sign in sign up
<<
Home , Blood pressure, Mean arterial pressure, Pulse pressure

Journal of Human (2001) 15, 775–779  2001 Nature Publishing Group All rights reserved 0950-9240/01 $15.00 www.nature.com/jhh ORIGINAL ARTICLE Determinants of mean arterial and pressure in chronic haemodialysis patients

AB Abdelfatah, G Motte, D Ducloux and JM Chalopin Department of Nephrology, Dialysis, and Renal Transplantation, Saint Jacques Hospital, Besanc¸on, France

Hypertension is highly prevalent in the dialysis popu- 1.84; 95% CI, 1.07–3.18, for 1% increase in IWG), and lation, and has been implicated in the pathogenesis of current smoking (RR, 2.59; 95% CI, 1.13–5.92) and nega- the observed excess of cardiovascular morbidity and tively with Hb concentration (RR, 0.92; 95% CI, 0.84– mortality in these patients. Nevertheless, there are no 0.99, for a 1 g/100 ml in Hb). Mean predialytic MAP was -reports on the clinical and biochemical determinants of 98 ؎ 15 mm Hg and significantly decreased after dialy -both (PP) and sis (91 ؎ 16 mm Hg; P Ͻ 0.0001). In multivariate analy (MAP) in dialysis populations. A total of 541 haemodia- sis, a 10 mm Hg increase in MAP was positively associa- lysed patients from 11 dialysis centres were included ted with parathyroid (PTH) (RR, 1.32; 95% CI, in the study. The demographic, clinical, and biological 1.15–1.6, for 50 ng/ml in PTH), erythropoietin (EPO) characteristics were recorded. Both pre- and post- treatment (RR, 1.09; 95% CI, 1.03–1.16), and current dialytic (systolic and diastolic) were smoking (RR, 1.87; 95% CI, 1.39–2.41). PP and MAP are measured. PP and MAP were calculated. Mean predia- associated with different clinical parameters. Most of -lytic PP was 67 ؎ 17 mm Hg and significantly decreased these factors are potentially reversible. Smoking cess after dialysis (60 ؎ 18 mm Hg; P Ͻ 0.0001). In multivari- ation, correction of anaemia and limitation of IWG ate analysis, a 10 mm Hg increase in PP was positively should be important challenges for physicians in care associated with age (RR, 2.01; 95% CI, 1.35–5.01, for a of dialysis patients. 10-year increase in age), mellitus (RR, 1.08; Journal of Human Hypertension (2001) 15, 775–779 95% CI, 1.04–1.14), interdialytic weight gain (IWG) (RR,

Keywords: mean arterial pressure; pulse pressure; haemodialysis

Introduction (increased sympathetic activity, fluid retention, acti- vation of the - axis, etc) have been (CVD) is the major cause of incriminated. Moreover, some clinical and bio- mortality both in the general population and in chemical parameters, including age, interdialytic patients with end-stage renal disease (ESRD) weight gain, erythropoietin (EPO) treatment, and accounting for approximately 40% of all deaths in 1,2 hyperparathyroidism have been also associated with both demographic groups. Nevertheless, com- hypertension in this patient category.4 pared to the general population, dialysis patients consists of both a steady compo- have disproportionately high rates of arteriosclerotic nent (mean arterial pressure, MAP) and a pulsatile outcomes, and CVD mortality in this patient cate- component (pulse pressure, PP). Major determinants gory is 10 to 20 times higher than in the general 3 of MAP are ventricular ejection and peripheral vas- population. Hypertension is highly prevalent in the 7,8 1,4 cular resistance. PP, the difference between sys- dialysis population, and has been implicated in tolic blood pressure (SBP) and diastolic blood press- the pathogenesis of the observed excess of cardio- 5,6 ure (DBP), is also made up of two major vascular morbidity and mortality in these patients. components, one due to ventricular ejection inter- The pathophysiology of dialysis-associated acting with the viscoelastic properties of the large hypertension is multifactorial. Several mechanisms and the other due to wave reflection.7,8 The distinction between the two components of blood pressure is clinically relevant because new evidence Correspondence: Dr Didier Ducloux, Department of Nephrology, Dialysis, and Renal Transplantation, Saint Jacques Hospital, suggest that PP is a more sensitive measure of car- Besanc¸on, France. E-mail: adjusyȰwanadoo.fr diovascular risk than other indices of blood pressure Received 14 November 2000; revised and accepted 19 June 2001 (DBP, SBP, and MAP).9 Moreover, arterial stiffening Determinants of MAP and PP in chronic haemodialysis patients AB Abdelfatah et al 776 is very pronounced in renal patients accounting for used in analyses were the means of all predialysis an increase in PP in this population compared to during 1 week. controls.10 Age, gender, dry weight, size, residual renal func- Nevertheless, there is no report on the clinical and tion, haemodialysis duration, weekly haemodialysis biochemical determinants of both PP and MAP in session duration, diabetes mellitus, smoking status, the dialysis population. Such a knowledge may help compliance, past history of cardiovascular events physicians in the comprehensive approach of hyper- (CVE), binephrectomy or parathyroidectomy, anti- tension in this patient category. We reported base- hypertensive drugs, and EPO use were assessed line data of a prospective survey study in 541 hemo- through anamnesis and medical records. dialysed patients. Blood pressure was measured and both PP and MAP were calculated to determine fac- Residual renal function: Data on estimated urine tors associated with the different components of volume were collected and reported as a dichot- blood pressure. omous variable, less than or greater than 200 ml/24- h. Patients with estimated urine output of less than Patients and methods 200 ml/24 h were considered to have lost their residual renal function (RRF) whereas patients with Patients characteristics estimated urine output greater than 200 ml/24 h A total of 541 patients with coronary disease were considered to have preserved RRF. were included in a prospective survey study of car- diovascular morbidity and mortality in April 2000. Compliance: Behavioural compliance was evalu- The demographic characteristics of the study popu- ated by physicians in each centre and categorised as lation are depicted in Table 1. All the dialysis good or moderate. centres in Bourgogne-Franche Comte´ were asked to participate in the study. The inclusion criteria Smoking behaviour: With respect to smoking included coronary heart disease for at least 3 months behaviour, subjects were categorised as current or and the absence of acute illness at the time of the former smokers and those who never smoked. study. Primary renal disease: Primary renal disease was Blood pressure assessed through medical records and categorised as chronic glomerulonephritis, chronic interstitial Predialytic blood pressure (SBP, DBP) was measured nephritis, polycystic disease, diabetes mel- by a nurse using a semi-automatic device, based on litus, nephroangiosclerosis, or undetermined. an oscillometric method with the patient in a sitting position after resting for more than 5 min before Nutritional status: Normalised protein catabolism starting the dialysis session. Postdialytic blood rate (nPCR) and albumin concentration were pressure was measured using the same method determined. Body mass index (BMI) was calculated 10 min after the end of the dialysis session. Pulse (weight (kg)/height (m)2). pressure (SBP − DBP) was calculated. MAP was defined as DBP plus one-third of PP. The values Dialysis adequacy: Kt/V was assessed using Dau- girdas 2 formula. Weekly dialysis time was also recorded. Table 1 Demographic, clinical and biological characteristics of the study population Ultrafiltration: Ultrafiltration was calculated as the Age (years) 66 ± 13 patient’s weight at the beginning of each session Gender (M/F) 340/201 minus the weight after the session, divided by the Diabetes 19% nephrologist’s determined dry weight. Ultrafil- Past history of cardiovascular disease 51% Binephrectomy 3.6% tration was calculated on the basis of the average of Current smokers 26% all measurements in the study week. BMI 24 ± 4.1 Haemodialysis duration (months) 41 ± 44 Haemodialysis session duration (hour/week) 12.1 ± 1.1 Statistical analysis UF rate 3.6 ± 1.4 Kt/V 1.47 ± 0.33 Results are given as mean Ϯ s.d. nPCR 1.04 ± 0.24 Initially, univariate analyses were carried out in Albumin (g/l) 39 ± 4 Hb (g/100 ml) 10.7 ± 1.9 order to examine the relationship between predialy- PTH (ng/ml) 236 ± 248 sis PP (or MAP), and a number of potential inde- Predialytic PP (mm Hg) 67 ± 17 pendent variables. Among the latter, covariates to Postdialytic PP (mm Hg) 60 ± 18 enter multivariate analyses were selected as follows: Predialytic MAP (mm Hg) 98 ± 15 Postdialytic MAP (mm Hg) 91 ± 16 • Continuous variables were tested using Student’s t-test or Mann–Whitney’s non-parametric test.

Journal of Human Hypertension Determinants of MAP and PP in chronic haemodialysis patients AB Abdelfatah et al 777 • Categorical variables were tested using Pearson’s Table 3 Factors associated with predialytic PP chi-square, or Fisher’s exact test. Parameter RR CI The alpha threshold for covariate selection was 0.20. Such variables linked with PP (or MAP) with P- Age (for a 10 years increase in age) 2.01 1.35–5.01 values less than 0.20 were retained for multivariate Diabetes mellitus 1.08 1.04–1.14 analyses, medical knowledge of the variables guid- IWG (for 1% increase in IWG) 1.84 1.07–3.18 Smoking status 2.59 1.13–5.92 ing the final choice. Selected covariates were Haemoglobin (for a 1 g/100 ml in Hb) 0.92 0.84–0.99 entered into stepwise multiple-regression analysis. Starting with as many covariates as were chosen (either statistically or otherwise), the software per- formed backwards stepwise selection, the limit to Mean arterial pressure remove a covariate being a P-value of 0.10. Mean predialytic MAP was 98 Ϯ 15 mm Hg and sig- nificantly decreased after dialysis (91 Ϯ 16 mm Hg; P Ͻ 0.0001). Results In monovariate analysis, age (P = 0.04), antihyper- = Demographic characteristics of the population tensive drugs (P 0.12), past history of cerebrovas- cular disease (P = 0.08), nPCR (P = 0.05), albumin Twenty-one patients with chronic arterial hypoten- level (P = 0.06), RRF (P = 0.12), parathyroid hormone sion (predialysis SBP Ͻ100 mm Hg and/or post- (PTH) adjusted for parathyroidectomy (P = 0.007), dialysis SBP Ͻ90 mm Hg) were excluded from EPO treatment (P = 0.01), BMI (P = 0.04), and current analysis. The demographic characteristics and the smoking (P = 0.002) were associated with MAP. laboratory values of the study population are shown In multivariate analysis, a 10 mm Hg increase in in Table 1. MAP was positively associated with PTH, adjusted for parathyroidectomy (RR, 1.32; 95% CI, 1.15–1.6, for a 50 ng/ml in PTH), EPO treatment (RR, 1.09; Pulse pressure 95% CI, 1.03–1.16), and current smoking (RR, 1.87; 95% CI, 1.39–2.41) (Table 3). Mean predialytic PP was 67 Ϯ 17 mm Hg and sig- nificantly decreased after dialysis (60 Ϯ 18 mm Hg; P Ͻ 0.0001). Discussion = In monovariate analysis, age (P 0.001), antihy- This study reports that, in a population of chronic, pertensive drugs (P = 0.13), past history of coronary stable haemodialysis patients. PP and MAP are asso- disease (P = 0.07), diabetes mellitus (P = 0.01), = ciated with different determinants. PP and MAP are haemodialysis duration (P 0.09), haemodialysis related to different physiological mechanisms. Pulse session duration (P = 0.02), interdialytic weight gain pressure, the difference between peak SBP and end- (IWG) (P = 0.002), albumin level (P = 0.05), RRF (P = = = DBP, represents the pressure increment over and 0.12), BMI (P 0.04), and current smoking (P above the existing DBP that results from ventricular 0.004) were positively associated with PP. Haemo- contraction and ejection of arterial blood into the globin concentration (P = 0.01) and binephrectomy = . At any given ventricular ejection and heart (P 0.18) were negatively correlated with PP. rate, large artery stiffness determines PP. Pulse In multivariate analysis, a 10 mm Hg increase in pressure, therefore, is the surrogate measurement of PP was positively associated with age (RR, 2.01; pulsatile opposition to blood flow during . In 95% CI, 1.35–5.01, for a 10-year increase in age), contrast, and peripheral vascular diabetes mellitus (RR, 1.08; 95% CI, 1.04–1.14), resistance influence the MAP equation. In the pres- IWG (RR, 1.84; 95% CI, 1.07–3.18, for 1% increase ence of a normal cardiac output, MAP is thought to in IWG), and current smoking (RR, 2.59; 95% CI, be the surrogate measure of static resistance to blood 1.13–5.92) and negatively with Hb concentration flow provided by during . Besides (RR, 0.92; 95% CI, 0.84–0.99, for a 1 g/100 ml in Hb) these physiological differences, our study outlines (Table 2). the role of different clinical and biochemical deter- minants on these two components of blood pressure. Previous studies have shown a progressive increase in blood pressure with aging, beginning in Table 2 Factors associated with predialytic MAP childhood and progressing throughout adulthood.11 The aging process in humans is associated with pro- Parameter RR CI gressive stiffening of the aortic and other elastic arteries. Age causes arterial elastin to decrease while PTH adjusted for PTx (for a 50 ng/ml 1.32 1.15–1.6 12 in PTH) collagen increases. These vascular changes cause a EPO use 1.09 1.03–1.16 substantial increase in ascending aortic PP. Systolic Current smoking 1.87 1.39–2.41 blood pressure increases with age due to aortic impedance and early return of wave reflections. In

Journal of Human Hypertension Determinants of MAP and PP in chronic haemodialysis patients AB Abdelfatah et al 778 the general population, aortic MAP also increases, level was the major determinant of blood pressure.28 resulting from increased peripheral resistance.13 Moreover, recent observations linking the correction Nevertheless, we do not observe any effect of aging of hyperparathyroidism by either admin- on MAP in our study population. istration29 or parathyroidectomy30 in chronic dialy- Current smoking was strongly associated with sis patients, resulting in a lower blood pressure, both MAP and PP in dialysis patients. Cigarette have supported this hypothesis. In our study, we smoking has been previously associated with hyper- found a correlation between PTH levels and predi- tension14 or resistance to hypertensive drugs15 in alysis MAP, but not PP. other populations. Moreover, adminis- We also found a positive relationship between tration has been associated with hypertension in recombinant human EPO (rhuEPO) therapy and rats.16 Several mechanisms are probably involved in MAP. Hypertension has been associated with the smoking-induced hypertension. Cigarette smoking treatment of anaemia with rhuEPO in chronic is associated with a release of the sympathetic neur- haemodialysis patients.31 Thus, the European Multi- otransmitter, noradrenaline, as well as the adreno- center Study32 reported that EPO (250 U/kg IV medullary hormone adrenaline.17 Furthermore, sev- weekly) increased MAP by 4 mm Hg at 6 months, eral clinical and experimental studies suggest that but not any further after 1 year. The Canadian Multi- smoking interferes with prostacyclin and thrombox- center Study33 also reported that EPO (300 U/kg IV ane A2 metabolism in the ,18,19 as well weekly) did not alter SBP at rest; however DBP as nitric oxide and -1.20,21 Moreover, the increased from 12 mm Hg after 6 months. These plasma concentration of the latter was found to be results are consistent with the association between increased in smokers.22 These mechanisms may con- rhuEPO and MAP, and not PP. The pathophysiology tribute to increase and thus of EPO-related hypertension is complex, but prim- MAP. On the other hand, smoking-related macrovas- arily involves an increase in vascular resistance. A cular disease and could account for direct vasopressor effect of EPO has been sug- the increased PP observed in smokers. gested,34 but other authors have reported that EPO is Remodelling of large arteries in ESRD patients is, not directly vasoconstrictive, but enhances vascular in part, associated with hyperkinetic circulation and responsiveness in vitro by increasing the release of increased cardiac output. In these patients the endothelin and vasoconstrictor prostanoids and haemodynamic effects of increased blood flow on decreasing the release of prostacyclin.35–37 the macrocirculation have been widely reported.10 The present observations have some limitations Our results confirm that factors associated with hyp- principally because of their cross-sectional and erkinetic circulation and fluid overload are also observational nature, and interventional studies are associated with increased PP. In our study, IWG was necessary to delineate the precise roles of various highly correlated with PP. The role of interdialytic factors associated with increased PP or MAP. weight gain and in the physiopa- Indeed, an association between two factors do not thology of hypertension in dialysis patients is con- imply a causal relationship and such a cross- troversial. Many studies have shown no correlation sectional analysis must be regarded as hypothesis- between IWG and blood pressure.23,24 Salem et al,25 generating. in an epidemiological study of change in blood Pulse pressure and MAP are associated with dif- pressure over 1 year, showed no relationship ferent clinical parameters. Most of these factors are between interdialytic weight gain and blood press- potentially reversible. In the view of a predominant ure. Cheigh et al,26 using 24-h ambulatory blood role for PP in cardiovascular risk, smoking cess- pressure , showed that the predialysis ation, correction of anaemia and limitation of IWG blood pressure change did not correlate with weight should be important challenges for physicians who gain during the interdialytic period. Dionisio et al27 are caring for dialysis patients. showed a strong correlation between total body water measured by bioelectrical impedance and mean 24-h blood pressure. The present study References strongly supports the contention that fluid overload 1 US Renal Data System: USRDS Annual Data Report, is an important factor affecting blood pressure in Bethesda, MD, National Institutes of , National haemodialysis patients. Haemoglobin concentration Institute of Diabetes and Digestive and Kidney Dis- was also found to be negatively associated with pre- eases, 1997. dialytic PP. Anaemia is associated with a hyperdyn- 2 US Department of Health and Human Services: Mor- amic circulation associated with increased bidity and Mortality: Chartbook on Cardiovascular, volume. The increase in ventricular ejection and , and Blood Diseases, Bethesda, MD, US Depart- may account for the role of anaemia in ment of Health and Human Services, 1996. 3 Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemi- increasing PP. ology of cardiovascular disease in chronic renal dis- There may be a correlation between an increase ease. Am J Kidney Dis 1998; 32 (Suppl 3): 112–119. in intracellular calcium levels induced by PTH 4 Rahman M, Fu P, Sehgal AR, Smith MC. Interdialytic excess and hypertension. In a recent study using weight gain, compliance with dialysis regimen, and ambulatory blood pressure monitoring, the PTH age are independent predictors of blood pressure in

Journal of Human Hypertension Determinants of MAP and PP in chronic haemodialysis patients AB Abdelfatah et al 779 patients. Am J Kidney Dis 2000; 35: 22 Haak T, Jungmann E, Raab C, Usadel KH. Elevated 257–65. endothelin-1 levels after cigarette smoking. Metab- 5 Foley RN et al. Impact of hypertension on cardiomy- olism 1994; 43: 267–269. opathy, morbidity and mortality in end-stage renal dis- 23 Coomer RW, Schulman G, Breyer JA, Shyr Y. Ambulat- ease. Kidney Int 1996; 49: 1379–1385. ory blood pressure monitoring in dialysis patients and 6 Zager PG et al. ‘U’ curve association of blood pressure estimation of mean interdialytic blood pressure. Am J and mortality in hemodialysis patients. Kidney Int Kidney Dis 1997; 29: 678–684. 1998; 54: 561–569. 24 Conion PJ. Predialysis systolic blood pressure corre- 7O’Rourke MF. Arterial function in health and disease. lates strongly with mean 24 hour systolic blood press- Churchill Livingstone: Edinburgh, UK, 1982. ure and left ventricular mass in stable hemodialysis 8 Nichols WW, Nicolini FA, Pepine CJ. Determinants of patients. J Am Soc Nephrol 1996; 7: 2658–2663. isolated in the elderly. J Hyper- 25 Salem MM, Davis M. Effect of one year of hemodia- tens 1992; 10 (Suppl 6): S73–S77. lysis on weight and blood pressure in 434 patients. 9 Franklin S. Ageing and hypertension: the assessment Artif Organs 1997; 21: 402–404. of blood pressure indices in predicting coronary heart 26 Cheigh JS et al. Hypertension is not adequately con- disease risk. J Hypertens 1999; 17 (Suppl 5): S29–S36. trolled in hemodialysis patients. Am J Kidney Dis 10 London GM et al. Cardiac and arterial interactions in 1992; 19: 453–459. end-stage renal disease. Kidney Int 1996; 50: 600–608. 27 Dionisio P et al.Influence of the hydration state on 11 Whelton PK. Blood pressure in adults and the elderly. blood pressure values in a group of patients on regular In: Bulpitt CJ (ed). Handbook of Hypertension. Vol 6. maintenance hemodialysis. Blood Purif 1997; 15:25– Elsevier: Amsterdam, 1985, pp 51–69. 33. 12 Nichols WW, Avolio AP, Kelly RP, O’Rourke MF. 28 Goldsmith DJ, Covic AC, Venning MC, Ackrill P. Effects of age and of hypertension on wave travel and Ambulatory blood pressure monitoring in renal dialy- reflections. In: O’Rourke MF, Safar M, Dzau V (eds). sis and transplant patients. Am J Kidney Dis 1997; 29: Arterial . Mechanisms and Therapy. 593–600. Lea & Febiger: Philadelphia, 1993. 29 Raine AE et al. Hyperparathyroidism, platelet intra- 13 O’Rourke MF. Arterial stiffness, systolic blood press- cellular free calcium and hypertension in chronic ure and logical treatment of arterial hypertension. renal failure. Kidney Int 1993; 43: 700–705. Hypertension 1990; 15: 339–347. 30 Goldsmith DJA, Covic AA, Venning MC, Ackrill P. 14 Omvik P. How smoking affects blood pressure. Blood Blood pressure reduction after parathyroidectomy for Pres 1996; 5:71–77. secondary hyperparathyroidism: further evidence 15 McNagny SE, Ahluwalia JS, Clark WS, Resnicow KA. implicating calcium in blood pressure Cigarette smoking and severe uncontrolled hyperten- regulation. Am J Kidney Dis 1996; 27: 819–825. sion in inner-city African Americans. Am J Med 1997; 31 Eschbach JW et al. Recombinant human erythropoietin 103: 121–127. in anemic patients with end-stage renal disease. Ann 16 Swislocki AL et al. Smokeless nicotine administration Int Med 1989; 111: 992–1000. is associated with hypertension but not with a deterio- 32 Sundal E, Kaeser U. Correction of of chronic ration in glucose tolerance in rats. Metabolism 1997; renal failure with recombinant human erythropoietin: 46: 1008–1012. safety and efficacy of one year’s treatment in a Euro- 17 Cryer PE, Haymond MW, Santiago JV, Shah SD. Nore- pean multicenter study of 150 haemodialysis-depen- pinephrine and epinephrine release and adrenergic dent patients. Nephrol Dial Transplant 1989; 4: 979– of smoking associated hemodynamic and 987. metabolic events. N Engl J Med 1976; 295: 573–577. 33 Canadian Erythropoietin Study Group. Association 18 Nadler JL, Velasco JS, Horton R. Cigarette smoking between recombinant erythropoietin and quality of life inhibits prostacyclin formation. Lancet 1983; 1: and capacity of patients receiving hemodia- 1248–1250. lysis. Br Med J 1990; 300: 573–578. 19 Welnnalm A et al. Relation between tobacco use and 34 Heidenreich S, Rahn K-H, Zidek W. Direct vasopressor urinary excretion of thromboxane A2 and prostacyclin effect of recombinant human erythropoietin on renal metabolites in young men. Circulation 1991; 83: resistance vessels. Kidney Int 1991; 39: 259–265. 1698–1704. 35 Bode-Bo¨ger SM et al. Endothelin release and shift of 20 Celermajer D, Sorensen KE, Georgakopoulos D. Ciga- prostaglandin balance are involved in the modulation rette smoking is associated with dose-related and of vascular tone by recombinant erythropoietin. J Car- potentially reversible impairment of endothelium- diovasc Pharmacol 1992; 20 (Suppl 12): S25–S28. dependent dilatation in healthy young adults. Circu- 36 Bode-Bo¨ger SM et al. Recombinant human erythropo- lation 1993; 88: 2149–2155. ietin enhances vasoconstrictor tone via endothelin-& 21 Kiowski W, Linder L, Stoschitzsky K. Diminished vas- and constrictor prostanoids. Kidney Int 1996; 50: cular response to inhibition of endothelium-derived 1255–1261. nitric oxide and enhanced to exogen- 37 Carlini RG et al. Recombinant human erythropoietin ously administered endothlin-1 in clinically healthy (rHuEPO) increases endothelin-1 release by endo- smokers. Circulation 1994; 90:27–34. thelial cells. Kidney Int 1993; 43: 1010–1014.

Journal of Human Hypertension