Full text of DOI:10.1007/s004170100333

Graefe’s Arch Clin Exp Ophthalmol
(2001) 239:673–677
C L I N I C A L I N V E S T I G AT I O N
DOI 10.1007/s004170100333
Gabriele Fuchsjäger-Mayrl
Kaija Polak
Alexandra Luksch
Elzbieta Polska
Retinal blood flow and systemic
blood pressure in healthy young subjects
Guido T. Dorner
Georg Rainer
Hans-Georg Eichler
Leopold Schmetterer
Abstract Background: The aim of
the present study was to investigate
the association between systemic
blood pressure and retinal blood
flow in healthy young subjects.
Methods: Three independent study
cohorts were included. A cross-
sectional study was performed in
420 young male subjects with
systolic blood pressure ≤160 mmHg
and diastolic blood pressure
≤100 mmHg. Retinal white blood
cell flux (n=210) and blood velocity
in the central retinal artery (n=210)
were measured. In addition, a longi-
tudinal study was performed in
Blood flow velocity in the central
retinal artery was measured by
Received: 27 February 2001
Revised: 6 June 2001
Accepted: 13 June 2001
Published online: 31 July 2001
© Springer-Verlag 2001
means of colour Doppler imaging.
Results: Retinal white blood cell flux
(r=0.262; P<0.001) and mean flow
velocity in the central retinal artery
(r=0.174, P=0.010) were significant-
ly associated with mean arterial pres-
sure in the cross-sectional study. In
the longitudinal study retinal white
blood cell flux and mean flow veloc-
ity in the central retinal artery were
also correlated with systemic blood
pressure. Conclusions: Our data in-
dicate a slight but significant in-
crease in retinal blood flow with
Proprietary interests: none
G. Fuchsjäger-Mayrl · K. Polak
A. Luksch · E. Polska · G.T. Dorner
H.-G. Eichler · L. Schmetterer (
)
✉
Department of Clinical Pharmacology,
Währinger Gürtel 18–20,
1090 Vienna, Austria
e-mail: Leopold.Schmetterer@univie.ac.at
Tel.: +43-1-404002981
Fax: +43-1-404002998
40 young male subjects in whom ret- blood pressure. Whether this is of
G.T. Dorner · G. Rainer
Department of Ophthalmology,
Vienna, Austria
inal and systemic haemodynamic
parameters were measured thrice
within 6 weeks. Retinal white blood
cell flux was measured with the
blue-field entoptic technique.
clinical relevance in eye diseases
with altered retinal perfusion, such
as diabetic retinopathy, remains to be
established.
L. Schmetterer
Institute of Medical Physics,
Vienna, Austria
flow on systemic blood pressure in young subjects with
normotension or mild untreated hypertension. Two meth-
ods were used to assess retinal perfusion in these volun-
teers. The blue-field entoptic technique was employed to
investigate leukocyte movement in perifoveal retinal
capillaries. In a separate study population, blood flow
velocities in the central retinal artery were measured
with colour Doppler imaging.
Introduction
It is well established that the retina has some autoregula-
tory capacity in response to acute changes in perfusion
pressure [2, 11, 14, 15]. However, in its strict sense reti-
nal autoregulation cannot be investigated in human clini-
cal trials, because changes in ocular perfusion pressure
cannot be induced without concomitant changes in other
vascular beds. In addition, it is currently unknown
whether long-term changes in ocular perfusion pressure,
as induced for example by systemic hypertension, may
induce alterations in retinal blood flow.
Materials and methods
Subjects
Hence, the relation between retinal blood flow and
systemic blood pressure is not well established. We
The study protocols were approved by the local ethics committee
therefore investigated the dependence of retinal blood and conformed with the ethical standards laid down in the declara-

674
tion of Helsinki. Three independent study cohorts were included. accurate. Subjects who did not reach a SD lower than 20% at any
Four hundred and twenty male non-smoking volunteers between matching trial during the study (protocol B) were excluded from
19 and 40 years of age were studied in protocol A. Retinal white the trial.
blood cell flux (n=210) and blood velocity in the central retinal ar-
Mean blood flow velocity (MFV), peak systolic flow velocity
tery (n=2 10) were measured. Forty male non-smoking volunteers (PSV), and end-diastolic flow velocity (EDV) were determined in
between 19 and 40 years of age participated in protocol B. All the right central retinal artery (CRA) with colour Doppler imaging
subjects had been drug free for at least 3 weeks at the time of as described in detail elsewhere [7]. MFV was measured manually
study and did not use any regular medication. The nature of the as time mean of the spectral outline. Measurements were per-
study was explained, and all subjects gave written consent to par- formed with a 7.5-MHz probe (CFM 750, Vingmed Sound, Hor-
ticipate. Each subject underwent screening that included medical ten, Norway). The resistive index (RI) in the CRA was calculated
history and physical examination, 12-lead electrocardiography and as RI=(PSV-EDV)/(PSV). All parameters were determined as
blood pressure measurement. Inclusion criteria were normal find- mean values over at least three cardiac cycles.
ings in the screening, systolic blood pressure ≤160 mmHg and dia-
stolic blood pressure ≤100 mmHg. Furthermore an ophthalmic ex-
amination, including slit-lamp biomicroscopy and indirect fundus- Data analysis
copy, was performed. Inclusion criteria were normal ophthalmic
findings and a refractive error of less than 3 dioptres in each eye. All statistical analyses were done using the Statistica software
In all subjects the right eye was studied.
package (Release 4.5, StatSoft, Tulsa, Okla., USA). The associa-
tion between ocular and systemic haemodynamic parameters as
obtained in protocol A was investigated using linear correlation
analysis. The three values obtained for each subject in protocol B
were averaged. A deviation from this mean was then calculated
for each value. For example the individual deviation in mean arte-
rial pressure DMAPi was calculated as MAPi–MAPmean. The asso-
ciation between these deviations from the mean were investigated
using linear correlation analysis. Data are presented as means
± SD. A two-tailed P <0.05 was considered the level of significance.
Study design
Protocol A
In 210 subjects retinal white blood cell flux (WBCF) was investi-
gated with the blue-field entoptic technique. In 210 other subjects
blood flow velocities in the central retinal artery were measured.
All subjects were studied after a resting period of at least 20 min.
Stable haemodynamic conditions were ensured by repeated blood
pressure measurement. The investigators who performed the ocular
haemodynamic measurements were not informed about the blood
pressure and pulse rate values obtained in the subjects under study.
Results
Protocol A
Protocol B
Subjects’ characteristics are shown in Table 1. Data are
presented separately for the subjects participating in the
blue-field measurements and for the subjects participat-
ing in the colour Doppler experiments. The age of the
subjects was comparable between the two groups.
In 40 subjects, who did not participate in protocol A, retinal
WBCF and blood flow velocities in the central retinal artery were
measured on three different days within a period of 6 weeks. All
measurements were performed between 8:00 and 12:00.
Study methods
Systolic, diastolic and mean blood pressures (SBP, DBP, MAP)
were measured on the upper arm using an automated oscillometric
device (HP-CMS patient monitor, Hewlett Packard, Palo Alto,
Calif., USA). Pulse rate was automatically recorded from a finger
pulse-oximetric device (HP-CMS patient monitor).
Table 1 Subjects’ characteristics (protocol A): 210 subjects un-
derwent measurement of retinal white blood cell flux with the
blue-field entoptic technique; 210 other subjects underwent colour
Doppler imaging of the central retinal artery. Data are presented as
means ± SD
Retinal WBCF was assessed with the blue-field entoptic tech-
nique. This non-invasive method is described in detail by Riva and
Petrig [12]. The blue-field entoptic phenomenon can be seen best
by looking into a blue light with a narrow optical spectrum at a
wavelength of approximately 430 nm. Under these conditions tiny
corpuscles can be seen flying around swiftly in an area with a ra-
dius of 10–15 degrees of arc centred at the fovea. Most likely this
phenomenon is caused by the fact that red but not white blood
cells absorb short-wavelength light. Thus, the passage of a white
blood cell is perceived as a flying corpuscle. For determination of
retinal haemodynamic parameters, a simulated particle field is
shown to the subjects under study. By comparison with their own
Blue-field entoptic technique (n=210)
Age (years)
27.7±4.8
131.9±14.1
70.1±14.1
90.8±13.1
64.5±8.1
Systolic blood pressure (mmHg)
Diastolic blood pressure (mmHg)
Mean arterial pressure (mmHg)
Pulse rate (beats/min)
White blood cell velocity (relative units)
White blood cell density (relative units)
White blood cell flux (relative units)
1.24±0.29
112.7±29.5
136.7±39.3
entoptic observation subjects can adjust the number of white blood Colour Doppler imaging (n=210)
cells (WBCD) and the mean flow velocity (WBCV). Retinal
Age (years)
27.8±5.1
130.9±13.7
68.1±12.7
89.0±12.6
63.2±7.8
6.4±0.6
WBCF is then calculated as the product of WBCD and WBCV.
Systolic blood pressure (mmHg)
All subjects had a training session which consisted of at least five
Diastolic blood pressure (mmHg)
matching tests. Only subjects who were able to provide reproduc-
Mean arterial pressure (mmHg)
ible values with a SD of less than 20% were included. For study
Pulse rate (beats/min)
purposes each measurement consisted of at least five matching
Mean flow velocity (cm/s)
tests, and the means of velocity and density were calculated.
Resistive index
0.72±0.04
Again, only values with a SD of less than 20% were accepted as

675
Fig. 1 Association between retinal white blood cell flux (WBCF) Fig. 2 Association between mean flow velocity (MFV) and mean
and mean arterial pressure (MAP). The regression line and the arterial pressure (MAP). The regression line and the 95% confi-
95% confidence interval are shown (n=210)
dence interval are shown (n=210)
We observed a significant direct correlation between
WBCF and MAP (Fig. 1) and between WBCV and MAP
(each P<0.001). The correlation coefficient between
WBCD and MAP was considerably lower (P=0.033).
The correlation coefficients between retinal WBCF
(r=0.34; P<0.001), retinal WBCV (r=0.37; P<0.001),
retinal WBCD (r=0.16; P=0.023) and SBP were in the
same range. Correlation coefficients between parameters
of retinal leukocyte movement and DBP were only
slightly smaller (WBCF: r=0.31; P<0.001; WBCV:
r=0.34; P<0.001; WBCD: r=0.14; P=0.047). By con-
trast, none of the parameters assessed with the blue-field
technique was dependent on pulse rate.
Mean flow velocity in the central retinal artery was
also dependent on MAP (P=0.010; Fig. 2), SBP (r=0.16;
P=0.018) and DBP (r=0.17; P=0.012), but correlation
coefficients were smaller than those observed for retinal
white blood cell flux and blood pressure. RI and blood
pressure were negatively correlated (MAP: r=–0.19;
P=0.006; SBP: r=–0.21; P=0.002; and DBP r=–0.21;
P=0.002), but again correlation coefficients were small.
Neither MFV nor RI was associated with pulse rate.
Fig. 3 Association between individual deviations in retinal white
blood cell flux (DWBCF) and mean arterial pressure (DMAP). The
regression line and the 95% confidence interval are shown (n=40)
Protocol B
Subjects’ characteristics are shown in Table 2. We ob-
served a significant direct correlation between DWBCF
and DMAP (r=0.256; P=0.005; Fig. 3) and DWBCV and
DMAP (r=0.336; P<0.001). The correlation between
retinal DWBCD and DMAP was not significant. DMFV
in the central retinal artery was dependent on DMAP
(r=0.204; P=0.026; Fig. 4), whereas DRI and DMAP
were not correlated. None of the retinal haemodynamic
variables were dependent on pulse rate. There was a
significant positive correlation between DMFV and
DWBCV (r=0.23; P=0.01).
Fig. 4 Association between individual deviations in mean flow
velocity (DMFV) and mean arterial pressure (DMAP). The regres-
sion line and the 95% confidence interval is shown (n=40)

676
Table 2 Subjects’ characteristics (protocol B). Data are presented
as means ± SD (n=40)
with this system if subjects are carefully selected [3, 9].
However, this technique assesses leukocyte movement,
and it is not entirely clear whether results can be directly
applied to volumetric blood flow in the retina. There is
some evidence that leukocytes move slower in retinal
capillaries than erythrocytes [1]. Whether erythrocyte
movement is differentially affected by hypertension in
the retina remains to be established. This may, in princi-
ple, be investigated using combined laser Doppler velo-
cimetry and vessel size determination [13]. For calcula-
tion of total blood flow with this technique it is neces-
sary to measure all veins or arteries entering the optic
nerve. This procedure is much more time-consuming
than measurements with the blue-field entoptic technique
and was therefore not employed in the present study.
In addition, it is widely accepted that leukocyte veloc-
ity can be adequately assessed with the blue-field tech-
nique, but it is unclear whether the number of leukocytes
as perceived in the blue light is an adequate measure of
vascular volume. To overcome the problem of inter-indi-
vidual variability we performed a study in which sub-
jects were measured thrice within 6 weeks (protocol B).
The results of this longitudinal study are, however, com-
parable to those with the cross-sectional approach.
Colour Doppler imaging can be applied to measure
blood flow velocities in retrobulbar vessels, but no quan-
titative information on vessel diameters can be obtained.
Hence, this method does not allow for quantification of
blood flow through the central retinal artery. Neverthe-
less, we observed a significant association between SBP
and MFV in the central retinal artery, which may indi-
cate increased blood flow through this artery with in-
creasing blood pressure. Again, the results in protocol B
are in good agreement with those observed in the cross-
sectional study.
Age (years)
24.2±3.9
124.6±10.0
61.7±9.4
Systolic blood pressure (mmHg)
Diastolic blood pressure (mmHg)
Mean arterial pressure (mmHg)
Pulse rate (beats/min)
82.4±9.8
59.9±8.3
White blood cell velocity (relative units)
White blood cell density (relative units)
White blood cell flux (relative units)
Mean flow velocity (cm/s)
Resistive index
1.22±0.41
116.3±30.7
138.9±42.6
6.1±0.7
0.73±0.03
Discussion
In the present study in subjects with normal or slightly
elevated blood pressure we observed an association be-
tween retinal WBCF and MAP. Our data also indicate
that perifoveal leukocyte velocity is dependent on MAP,
whereas retinal leukocyte density shows only little blood
pressure dependence. The hypothesis that retinal blood
flow is dependent on blood pressure is further supported
by our observation that MFV in the central retinal artery
is correlated with MAP.
Generally the dependence of retinal haemodynamic
variables on blood pressure in the present study was
small. Our data indicate that WBCF increases between
3.6% (protocol B) and 5.6% (protocol A) per 10-mmHg
increase in MAP. Accordingly, MFV increases between
1.3% (protocol A) and 5.8% (protocol B) per 10-mmHg
increase in MAP. Whether this is of clinical relevance re-
mains to be shown. However, even a small hypertension-
induced elevation of retinal blood flow may increase the
risk of onset or progression of ocular vascular disease [6].
In the present study we focused on young male sub-
jects because retinal WBCF decreases with age [4].
Patients with any systemic or ocular pathology likely to
influence retinal blood flow were excluded from the
present study. In addition, we included only subjects who
did not receive any regular medication. We considered
this important because any type of systemic vasoactive
medication may influence retinal blood flow [5, 8].
Measurement of retinal blood flow with the blue-field
entoptic technique is a subjective method. Therefore, on-
ly subjects whose results displayed little variability were
included in the present study. There is evidence from
previous trials that reproducible data can be obtained
The present study’s findings regarding RI are difficult
to interpret. We have previously shown that RI is not an
adequate measure of vascular resistance in the retina
[10]. Hence, it seems very unlikely that the correlation
between RI and blood pressure observed in the present
study contains any information on the dependence of ret-
inal perfusion on ocular perfusion pressure.
Concluding, our data indicate a slight but significant
increase in retinal blood flow with increasing blood pres-
sure. Whether this is of clinical relevance in eye diseases
with altered retinal perfusion, such as diabetic retinopa-
thy, remains to be established.
References
1. Ben-Nun J (1996) Comparative flow
2. Dumsky MJ, Eriksen JE, Dore CJ,
Kohner EM (1996) Autoregulation in
the human retinal circulation: assess-
ment using isometric exercise, laser
Doppler velocimetry, and computer
assisted image analysis. Microvasc Res
51:378–392
3. Grunwald JE, Zinn H (1992) The acute
effect of oral acetazolamide on macular
blood flow. Invest Ophthalmol Vis Sci
33:504–507
velocity of erythrocytes and leukocytes
in feline retinal capillaries. Invest
Ophthalmol Vis Sci 37:1854–1859

677
4. Grunwald JE, Piltz J, Patel N, Bose S,
Riva CE (1993) Effect of ageing on
retinal macular microcirculation: a blue
field simulation study. Invest
Ophthalmol Vis Sci 34:3609–3613
5. Grunwald JE, DuPont J, Dreyer EB
(1997) Effect of chronic nitrate treat-
ment on retinal vessel caliber in open-
angle glaucoma. Am J Ophthalmol
123:753–758
6. Kohner EM, Patel V, Rassam SMB
(1995) Role of blood flow and im-
paired autoregulation in the pathogene- 10. Polska E, Kircher K, Ehrlich P, Vecsei
sis of diabetic retinopathy. Diabetes
44:603–607
7. Lieb WE, Cohen SM, Merton DA,
Shields JA, Mitchell DG, Goldberg BB
(1991) Color Doppler imaging of the
eye and the orbit. Arch Ophthalmol
109:527–531
8. Patel V, Rassam SM, Chen HC, Jones
M, Kohner EM (1998) Effect of angio-
tensin-converting enzyme inhibition
with perindopril and beta-blockade
with atenolol on retinal blood flow in
hypertensive diabetic subjects.
12. Riva CE, Petrig BL (1990) Retinal
blood flow: laser Doppler velocimetry
and blue field simulation technique.
In: Masters B (ed) Noninvasive diag-
nostic techniques in ophthalmology.
Springer, New York Berlin Heidelberg,
pp 390–409
Metabolism 47 [Suppl 1]:28–33
9. Polak K, Dallinger S, Polska E, Findl
O, Eichler HG, Wolzt M, Schmetterer
L (2000) Effects of insulin on retinal
and pulsatile choroidal blood flow in
humans. Arch Ophthalmol 118:55–59
13. Riva CE, Grunwald JE, Sinclair SH,
Petrig BL (1985) Blood velocity and
volumetric flow rate in human retinal
vessels. Invest Ophthalmol Vis Sci
26:1124–1132
14. Riva CE, Grunwald JE, Petrig BL
(1986) Autoregulation of human retinal
blood flow. Invest Ophthalmol Vis Sci
27:1706–1712
15. Robinson F, Riva CE, Grunwald JE,
Petrig BL, Sinclair SH (1986) Retinal
blood flow autoregulation in response
to an acute increase in blood pressure.
Invest Ophthalmol Vis Sci 27:722–726
VP, Schmetterer L (2001) Resistive in-
dex in central retinal artery as assessed
by color Doppler imaging does not cor-
respond to retinal vascular resistance.
Am J Physiol (in press)
11. Rassam SMB, Patel V, Kohner EM
(1995) The effect of experimental hy-
pertension on retinal vascular autoreg-
ulation in humans: a mechanism for the
progression of diabetic retinopathy.
Exp Physiol 80:53–68