Daily dietary salt intake exceeds recommended maximum levels,  and has been linked to hypertension and increased cardiovascular risk . However, not everyone responds to high-salt intake with a blood pressure (BP) increase . The trait characterized by a BP increase following high-salt intake is known as salt-sensitivity, as opposed to salt-resistance . The pathophysiology of salt-sensitivity remains to be elucidated. Until recently it was thought that in salt-sensitive individuals hypertension is caused via renal salt retention, leading to an increase in extracellular fluid volume (ECFV) and cardiac output (CO). Contrarily, recent studies have demonstrated that an increase in CO is also seen in salt-resistant individuals, but that compensatory reduction in systemic vascular resistance (SVR) is impaired in salt-sensitive individuals [5,6]. Nitric oxide (NO)-mediated pathways might explain this differential response in SVR, as activity of this endogenous vasodilator is generally decreased after high-salt intake, but is more extensively reduced in salt-sensitive individuals [7,8]. Theoretically this NO response could lead to a decrease in microvascular density that is demonstrated both in hypertensive patients [9–12] and following high-salt intake in both hypertensive and normotensive individuals [13–16]. This decrease in microvascular density can either be structural (anatomical absence of vessels) or functional (vessels are anatomically present yet not perfused), or both . As reduction of microvascular density increases SVR , this phenomenon could be the missing link in understanding the mechanism of salt-sensitivity. It is yet unknown whether high-salt intake causes reduction of microvascular density in normotensive patients, and therefore the causal relationship with hypertension remains unclear. An alternative explanation for the differential response in SVR could involve recent findings of sodium compartmentalization. Sodium can be stored in various tissues without commensurate water retention and subsequent expansion of the ECFV . Whereas salt-resistant individuals have this capacity for nonosmotic sodium storage, this seems perturbed in those who are salt-sensitive , resulting in weight gain and high SVR following salt loading via mechanisms not well understood . So far, it has not been studied whether ECFV changes and microvascular alterations are related.
Improvements of imaging modalities grant the opportunity to have a closer look into the relationship between high-salt intake and microcirculatory changes. The aim of this study was to determine whether high-salt intake causes reduction of sublingual microvascular density in normotensive individuals. Though not the primary aim of this study, in light of newly emerged evidence regarding nonosmotic sodium storage, we also assessed the effect of sodium-induced body weight changes on the relationship between salt intake and microvascular density.
We studied healthy male volunteers between 18 and 40 years old in an open label randomized crossover trial. The study was performed at the Academic Medical Center Amsterdam between October 2016 and April 2017 according to the principles of the Declaration of Helsinki . Eligibility and exclusion criteria are presented in the Supplemental material (S1), http://links.lww.com/HJH/B43. Volunteers were recruited via local advertisement and provided informed written consent. The protocol was approved by the local ethics committee and registered at the Netherlands trial registry (NTR4785; http://www.trialregister.nl/trialreg/admin/rctview.asp?TC=4785).
All participants were asked to subsequently adhere to a high-salt diet (HSD) (>12 g/day) and low-salt diet (LSD) (<3 g/day) for 14 days each in randomized order. Randomization was performed by the research coordinator via sealed, opaque envelopes in blocks of four after assessment of eligibility and signing informed consent. There was no washout period between diets. Dietary compliance was verified at days 7 and 11 with collection of 24-h urine. On day 15 after an overnight fast participants visited our research department for measurement of microvascular density, hemodynamic parameters and laboratory testing.
The primary aim of this study was to assess a difference in sublingual microvascular density after a HSD compared with a LSD, with participants serving as their own controls. Microvascular densities were measured sublingually with sidestream dark-field (SDF) videomicroscopy (Microscan; Microvision Medical B.V. Amsterdam, The Netherlands) that captures hemoglobin (Hb) in passing red blood cells with green light-emitting diodes (540 nm). Therefore, vessels filled with red blood cells are captured, but nonperfused vessels are not. To visualize and maximize the residual capacity of the sublingual microcirculation (structural vessel density), vasodilation was induced via one dose of 0.4-mg sublingual nitroglycerin (NTG) .
Videos were assessed for sufficient quality (Supplemental methods S3, http://links.lww.com/HJH/B43) . Video-image analysis was performed with the operator blinded for the characteristics of the participants using a semiautomated analysis program automated vessel analysis (AVA) 3.2 and in concordance with the 2007 consensus statement . With AVA 3.2 densities are measured as vessel-length per surface (mm/mm2) for vessels with diameters less than 20 μm (TVDsmall) and all vessels with diameters less than 150 μm (TVDallvessels). Microvessels with diameters less than 20 μm are mostly capillaries, therefore we used TVDsmall to answer our hypothesis. TVDallvessels mostly consists of venules and is considered a quality check . The video analyst ranked the flow of erythrocytes through the vessels from no flow to continuous flow. The proportion of vessels with flow is expressed as PPVsmall (% of vessels with diameters <20 μm with flow) and PPVallvessels (% of all vessels with flow). This proportion is multiplied with TVD to assess density of vessels with flow: perfused vessel density (PVD)small (density of vessels with diameters <20 μm with flow) and PVDall vessels (density of all vessels with flow). Finally microvascular flow index was measured for all vessels and for small vessels, in which flow is ranked per video quadrant. More detailed methods regarding SDF analysis are provided in the Supplement (S3), http://links.lww.com/HJH/B43.
Hemodynamic parameters were measured after both diets. Seated SBP, DBP and heart rate (HR) (mean of the last two measurements) were done after 5-min rest at the nondominant arm with an automated device (Omron M4 oscillometric device; OMRON Healthcare Europe B.V., Hoofddorp, The Netherlands). On day 14 of both diets 24-h ambulatory SBP and DBP, mean arterial pressure and HR were recorded at 15-min daytime intervals and 30-min night-time intervals (Mobil-O-Graph 24 h PWA Monitor; I.E.M. GmbH, Stolberg, Germany). The cuff was secured at the nondominant arm. When arm circumference exceeded 32 cm a large cuff was used.
Laboratory testing included plasma sodium, potassium and creatinine, and analysis of 24-h urinary sodium, potassium and creatinine levels. All biochemical tests were performed on a COBAS C8000 Modular Analyzer (Roche Diagnostics GmbH, Mannheim, Germany).
Participants were stratified to groups with a large and small salt-induced increase in weight, to assess the effect of sodium-induced body weight changes on the relationship between salt intake and microvascular density.
Sample size calculations were based on results of a previously conducted pilot study, showing that 18 participants were needed to detect a mean difference of 8.4% (SD of 12.0) in microvascular density measured with SDF-videomicroscopy, with 80% power using a two-sided paired t test at the 0.05 significance level. Continuous variables are reported as mean and SD, or as median and interquartile range if the data were not normally distributed. We checked for period and carry-over effects (Supplemental methods S4, http://links.lww.com/HJH/B43) .
To compare the outcomes between diets, paired t tests or Wilcoxon signed ranks tests were used. Pearson correlation was used to test the correlation between the change in sodium excretion and the change in microvascular density between diets. Participants were stratified by median-split for amount of weight change. One-sample t tests or Wilcoxon rank sum tests were used to compare salt-induced changes in microvascular density between groups. As SDF imaging captures Hb, we suspected that salt-induced differences in hematocrit levels could influence our results. Therefore, we repeated these tests with a correction for hematocrit levels in a linear mixed model. A two-tailed P less than 0.05 was considered statistically significant. Statistical analyses were done in Rstudio (Version 1.0.136; RStudio, Inc., Boston, Massachusetts, USA).
Between October 2016 and April 2017 we included 18 individuals; 10 of whom were randomized to start with the HSD, and eight to start with the LSD. The mean (SD) age was 29 (5) years, and baseline BP was 118 (8)/73 (5) mmHg. Baseline characteristics are shown in Table 1.
Mean 24-h urinary sodium and creatinine excretion indicated overall compliance to both diets and complete urine sampling. Salt excretion was 15.2 g/day following high-salt intake and 2.3 g/day after low-salt intake (Table 2). An unintended effect of salt intake reduction was decreased 24-h potassium excretion (P = 0.02; Table 2). The difference in weight between diets was 1.8 (0.9) kg (P < 0.0001).
No differences in BP were observed (Table 2), irrespective of diet sequence. We found no differences in microvascular density between diets (Supplemental file S5, http://links.lww.com/HJH/B43). Mean microvascular density following NTG was slightly higher following the LSD, but the difference between diets was NS (P = 0.31, S5, http://links.lww.com/HJH/B43). No period or carry-over effects were detected. Adjustment for plasma hematocrit did not affect these results (data not shown).
There was no correlation between the increase in 24 h urine sodium excretion (i.e. salt exposure) and the change in microvascular density before administration of NTG (Fig. 1a and b), but there was a significant correlation between the increase of 24-h urine sodium excretion and the decrease in microvascular density following NTG administration (Fig. 1c and d).
When participants with a small change in weight were compared with participants with a larger change in weight, the change in microvascular density (TVDsmall) before administration of NTG differed significantly between groups (Δ1.63 (1.43), P = 0.031, Fig. 2a and b). Whereas we found an increase in microvascular density following the HSD in the group with a small salt-induced change in weight (P = 0.008), a decrease in microvascular density was present in the group with a larger salt-induced change in weight (P = 0.054). In line with these results, the weight change of all participants demonstrated significant correlation with the change in microvascular density (TVDsmall; Fig. 3). There was no correlation between weight change and the change in microvascular density following NTG (Fig. 3). Baseline characteristics, 24-h urine excretion and microvascular density following NTG did not differ significantly between participants with a small change in weight and those with larger change in weight (Table 3).
In this study, we aimed to obtain insight into the effect of salt intake on the sublingual microcirculation in healthy men. Overall no microvascular differences were found between the LSD and HSD. However, the increase in salt consumption from the LSD to the HSD significantly correlated with a lower recruitment rate of sublingual capillaries after administration of NTG, which indicates lower structural microvascular density. In individuals with larger body weight increase following high-salt intake, we observe significantly lower rates of perfused capillaries reflecting impaired functional microvascular density. As both phenomena occur independently of BP effects, our study indicates that high-salt load as such contributes to microvascular dysfunction, which is generally considered as an early feature of end-organ damage in various cardiovascular risk patients, including hypertensive patients.
To our knowledge this is the first study that has investigated the effect of salt on microvascular densities in normotensive patients using in-vivo sublingual imaging. With the use of other in-vivo techniques and in different microvascular tissues, previous studies have demonstrated reduction of capillary density in hypertensive [9,10,24], borderline hypertensive individuals [25,26] and offspring of hypertensive individuals [27,28], suggesting that reduction of microvascular density is an early feature of increased BP. However, these studies did not take sodium intake nor sodium-sensitivity into account as possible contributors to the BP associated microvascular changes. Other studies that did explore the combined effects of sodium, BP and the microcirculation, have shown that sodium-sensitive individuals with borderline hypertension or normal BP had significantly lower capillary density in the conjunctival microvasculature  and demonstrate an inverse association of nailfold capillary recruitment and the sodium-sensitive BP response among hypertensive and normotensive individuals . As these studies had a cross-sectional design and did not report dietary sodium status at time of measurements, direct effects of sodium reduction and associated BP response on the microcirculation were not evaluated. So far, data on effects of sodium reduction on microvascular networks are therefore currently available in untreated hypertensive individuals, in either the conjunctival vascular bed  or skin capillaries . In contrast to these studies we could not demonstrate salt-induced microvascular changes when comparing LSD vs. HSD in a cross-over fashion. Yet, there was a significant correlation between the increase in salt intake and decrease in microvascular density following administration of NTG. This suggests that salt exposure has negative effects on the recruitment of sublingual microvasculature, as they move together in a linear fashion. Our results are in line with studies in the cremaster muscle of rats in which high-salt intake led to decrease in microvascular densities [29,30]. However, we were unable to detect differences in microvascular density between diets. The absence of BP effects might be explanatory for our observation that there was no difference in microvascular density when comparing the HSD and LSD. He et al. have demonstrated an increase in both functional and structural microvascular density following sodium reduction among hypertensive patients. This may be related to the fact that He et al. reported an increase in microvascular density with a decrease in BP, while we did not find a change in BP.
The data of our stratified analysis are of interest in light of the recently rediscovered concept of nonosmotic sodium storage and the effect of salt on microvascular density. In the current study, we observed that the amount of weight change between diets differed substantially between participants that, considering the duration of the salt intervention periods, can be attributed to changes in ECFV. Although there was a difference in salt-induced fluid expansion between individuals, there was no difference in salt excretion. This is in line with previous observations of Laffer et al. who demonstrated that with the same amount of total body sodium, salt-sensitive individuals had an increase in body weight, whereas salt-resistant individuals did not. Our observations with a wide range of weight change for similar salt levels challenge traditional beliefs that salt retention induces iso-osmolar water retention. Via 23Na MRI it was shown that high amounts of nonosmotic sodium storage in striated muscle were associated with hypertension . This was further substantiated in studies in mice in which disruption of salt efflux from the nonosmotic storage compartment led to high BP . One may hypothesize that saturation of the nonosmotic storage compartment leads to a subsequent smaller capacity for nonosmotic storage of added salt. The following increase in BP may be related to microcirculatory alterations.
We did not observe a difference in BP response and therefore we can only assume that those who showed a larger salt-induced fluid expansion might become more salt-sensitive in terms of BP response at an older age. Furthermore, after administration of NTG microvascular densities in the group with large salt-induced weight gain increased to similar levels of vessel densities measured in the group with small weight changes, dissolving the correlation between weight change and vessel densities. This suggests a relatively larger response to exogenous NO in the group with a larger salt-induced fluid expansion, indicating a decrease of NO activity in those individuals. These results are in line with studies among salt-sensitives. Schmidlin et al. demonstrated an increase in asymmetric dimethylarginine (ADMA), a NO inhibitor, in salt-sensitives but not in those salt-resistant. Another study rendered similar results measuring plasma NOx (NO metabolites nitrate and nitrite) concentrations . Our results add to their findings and show that NO activity might play a role in the interaction between microvascular changes and a smaller capacity for nonosmotic sodium storage.
Our study has some limitations. First, we only assessed sublingual microcirculatory parameters. It remains uncertain if our findings can be extrapolated to other microvascular beds. Also, though SDF imaging obtains high-resolution images, incident dark field imaging (IDF) is considered to provide images with higher resolution . However, images were graded for sufficient quality, and discarded if necessary. Also, the fact that administering NTG led to significant increase in vessel density suggests that our image quality was sufficient for detecting differences between groups. Finally, our analyses generated similar results in comparison to a study using IDF imaging that also used NTG . Another limitation might be that we did not precisely measure the ECFV, but used an indirect measurement (i.e. body weight). Given the short-time frame of our intervention, most of the differences in body weight are likely to be attributable to changes in body water, also because 24-h creatinine excretion levels, reflecting muscle mass, remained similar. Also, we studied healthy male individuals to exclude the influence of menstrual cycle related hormonal changes, and therefore one might question whether our results are applicable to other patient categories. Finally, we found that potassium excretion was also significantly different between both intervention periods. Considering the effects of potassium intake on BP, this may explain why no BP effects were observed. Yet, potassium excretion levels were above WHO recommendations of 90 mmol/day in both interventions, and no associations between potassium excretion and BP or sodium-to-potassium ratio and BP were seen.
We show that increment of dietary salt intake is associated with a reduction in sublingual microvascular density following administration of NTG among healthy men. Our results suggest that the ability of our healthy male volunteers to maintain sufficient microvascular density is vital in maintaining normal BP. We demonstrate that in vivo a decrease of sublingual microvascular density is present in individuals with a larger change in body weight which may reflect smaller capacity for nonosmotic sodium storage. Furthermore, when exogenous NO in the form of NTG was administered, no change in microvascular density was present, suggesting that impaired activity of NO may play a role in the link between nonosmotic sodium storage and microvascular alterations. This also may imply that healthy individuals may benefit from dietary sodium reduction before hypertension or microvascular end-organ damage becomes apparent. More research is needed to further assess underlying pathophysiological mechanisms and longitudinal consequences of the interaction between the microcirculation and salt intake.
The authors thank the volunteers for participation in this study. The assistance of Microvision Medical B.V. for providing materials (SDF videomicroscope and disposables) and advice regarding image analysis is gratefully acknowledged. The authors acknowledge the efforts of Yasin Ince (Department of Translational Physiology, Amsterdam UMC, Academic Medical Center, Amsterdam) who helped with image analysis.
Conflicts of interest
There are no conflicts of interest.
1. Brown IJ, Tzoulaki I, Candeias V, Elliott P. Salt intakes around the world: implications for public health. Int J Epidemiol
2. Mente A, O’Donnell MJ, Rangarajan S, McQueen MJ, Poirier P, Wielgosz A, et al. Association of urinary sodium
and potassium excretion with blood pressure
. N Engl J Med
3. Elijovich F, Weinberger MH, Anderson CA, Appel LJ, Bursztyn M, Cook NR, et al. Salt sensitivity
of blood pressure
: a scientific statement from the American Heart Association. Hypertension
4. Weinberger MH, Fineberg N S, Fineberg S E, Weinberger M. Salt sensitivity
, pulse pressure, and death in normal and hypertensive humans. Hypertension
2001; 37 (2 Pt 2):429–432.
5. Schmidlin O, Sebastian AF, Morris RC Jr. What initiates the pressor effect of salt in salt-sensitive humans? Observations in normotensive blacks. Hypertension
6. Laffer CL, Scott RC 3rd, Titze JM, Luft FC, Elijovich F. Hemodynamics and salt-and-water balance link sodium
storage and vascular dysfunction in salt-sensitive subjects. Hypertension
7. Fujiwara N, Osanai T, Kamada T, Katoh T, Takahashi K, Okumura K. Study on the relationship between plasma nitrite and nitrate level and salt sensitivity
in human hypertension
: modulation of nitric oxide synthesis by salt intake. Circulation
8. Schmidlin O, Forman A, Leone A, Sebastian A, Morris RC Jr. Salt sensitivity
in blacks: evidence that the initial pressor effect of NaCl involves inhibition of vasodilatation by asymmetrical dimethylarginine. Hypertension
9. Serne EH, Gans RO, ter Maaten JC, Tangelder GJ, Donker AJ, Stehouwer CD. Impaired skin capillary recruitment in essential hypertension
is caused by both functional and structural capillary rarefaction. Hypertension
10. Antonios TF, Singer DR, Markandu ND, Mortimer PS, MacGregor GA. Structural skin capillary rarefaction in essential hypertension
11. Prasad A, Dunnill GS, Mortimer PS, MacGregor GA. Capillary rarefaction in the forearm skin in essential hypertension
. J Hypertens
12. Kanoore Edul VS, Ince C, Estenssoro E, Ferrara G, Arzani Y, Salvatori C, et al. The effects of arterial hypertension
and age on the sublingual microcirculation
of healthy volunteers and outpatients with cardiovascular risk factors. Microcirculation
13. He FJ, Marciniak M, Markandu ND, Antonios TF, MacGregor GA. Effect of modest salt reduction on skin capillary rarefaction in white, black, and Asian individuals with mild hypertension
14. Sullivan JM, Prewitt RL, Ratts TE, Josephs JA, Connor MJ. Hemodynamic characteristics of sodium
-sensitive human subjects. Hypertension
15. de Jongh RT, Serne EH, RG IJ, Stehouwer CD. Microvascular function: a potential link between salt sensitivity
, insulin resistance and hypertension
. J Hypertens
16. Houben AJ, Willemsen RT, van de Ven H, de Leeuw PW. Microvascular adaptation to changes in dietary sodium
is disturbed in patients with essential hypertension
. J Hypertens
17. Greene AS, Tonellato PJ, Lui J, Lombard JH, Cowley AW Jr. Microvascular rarefaction and tissue vascular resistance in hypertension
. Am J Physiol
1989; 256 (1 Pt 2):H126–H131.
18. Machnik A, Neuhofer W, Jantsch J, Dahlmann A, Tammela T, Machura K, et al. Macrophages regulate salt-dependent volume and blood pressure
by a vascular endothelial growth factor-C-dependent buffering mechanism. Nat Med
19. World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA
20. Hilty MP, Pichler J, Ergin B, Hefti U, Merz TM, Ince C, et al. Assessment of endothelial cell function and physiological microcirculatory reserve by video microscopy using a topical acetylcholine and nitroglycerin challenge. Intensive Care Med Exp
21. Massey MJ, Larochelle E, Najarro G, Karmacharla A, Arnold R, Trzeciak S, et al. The microcirculation
image quality score: development and preliminary evaluation of a proposed approach to grading quality of image acquisition for bedside videomicroscopy. J Crit Care
22. De Backer D, Hollenberg S, Boerma C, Goedhart P, Buchele G, Ospina-Tascon G, et al. How to evaluate the microcirculation
: report of a round table conference. Crit Care
23. Hills M, Armitage P. The two-period cross-over clinical trial. Br J Clin Pharmacol
24. Gasser P, Buhler FR. Nailfold microcirculation
in normotensive and essential hypertensive subjects, as assessed by video-microscopy. J Hypertens
25. Antonios TF, Singer DR, Markandu ND, Mortimer PS, MacGregor GA. Rarefaction of skin capillaries in borderline essential hypertension
suggests an early structural abnormality. Hypertension
1999; 34 (4 Pt 1):655–658.
26. Sullivan JM, Prewitt RL, Josephs JA. Attenuation of the microcirculation
in young patients with high-output borderline hypertension
27. Antonios TF, Rattray FM, Singer DR, Markandu ND, Mortimer PS, MacGregor GA. Rarefaction of skin capillaries in normotensive offspring of individuals with essential hypertension
28. Noon JP, Walker BR, Webb DJ, Shore AC, Holton DW, Edwards HV, et al. Impaired microvascular dilatation and capillary rarefaction in young adults with a predisposition to high blood pressure
. J Clin Invest
29. Greene AS, Lombard JH, Cowley AW Jr, Hansen-Smith FM. Microvessel changes in hypertension
measured by Griffonia simplicifolia I lectin. Hypertension
1990; 15 (6 Pt 2):779–783.
30. Hansen-Smith FM, Morris LW, Greene AS, Lombard JH. Rapid microvessel rarefaction with elevated salt intake and reduced renal mass hypertension
in rats. Circ Res
31. Kopp C, Linz P, Dahlmann A, Hammon M, Jantsch J, Muller DN, et al. 23Na magnetic resonance imaging-determined tissue sodium
in healthy subjects and hypertensive patients. Hypertension
32. Gilbert-Kawai E, Coppel J, Bountziouka V, Ince C, Martin D. A comparison of the quality of image acquisition between the incident dark field and sidestream dark field video-microscopes. BMC Med Imaging
* Nienke M.G. Rorije and Emma Rademaker contributed equally to the article.