Episodic dopamine discharge in paroxysmal hypertension. Page's syndrome revisited.

Dopamine concentration, a marker of the sympathetic discharge additional to norepinephrine and epinephrine levels, was determined in 31 patients. These patients, mostly women, had essential hypertension and hypertensive episodes that mimicked pheochromocytoma, except that the patients were rather plethoric (instead of pale) and often had associated nausea, epigastric discomfort, and polyuria. During and after hypertensive paroxysms, plasma free norepinephrine and epinephrine levels did not increase, but we found a mean eightfold and 16-fold increase of free and sulfated plasma dopamine levels, respectively, and similar although less marked dopamine level increases in the urine collected following the paroxysm. The hypertensive paroxysms, spontaneous or precipitated by stimulation of the autonomic nervous system, were similar to those described by Page as simulating diencephalic stimulation. Dopamine level may be a marker of the sympathetic discharge, undetected by measurements of free norepinephrine level, and may explain some clinical features of Page's syndrome.

Patient self-management of continuous subcutaneous insulin infusion.

Continuous subcutaneous insulin infusion (CSII) is one of the ways to control blood glucose for prolonged periods. This study was undertaken to establish the long-term feasibility and efficacy of CSII with patient self-management. Patients were instructed to maintain their calorie and carbohydrate intake. Basal infusion of insulin, representing 50% of the total pre-CSII dose, was supplemented by boluses of insulin based on carbohydrate intake for each meal. With this type of regimen, blood glucose and M-values were easily normalized during the physician-directed periods. This study demonstrated that near-normalization of blood glucose, M-values, and glycosylated hemoglobin was maintained after a 1 1/2-yr period of patient self-management. We attributed this successful management in part to the protocol used, in which boluses were related solely to carbohydrate intake while basal insulin was adjusted according to fasting blood glucose. The chronic normalization of blood glucose resulted in improvement of platelet function as witnessed by responsiveness to antiaggregating (PGE1) and aggregating (epinephrine) agents. An improvement was noticed in doppler measurement of ankle-arm blood pressure and a near-normalization of nerve latency and conductivity was observed.

Effects of a beta-blocker or a converting enzyme inhibitor on resistance arteries in essential hypertension.

Seventeen male untreated mild essential hypertensive patients aged 41 +/- 2 years agreed to participate in a double-blind randomized trial to test the effects of antihypertensive treatment on the structure and function of subcutaneous resistance arteries. Patients were treated with either 50 to 100 mg/d atenolol or 2.5 to 5 mg/d cilazapril. Blood pressure before treatment was 148 +/- 6/99 +/- 1 and 147 +/- 2/99 +/- 1 mm Hg, respectively. At 1 year of treatment blood pressure was 131 +/- 4/85 +/- 2 and 132 +/- 2/87 +/- 1 mm Hg, respectively. Resistance arteries (200 to 400 microns lumen diameter) dissected from subcutaneous gluteal biopsies obtained before treatment and at 1 year showed that the media-lumen ratio of arteries from patients treated with cilazapril was reduced to 6.31 +/- 0.21% from 7.54 +/- 0.31% before treatment (P < .05), still slightly but significantly larger (P < .05) than the media-lumen ratio of resistance arteries of normotensive control subjects (5.15 +/- 0.30%). In contrast, in arteries from patients treated with atenolol there was no significant change with treatment (7.97 +/- 0.60% before and 8.07 +/- 0.45% after 1 year of treatment). Active wall tension responses to endothelin-1 were blunted in hypertensive patients and normalized in the cilazapril-treated patients. Depressed active media stress responses to norepinephrine, arginine vasopressin, and endothelin-1 were accordingly normalized in the patients receiving cilazapril as the media width became thinner but were unchanged in those taking atenolol.(ABSTRACT TRUNCATED AT 250 WORDS)

The atrial natriuretic factor in hypertension. State of the art lecture.

Studies were conducted to assess the effects of bolus injections and infusions of human atrial natriuretic factor (ANF) in control subjects and patients with mild essential hypertension, and to measure plasma immunoreactive ANF (irANF) concentration in a large group of patients with essential hypertension. The results are compared with those obtained by other groups on the measurements of plasma irANF in hypertensive patients. It appears that plasma irANF concentrations are not increased in patients with mild essential hypertension despite the evidence of increased preload and of atrial distention as reported by others. This suggests a hyporesponsiveness of the atria to release ANF.

Identification and plasma concentrations of the N-terminal fragment of proatrial natriuretic factor in man.

A specific RIA was developed to measure plasma atrial natriuretic factor (ANF) N-terminal immunoreactivity in man. Antibodies raised in rabbits against a rat ANF N-terminal fragment [ANF-(11-37)] had 100% cross-reactivity with human ANF-(1-30) and purified plasma N-terminal ANF immunoreactivity. The ED80 and ED50 of standard curves prepared using [125I]human ANF-(1-30) and human ANF-(1-30) were 31.5 +/- 5.4 (+/- SD) and 132.5 +/- 20.4 fmol/tube, respectively. The plasma ANF N-terminal peptide concentrations were assayed directly, without extraction, since dilution of plasma and addition of standard to plasma yielded parallel dose-responses in the RIA and virtually 100% recovery of ANF-(1-30) added to plasma. Purification of ANF N-terminal immunoreactivity from 1.5 L human plasma by affinity chromatography and amino acid sequencing suggested that it was closely related to ANF-(1-98), although some degraded peptides were also detected. The mean basal plasma ANF N-terminal peptide level measured in 34 normal subjects was 420 +/- 157 (+/- SD) pmol/L. The values were higher in plasma from patients with congestive heart failure (grades III and IV; 7,041 +/- 6,136 pmol/L; n = 13) or chronic renal failure (10,079 +/- 4,942 pmol/L; n = 20). In 9 patients with chronic renal failure, hemodialysis resulted in a 30% (P less than 0.05) decrease in plasma ANF-(99-126) levels, from 34.7 +/- 12.3 (+/- SD) to 23.2 + 6.1 pmol/L, but no changes in plasma ANF N-terminal peptide concentrations. These data indicate that the N-terminal portion of pro-ANF is cosecreted with ANF-(99-126). Its higher plasma levels in the basal state and during chronic renal failure suggest a different process of elimination than that of ANF-(99-126), which may be partly mediated by the kidney.

Effect of acute and chronic administration of calcitonin on serum glucose in patients with Paget's disease of bone.

The effects of an acute injection of synthetic salmon calcitonin (sCT) and human CT (hCT) and of long term (4-month) administration of sCT on serum glucose levels were investigated in eight patients with Paget's disease of bone. The results obtained demonstrate a small but statistically significant rise in serum glucose after a single sc injection of synthetic hCT. However, the serum glucose level was not increased after 4 months of daily administration of synthetic sCT to our pagetic patients. Our results also substantiate the clinical observation that long term administration of CT does not cause clinical diabetes or significantly change fasting blood glucose concentration. Our results are also consistent with the view that the effect of CT administration on glucose metabolism is related to the secondary hypocalcemia.

The hemodynamic relevance of free and conjugated dopamine in pheochromocytoma.

Two patients with adrenomedullary hypersecretion (confirmed pheochromocytoma and adrenomedullary hyperplasia) presented 15 spontaneous crises associated with hypertension or hypotension with or without tachycardia. Correlation coefficients calculated between extreme values of pulse rates and of systolic and diastolic blood pressures on the one hand and plasma free and conjugated norepinephrine, epinephrine, and dopamine (DA) sampled at the height of the crises on the other, showed no relationships between free or conjugated norepinephrine or epinephrine and blood pressure or pulse rate. However, plasma conjugated DA was negatively correlated with systolic blood pressures (P less than 0.02) and diastolic blood pressures (P less than 0.03) and free plasma DA was negatively correlated with pulse rates (P less than 0.001). These data suggest that the extremely high circulating level of conjugated DA in pheochromocytoma may, in the presence of high circulating conjugated norepinephrine and possibly high alpha-adrenergic receptor occupancy, decrease blood pressure by its predominant action on dopaminergic receptors while elevated free DA may decrease the tendency to tachycardia, possibly by lowering the venous return.

Prazosin plasma concentration and blood pressure reduction.

Prazosin was administered to 16 patients with essential hypertension in an initial dose of 0.5 mg, after which the blood pressure (BP), pulse, and plasma concentrations of prazosin were measured at 0, 0.5, 1, 1.5, 2, 2.5, 3, 4, 6, 8, and 24 hours. The dose of prazosin was then increased over 16 to 20 weeks, and similar sequences of measurements were obtained twice. Eleven patients completed the 20-week course. All patients did not respond in a similar way; two distinct patterns of BP and pulse response emerged, although there was no significant difference in the pharmacokinetic parameters, namely, absorption rate constant (Ka), maximum plasma concentration (Cpmax), time to reach the maximum concentration (Tmax), prazosin plasma half-life (T 1/2), elimination rate constant (kel), prazosin plasma concentration-time curve (AUC), and clearance. Patients in Group 1 had a marked reduction (52/30 mm Hg) of BP after the first dose of prazosin, no pulse increase, and needed a small dose of prazosin to maintain an adequate BP response. Patients in Group 3 had a minimal reduction in BP (14/13 mm Hg) after a first dose, a significant pulse increase, and needed a high dose of prazosin to control their BP. We conclude that this effect might be due to a different drug-receptor interaction, and the BP response and dose could be predicted from the response of the first dose of prazosin.

Regional sources of free and sulfoconjugated catecholamines in hypertension.

To elucidate the sources of free catecholamines (CA) and their sulfates in hyperadrenergic essential hypertensives (EH), their arteriovenous differences were determined radioenzymatically and by sulfatase hydrolysis (with correction for cross-contamination) across several organs and regions in 16 hyperadrenergic essential hypertensive patients. Comparison with arterial concentrations showed that: the adrenal venous outflow contains 240 times more free epinephrine (E), 55 times more free norepinephrine (NE), and 7 times more free dopamine (DA) concentrations, but E, NE, and DA sulfates are not different; free E concentrations are lower in the peripheral venous blood; NE sulfate concentrations are higher in the superior vena cava (p less than 0.05 for all differences noted). The data suggest the following conclusions for hyperadrenergic EH patients: with the exception of NE sulfate added into the superior vena cava region, no other organ or region can be associated with a net DA or NE sulfate release. The proportional adrenal vein concentrations of DA:NE:E are approximately 1:10:50, which are very close to those seen in other studies performed under different degrees of stress. Free E is extracted in peripheral tissues. The DA surges in hyperadrenergic EH patients probably result from the pulsatile, predominantly adrenal, release of free DA.

Dopamine surges in hyperadrenergic essential hypertension.

From a total of 61 referred hypertensive patients, 21 were clinically suspected of pheochromocytoma but in none was this diagnosis confirmed. Instead we found nine of the 21 patients had surges of conjugated dopamine during hyperadrenergic periods unaccounted for by rise in norepinephrine (NE) or epinephrine (E). Overall, essential hypertensive (EH) patients had in plasma (ng/ml) higher conjugated dopamine (DA) (2.3 +/- 0.2 vs 1.0 +/- 0.1, p less than 0.01), increasing with age (p less than 0.01), lower conjugated NE + E (0.6 +/- 0.1 vs 1.2 +/- 0.2, p less than 0.01), and higher free E (p less than 0.007), lower urinary free DA and total DA but higher free NE + E excretions (each p less than 0.05) than 24 control subjects. Following the DA surges, a short-lived urinary overflow of total DA occurred. The patients with DA surges were older, had a higher incidence of low conjugated NE + E (less than 0.23 ng/ml), a higher proportion of arterial free DA, and higher venous baseline conjugated plasma DA than the rest of the patients. Patients with low conjugated NE + E had in turn higher plasma DA concentrations at several regional sampling sites than patients with normal conjugated NE + E. High conjugated DA in EH probably results from pulsatile DA surges leading to a rise of baseline plasma conjugated DA. In the short run DA pulses can result in temporary alpha- and beta-adrenergic actions of huge arterial free DA concentrations prior to DA conjugation; in the long run the excessive high affinity DA conjugation may take preference to the lower affinity NE and lowest affinity E conjugation and free E increases. Both result in an acute or chronic increase of sympathetic tone.

Unconjugated hyperepinephrinemia: a hallmark of hypertension imitating pheochromocytoma?

Hypertensive patients with elevated and hyperresponsive plasma norepinephrine and epinephrine (NE + E) associated with low conjugated NE + E were previously identified by determination of the sum of NE + E. Because of their excessive E but not NE responses to glucagon and also hypertension corresponding to E excess, we explored whether an elevated unconjugated E resulting from a selective E conjugation defect could be obscured by the sum of NE + E. We found that nine patients with elevated E (reflected by the normal 4:1 ratio of plasma NE to E reversed in favor of E), had, when compared to 31 patients with plasma NE exceeding E:1) lower plasma conjugated E (mean 0.03 vs 0.27 ng/ml, p less than 0.01), lower degree of E conjugation (8 vs 51%, p less than 0.01), and a higher maximum systolic (p less than 0.05), pulse pressure (p less than 0.02) and higher pulse rates (p less than 0.04), but no differences in the unconjugated and conjugated proportions of plasma NE; and 2) an absence of conjugated E throughout the circulation and relative preponderance of E over NE at sampling points close to the peripheral venous blood (p less than 0.05). The absolutely and relatively decreased plasma conjugated E in patients with E exceeding NE (without difference in conjugated NE) is a preliminary indication that a selective sulfoconjugating defect of E results in plasma E higher than NE in accordance with the hyper-beta-adrenergic features of their hypertension. Epinephrine, a circulating hormone, is more dependent on conjugated E reflect better this defect than those measuring the sum of NE and E.

Role of tumor necrosis factor-alpha gene locus in obesity and obesity-associated hypertension in French Canadians.

Obesity represents a serious risk factor for the development of cardiovascular diseases, including hypertension. Segregation studies suggest that obesity and obesity-associated hypertension may share some genetic determinants. The results of the present candidate gene investigation suggest that in hypertensive pedigrees of French-Canadian origin, one such determinant is the tumor necrosis factor (TNF)-alpha gene locus. Gender-pooled quantitative sib-pair analysis demonstrated a significant effect of the gene locus on 3 global and 7 regional measures of obesity (P=0.05 to 0.0004). Gender-separate quantitative sib-pair analyses showed that the impact of the locus on obesity is most significant in the abdominal region in men and in the thigh region in women. Furthermore, the haplotype relative-risk test demonstrated a significant association between the TNF-alpha gene locus and both obesity (P=0.006) and obesity-associated hypertension (P=0.02). These effects were most significant in individuals with nonmorbid obesity. In conclusion, the results of linkage and association analyses suggest that in hypertensive pedigrees of French-Canadian origin, the TNF-alpha gene locus contributes to the determination of obesity and obesity-associated hypertension. In addition, the data indicate that gender modifies the effect of the locus on the regional distribution of body fat.

Enhancement of nadolol elimination by activated charcoal and antibiotics.

This study was carried out to assess whether nadolol undergoes enterohepatic circulation. Eight healthy subjects received 80 mg nadolol orally on three occasions at least 2 wk apart. The first experiment was a control. The second consisted of nadolol followed in 3 hr by 3 gm activated charcoal given over a 9-hr period. In the third, the subjects received 0.5 gm erythromycin base and 0.5 gm neomycin four times a day orally for 2 days before nadolol. After the activated charcoal, the nadolol AUC fell from 2455 +/- 155 to 1355 +/- 123 ng . hr/ml (mean +/- SE), as did the percentage nadolol recovered in urine (15.4 +/- 1.4 to 10.2 +/- 0.7%) and the nadolol t1/2 (17.3 +/- 1.7 to 11.8 +/- 1.6 hr). These data suggest that nonrenal elimination increased. After the antibiotics, nadolol AUC was constant, percentage of nadolol recovered in urine fell to 12.7 +/- 1.7%, nadolol t1/2 fell to 11.6 +/- 1.3 hr, and mean peak nadolol concentration rose from 146 +/- 15 to 397 +/- 52 ng/ml. These results suggest that there is an enterohepatic circulation for nadolol, that activated charcoal may decrease nadolol bioavailability, and that antibiotics may increase the nadolol effect.

Effects of phenobarbital and tobacco smoking on furosemide kinetics and dynamics in normal subjects.

This study was carried out to determine whether furosemide (F) kinetics and dynamics were influenced by phenobarbital and tobacco smoking. Our subjects were 10 normal men: five nonsmokers (NS) and five smokers (S). They received a single intravenous F injection of 40 mg. Regular serum and urine collections were made. In a second study, the NS group received 100 mg phenobarbital orally for 15 days and then a second dose of F. Cumulative 8-hr urinary excretion of sodium was identical for NS, NS with phenobarbital, and S at 345 +/- 30, 357 +/- 29, and 353 +/- 25 mmol. Diuresis was smaller by 800 ml (20%) in S than in NS. F increased endogenous creatinine clearance from 117 +/- 13 to 196 +/- 17 ml/min in NS and from 110 +/- 12 to 222 +/- 30 ml/min in NS with phenobarbital. In the S group, endogenous creatinine clearance showed a tendency to increase only slightly, from 136 +/- 23 to 180 +/- 34 ml/min. The increase in free water clearance caused by F was smaller in the S group than in the NS group (P less than 0.05). Protein binding and distribution of F were not affected by phenobarbital or tobacco smoking. F clearance was slightly higher in S than in NS, which was primarily the result of a slight increase in extrarenal F clearance. In the NS group, F clearance remained constant after phenobarbital. It is concluded that tobacco smoking in normal subjects affects the diuretic response to F without modifying kinetics.

Tixocortol pivalate, a corticosteroid with no systemic glucocorticoid effect after oral, intrarectal, and intranasal application.

Tixocortol pivalate is a corticosteroid with topical anti-inflammatory activity equal to that of hydrocortisone. It was evaluated in a group of 18 normal subjects to determine whether it exerted any systemic glucocorticoid activity after single oral or intrarectal doses and after short-term dosing by the intranasal route. Effects of tixocortol pivalate were compared to those of oral dexamethasone and intrarectal betamethasone 21-phosphate. By the three routes, tixocortol pivalate does not induce any changes in plasma cortisol, leukocyte counts (neutrophils, lymphocytes, monocytes, eosinophils), blood glucose, or 24-hr urinary excretion of sodium and potassium, whereas there were changes after dexamethasone and betamethasone. Tixocortol pivalate, however, increased urinary free cortisol-like substances. It is concluded that tixocortol pivalate given for short periods by nonparenteral routes does not induce a measurable systemic glucocorticoid effect.

Basis of false-positive glucagon tests for pheochromocytoma.

In more than half of 67 patients suspected of having pheochromocytoma, glucagon stimulation increased plasma free norepinephrine (NE) and epinephrine (E) 50% or more, with rising blood pressure or pulse rate; only three patients, however, harbored a pheochromocytoma. A low degree of catecholamine conjugation accounts for most of the false-positive results. In patients with low conjugated NE +E there was a greater rise in free NE +E and free E as well as in pulse rate after glucagon stimulation than in those with normal levels of conjugated NE+E. Glucagon-sensitive adenylate cyclase was found in pheochromocytomas but not in a functional adrenocortical adenomas. After sham administration of glucagon, there were rises in blood pressure but not in free NE or E in four patients. The glucagon-induced catecholamine test can be false-positive in hyperadrenergic essential hypertensive patients with abnormally low conjugated NE +E. Saline alone in a sham glucagon test in susceptible patients raises systolic blood pressure and pulse rate, and therefore, if plasma free NE and E are measured and found not to rise this type of false-positive result can be eliminated.

Effect of metoclopramide on plasma catecholamine release in essential hypertension.

The catecholamine (CA)-releasing action of metoclopramide (MCP) observed in patients with pheochromocytoma was tested in 20 subjects with essential hypertension and compared with the same effect of glucagon in 10 of them. We found that even in the absence of pheochromocytoma, MCP is a CA-releasing substance, moderately increasing systolic blood pressure and pulse rate. The release of CA is reflected by an increase in concentrations of free norepinephrine and total (free plus sulfated) epinephrine 3 minutes and of total dopamine and norepinephrine 10 minutes after the MCP bolus dose, whereas glucagon had an effect on the release of free epinephrine. Regional catheterization before and after MCP dosing in one subject showed a considerable increase in adrenal epinephrine and norepinephrine concentrations 45 seconds after the MCP bolus dose. MCP has a free CA-releasing potency much like that of glucagon. Because the released free CA is readily sulfoconjugated, the effect on CA release can be more easily detected when conjugated CA is determined. MCP should thus be used with caution in pheochromocytoma as well as in other forms of hypertension.

Single-dose pharmacokinetics of 14C-lovastatin in chronic renal failure.

An open study on the pharmacokinetics of lovastatin was conducted in six patients with chronic renal failure (mean creatinine clearance, 0.40 ml/sec; range, 0.20 to 0.65 ml/sec) and seven healthy subjects. Plasma levels of 3-hydroxy-3-methylglutaryl-coenzyme reductase inhibitory activity (total and active) and total radioactivity were determined over 168 hours after a single dose of 80 mg 14C-lovastatin. The mean area under the plasma concentration-time curve for active inhibitors were 606 +/- 346 and 282 +/- 138 ngEq.hr/ml (p = 0.04) in patients and control subjects, respectively. Total inhibitors in plasma and total radioactivity were similarly elevated in patients with chronic renal failure. Results indicate that patients with severe renal dysfunction have altered elimination kinetics of lovastatin. Current ongoing clinical studies in patients with renal dysfunction will allow better assessment of the pharmacodynamic meaning of our observations.

Effect of aminoglycosides on the disposition of thyroid hormones and thyroglobulin.

Our study was designed to confirm the potential effects of three aminoglycosides on the disposition of thyroid hormones. Twenty-seven patients diagnosed with either cellulitis (n = 19), chronic osteitis (n = 4), or an abscess (n = 4) were selected. Thirteen patients received tobramycin, 60 to 100 mg iv q. 8 h., plus cloxacillin, 1 gm iv q. 4 h.; seven patients received netilmicin, 40 to 120 mg iv q. 8 h., plus cloxacillin, 1 gm iv q. 4 h.; and seven patients received either cloxacillin, 1.5 gm iv q. 4 h., or cefoperazone, 2 to 4 gm iv q. 12 h. for at least 7 days. Another group of six normal subjects received neomycin, 0.5 gm po q. 6 h. for 7 days. All these subjects had normal thyroid function before antibiotic dosing and none had thyroid function abnormalities. Tobramycin and cloxacillin/cefoperazone did not influence thyroid function. Netilmicin decreased the total serum concentrations of triiodothyronine (T3) from 114 +/- 9 to 75 +/- 7 ng/dl (P less than 0.01), probably because of increased clearance, as the T3 free fraction increased from 0.43% +/- 0.02% to 0.49% +/- 0.02% (P less than 0.05). Thyroxine (T4) and reverse T3 (rT3) levels were not affected. Neomycin decreased T3 levels from 104 +/- 8 to 92 +/- 7 ng/dl (P less than 0.05) and the serum concentrations of thyroglobulin from 17.3 +/- 2.0 to 11.7 +/- 2.0 ng/ml (P less than 0.001). Because T4 and rT3 levels did not change, our results suggest that neomycin may have directly affected the gland. We conclude that some aminoglycosides can alter the disposition of thyroid hormones.

Blunted effects of endothelin upon small subcutaneous resistance arteries of mild essential hypertensive patients.

OBJECTIVE: In experimental models of hypertension in the rat, resistance arteries present a blunted response to endothelin, a potent vasoconstrictor peptide. The primary objective of this study was to investigate whether, as in hypertensive rat blood vessels, the response of human resistance arteries to endothelin was altered in essential hypertensive patients, in order to further understand the possible physiopathological involvement of this peptide in human hypertension. DESIGN: Normotensive male subjects and sex- and age-matched mild essential hypertensive patients who had not received antihypertensive drugs for more than 6 months were investigated. METHODS: Small arteries were dissected from gluteal subcutaneous biopsies and mounted on a wire-myograph. Blood vessels were measured and dose-response curves to different agents tested. RESULTS: The external diameter of blood vessels of the hypertensive patients tended to be smaller and the width of their media tended to be thicker, but the cross-sectional area of the wall was similar in both groups. Lumen diameters were significantly smaller in hypertensives and the media:lumen ratio was significantly increased in hypertensive patients. Active tension responses and sensitivity to norepinephrine, arginine vasopressin and angiotensin II were similar in both groups, but calculated active pressure responses were enhanced in hypertensives due to the smaller blood vessel lumen. Tension responses to endothelin-1 at increasing concentrations of 0.1 to 100 nmol/l were lower in hypertensive patients, but the calculated transmural active pressure developed was not significantly different at or above 10 nmol/l. CONCLUSION: These results suggest that gluteal subcutaneous small resistance arteries of male essential hypertensive patients exhibit a decrement in responsiveness to endothelin-1. The altered design of the hypertensive blood vessels enhanced calculated pressure responses, which may contribute to the maintenance of elevated blood pressure.