study. Gugleta et al. used NFC to confirm subjects with or without vasospastic propensity, and then analyzed the difference in retinal vasodilation after flicker light. The results showed people with vasospastic propensity had lower maximum dilatory amplitude and missed augmentation of the maximum vasodilation, which suggested the diseases involved vasospasm not only in the affected organs, but also in the whole body (Gugleta et al., 2006). Gasser et al. testified, in glaucoma patients, the blood flow velocities in nailfold capillary were associated with the velocities ophthalmic artery (Gasser et al., 1999). For retinal vascular diameter, Branca et al. (2006) did not find the retinal arteriole or retinal venule diameters in subjects with vasospastic propensity were smaller than retinal vessels in subjects without vasospastic propensity. Kochkorov et al. (2006) got the similar results about retinal vascular diameter, whereas they further attested healthy vasospastic subjects had higher retinal spatial irregularity and an increased vessel diameter variation in heartbeat. Here, our results showed subjects with nailfold microcirculation abnormalities had lower retinal VD, further supported the relationship between peripheral and ocular circulation. We did not found any associations among idVD, fVD, and nailfold microcirculation. These two parameters could be lack of clinical significance with less practical application in previous studies. Besides, there was not significant association between retinal microcirculation and nailfold capillary hemorrhages. The reason could be limited sample size.

In analysis of systemic factors for nailfold microcirculation, we found age, gender, and DBP could be involved in nailfold abnormalities. Previous studies have showed nailfold capillary density increased with age of the subjects (Terreri et al., 1999; Hoerth et al., 2012). Here, even though the difference in age of subjects with different levels of naildfold capillary density was not significant, the Spearman rank correlation test showed age was positively correlated with nailfold capillary density (r = 0.324, p = 0.022). In this study, the proportion of hemorrhages of healthy subjects was within the range of other results (Pasquale et al., 2015; Chen et al., 2018). And, 46% of subjects had avascular zones, which was smaller when compared with the other result aimed at Chinese, 60.8% (Chen et al., 2018). The reason could be the subjects in this study were younger than the previous study. Besides, the avascular zone was encountered more often in women than in men and affected by DBP, which was in keeping with previous researches (Andrade et al., 1990; Ruaro et al., 2016). As to dilated capillaries, that we took a stricter definition, which could be a reason that the subjects with dilated capillaries in the present study were more than the results reported in the previous literature (Pasquale et al., 2015; Chen et al., 2018).

After the affecting factors controlled, thicker RNFL thickness was associated with higher nailfold capillary density, and the occurrence of nailfold avascular zones was associated with thinner RNFL and less dense RPCs. Similarly, the appearance of dilated nailfold capillaries was associated with a decline in retinal superficial capillary density. Previous studies also confirmed the relationship indirectly, which demonstrated ocular diseases accompanying with nailfold abnormalities as well as systemic diseases accompanying with retinal abnormalities. Pasquale et al. (2015) have performed NFC on 199 patients with POAG and 124 controls and found dilated capillaries, hemorrhages, and avascular zone were associated with POAG after eliminated age, gender, and other confounding elements. Philip et al. (2019) showed patients with exfoliation glaucoma and normal-tension glaucoma presented lower resting peripheral capillary blood flow when compared with subjects without glaucoma. The study by Chen et al. (2018) demonstrated patients with uveitis had higher tortuosity ratings, lower density, and dilated capillaries in nailfold. Vadalà et al. (2019) showed hypertensive patients with worse renal function had retinal thinning and less capillary density in OCTA. Previous studies also attested the relationship between systemic microcirculation and ocular diseases from other perspectives. Michalska and Ludmiła (2012) found significant rheological changes occurred in glaucoma, like enhanced erythrocyte aggregation and decreased erythrocyte deformability. The previous studies also showed elevated plasma endothelin-1 was associated retinal vein occlusions, anterior ischemic optic neuropathy, and glaucoma (Iannaccone et al., 1998; Flammer et al., 2001; Li et al., 2016). And the higher degree of carotid arterial stenosis contributed to the thinner RNFL thickness (Wang et al., 2017). However, at present, the study of nailfold microcirculation in ocular diseases is limited. Besides, the causal relationship between retinal vascular changes and ocular diseases has not yet been determined. Observing the occurrence and development of diseases from a systemic perspective helps us understand the pathogenesis of diseases. To our knowledge, this is the first time to directly link the manifestations of nailfold microcirculation with RNFL thickness and superficial retinal VD measured by OCTA, which reflect the structure and function of retina, together and further demonstrate the relationship between nailfold microcirculation and retinal microcirculation.

The mechanism of the relationship was still obscure. Anatomically, nailfold capillaries, macular capillaries, and RPCs are all terminal capillaries with commonalities including a single layer of endothelial cell vascular wall. Capillaries in the nailfold and retina are mainly regulated by vascular endotheliocytes and regulatory factors, not by autonomic nerve (Gugleta et al., 2006; Mozaffarieh et al., 2008). When there is minimal vascular autoregulation in these small vessels, hemorheological features play a more significant role in blood flow regulation. RPCs are the terminal arterioles following the path of the retinal nerve fiber (Henkind, 1967; Fouquet et al., 2017). There is also generally a strong correlation between RPCs density and RNFL thickness (Yu et al., 2014). Based on this, the relationship between nailfold microcirculation and RNFL thickness is receivable. For nailfold abnormalities, one of the possible contributing factors is serum matrix metalloproteinase (MMP)-9 levels, an essential substance involved in hydrolyzing extracellular matrix. Higher MMP-9 levels might also degrade the basement membrane of capillaries, deform the vessels, and lead to formation of avascular zone in nailfold capillaries (Lee et al., 2015; Pasquale et al., 2015). Circulating endothelial progenitor cells (EPCs) play a role in ischemia induced angiogenesis (Gurtner et al., 2008). Having fewer EPCs could affect the vascular genesis and contribute to avascular zone formation. It follows that a decreasing number of EPCs could also contribute to a decreased density of superficial retinal capillaries. Nailfold hemorrhages may be related to the presence of super activated platelets (SAPs), which accelerate platelet microparticles formation and create a prothrombotic environment (Morel et al., 2011), as well as decreased EPCs, which have anticoagulation properties (Hristov et al., 2003; Smadja et al., 2008). Besides, the function lessening of endothelium-dependent flow-mediated vasodilation could facilitate the occurrence of abnormal capillary dilation. Subjects with abnormalities in hemorheology, such as increased blood viscosity, erythrocyte aggregation, and slower flow velocity, could suffer more frequent appearance of dilated capillaries.

In this study, we connected nailfold microcirculation with retinal microcirculation in healthy subjects through two commonly used technologies, NFC and OCTA. Our findings are consistent with previous studies and further establish the clinical significance of the nailfold capillary study in ophthalmology investigation. There are some differences between nailfold capillaries and retinal capillaries, so we do not support to replace OCTA completely with NFC. The nailfold capillary is easily affected by external factors, like temperature, and more susceptible to trauma. On the contrary, retinal microcirculation could be affected by IOP, axial length, refractive error, and changes in the thickness and function of RNFL as well as ganglion cells (Yu et al., 2014; Shiihara et al., 2018; Chao et al., 2019). However, the OCTA is more objective. The NFC could provide more information about function of systemic microcirculation at dynamic state. There could be mutual assistance and complementation in these two examinations. Based on the results of this research and previous studies, we believe the nailfold condition could be a biomarker in ocular diseases involved in microcirculation abnormalities. The retinal microcirculation could be a biomarker in systemic microcirculation disorders.

Some limitations should be noted. The sample in the study was limited, a larger population study could provide a more powerful conclusion for relationship between retinal and nailfold microcirculation. As a cross-sectional study, our study did not provide any insight into changes in nailfold or retinal microvascular over time. More parameters, like nailfold capillary blood flow, morphology, and choroid thickness, could be analyzed for enhancing the perception of the relationship. Additionally, according to previous studies, we strictly controlled the influencing factors, like menstruation for women, smoking, alcohol consumption, to get a relatively clear association between nailfold and retinal microcirculation (Gugleta et al., 2006; Kochkorov et al., 2006). A study covering more confounding factors, which is to explore the relationship in subjects with different characteristics and influence, should be performed. Besides, when suffering the changes or stimuli in the external environment, whether the nailfold and retinal microcirculation will respond in a consistent or opposite manner is need to clarify.

From the above, the study showed, in healthy subjects, there was directly relationship between nailfold and retinal microcirculation. The results provide a theoretical foundation for the previous studies showing nailfold microcirculation in ocular diseases with microvascular abnormalities and testified the significance of the conclusions. Besides, at present, the application of NFC in ophthalmic research is limited. We hope the study could inspire more ideas and contribute to pathogenesis understanding in related diseases. Both NFC and OCTA have the potential to identify risk factors. Combined use of them may enhance the understanding of patient’s condition, improve accuracy of the early diagnosis, and contribute to therapy formulation in ocular, even systemic diseases with any microvascular component. A longitudinal study on relationship between nailfold and retinal microcirculation may provide more insight on microcirculation diseases.

All datasets presented in this study are included in the article/supplementary material.

The studies involving human participants were reviewed and approved by the Ethics Committee of Beijing Tongren Hospital. The patients/participants provided their written informed consent to participate in this study.

JT, JS, HW, and NW contributed to study concept and design. JT, YX, YY, and YSu performed the study. JT and YSh drafted the manuscript. JT, KC, and YH performed statistical analysis. CX, YSh, YH, and JO revised the manuscript. NW contributed to administrative, technical, or material support and study supervision. All authors participated in and provided help to the study.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.