Plasma Norepinephrine in Essential Hypertension
A Study of the Studies
D A V I D S. GOLDSTEIN, M.D., P H . D .
Downloaded from http://hyper.ahajournals.org/ by guest on June 17, 2017
SUMMARY Of 32 studies comparing plasma norepinephrine concentrations in hypertensive and normotensive groups, 28 (88%) reported higher levels in the hypertensive group. However, only 13 (41%) of the studies
reported statistically significant hypertensive-nonnotensive differences in norepinephrine, leading the present
attempt to identify factors differentiating "positive" studies (those reporting significant hypertensivenormotensive differences) from "negative" studies (those reporting nonsignificant differences). Hypertensive
norepinephrine levels were similar in positive and negative studies (281 vs 288 pg/ml), but normotensive levels
were lower in the positive studies (177 vs 269 pg/ml). When compared with the fluorimetric technique, the
radioenzymatic type of assay was associated both with a lower frequency of positive results (25% vs 100%) and
greater intrastudy standard deviations (152 vs 72 pg/ml). Hypertensive-normotensive differences varied inversely with age (> = —0.37). Resolution of the persisting controversy about norepinephrine levels in essential
hypertension will require more attention to the causes of variability associated with the assay technique, to the
sources, characteristics, and treatment of the normotensive controls, and to the age of the patient population.
(Hypertension 3: 48-52, 1981)
KEY WORDS
• norepinephrine
• catecholamines
T
• hypertension
primary data sources independent of data from other
sources; 3) they specifically involved plasma
norepinephrine, not total catecholamines; 4) they used
a sensitive and reliable fluorimetric technique, such as
that of Renzini, et al.;1 a radioenzymatic assay; or
high-pressure liquid chromatography with electrochemical detection; and 5) they reported resting,
supine plasma norepinephrine levels in both a normotensive control group and an entire hypertensive
group (as opposed to a subgroup selected for certain
characteristics).
The large number of studies culled using these
criteria — 32 — allowed a statistical approach in
which the mean hypertensive and normotensive values
from each study provided single data points. When
mean group values had not been reported, they were
derived from the weighted contributions of the subgroups in the study; and when standard deviations
(SDS) had not been reported, they were calculated from
standard errors of the mean (SEMS) by multiplying the
SEMS by the square root of the number of observations.
Statistical analyses used two-tailed independentmeans and dependent-means t tests, chi-squared tests,
and Pearson correlation coefficients." All studies,
despite their probably variable reliability, were accorded equal weight.
HE introduction about a decade ago of sensitive techniques for measuring plasma norepinephrine,'• a and indications that those levels
reflect sympathetic neural activity,' promised to help
answer a perennial question in hypertension research:
Does sympathetic hyperactivity cause essential hypertension? Despite the subsequent publication of more
than 30 comparative studies in hypertensives and normotensive controls,*"" controversy persists about
whether patients with hypertension exhibit elevated
levels of norepinephrine. This report reexamines these
studies in an attempt to identify the causes of the inconsistencies in results.
Methods
The studies considered in this review satisfied the
following criteria: 1) they were published since 1973,
in English; 2) they were not merely abstracts, but were
From the Hypertension-Endocrine Branch, National Heart,
Lung, and Blood Institute, National Institutes of Health, Bethesda,
Maryland.
Address for reprints: David S. Goldstein, M.D., Ph.D.,
Hypertension-Endocrine Branch, National Heart, Lung, and Blood
Institute, National Institutes of Health, Building 10, Room 7N246,
Bethesda, Maryland 20205.
Received January 8, 1980; revision accepted June 18, 1980.
48
NOREPINEPHRINE IN HYPERTENSION /Goldstein
Results
Table 1 displays hypertensive and normotensive
mean norepinephrine levels from the 32 studies.
Twenty-eight (88%) reported higher levels in the
hypertensives, by an average of 54 pg/ml (t = 4.06, p
< 0.001). Only 13 reported statistically significant
hypertensive-normotensive differences in norepinephrine. Since about two-fifths of the studies were
"positive" (i.e., reporting a significant hypertensivenormotensive difference) and about three-fifths were
"negative" (reporting no significant difference), one
can appreciate that any generalizations from the literature might be questioned.
What made the "positive" studies positive, and the
"negative" negative? If the cause were a variably
represented "hyperadrenergic" subgroup of hypertensives,7 then studies with a preoponderance of these
patients should have been positive due to the excessive
mean norepinephrine levels in the hypertensive
groups. However, mean hypertensive levels were not
significantly higher in the positive than the negative
1. Plasma Norepinephrine Levels in Patients with Essential Hypertension (H) and
tensive Controls (N)
H/N
MAP H/N
Age H/N
(JTB)
(no.)
Author
Assay
NEH/N
(mm Hg)
R
Bertel et al.4
24/20
41/40
112/93
263/250
7
7
F
Brecht et al.B
257/135*
59/15
7
F
125/107
Brecht and Schoeppe6
40/42
201/128*
7
R
Cousineau et al.7
46/28
332/226*
41/39
DeLeeuw et al.8
R
120/95
69/22
270/240
45/41
Eide et al.6
F
240/167*
7/7
109/89
40/36
R
Eng et al.10
20/17
114/90
390/250*
47/34
?
7
Esleret al."
F
193/138*
21/11
12
R
Franco-Morselli et al.
43/45
116/90
19/11
269/248
7
Franco-Morselli et al.13
104/87
27/12
R
277/2501
7
7
Geffen et al.14
400/160*
R
20/8
Henry et al.15
R
41/27
32/93
151/144
119/88
Hjemdahl and Eliasson"
101/87T
H
436/353
7/7
35/35
Hofman et al.17
351/248*
97/90
R
18/18
19/18
7
Jones et al.18
410/354
R
31/28
47/38
R
Kafka et al.19
112/91
265/289
15/18
50/50
80
7
Kiowski et al.
44/34
289/283
R
43/43
Kobayashi et al.21
R
291/224*
111/81
27/21
42/42
Lake et al.22
R
112/86
67/84
339/304
44/33
R
Lake23
112/88
151/117
297/294
43/45
2
7
7
Louis et al. *
24/7
R
390/160t*
Miura et al.26
112/87
F
120/30
188/130*
35/30J
Miura et al.26
120/49
112/83
F
209/160*
35/33
Pedersen and Christensen 27
142/97
R
242/254
19/32
41/40
28
7
Philipp et al.
29/29
216/173*
F
38/33
Robertson et al.29
226/196
101/86
9/10
R
25/27
Sever et al.30
411/403
124/92
R
56/59
46/46
3
352/372
R
Sever et al. '
121/95
100/48
45/47
R
Taylor et al.32
123/89
51/26
46/40
240/260T
R
256/205
Vlachakis 33
38/14
48/49
119/92f
Vlachakis and Aledort34
?
277/234
R
22/13
50/44
3
Weidmann et al. "
R
123/89
202/169
79/90
46/37
TABLE
49
in Normo-
Downloaded from http://hyper.ahajournals.org/ by guest on June 17, 2017
SDH/N
98/121
147/59
7
198/106
150/100
99/31
175/62
77/36
122/186
286/364f
223/113
107/105
156/94
110/123
223/160
194/136
98/115
94/128
188/183
143/159
7
88/60
7
7
72/45
7
197/184
178/171
230/257t
139/60
89/90
128/91
148/124
114/89*
285/231*
1496/1085
41/38*
•Significant hypertensive-normotensive difference with p < 0.05.
tTaken from figure.
^Approximate ages.
R = radioenzymatic; F =fluorimetric;H = high-pressure liquid chromatography with electrochemical
detection; MAP = group mean blood pressure; NE = group mean norepinephrine level; SD = standard
deviation.
Average/Sum
50
HYPERTENSION
studies (281 vs 288 pg/ml). In contrast, the normotensive control levels were significantly lower in the
positive studies (177 vs 269 pg/ml, / = 4.27, p <
0.001). Standard deviations of normotensive levels
were also smaller in the positive studies (76 vs 152
pg/ml, / = 3.03, p < 0.005).
These results emphasize the need for close attention
to sources of variability in obtained norepinephrine
levels, and to characteristics of the normotensive controls. Several of these factors are considered below.
Number of Subjects
Downloaded from http://hyper.ahajournals.org/ by guest on June 17, 2017
It would be inappropriate to weight studies with
different numbers of patients equally, if small studies,
perhaps involving selected series of patients and therefore smaller standard deviations of results, tended to
be positive while larger studies, involving more
generalized patient populations and larger standard
deviations, tended to be negative. In fact, however,
seven of 15 (47%) studies with less than 50 subjects
were positive, and six of 17 (35%) with more than 50
subjects were positive, a nonsignificant difference in
proportions. Average standard deviations of
norepinephrine values were also similar in the two
types of studies (134 vs 138 pg/ml respectively). The
size of the study was therefore unrelated to the frequency of positive results.
Type of Assay
Of seven studies using a fluorimetric assay, all seven
were positive, while only six of 24 (25%) using a
radioenzymatic assay were positive (x 2 = 3.42, p <
0.001), despite roughly comparable hypertensivenormotensive mean differences (68 and 49 pg/ml
respectively). The average standard deviation of
norepinephrine levels was significantly greater with
the radioenzymatic technique, for both hypertensive
(160 vs 97 pg/ml, t = 2.46, p < 0.05) and normotensive (144 vs 46 pg/ml, t = 3.12, p < 0.001) groups.
The definition of a "normal" norepinephrine level
depended on the type of assay. For the fluorimetric
type, the mean normotensive level was 147 pg/ml,
compared with 251 pg/ml for the radioenzymatic (t =
4.10, p < 0.001). Of the 22 studies using a radioenzymatic technique and specifying the enzyme, nine
used
phenylethanolamine-N-methyl-transferase
(PNMT) and 13 used catechol-O-methyl-transferase
(COMT). For the PNMT type of radioenzymatic
assay, the mean normotensive level was 283 pg/ml,
compared with 222 pg/ml for the COMT type (f =
2.32, p < 0.05).
Patient Age
Lake, et al.22 have suggested that systematic
differences in group mean age introduced artifactual
effects in prior positive studies, because plasma
norepinephrine varied with age and because hypertensives were older than normotensives. Across the 27
VOL 3, N o
1, JANUARY-FEBRUARY
1981
studies that listed mean group ages between 20 and 50
years, mean group norepinephrine levels did correlate
significantly with mean group age for the normotensives (r = 0.42, p < 0.05) but not the hypertensives (r
= 0.28). The hypertensives were also older than the
normotensives, by an average of 3.4 years (/ = 3.74, p
< 0.001).
Nevertheless, this artifactual effect of age would
require that positive studies include older hypertensives than normotensives, and negative studies include
groups of similar age. The correlation between the
difference in group mean ages and the difference in
group mean norepinephrine levels, 0.15, was negligible.
In contrast, hypertensive-normotensive differences
correlated negatively with mean patient age (r =
-0.37, p < 0.05). Seven of 11 (64%) studies with a
mean patient age less than 40 years were positive,
while three of 17 (18%) with a mean patient age
greater than or equal to 40 years were positive (x 2 =
4.26, p < 0.05).
Severity of Hypertension
If norepinephrine levels varied with the severity of
hypertension, then positive results could have derived
from the inclusion of hypertensives with excessively
high blood pressures. Across the 44 groups in which
mean pressures were reported or calculable, a significant correlation of 0.36 (p < 0.05) was obtained
between mean pressures and mean norepinephrine
levels. However, the mean pressures of the hypertensives in the negative studies were actually higher than
those in the positive studies (116 vs 109 mm Hg),
though not significantly so.
Constitution of Control Groups
The available literature provides little information
about whether the controls were inpatients or outpatients, laboratory workers or newcomers to a
hospital setting, whether they had previously undergone phlebotomy, smoked, drank alcohol or coffee
recently, whether they were paid, or how they were
recruited.
The sources of control subjects varied from patients
with "various neurological disorders'" to patients
recovering from uterine hemorrhage, respiratory infection, or malnutrition,12 to laboratory workers,13 to
patients detected on population screening," to clinic
frequenters with minor complaints.25 One study
reported lower norepinephrine levels in normotensive
laboratory staff than in other normotensive controls. 1 '
Another found higher levels in normotensive outpatients than inpatients.21
The majority of studies did not specify blood
pressure — or any other — criteria for inclusion in the
normotensive group. Across the restricted interstudy
range of mean normotensive pressure, plasma
norepinephrine correlated 0.34 with mean normotensive pressure — a suggestive but not statistically
significant relationship.
NOREPINEPHRINE IN HYPERTENSION/Go/drto'n
Drug Treatment Status
Twenty-seven of the 32 studies described the drug
treatment status of the hypertensives. Five studies included only patients who had not previously been
treated with an antihypertensive medication. In the
remaining 22 studies, medication had been discontinued 1 to 6 weeks prior to phlebotomy. Across these
22 studies, there was no relationship between mean
hypertensive norepinephrine levels and amount of
time off medication. The mean norepinephrine level of
the previously untreated hypertensives (288 pg/ml)
was similar to that of the previously treated hypertensives (263 pg/ml).
Sample Handling and Collection
Downloaded from http://hyper.ahajournals.org/ by guest on June 17, 2017
Among the 24 studies reporting their phlebotomy
technique, the proportion of positive studies using
venepuncture was not significantly different from that
using an indwelling catheter. To the extent data were
available, the time of day when sampling was done
and the length of time the patients were supine prior to
sampling were unrelated to the frequency of significant hypertensive-normotensive differences.
Only five studies included information about the
kinds of collection tubes used, additives, and length
and temperature of storage prior to performing the
assay. Yet these mundane points may crucially influence the outcome of a study. For instance, since
catecholamine levels may decrease with time in
storage," obtained
hypertensive-normotensive
differences may be artifactually decreased when the
investigator samples a preponderance of elderly normotensives near the end of the study in order to
achieve age matching with the usually older hypertensives.
Discussion
The vast majority of studies of norepinephrine in
patients with essential hypertension and in normotensive controls have reported higher mean levels in the
hypertensives. To the extent plasma norepinephrine
reflects sympathetic activity, this finding supports the
hypothesis that sympathetic activity is increased in essential hypertension. Only about two-fifths of the
studies, however, reported statistically significant
hypertensive-normotensive differences, leading to the
effort, in the present study, to identify factors differentiating positive from negative studies.
Hypertensive norepinephrine levels in the positive
studies did not differ from those in the negative
studies. This argues against the influence of a variably
represented hyperadrenergic subgroup. In contrast,
normotensive levels were lower and less variable in the
positive studies. Unfortunately, the published
literature has virtually ignored the sources,
characteristics, and treatment of the normotensive
controls.
The likelihood of positive results depended on the
type of assay used to measure plasma norepinephrine.
51
The radioenzymatic assay was associated both with a
lower frequency of positive results and greater intrastudy standard deviations than the fluorimetric.
"Normal" norepinephrine levels varied by up to 100%
across assay types. The available literature provides
no clues about why the results should have been so
assay-dependent.
The age factor has obscured the role of plasma
norepinephrine, since norepinephrine tends to increase
with age and since hypertensives have been older than
normotensives. Hypertensive-normotensive differences in norepinephrine did not derive from the artifactual effects of poor age matching. On the other
hand, the largest hypertensive-normotensive differences occurred when the patients — and controls —
were relatively young. Further, the age-norepinephrine relationship was statistically significant across the
normotensive but not the hypertensive groups. These
two findings are consistent with the hypothesis that
elevated resting plasma norepinephrine levels characterize a proportion of young patients with essential
hypertension.
The available literature has paid inadequate attention to the interactions among several potential intervening variables that may mediate the relationship
between plasma norepinephrine and blood pressure.
These include thyroid function, the renin-angiotensinaldosterone axis, plasma volume, sodium balance,
alpha- and beta-receptor sensitivity, and the psychological stress associated with hospitalization,
diagnostic maneuvers, and the patient's awareness of
the disease. A comprehensive understanding of
norepinephrine in essential hypertension will require
measurement of or control for these factors.
In summary, three factors differentiated positive
from negative studies: 1) normotensive control values;
2) type of assay; and 3) age of the patient population.
How and why these factors exert their effects are
matters for future research.
References
1. Renzini V, Brunori CA, Valori C: A sensitive and specific
fluorimetric method for the determination of noradrenalin and
adrenalin in human plasma. Clin Chim Acta 30: 587, 1970
2. Engclman K, Portnoy B, Lovenberg W: A sensitive and specific
double-isotope derivative method for the determination of
catecholamines in biological specimens. Am J Med 255: 259,
1968
3. Lake CR, Ziegler MG, Kopin IJ: Use of plasma
norepinephrine for evaluation of sympathetic neuronal function
in man. Life Sci 18: 1315, 1976
4. Bertel O, Buhler FR, Kiowski W, Lutold BE: Decreased betaadrenoreceptor responsiveness as related to age, blood pressure,
and plasma catecholamines in patients with essential hypertension. Hypertension 2: 130, 1980
5. Brecht HM, Banthien F, Ernst W, Schoeppe W: Increased
plasma noradrcnaline concentrations in essential hypertension
and their decrease after long-term treatment with a betareceptor blocking agent (prindolol). Clin Sci Mol Med 51:485s,
1976
6. Brecht HM, Schoeppe W: Relation of plasma noradrenaline to
52
7.
8.
9.
10.
11.
12.
Downloaded from http://hyper.ahajournals.org/ by guest on June 17, 2017
13.
14.
15.
16.
17.
18.
19.
20.
21.
HYPERTENSION
blood pressure, age, sex, and sodium balance in patients with
stable essential hypertension and in normotensive subjects. Clin
Sci Mol Med 55: 81s, 1978
Cousineau D, deChamplain J, Lapointe L: Circulating
catecholamines and systolic time intervals in labile and
sustained hypertension. Clin Sci Mol Med 55: 65s, 1978
DeLecuw PW, Wester A, Punt R, Falke HE, Birkenhager WH:
Noradrenaline levels in essential hypertensives and normotensive controls. Neth J Med 22: 145, 1979
Eide I, Kolloch R, DcQuattro V, Miano L, Dugger R, Van der
Meulen J: Raised cerebrospinal fluid norepinephrine in some
patients with primary hypertension. Hypertension 1: 255, 1979
Eng FWHT, Hubcr-Smith M, McCann DS: The role of sympathetic activity in normal renin essential hypertension.
Hypertension 2: 14, 1980
Esler M, Zweifler A, Randall O, Julius S, DcQuattro V: Agreement among three different indices of sympathetic nervous
system activity in essential hypertension. Mayo Clin Proc 52:
379, 1977
Franco-Morselli R, Elghozi JL, Joly E, DiGiuliu S, Meyer P:
Increased plasma adrenaline concentrations in benign essential
hypertension. Br Med J 2: 1251, 1977
Franco-Morselli R, Baudouin-Lcgros M, Meyer P: Plasma
adrenaline and noradrenaline in essential hypertension and
after long-term treatment with beta-adrenoreceptor-blocking
agents. Clin Sci Mol Med 55: 97s, 1978
Geffen LB, Rush RA, Louis WJ, Doyle AE: Plasma dopamine
beta-hydroxylase and noradrenaline amounts in essential
hypertension. Clin Sci 44: 617, 1973
Henry DH, Luft FC, Weinberger MH, Fineberg NS, Grim CE:
Norepinephrine in urine and plasma following provocative
maneuvers in normal and hypertensive subjects. Hypertension
2: 20, 1980
Hjemdahl P, Eliasson K: Sympatho-adrenal and cardiovascular
response to mental stress and orthostatic provocation in latent
hypertension. Clin Sci 57: 1895, 1979
Hofman A, Boomsma F, Schalekamp MADH, Valkenburg
HA: Raised blood pressure and plasma noradrenaline concentrations in teenagers and young adults selected from an open
population. Br Med J 1: 1536, 1979
Jones DH, Hamilton CA, Reid JL: Plasma noradrenaline, age
and blood pressure: a population study. Clin Sci Mol Med 55:
73s, 1978
Kafka MS, Lake CR, Gullncr HG, Tallman JF, Bartter FC,
Fujita T: Adrenergic receptor function is different in male and
female patients with essential hypertension. Clin Exp Hypertension 1: 613, 1979
Kiowski W, VanBrummclen P, Buhler FR: Plasma
noradrenaline correlates with a-adrenoreceptor-mediated
vasoconstriction and blood pressure in patients with essential
hypertension. Clin Sci 57: 177s, 1979
Kobayashi K, Kolloch R, DeQuattro V, Miano L: Increased
plasma and urinary normetanephrine in young patients with
VOL 3, No
1, JANUARY-FEBRUARY
1981
primary hypertension. Clin Sci 57: 173s, 1979
22. Lake CR, Ziegler MG, Coleman MD, Kopin IJ: Age-adjusted
plasma norepinephrine levels are similar in normotensive and
hypertensive subjects. New Engl J Med 296: 208, 1977
23. Lake CR: Relationship of sympathetic nervous system tone and
blood pressure. Nephron 23: 84, 1979
24. Louis WJ, Doyle AE, Anavekar S: Plasma norepinephrine
levels in essential hypertension. New Engl J Med 288: 599, 1973
25. Miura Y, Kobayashi K, Sakuma H, Tomioka H, Adachi M,
Yoshinaga K: Plasma norepinephrine levels and hemodynamics
in young patients with essential hypertension. Jap Circ J 42:
609, 1978
26. Miura Y, Kobayashi K, Sakuma H, Tomioda H, Adachi M,
Yoshinaga K: Plasma noradrenaline concentrations and
haemodynamics in the early stage of essential hypertension.
Clin Sci Mol Med 55: 69s, 1978
27. Pcdersen EB, Christensen NJ: Catecholamines in plasma and
urine in patients with essential hypertension determined by
double-isotope derivative techniques. Acta Med Scand 198:
373, 1975
28. Philipp T, Distler A, Cordes U: Sympathetic nervous system
and blood pressure control in essential hypertension. Lancet 2:
959, 1978
29. Robertson D, Shand DG, Hollifield JW, Nies AS, Frolich JC,
Oates JA: Alterations in the responses of the sympathetic nervous system and renin in borderline hypertension. Hypertension
1: 118, 1979
30. Sever PS, Osikowska B, Birch M, Tunbridge RDG: Plasmanoradrcnalin in essential hypertension. Lancet 1: 1078, 1977
31. Sever PS, Peart WS, Davies IB, Tunbridge RDG, Godron D:
Ethnic differences in blood pressure with observations on
noradrenaline and renin. 2. A hospital hypertensive population.
Clin Exp Hyper 1: 745, 1979
32. Taylor AA, Pool JL, Lake CR, Ziegler MG, Rosen RA,
Rollins DE, Mitchell JR: Plasma norepinephrine concentrations: no differences among normal volunteers and low, high,
and normal renin hypertensive patients. Life Sci 22: 1499, 1978
33. Vlachakis ND: Blood pressure and catecholamine responses to
sympathetic stimulation in normotensive and hypertensive subjects. J Clin Pharmacol 19: 458, 1979
34. Vlachakis ND, Alcdort L: Platelet aggregation in relationship
to plasma catecholamines in patients with hypertension.
Atherosclerosis 32: 451, 1979
35. Weidmann P, Beretta-Piccoli C, Ziegler WH, Kcusch G, Gluck
Z, Reubi FC: Age versus urinary sodium for judging renin,
aldosterone, and catecholamine levels: Studies in normal subjects and patients with essential hypertension. Kid lnternat 14:
619, 1978
36. Edwards AL: Statistical Methods. New York, Holt, Rinehart,
and Winston, 1967
37. Carruthers M, Taggart P, Conway N, Bates D, Somerville W:
Validity of plasma-catecholamine estimations. Lancet 2: 62,
1970
Plasma norepinephrine in essential hypertension. A study of the studies.
D S Goldstein
Hypertension. 1981;3:48-52
doi: 10.1161/01.HYP.3.1.48
Downloaded from http://hyper.ahajournals.org/ by guest on June 17, 2017
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1981 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/3/1/48
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/