From Turku PET Center (U.R., V.O., M.J.K.) and the Departments of
Medicine (H.L., P.N.), Clinical Physiology (M.L.), Clinical Chemistry (M.R.),
and Anesthesiology (O.K.) of the University of Turku (Finland); the Department
of Electronics and Informatics (P.V., C.C.), University of Padova (Italy); and
the Department of Medicine (T.U., H.Y.-J.), Division of Endocrinology and
Diabetology, University of Helsinki (Finland).
Correspondence to Hanna Laine, MD, Department of Medicine, University of Turku, Kiinamyllynk 4–8, FIN-20520 Turku, Finland. E-mail hannal{at}pet.tyks.fi
Methods and Results—Positron emission tomography, combined with
[15O]H2O and [15O]CO as tracers
for direct measurement of blood flow and volume in skeletal muscle, and
a new bayesian iterative reconstruction algorithm allowing
pixel-by-pixel quantitation of blood flow and flow dispersion, were
used. Measurements were performed basally after an overnight fast and
under normoglycemic hyperinsulinemic conditions in 11
newly diagnosed, untreated mildly hypertensive men (age, 35±1 years;
body mass index, 25.2±0.4 kg/m2, blood pressure
141±4/96±2 mm Hg, mean±SE) and 11 matched normotensive
men. Insulin-stimulated whole body glucose uptake was significantly
decreased in the hypertensive men (41±4 µmol/kg per minute)
compared with the normotensive (59±4 µmol/kg per minute,
P<0.005) men. Mean blood flow in skeletal muscle was
significantly lower in the hypertensive than the normal subjects
basally (1.7±0.2 versus 2.7±0.4 mL/0.1 kg per minute,
P<0.05) and during hyperinsulinemia
(2.3±0.2 versus 4.2±0.8, P<0.05). The flow response
to insulin (0.6±0.2 versus 1.9±0.5 mL/0.1 kg per minute, hypertensive
versus normal subjects, P<0.05) was also significantly
blunted. Muscle blood volume was significantly lower in the
hypertensive than in the normal subjects, both basally (3.0±0.2 versus
3.5±0.2 mL/0.1 kg, P<0.05) and during
hyperinsulinemia (3.1±0.2 versus 4.0±0.2 mL/0.1
kg muscle, P<0.02). The increase in muscle blood volume
by insulin was significant in the normal (P<0.05) but
not the hypertensive subjects. Regional pixel-by-pixel analysis
within femoral muscles revealed significant spatial
heterogeneity of blood flow. Insulin increased absolute
dispersion of blood flow significantly more in the normal subjects than
in the hypertensive subjects (P<0.05).
Conclusions—True flow heterogeneity, as judged
from the coefficients of variation (relative dispersion), was
comparable between the groups basally and during
hyperinsulinemia. We conclude that mean flow, its
absolute dispersion, and blood volume exhibit insulin resistance in
patients with essential hypertension.
Surprisingly, however, with the exception of the subgroup of
hypertensive patients with high Na/Li countertransport activity in
erythrocytes, insulin-stimulated blood flow has been found to be
unaltered in essential hypertension in all seven studies hitherto
performed.15 16 17 18 19 20 21 In normal subjects, however,
inverse relations have been found in two studies between mean
arterial blood pressure and insulin-stimulated blood
flow.22 23 In the latter
studies,22 23 the insulin concentrations were
high enough to clearly (on the average by 80% to 117%) increase blood
flow above basal values, whereas in the studies performed in
hypertensive subjects,15 16 17 18 19 20 21 infusion of insulin
either did not increase blood flow significantly or the increase was
modest (10% to 30%). We hypothesized that if defects in
insulin-stimulated blood flow do indeed characterize patients with
essential hypertension, the stimulus, that is, the dose of insulin or
duration of the insulin infusion, must be high enough to clearly
increase blood flow in normal subjects.
Even if a defect did exist in insulin-stimulated blood flow, its
significance for insulin-stimulated nutrient delivery would remain
unclear. If flow increased because of an increase in linear blood flow
velocity through functional arteriovenous shunts, both blood volume and
substrate availability in nutritive capillaries remain
unchanged.24 25 However, insulin recently has
been shown to increase capillary recruitment in hindleg muscles of
anesthetized rats.26 These new data
combined with our previous data showing that insulin increases not only
blood flow but also blood volume in normal
subjects27 classifies insulin as a true
vasodilatory hormone. A defect in the ability of insulin to increase
blood volume in patients with essential hypertension would therefore be
consistent with diminished capillary recruitment. The effect of
insulin on muscle blood volume has not yet been studied in essential
hypertension.
Capillary exchange of nutrients is also dependent on the distribution
of blood among exchange vessels.28 A nonuniform
distribution of flow among vessels, which can be defined as some
vessels receiving more and some less of their appropriate fraction of
total,28 has been invoked to explain phenomena
such as flow-limited muscular
performance29 and suboptimal capillary
transport of small solutes.30 In animals, various
experimental maneuvers have opposite effects on flow and capillary
exchange.28 For example, infusion of vasodilators
may increase flow but reduce capillary transport of diffusible
indicators in skeletal muscle.30 31 Currently no
data are available regarding the effect of insulin on blood flow
heterogeneity in patients with essential
hypertension.
Use of positron emission tomography (PET),
[15O]-labeled water
([15O]H2O), and
[15O]-labeled carbon monoxide
([15O]CO) enables the quantitation of skeletal
muscle blood flow and volume in vivo in
humans.27 32 Recent refinements in reconstruction
methods of PET images33 enabled us to accurately
perform pixel-by-pixel quantitation of regional muscle blood flow and
analyze the absolute and relative dispersion of blood flow and
the shape of its frequency distribution as indexes of flow
heterogeneity.28 In our study, we
used this technique to compare effects of insulin on mean muscle blood
flow and volume in normal subjects and patients with essential
hypertension. We also examined whether insulin changes the absolute or
relative distribution of flow in normal subjects and whether these
actions of insulin are preserved in essential hypertension. Our results
demonstrate defects in the vascular actions of insulin in the
stimulation of mean flow, absolute flow dispersion, and blood volume
but preserved relative dispersion in patients with essential
hypertension.
Study Design
Production of Positron Emitting Tracers, Scanning
Measurement of Muscle Blood Flow
Measurement of Muscle Blood Volume
Regions of Interest
Heterogeneity Analysis of Blood
Flow
Other Measurements
Statistical Methods
During insulin infusion, systolic blood pressure (Table
Muscle Blood Flow
Heterogeneity of Blood Flow
Muscle Blood Volume
Flow Heterogeneity
Flow heterogeneity, as defined by Duling and
Damon,28 can be defined simply as an uneven
distribution of flow among perfused vessels. Relative dispersion serves
as a true measure of flow heterogeneity and reflects
the extent to which some vessels receive more and some less flow than
their appropriate fraction of total.28 We
quantitated relative blood flow dispersion in the basal state and under
normoglycemic hyperinsulinemic conditions in normal and
hypertensive subjects. Relative blood flow dispersions and normalized
flow distribution histograms remained unchanged in both groups. This
suggests that the fractional distribution of blood flow remained
constant as the mean blood flow increased. These data are
consistent with previous studies demonstrating that changes in
mean blood flow by infusion of glucose-insulin-potassium in the dog
heart,45 exercise,46 or
sympathetic nerve activity47 in skeletal muscle
of the rabbit do not change relative flow or the shape of flow
distributions, whereas vasodilatation with adenosine increases
both mean flow and its relative dispersion in the
heart.48 The magnitude of relative dispersion,
0.38 in both the normal and hypertensive subjects basally, is within
the range of values reported with the use of
microspheres,49 number of ink-filled
capillaries in histologic sections,50 and
tritiated water51 in dog,51
rat (0.56, [Reference 4949 ]), and rabbit50
skeletal muscle preparations. In contrast to relative dispersion,
insulin did increase absolute dispersion (standard deviation of blood
flow) significantly in the normal subjects. The increase in absolute
dispersion by insulin was not statistically significant in the
hypertensive subjects. Although one may predict absolute dispersion to
change as a function of mean blood flow, the small standard error of
blood flow and the relatively small number of hypertensive subjects
studied could have contributed to the failure to observe a significant
increase in absolute dispersion in the hypertensive subjects.
Variability of blood flow among vessels has been termed the spatial
heterogeneity and the variability with time within each
vessel as temporal
heterogeneity.52 There are no
perfect methods for measuring flow heterogeneity, and
each method is subject to significant and different sampling errors as
a result of temporal and spatial variations in
flow.28 In our study, images were integrated from
data collected over 4 minutes (250 seconds). In resting frog sartorius
muscle, in which spatial and temporal heterogeneity
have been separately quantitated, overall flow
heterogeneity measured during a 10-second period was
similar to spatial heterogeneity measured over 10
minutes, although temporal heterogeneity made a small
contribution to overall
heterogeneity.53 In the dog
heart, spatial heterogeneity of myocardial blood flow
remains stable for hours.54 These data suggest
that the heterogeneity measured in the present
study predominantly reflects spatial heterogeneity.
Considering the diameter of a capillary (5 to 7 µm) relative to
the width of a pixel (3 mm) or the smallest surface area that can
be quantitated after the reconstruction process (5x5
mm2), values of flow
heterogeneity in this study do not measure
heterogeneity at the capillary level.
Heterogeneity in capillaries, when measured with direct
visualization techniques, have given higher coefficients of variation
(mean, 0.61, see Reference 2828 for review) than those observed in the
present or in whole organ studies. From the view of microvessel
organization, regions of groups of capillaries with dimensions of cubic
millimeters are controlled by an arteriole,55 56
but it is unknown whether the present analysis of flow
heterogeneity reflects heterogeneity at
this level. Thus we can only conclude that flow
heterogeneity, when measured in areas of
Insulin and Skeletal Muscle Blood Volume
Possible Causes of Altered Blood Flow and Volume Responses to
Insulin in Essential Hypertension
In this study, we did not compare the blood flow response to insulin
with a response to another vasoactive agent. Thus the blunted vascular
response could have been limited to insulin, or it might reflect other
vascular abnormalities present in patients with mild essential
hypertension. In addition to sympathetic overactivity, subnormal nitric
oxide synthesis or action might underlie the defect in insulin-induced
vasodilatation. A defect in nitric oxide production could
explain blunted responses to several vasoactive agents including
insulin,21 ß-agonists,59
and classic endothelium-dependent agents such as
acetylcholine.8 In studies specifically examining
the integrity of nitric oxide–dependent vasodilatation, abnormal
responses have been found in many8 9 10 11 but not
all12 studies. It has also been recently
demonstrated that forearm ß-adrenergic receptor–mediated
vasodilatation is impaired, without alteration in local forearm
norepinephrine spillover, in borderline
hypertension.59 The vasodilatory effect of
ß-stimulation in the human forearm can be reduced by inhibition of
nitric oxide synthesis.60 Considering that
insulin-induced vasodilatation also can be abolished by inhibition of
nitric oxide synthesis3 4 5 and that insulin
potentiates isoproterenol-induced
vasodilatation,61 the flow defect in the
hypertensive subjects could reflect impairment in
endothelial synthesis of nitric oxide or resistance to
nitric oxide action.
In conclusion, we examined multiple aspects of the vascular actions of
insulin in skeletal muscle and found patients with essential
hypertension to be resistant to stimulation of mean flow, its
absolute dispersion, and blood volume. It is presently unclear
whether these abnormalities are specific for insulin or merely reflect
some generalized defect in endothelial nitric oxide
synthesis or sympathetic overactivation in these patients. It also
remains to be established whether the resistance to insulin in blood
vessels is a consequence of hypertension or a primary contributor to
increased peripheral resistance.
Received October 7, 1997;
revision received January 7, 1998;
accepted January 16, 1998.
© 1998 American Heart Association, Inc.
Clinical Investigation and Reports
Preserved Relative Dispersion but Blunted Stimulation of Mean Flow, Absolute Dispersion, and Blood Volume by Insulin in Skeletal Muscle of Patients With Essential Hypertension
Abstract
Top
Abstract
Introduction
Methods
Results
Discussion
References
Background—We examined the integrity
of the effects of insulin on mean muscle blood flow, flow
heterogeneity, and blood volume in essential
hypertension.
Key Words: blood pressure • insulin • glucose • tomography
Introduction
Top
Abstract
Introduction
Methods
Results
Discussion
References
Insulin has dual
effects on skeletal muscle blood flow. It increases blood flow in a
time- and concentration-dependent fashion1 2
through a mechanism that can be abolished by inhibiting nitric oxide
synthesis.3 4 5 Insulin also stimulates, at
physiological concentrations, sympathetic nerve
activity, an effect counteracting insulin-induced
vasodilatation.6 In normal subjects, the
vasodilatory effect of insulin predominates.1 7
In patients with essential hypertension, blunted nitric
oxide–dependent vasodilatory responses have been reported in
some8 9 10 11 but not all12
studies. This possible defect, vascular rarefaction in biopsies of
skeletal muscle,13 lack of attenuation of
vasoconstrictive responses to angiotensin
II by insulin,14 and data demonstrating
overactivation of the sympathetic nervous system by
insulin,15 raise the possibility that the
vascular effects of insulin also are impaired in these patients.
Methods
Top
Abstract
Introduction
Methods
Results
Discussion
References
Subjects
Eleven men with newly diagnosed, untreated essential
hypertension and 11 normotensive men were studied
(Table
). The hypertensive subjects were recruited among
clients of an occupational health service on the basis of the following
criteria: (1) age 20 to 50 years, (2) diastolic blood
pressure >95 mm Hg on at least three separate occasions, (3)
no signs of other cardiovascular disease, and (4) no
regular medication. Secondary forms of hypertension were excluded by
history and physical examination and standard laboratory tests. All
subjects had normal electrocardiograms. At least 3 days
before the study, the subjects consumed a weight-maintaining diet
containing 200 g carbohydrate per day. The nature, purpose, and
potential risks of the study were explained to all subjects before they
gave their voluntary consent to participate. The study protocol was
approved by the Ethical Committee of the Turku University
Hospital.
View this table:
[in a new window]
Table 1. Characteristics of the Study Groups
The subjects were studied in the supine position after an
overnight 10- to 12-hour fast. Two catheters were inserted: one in an
antecubital vein for infusion of glucose and insulin and injection of
[15O]H2O and one in the
opposite radial artery for blood sampling. Each study consisted of a
40-minute basal period (-40 to 0 minutes) and a 100-minute (0 to 100
minutes) hyperinsulinemic period (Figure 1). Blood volume and flow were measured
in cross sections of the femoral region during the basal period as
described in detail below. Serum insulin was thereafter increased with
a primed, continuous infusion of insulin at a rate of 5 mU/kg per
minute (Velosulin, Novo Nordisk A/S). Normoglycemia was maintained with
an infusion of 20% glucose on the basis of arterial plasma
glucose concentration measurements,34 which were
performed every 5 to 10 minutes. After 50 to 60 minutes of
hyperinsulinemia, measurements of blood volume and
flow were repeated (Figure 1). Whole body glucose uptake was calculated
from the glucose infusion rate (during 60 to 100 minutes) after
correction for changes in the glucose pool
size.35 Arterial blood samples for
measurement of the concentrations of serum insulin were taken at
30-minute intervals.36 Blood
[15O]H2O and
[15O]CO radioactivities were measured as
detailed below.
View larger version (13K):
[in a new window]
Figure 1. Design of the study. Arrow denotes injection times
of the tracers [15O]CO and
[15O]H2O for measurement of blood flow and
volume. Shaded rectangle denotes the period of positron emission
tomography (PET) scanning and blood sampling after tracer
injections.
[15O]H2O32 37
and [15O]CO27 were
produced as previously described. An eight-ring ECAT 931/08–12
tomograph (Siemens/CTI) was used for scanning.38
The subject was positioned in the tomograph with the femoral region
within the gantry. Before emission scanning, a transmission scan for
correction of photon attenuation in the body was performed for 15
minutes with a removable ring source containing
68Ge (total counts, 15 to
30x106 in a plane). Before the images were
reconstructed, the data were corrected for decay of
[15O]. Dead time was corrected for in each
plane and frame separately, and correction for photon attenuation was
done with the use of data obtained from a transmission scan. The data
were then reconstructed into a 128x128 matrix with the use of a
recently developed bayesian iterative reconstruction algorithm with
median root prior (the MRP method) with 150 iterations and a bayesian
coefficient of 0.3 in the flow studies and 0.9 in the blood volume
studies.33 The pixel-by-pixel inhomogeneity and
reconstruction artifacts were defined with a cylindrical phantom in
which the relative radioactivity concentrations were chosen to
represent the levels in muscle and blood vessels.
Inhomogeneity, as estimated from the coefficient of variation (%),
averaged 
10% using the former reconstruction method (filtered
back-projection, the FBP method39) and 6%
using the newly developed iterative
algorithm.33
[15O]H2O (30 to 45
mCi) was infused intravenously (30 seconds), and a dynamic
scan for 6 minutes was started simultaneously. In three
subjects, insulin-stimulated blood flow could not be measured because
of technical problems. To obtain the input function,
arterial blood was withdrawn with a pump at a speed of 6
mL/min from the radial artery, and the radioactivity concentration was
measured with a two-channel detector system (Scanditronix), which was
calibrated to the well counter (Bicron 3MW3/3, Bicron Inc) and the PET
scanner. The delay between the input curve and the tissue curve was
solved by fitting.40 Blood flow was quantitated
with a single compartmental model and an autoradiographic
method as previously described in detail.41 We
have recently validated this method for quantitation of low blood flows
in skeletal muscle by using
[15O]H2O, PET, and
plethysmography.32 37 Blood flow was calculated
pixel by pixel into parametric flow images with a 250-second
tissue integration time as previously
described.32 37 The size of a pixel in the
parametric flow image was 9 mm2
(3x3 mm).
The subjects inhaled [15O]CO (0.14% CO
mixed with room air) (95 to 135 mCi) for 2 minutes. A static scan was
started 2 minutes after the end of the inhalation, as previously
described.27 The inhaled CO combines with
hemoglobin to form carboxyhemoglobin, which becomes distributed in the
volume of red blood cells. During the scan period, three blood samples
were taken and their radioactivity concentration was measured with a
well counter (Bicron 3MW3/3). Regional muscle blood volume was
calculated under steady-state conditions by dividing the concentration
of tissue [15O]CO radioactivity by the
concentration of [15O]CO in the blood, as
previously described.27 This ratio was then
divided by the regional tissue–to–large vessel hematocrit, which was
assumed to equal 0.91 on the basis of studies of Chaplin et
al,42 who showed the ratio of total body
hematocrit to venous hematocrit to be constant over a wide range of
peripheral venous hematocrits. There are no published data
on the effect of insulin on the hematocrit in a deep vein draining
muscle tissue. However, hematocrits were determined and remained
unaffected by insulin or glucose in a large study in which forearm
glucose uptake was measured over a range of insulin concentrations from
basal to 1800 mU/L and glucose of 5 to 22
mmol/L.43 We have previously shown the
reproducibility of two repeated measurements of blood volume in the
same subject to be 3.0±1.8%.27
Regions of interest (ROIs) were drawn in the posterior,
anterolateral, and anteromedial muscular compartments of the femoral
region in four transaxial slices in both legs, carefully avoiding the
great vessels. The localization of the muscle compartments was verified
by comparing the CO images with the transmission image, which provides
a topographic distribution of tissue density. The ROIs outlined in the
CO images were copied to the flow images to quantitate flow and volume
in identical regions.
Distributions of blood flow were constructed from the
pixel-by-pixel images. ROIs were drawn onto the muscle areas as
described above and used for heterogeneity
analysis. Each ROI comprised on the average 23.3±1.5
cm2 . The standard deviation (SD) of flow values
was used to characterize the absolute dispersion of
flow.28 The coefficient of variation (CV) for
blood flow was calculated by dividing the SD by its respective mean
value (SD/mean flow).28 The CV is a measure of
relative dispersion or true flow heterogeneity.
However, it is possible that the coefficient of variation remains
constant in response to an intervention, whereas the shape of the
distribution changes. The latter would also indicate redistribution of
blood flow.28 This possibility was evaluated by
visually analyzing the shape of the histograms depicting relative
flow.
Maximal aerobic power (VO2max)
was determined by use of an electrically braked cycle ergometer
(Ergoline 800 S, Mijnhardt). The criteria used to establish the
VO2max were a plateau in
VO2 with increasing intensity and a
respiratory quotient of >1.10. Blood pressure and heart rate were
followed every 15 minutes during the study.
Comparison of normotensive and hypertensive subjects and
differences between basal and insulin-stimulated measurements was
performed with two-way ANOVA for repeated measures. Analysis of
frequency histograms has been described above. Correlation
analysis was calculated with Spearman's rank correlation
coefficient. The abbreviation "kg" after muscle flow and volume
measurements refers to kilograms of muscle. Probability values of <.05
were considered statistically significant. All data are shown as
mean±SEM.
Results
Top
Abstract
Introduction
Methods
Results
Discussion
References
Whole Body Glucose Uptake
Fasting plasma glucose and insulin concentrations are given in the
Table
. During the insulin infusion, serum-free insulin (394±22 versus
453±28 mU/L, normotensive versus hypertensive subjects, NS) and plasma
glucose (5.5±0.1 versus 5.4±0.1 mmol/L, respectively)
concentrations were comparable. Insulin-stimulated whole body glucose
uptake was 32% lower in the hypertensive men (41±4 µmol/kg
body wt per minute) than in the normotensive men (59±4 µmol/kg
body wt per minute, P<0.05). )
remained unchanged in both groups, whereas diastolic blood
pressure decreased from 79±3 to 66±7 mm Hg (P<0.05)
in the normal subjects and from 96±2 to 87±2 mm Hg
(P<0.05) in the hypertensive subjects. Basal heart rate was
significantly higher in the hypertensive subjects (68±2 bpm) than in
the normal (58±3 bpm) subjects. During
hyperinsulinemia, heart rate increased to 62±3 bpm
in the normal subjects (P<0.05) and to 70±3 bpm in the
hypertensive subjects (NS).
Basal muscle blood flow was significantly lower in the
hypertensive men than in the normotensive men (Figure 2). Insulin increased mean muscle blood
flow significantly in both groups, by 55%, from 2.7±0.4 to 4.2±0.8
mL/0.1 kg per minute in the normal subjects (P<0.01), and
by 35%, from 1.7±0.2 to 2.3±0.2 mL/0.1 kg per minute in the
hypertensive subjects (P<0.01). Both basal
(P<0.05) and insulin-stimulated flow (P<0.05)
and the increments in blood flow by insulin (0.6±0.2 versus 1.9±0.5
mL/0.1 kg per minute, P<0.05) were significantly lower in
the hypertensive than the normotensive subjects.
View larger version (32K):
[in a new window]
Figure 2. Basal (open bars) and insulin-stimulated (shaded
bars) skeletal muscle blood flow (top) and volume (bottom) in
hypertensive patients and normal subjects. **P<0.01
basal vs insulin, +P<0.05 for hypertensive vs normal
subjects.
Examples of frequency distributions of absolute flows from one
hypertensive and one normal subject are shown in Figure 3. In response
to insulin, absolute dispersion of flow increased significantly in the
normal subjects from 1.11±0.25 to 1.47±0.35 mL/0.1 kg per minute
(P<0.03) but not in the hypertensive subjects (0.63±0.06
versus 0.75±0.05, basal versus insulin, NS). Relative dispersion was
not significantly different between the groups either basally
(0.355±0.03 versus 0.373±0.01, normal versus hypertensive subjects)
or during hyperinsulinemia (0.332±0.03 versus
0.336±0.01, respectively, Figure 4).
Also, the responses to insulin did not differ significantly between the
groups (-0.037±0.01 versus -0.022±0.02, respectively, NS).
View larger version (22K):
[in a new window]
Figure 3. Frequency histograms of muscle blood flow values
in a hypertensive ( 
, HTN) and a normal (
, CONT) subject after an
overnight fast (BASAL) and under normoglycemic
hyperinsulinemic conditions (INSULIN). Absolute flow
values (x-axis) are plotted against their frequency
(y-axis).
View larger version (19K):
[in a new window]
Figure 4. Frequency histograms of mean relative muscle blood
flow values in normal ( , CONT) and hypertensive (, HTN) subjects
after an overnight fast (BASAL) and under normoglycemic
hyperinsulinemic conditions (INSULIN). Relative flow
(x-axis) was calculated by dividing the standard
deviation of flow by mean flow, and this value was plotted against its
frequency of all relative flows (y-axis).
Insulin increased muscle blood volume by 14%, from 3.5±0.2
to 4.0±0.2 mL/0.1 kg in the normal subjects (P<0.01), but
did not change blood volume significantly in the hypertensive subjects
(3.0±0.2 versus 3.1±0.2 mL/0.1 kg, NS, Figure 2). Muscle blood volume was lower both
basally (P<0.05) and during
hyperinsulinemia (P<0.02) in the
hypertensive subjects compared with the normal subjects (Figure 2).
Discussion
Top
Abstract
Introduction
Methods
Results
Discussion
References
Insulin and Mean Muscle Blood Flow in Essential
Hypertension
The primary aim of this study was to establish whether blood flow
responses to insulin are abnormal in patients with essential
hypertension and if so, whether a defect in blood flow is associated
with a blunted increase in blood volume or abnormal distribution of
blood flow in skeletal muscle. We chose conditions that would maximize
the likelihood of detecting differences in the
hemodynamic effects of insulin. Insulin is known to be
a slow-acting vasodilator in human skeletal
muscle.1 7 In contrast to glucose extraction,
which is maximally and by 10- to 20-fold stimulated within 30 minutes
of insulin exposure in vivo, blood flow increases continuously over
several hours in a time-dependent and insulin concentration–dependent
fashion.1 7 These data imply that defects in the
vascular effects of insulin cannot be found if studies are performed
under conditions in which insulin does not increase flow or flow
increases only marginally even in normal subjects. This may explain why
defects in blood flow have not been observed in previous studies
performed in patients with essential
hypertension.15 16 17 18 19 20 21 Consistent with this
hypothesis, Baron et al23 and Utriainen et
al,22 who used 8 to 35 times higher doses of
insulin than those used to study patients with essential hypertension,
did find correlations between mean arterial pressure and
leg blood flow in normal subjects. Our finding of a defect in mean
blood flow in patients with essential hypertension at an insulin
concentration of 450 mU/L is in line with the latter observations and
provides the first evidence of a defect in the vasodilatory effect of
insulin in these patients. Because blood flow was quantitated directly
in skeletal muscle, artifactually low blood flow values caused by
possible differences in limb fat content between the groups were
avoided. Large differences in body composition are also unlikely
because of comparable body mass indexes between the groups. It should
be emphasized, however, that studies performed with high doses or
prolonged infusions of insulin should be viewed merely as attempts to
unravel vascular abnormalities in hypertension in a manner analogous to
the studies measuring blood flow responses to other vasoactive
agents.12 44
Use of PET and
[15O]H2O enables
quantitation of low flow rates in tissues such as resting skeletal
muscle.32 It enables presentation of
blood flows in pixel-by-pixel parametric images. In resting
skeletal muscle, the relation between tissue radioactivity after a
bolus injection of
[15O]H2O and blood flow
is linear.32 Therefore, the calculated flow in
the volume of measurement represents true average flow. The MRP
method is a new reconstruction algorithm that reduces noise and
improves image quality.33 In our study,
statistical noise was reduced by use of a long integration time (250
seconds) and the MRP method. The better image quality enabled
construction of absolute and relative flow distributions and made it
possible to examine, for the first time in vivo in humans, blood flow
heterogeneity in skeletal muscle of patients with
essential hypertension. 5x5
mm2 in size in the femoral region, is comparable
between normal subjects and patients with essential hypertension.
Despite this, the present methodology has advantages compared with
some previous approaches. First, the method is suitable for use in
humans. Second, it involves neither invasive manipulation of tissue nor
its innervation, factors that profoundly affect flow
heterogeneity.51
Theoretically, a defect in insulin-stimulated blood flow could be
due to a defect in blood flow velocity or blood volume (capillary
recruitment) or both.24 The former is thought to
reflect nonnutritive flow (functional vascular shunting) and the latter
nutritive flow.25 Capillary recruitment can lead
to a large increase in the number of capillaries open at any given
moment and thereby facilitate the access of nutrients to tissue. This
mechanism of flow increase might be expected to be accompanied by an
increase in muscle blood volume, whereas a simple increase in linear
blood flow velocity through functional vascular shunts would not alter
blood volume.25 In our study we found significant
decreases in blood volume basally and during insulin stimulation in the
hypertensive subjects. Furthermore, the insulin-induced increase in
blood volume was blunted in the hypertensive subjects. The difference
in blood volume between the hypertensive and normal subjects in this
study (3.0 versus 3.5 mL/0.1 kg muscle) may appear small compared with
our previous report of a blood volume of 3.3 mL/0.1 kg muscle in normal
subjects. It should be considered, however, that (1) a small difference
in blood volume translates into a much larger difference in blood flow
because flow is proportional to the fourth power of the radius of a
blood vessel, at least theoretically according to Poiseuille's law,
(2) a different reconstruction method was used in the present study
compared with the previous study, and (3) although mean blood volume in
the hypertensive subjects was within the previously found normal range
(2.3 to 4.1 mL/0.1 kg), mean blood volumes were significantly different
between the hypertensive and normal subjects in the present study.
Previous data on muscle blood volume in hypertension are sparse because
of methodological difficulties in accurately quantitating this
parameter in humans. Schmieder et
al57 measured total blood volume by using
125iodine-labeled albumin and hematocrit
and central blood volume by indocyanine green dilution and found
central blood volume to be 20% increased in 40 patients with
borderline hypertension. In these patients, total blood volume was
unchanged, implying that peripheral blood volume was
subnormal. The data from this study provide direct support for the
latter possibility.
Previous studies have documented multiple abnormalities in
neurohumoral control of blood flow as well as structural abnormalities
such as capillary rarefaction13 in blood vessels
in patients with essential hypertension. Such changes, if present
in our study subjects, could contribute to the observed defects in
insulin-stimulated blood flow and volume and to blunted responses of
these parameters to insulin. Sympathetic overactivity is
well documented, especially in young patients with borderline
hypertension.58 In this study, basal heart rate
was significantly higher in the hypertensive subjects than in the
normal subjects, suggesting, in view of comparable physical fitness of
the groups (Table), that sympathetic activity might have been increased
in our study subjects. In patients with essential hypertension, insulin
increases sympathetic activity more than in normotensive
subjects.15 This activation and
peripheral insulin resistance are observed when insulin is
given systemically rather than locally, suggesting that central
activation of the sympathetic nervous system is responsible for insulin
resistance.15
Acknowledgments
This study was supported by grants from the Academy of Finland
(H.Y.-J., P.N.), the Ahokas (H.Y.-J., P.N.), Novo Nordisk (H.Y.-J.,
P.N.), Turku University (H.L.), and the Maud Kuistila (H.L.)
Foundations. We thank Mika Teräs, MSc, Hannu Sipilä, MSc,
Tuula Tolvanen, MSc, and all the personnel in the Turku PET Center for
help during the studies.
References
Top
Abstract
Introduction
Methods
Results
Discussion
References
This article has been cited by other articles:
 |
|
 |
|
  R. D. Brook, L. Glazewski, S. Rajagopalan, and R. L. Bard Hypertension and Triglyceride Catabolism: Implications for the Hemodynamic Model of the Metabolic Syndrome J. Am. Coll. Nutr., August 1, 2003; 22(4): 290 - 295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Laaksonen, K. K. Kalliokoski, H. Kyrolainen, J. Kemppainen, M. Teras, H. Sipila, P. Nuutila, and J. Knuuti Skeletal muscle blood flow and flow heterogeneity during dynamic and isometric exercise in humans Am J Physiol Heart Circ Physiol, March 1, 2003; 284(3): H979 - H986. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Coggins, J. Lindner, S. Rattigan, L. Jahn, E. Fasy, S. Kaul, and E. Barrett Physiologic Hyperinsulinemia Enhances Human Skeletal Muscle Perfusion by Capillary Recruitment Diabetes, December 1, 2001; 50(12): 2682 - 2690. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Peltoniemi, H. Yki-Jarvinen, H. Laine, V. Oikonen, T. Ronnemaa, K. Kalliokoski, O. Raitakari, M. J. Knuuti, and P. Nuutila Evidence for Spatial Heterogeneity in Insulin- and Exercise-Induced Increases in Glucose Uptake: Studies in Normal Subjects and Patients with Type 1 Diabetes J. Clin. Endocrinol. Metab., November 1, 2001; 86(11): 5525 - 5533. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. K. Kalliokoski, V. Oikonen, T. O. Takala, H. Sipila, J. Knuuti, and P. Nuutila Enhanced oxygen extraction and reduced flow heterogeneity in exercising muscle in endurance-trained men Am J Physiol Endocrinol Metab, June 1, 2001; 280(6): E1015 - E1021. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. H Serne, R. O.B Gans, J. C ter Maaten, P. M ter Wee, A. J.M Donker, and C. D.A Stehouwer Capillary recruitment is impaired in essential hypertension and relates to insulin's metabolic and vascular actions Cardiovasc Res, January 1, 2001; 49(1): 161 - 168. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Peltoniemi, P. Lonnroth, H. Laine, V. Oikonen, T. Tolvanen, T. Gronroos, L. Strindberg, J. Knuuti, and P. Nuutila Lumped constant for [18F]fluorodeoxyglucose in skeletal muscles of obese and nonobese humans Am J Physiol Endocrinol Metab, November 1, 2000; 279(5): E1122 - E1130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Dawson, M. A. Vincent, E. J. Barrett, S. Kaul, A. Clark, H. Leong-Poi, and J. R. Lindner Vascular recruitment in skeletal muscle during exercise and hyperinsulinemia assessed by contrast ultrasound Am J Physiol Endocrinol Metab, March 1, 2002; 282(3): E714 - E720. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||