Prevalence of hypertension in 1,795 subjects with chronic renal disease: The modification of diet in renal disease study baseline cohort

Prevalence of Hypertension in 1,795 Subjects With Chronic Renal Disease: The Modification of Diet in Renal Disease Study Baseline Cohort Vardaman M. Buckalew, Jr, MD, Richard L. Berg, MS, Shin-Ru Wang, MS, Jerome G. Porush, MD, Sally Rauch, MS, RD, and Gerald Schulman, MD, and the Modification of Diet in Renal Disease Study Group* • The Modification of Diet in Renal Disease Study was a multicenter trial of the effect of protein restriction and strict blood pressure control on the progression rate of chronic renal failure of multiple causes. At the first baseline visit, 1,795 screened patients with renal disease had blood pressure measured, antihypertensive medications recorded, glomerular filtration rate (GFR) determined by 12Sl-iothalamate clearance, a nutritional assessment, and a 24-hour urine collection to determine sodium and potassium levels. A total of 1,494 patients in this cohort were classified as hypertensive (83%) and the remainder (301 patients) as nonhypertensive. Ninety-one percent of the hypertensive subjects were on treatment, 54% being controlled to a blood pressure of -< 140/90 mm Hg. To better understand the factors that contribute to the development of hypertension in chronic renal disease, some determinants of the prevalence of hypertension in this cohort were investigated. Compared with normotensive subjects, hypertensive patients were older (51.2 +_ 12.7 years v 46.6 _+ 13.1 years [mean +_ SD]), had a higher body mass index (BMI; 27.5 _+ 4.7 kg/m 2 v 25.4 + 4.2 kg/m2), and had a lower GFR (37.8 _+ 19.6 mL/min/1.73 m 2 v 50.1 +_ 25 mL/min/1.73 m2). All these differences were significant (P < 0.01). The prevalence of hypertension was significantly higher for men than for w o m e n (86% v 80%; P = 0.001), and for blacks than for whites (93% v 81%; P < 0.001). The prevalence of hypertension was higher in subjects with glomerular disease than in those with tubulointerstitial disease (85% v 62.6%; P < 0.001). The prevalence of hypertension varied inversely with GFR (from 66% at a GFR of 83 mL/min/1.73 m 2 to 95% at a GFR of 12 mL/min/1.73 m2). The prevalence of hypertension varied directly with BMI (from 70% with a BMI at the 10th percentile to 94% with a BMI at the 97th percentile). This relationship was independent of GFR. Multiple logistic regression analysis showed five predictors in decreasing order of significance as determined by chi-square values: GFR, 83.2; BMI, 36.7; black race, 19.9; increasing age, 14.5 (all P < 0.001); and male gender, 5.1 (P = 0.024). Salt intake was not a determinant of blood pressure status. These results confirm previous reports indicating that hypertension in renal disease is determined by the level of renal function. For the first time, three factors known to predict blood pressure levels in populations with normal renal function were also shown to be determinants of blood pressure in renal disease: BMI, black race, and age. In addition, the data suggest that hypertension is inadequately treated in more than half of patients with chronic renal disease in the United States.

© 1996 by the National Kidney Foundation, Inc. INDEX WORDS: Hypertension; chronic renal disease; risk factors.

HRONIC renal parenchymal disease is the most common cause of secondary hypertension, accounting for an estimated 5% of all cases. 1 Previous studies have shown that the prevalence of hypertension in chronic renal disease increases as overall renal function decreases, 2-7 approaching 100% at end stage in patients with glomerular disease and 75% to 80% in patients with tubulointerstitial disease. 4 In these earlier studies, the glomerular filtration rate (GFR) was estimated by serum creatinine or creatinine clearance so that the relationship between hypertension and GFR was not precisely determined. Moreover, very little is known about the effect of factors (other than renal function) that influence blood pressure in the general population on the prevalence of hypertension in patients with chronic renal disease. The Modification of Diet in Renal Disease (MDRD) Study offers a unique opportunity to investigate these issues. The present cross-sectional study is based on

C

1,795 patients with a variety of kidney diseases admitted to baseline in the M D R D Study and evaluated for the presence of hypertension as well as a variety of other demographic and nutritional characteristics, such as age, gender, race, From the National Institute of Diabetes, Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD. Received February 28, 1996; accepted in revised form August 13, 1996. *The institutions and investigators who participated in the Modification of Diet in Renal Disease Study are listed in a previous publication (N Engl J Med 330:877-884, 1994). Supported by the National Institute of Diabetes, Digestive and Kidney Diseases and the Health Care Financing Administration. Address reprint requests to Gerald J. Beck, PhD, MDRD Data Coordinating Center, Department of Biostatistics and Epidemiology, P88, The Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195. © 1996 by the National Kidney Foundation, Inc. 0272-6386/96/2806-000453.00/0

American Journal of Kidney Diseases, Vol 28, No 6 (December), 1996: pp 811-821

811

812

VARDAMAN M. BUCKALEW

s o d i u m a n d p o t a s s i u m i n t a k e , a n d b o d y m a s s index (BMI), and who underwent GFR measurem e n t b y 12sI-iothalamate c l e a r a n c e . S t r i k i n g l y , t h e p r e v a l e n c e o f h y p e r t e n s i o n in this c o h o r t w i t h c h r o n i c r e n a l d i s e a s e w a s n o t o n l y i n v e r s e l y rel a t e d to G F R , b u t a l s o w a s p r e d i c t e d b y B M I , r a c e , a n d age, t h r e e c h a r a c t e r i s t i c s t h a t i n f l u e n c e b l o o d p r e s s u r e in p o p u l a t i o n s w i t h o u t r e n a l d i s ease.

MATERIALS AND METHODS A complete description of the MDRD Study, including design, inclusion and exclusion criteria, recruitment methods, data collection methodology, laboratory assays, anthropometries, and dietary assessment, has appeared elsewhere.8'9 Briefly, the MDRD Study had a two-by-two factorial design that tested the effect of dietary protein restriction and strict blood pressure control on the rate of progression of renal disease. Baseline entry criteria included age between 18 and 70 years, documented evidence of renal disease, and serum creatinine of 1.2 to 7.0 mg/dL in women, 1.4 to 7.0 mg/dL in men, or creatinine clearance less than 70 mL/min/1.73 m 2. Exclusion criteria included pregnancy, body weight less than 80% or greater than 160% of standard, type I diabetes, insulin-dependent type II diabetes, glomerulonephritis due to autoimmune diseases such as systemic lupus erythematosus, obstructive uropathy, renal artery stenosis, proteinuria greater than 10 g/d, mean arterial pressure greater than 125 mm Hg, and prior renal transplantation. The presence or absence of hypertension was not considered in the inclusion or exclusion criteria. The qualifying upper and lower limits of GFR were 55 and 13 mL/min/1.73 m s, respectively. Potential subjects were identified by a variety of methods, including chart screening, physician referral, and self referral in response to mass media advertisements. 9 A total of 1,795 subjects met initial screening criteria and were enrolled in the 3-month baseline period during which their GFR, nutritional status, and ability to comply with study requirements were determined. This cohort forms the study population on which the current cross-sectional study is based. Only a subset (840 subjects) of this larger cohort was ultimately randomized to the trial as many of those admitted to baseline did not meet GFR or other criteria. Methods of particular importance to this report are described below.

Blood Pressure Classification and Measurement At the initial baseline visit, patients were classified by MDRD Study physicians as "hypertensive" or "not hypertensive" based on patient history (including the use of antihypertensive drugs) and review of medical records. The blood pressure measurement itself was not used as a criterion for defining hypertension since baseline blood pressure levels were recorded with patients taking their usual antihypertensive medication. Blood pressures were measured with a Hawksley random zero sphygmomanometer according to a protocol using a technique adopted from the Fifth Joint National Committee report. 1° After an initial 5-minute rest period with the patient in the seated position, three consecutive

measurements were obtained from the patient's fight arm, which was resting comfortably on a table at approximately heart level. Between measurements, the patient's arm was raised overhead for 15 seconds by the blood pressure technician. The average of the second and third measurements was recorded as the pressure for the visit. Mean arterial pressure was calculated as one third systolic blood pressure + two thirds diastolic blood pressure.

Glomerular Filtration Rate The GFR was measured as the renal clearance of ~25Iiothalamate following subcutaneous injection without epinephrine after water loading. 11Urine and serum samples were obtained in four consecutive periods and were measured for radioactivity at the Central GFR Laboratory. The GFR values reported in this communication are the result of a single determination for each patient at entry into baseline.

Body Weight and Anthropometry Body weight and skinfold thickness were measured in each clinic by trained personnel according to a standard protocol. Standard body weight was calculated as the 50th percentile of the weight for a given height and frame size for each gender using tables based on National Center for Health Statistics National Health and Nutrition Examination Surveys I and II data] 2 BMI was defined as weight in kilograms divided by the square of height in meters standardized for age and gender with National Center for Health Statistics National Health and Nutrition Examination Surveys tables. Percentage body fat was determined from (1) a regression equation for the prediction of body density from the triceps, biceps, and subscapular skinfolds; and (2) an equation using the known relationship between body density based on the skinfold measurements and the proportion of fat in the body. lzA3

Urinary Electrolyte Excretion A 24-hour urine collection to determine creatinine clearance and urea nitrogen excretion, as well as sodium and potassium excretion, was obtained from all patients. Patients were not asked to alter their diets or fluid intake prior to collecting the urine.

Renal Diagnoses The causes of renal disease were assessed by MDRD Study physicians based on all available clinical data (including biopsy in some cases) at the initial baseline visit. Initial classification included 23 diagnostic categories, which were subsequently collapsed into nine groups: glomerular diseases, polycystic kidney disease, hypertensive nephrosclerosis, tubulointerstitial disease, urinary tract disease, absence of one kidney, diabetic nephropathy, hereditary nephritis, and other or unknown. 14 Analyses using these nine disease categories were performed for this communication; however, to simplify data presentation these nine diagnostic groups were further collapsed into three groups: glomerular diseases (including hereditary nephritis and diabetic nephropathy), polycystic kidney disease, and other (Tables 1 and 2). The classification used in this report differs from that used in other MDRD Study publications in two respects. First, the other or unknown group was subsequently divided into two

813

HYPERTENSION IN CHRONIC RENAL DISEASE Table 1. Characteristics of the Baseline Cohort

Race White Black Other Total Renal diagnosis GIomerular diseases* Polycystic Kidney Dis. Other

Males (%)

Females (%)

All Patients (%)

880 (81.1) 131 (12.1) 74 (6.8) 1,085

549 (77.3) 96 (13.5) 65 (9.2) 710

1,429 (79.6) 227 (12.6) 139 (7.7) 1,795

394 (36.3) 199 (18.3) 492 (45.3)

185 (26.1) 197 (27.7) 328 (46.2)

579 (32.3) 396 (22.1) 820 (45.7)

* Including hereditary nephritis, diabetic nephropathy, and glomerular diseases.

categories based on protein excretion (1 to 3 g/d and <1 g/ d)./4 Second, information from later baseline assessments was incorporated so that the categories of hypertensive nephrosclerosis, glomerular diseases, and tubulointerstitial diseases were restricted to patients in whom the diagnosis was confirmed by biopsy or in whom the diagnosis was "established" by other objective evidence. ~4

pressure classification with GFR and other variables known to be important in essential hypertension: BMI, race, age, and gender. All probability values reported are for two-sided tests, and results are deemed statistically significant if P < 0.05.

RESULTS

Demographics and Renal Diagnoses

Left Ventricular Hypertrophy Twelve-lead electrocardiograms were performed by standard techniques. Tracings were interpreted separately and then jointly in a central laboratory by two blinded readers. A consensus reading obtained jointly became the official interpretation. Left ventricular hypertrophy was considered present when a repolarization abnormality was found in V5 or V6 as follows: depressed ST segment and descending limb of T-wave upwardly convex with T diphasic or inverted, sometimes associated with a late intrinsicoid deflection.

Statistical Analyses Results for continuous variables are summarized as mean values with associated standard deviations. Comparisons of subgroups for continuous variables are based on ANOVA and two-sample t-tests. Comparisons of categorical variables are based on chi-square procedures. Multiple logistic regression ~5 was used to jointly assess the relationship of blood

T h e d e m o g r a p h i c characteristics and renal dia g n o s e s o f the c o h o r t are s h o w n in T a b l e 1. T h e m a j o r i t y o f subjects w e r e m a l e ( 6 0 . 4 % ) and w h i t e (79.6%). O n l y 12.6% o f the c o h o r t w e r e A f r i c a n - A m e r i c a n s , 5 . 2 % w e r e H i s p a n i c , and 2 . 5 % w e r e o f o t h e r races. U s i n g o n l y three diagnostic c a t e g o r i e s , 3 2 . 3 % o f the c o h o r t h a d glom e r u l a r disease, 2 2 . 1 % h a d p o l y c y s t i c k i d n e y disease, and 4 5 . 7 % h a d o t h e r or u n k n o w n diseases. U s i n g less s t r i n g e n t criteria than t h o s e u s e d in the M D R D Study, h y p e r t e n s i v e n e p h r o sclerosis was d i a g n o s e d in 16.7% o f the cohort. In the p r e s e n t study, subjects w i t h h y p e r t e n s i v e n e p h r o s c l e r o s i s w e r e i n c l u d e d in the " o t h e r " category.

Table 2. Prevalence of Hypertension by Diagnosis, Race, and Gender in the Modification of Diet in Renal Disease Study Baseline Cohort

White Males (%) Renal diagnosis Polycystic kidney disease Glomerular diseases1. Other Totals

165 267 306 738

(92.2) (85.3) (78.9) (83.9)

Black Females (%)

137 97 188 422

(80.6) (78.2) (73.7) (76.9)

Males (%)

10 43 71 124

(100.0) (93.5) (94.7) (94.7)

All Patients* Females (%)

14 31 43 88

* Includes all races. 1 Including hereditary nephritis, diabetic nephropathy, and glomerular diseases.

(87.5) (93.9) (91.5) (91.7)

Males (%)

184 342 402 928

(92.5) (86.8) (81.7) (85.5)

Females (%)

161 151 254 566

(81.7) (81.6) (77.4) (79.7)

814

Blood Pressure Classification by Gender, Race, and Diagnosis Table 2 summarizes the blood pressure classification by gender, race, and renal diagnosis. Eighty-three percent of the cohort was classified as having hypertension. The prevalence of hypertension was significantly higher for men than for women (86% v 80%; P = 0.001) and for blacks than for whites (93% v 81%; P < 0.001). This racial difference was present in both genders and in all diagnostic categories. Using the three collapsed diagnostic categories shown in Table 11 the prevalence of hypertension was lower in the "other" category compared with polycystic and glomerular diseases. Other differences included a higher prevalence of polycystic kidney disease in females (27.7% v 18.3% in males; P < 0.001) and in whites compared with blacks (24% v 11%; P < 0.001) (data not shown). When the extended list of nine diagnostic categories was analyzed (data not shown), the highest prevalence of hypertension was in the hypertensive nephrosclerosis group (98%), as expected. As this prevalence figure indicates, not all patients with a diagnosis of hypertensive nephrosclerosis were classified as hypertensive, a discrepancy confined exclusively to white subjects. The prevalence of hypertension in other disease categories varied from 93.3% in diabetic nephropathy to 56.5% in subjects with absence of one kidney. As previously reported,4 the prevalence of hypertension in subjects with glomerular diseases (85%) was higher than in those with tubulointerstitial diseases (62.6%) (P < 0.001). This difference was present in both males and females.

Blood Pressure of Cohort by Blood Pressure Classification Blood pressure levels obtained at the initial baseline visit in 1,793 subjects are shown in Table 3. Average (_+SD) systolic and diastolic blood pressures of the hypertensive cohort were 136 _+ 19.5 mm Hg and 83.2 + 11.1 mm Hg, respectively, compared with 123.4 + 16.5 mm Hg and 77.4 _+ 10.1 mm Hg in the nonhypertensive group. Six hundred eighty-seven (46%) of the hypertensive cohort had blood pressure levels greater than Fifth Joint National Committee criteria for hypertension (> 140 mm Hg systolic and/ or > 90 mm Hg diastolic). 1° Two hundred sixty of these (17.4% of entire cohort) had both sys-

VARDAMAN M. BUCKALEW

tolic and diastolic blood pressure levels above the normal range. In the nonhypertensive cohort, 52 (17.3%) had blood pressure levels above the normal range. Of these, 15 (5.0% of the entire cohort) had elevations of both systolic and diastolic blood pressure. To further evaluate the prevalence of elevated blood pressure in the group classified as nonhypertensive, blood pressure levels recorded at later baseline visits were evaluated. One hundred seventy-three of the 301 nonhypertensive patients had at least three baseline visits. Of these, 18 (10%), seven (4%), and eight (5%) had mean systolic, mean diastolic, and mean arterial pressure for all three visits greater than 140, 90, and 107 mm Hg, respectively. Thus, the prevalence of potential misclassification is slightly lower when repeated blood pressure measurements at subsequent visits are considered.

Left Ventricular Hypertrophy by Blood Pressure Classification Electrocardiograms were available for interpretation in 946 subjects classified as hypertensive and 160 classified as nonhypertensive. Forty-three (4.5%) of the hypertensive cohort had this end-organ complication compared with three (1.9%) of the nonhypertensive group (P = 0.12) (Table 3). Only one of the three patients in the nonhypertensive group with left ventricular hypertrophy had an elevated blood pressure at the initial baseline visit.

Antihypertensive Drug Use by Blood Pressure Classification The prevalence of antihypertensive drug use by blood pressure classification is also shown in Table 3. Seventy-six percent of the entire cohort were taking drugs that lower blood pressure (78.5% of males and 72.5% of females). Ninetyone percent of subjects classified as hypertensive were taking antihypertensive medications compared with 9.6% in the nonhypertensive group. The nonhypertensive patients were taking these medications for indications other than hypertension (eg, beta blockers and calcium channel blockers for angina).

Antihypertensive Drug Use in the Hypertensive Cohort Table 4 shows the frequency with which the major classes of antihypertensive medications

815

HYPERTENSION IN CHRONIC RENAL DISEASE

Table 3. Blood Pressure and Other Characteristics of Cohort by Blood Pressure Classification Hypertensive Subjects (n = 1,494)

Systolic, mm Hg (mean +_ SD) Diastolic, mm Hg (mean +_ SD) MAP, mm Hg (mean + SD) No. of patients with systolic > 140 and diastolic > 90 mm Hg (%) No. of patients with systolic > 140 or diastolic > 90 mm Hg (%) No. of patients on antihypertensive drugs (%) No. of patients with left ventricular hypertrophy (%)

Males (n - 926)

Females (n = 566)

All Patients* (N = 1,492)

136.9 _+ 18.51. 84.2 -- 11.31101.8 _+ 12.01.

135.6 _+ 2 1 . 0 i 81.5 +_ 10.5j99.5 _+ 12.01.

136.4 _+ 19.5183.2 _+ 11.11100.9 _+ 12.01-

182 (19.7)

78 (13.8)

260 (17.4)1-

448 (48.4)

239 (42.2)

687 (46.0)1-

852 (91.8)

515 (91.0)

1367 (91.5)1.

13 (3.6)

43 (4.5)

30 (5.1)

Nonhypertensive Subjects (n = 301) Males (n = 157) 126.2 _+ 15.9 79.1 + 10.0 94.8 _+ 10.4

Systolic, mm Hg (mean _+ SD) Diastolic, mmHg (mean _+ SD) MAP, mm Hg (mean + SD) No. of patients with systolic > 140 and diastolic > 90 mm Hg No. of patients with systolic > 140 or diastolic > 90 mm Hg (%) No. of patients on antihypertensive drugs (%) No. of patients with left ventricular hypertrophy (%)

Females (n = 144) 120.2 _+ 16.6 75.5 +_ 10.0 90.4 +_ 11.2

All Patients* (N = 301) 123.4 _+ 16.5 77.4 +_ 10.1 92.7 _+ 11.0

8 (5.1)

7 (4.9)

15 (5.0)

34 (21.7)

18 (12.5)

52 (17.3)

12 (7.6)

17 (11.8)

29 (9.6)

1 (1.2)

2 (2.7)

3 (1.9)

Abbreviation: MAP, mean arterial pressure. * Number of patients who had blood pressure measurements. 1. P < 0.01 for comparison with nonhypertensive patients. Mean _+ SD.

were used in subjects classified as hypertensive. More subjects were taking diuretics (45.6%) than any other antihypertensive medication. The frequency of use of the other major antihypertensive Table 4. Percentage of Antihypertensive Drug Use in Subjects Classified as Hypertensive by Major Drug Classes

Beta blockers Calcium channel blockers Angiotensin-converting enzyme inhibitors Diuretics Other antihypertensives Total on antihypertensive medications

% Males (n =928)

% Females (n=566)

% All Patients (N =1,494)

37.2

32.7

35.5

31.7

28.3

30.4

39.8 43.2 22.6

37.8 49.5 17.3

39.0 45.6 20.6

91.8

91.0

91.5

drug classes was remarkably similar. The proportion of patients taking one, two, or three drugs was 37%, 31%, and 15%, respectively. Six percent were receiving four or more drugs. A total of 8.5% of subjects classified as hypertensive were not taking antihypertensive drugs.

Nutritional Status and Renal Function by Blood Pressure Classification Table 5 shows the age, body weight, BMI, GFR, serum creatinine, and sodium and potassium excretion by blood pressure status of the study cohort. The mean age of the hypertensive subjects (51.2 _+ 12.7 years) was higher than that of the nonhypertensive subjects (46.6 + 13.1 years) (P < 0.01). The hypertensive group was heavier both by absolute weight and by percent standard weight than the nonhypertensive group. Hypertensive subjects were 111.7% + 17.0% of standard body weight compared with nonhyper-

816

VARDAMAN M. BUCKALEW

Table 5. Renal Function and Nutritional Assessment of Cohort by Blood Pressure Classification Males

Hypertensive subjects Age Weight (kg) Percentage of standard weight BMI (kg/m 2) Percentage of body fat GFR (mL/min/1.73 m 2) Serum creatinine (mg/dL) Sodium excretion (mEq/d) Potassium excretion (mEq/d) Sodium/potassium Nonhypertensive subjects Age Weight (kg) Percentage of standard weight BMI (kg/m 2) Percentage of body fat GFR (mL/min/1.73 m 2) Serum creatinine (mg/dL) Sodium excretion (mEq/d) Potassium excretion (mEq/d) Sodium/potassium

Females

All Patients

No.

Mean _+ SD

No.

Mean _+ SD

No.

Mean _+ SD

928 926

51.7 -- 12.7" 86.4 ± 14.9"

566 563

50.3 _+ 12.5" 71.6 _+ 15.3"

1,494 1,489

51.2 _+ 12.7" 80.8 ± 16.7"

926 925 610 923 913 877

110.6 27.8 27.2 38.7 2.5 164.8

± ± ± ± + _

15.3" 4.2* 6.01 19.8" 1.2" 69.6

563 563 372 563 559 532

113.6 27.2 35.2 36.2 2.1 129.4

___19.3" ___5.4* ___6.1" + 19.2" + 1.1" ± 54.7

1,489 1,488 982 1,486 1,472 1,409

111.7 27.5 30.3 37.8 2.3 151.4

_+ 17.0" ± 4.7* _+ 7.2" ± 19.6" ± 1.2" _+ 66.6

877 877

66.2 + 26.2 2.8 ± 1.4

532 532

52.3 ___20.8 2.8 ± 1.4?

1,409 1,409

60.9 ± 25.2 2.8 ± 1.4

157 156

48.0 ___ 13.5 82.3 + 13.4

144 143

45.0 ± 12.5 65.2 ___13.5

301 299

46.6 _+ 13.1 74.1 _+ 15.9

156 156 100 156 157 147 147 147

105.1 26.2 25.6 48.8 2.0 169.0

___ 13.5 + 3.6 + 5.8 ± 21.9 ± 1.0 ___69.1

143 143 81 143 143 139

63.0 ± 23.1 2.9 ± 1.4

139 139

104.6 24.5 31.9 51.6 1.6 121.1

± 16.5 ___4.6 + 5.8 + 27.9 ± 0.8 ___54.8

299 299 181 299 300 286

53.3 ± 23.3 2.5 ± 1.2

286 286

104.9 25.4 28.4 50.1 1.8 145.7

_+ 15.0 ± 4.2 ± 6.6 _+ 25.0 ± 0.9 ± 66.9

58.3 ± 23.6 2.7 _+ 1.3

* P < 0.01 for comparison with nonhypertensive patients. 1 P < 0.05 for comparison with nonhypertensive patients.

tensive subjects, who were 104.9% _+ 15% of standard weight (P < 0.01). The standard weight of hypertensive females (113.6% + 19.3%) was significantly higher than that of hypertensive males (110.6% _+ 15.3%) (P = 0.002), but standard weight was similar in male and female nonhypertensive subjects. Hypertensive subjects had a higher percent body fat (30.3% _+ 7.2%) than nonhypertensive subjects (28.4% _ 6.6%) (P < 0.01). Females had a higher percent body fat than males in both the hypertensive and nonhypertensive groups. Mean GFR of the hypertensive patients (37.8 _+ 19.6 mL/mirdl.73 m 2) was lower than that of the nonhypertensive patients (50.1 _+ 25 mL/min/ 1.73 m 2) (P < 0.01). Mean serum creatinine in the hypertensive patients (2.3 _+ 1.2 mg/dL) was higher than in the nonhypertensive patients (1.8 _+ 0.9 mg/dL) (P < 0.01). The GFR was significantly higher in the hypertensive males than in the hypertensive females (P = 0.02). Among the

nonhypertensive patients, mean GFR was lower in males, although the difference was not statistically significant (P = 0.34). Mean sodium excretion was not significantly different between the hypertensive and nonhypertensive groups (151.4 mEq/d v 145.7 mEq/d; P = 0.18). However, the ratio of urine sodium excretion to potassium excretion (Na/K) was higher in the hypertensive females than in the nonhypertensive females (2.8 v 2.5; P < 0.05).

Progression or Nonprogression by Blood Pressure Classification Using serum creatinine values prior to entry into the MDRD Study, subjects were classified as "progressors" (n -- 680) or "nonprogressors" (n = 741) according to whether the slope of 1/SCr versus time was less than or greater than -0.003 dL/mg/mo. The hypertensive patients were more likely to show progression than the nonhypertensive patients (47.9% v

HYPERTENSION IN CHRONIC RENAL DISEASE

817

100 -

100

100

9O

90

90-

so

SO

80

100

-

90 80 ..,, '

70

70

60-

60

50

50

40-

40

30

70 o3

70

~

60

60

50-

50

40-

40

30

30

30

:20-

20

20

20

10

10

10

10

0

0

O10

20

30

40

50

60

70

80

1-

0 0

90

20

39.2%; P < 0.001). Furthermore, the slope of the progression rate for hypertensive patients was -0.0004 _+ 0.0423 dL/mg/mo compared with 0.0018 _+ 0.0445 dL/mg/mo in the nonhypertensive patients (P < 0.0001).

Factors Affecting Prevalence of Hypertension Figure 1 shows the relationship between the prevalence of hypertension and the level of GFR. Subjects were ranked by level of GFR into 10 groups of approximately equal numbers. The prevalence of hypertension varied inversely with GFR (from 66% at a GFR of 83 mL/min/1.73 m 2 to 95% at a GFR of 12 mL/min/1.73 m2). Figure 2 shows the relationship of BMI to the prevalence of hypertension. As in Fig 1, subjects were ranked by mean BMI into 10 equal groups. The prevalence of hypertension varied directly with BMI (from 70% with a BMI at the 10th percentile to 94% with a BMI at the 97th percentile). Figure 3 shows the separate effects of GFR and BMI on the prevalence of hypertension. As in Fig 1, subjects were divided by GFR into 10 groups, but separately for those in the highest 30% of BMI and those in the lowest 30%. In both the high and low BMI groups, the linear relationship between GFR and prevalence of hypertension was observed. Furthermore, prevalence was higher at every level of GFR in the

60

80

100

Mean Body Mass Index (NHS Percentile)

M e a n G F R ( r n l / m i n / 1 , 7 3 m 2)

Fig 1. Prevalence of hypertension by level of GFR. Patients were ranked by GFR into 10 groups, each containing 179 to 180 patients. Data are presented as mean values ± SE.

40

Fig 2. Prevalence of hypertension by BMi. Patients were ranked into 10 groups by age- and gender-adjusted BMI percentiles based on the National Health Survey (series 11, no. 238). Data are presented as mean values _+ SE.

high BMI group than in the low BMI group. Thus, the association of BMI with prevalence of hypertension was distinct from the effect of GFR. Table 6 shows the results of multiple logistic regression analysis with blood pressure status as the outcome, and with GFR, age, gender, BMI, race, sodium and potassium excretion, and the

100

L

100

90

9O

80

SO

70-

70

60-

60

50 =

50

40

40

3o-

30

20-

20

lO- @Lowest OHighest OO

10

30% BMI/ 30% BMI

20

30

40

Mean GFR

10

J

0 50

60

(ml/minPl.73m

70

ao

90

2)

Fig 3. Prevalence of hypertension by level of GFR and BMI. The lowest and highest 30% of patients according to BMI are ranked by GFR into 10 groups each. Data are presented as mean values ~ SE.

818

VARDAMAN M. BUCKALEW Table 6. Logistic Regression Model for Determinants of Blood Pressure Classification

GFR (mL/min/1.73 m2) BMI (National Health Survey percentile) Black race Age (yr) Male Potassium excretion (mEq/d)

Estimate

SE

Chi-square

ProbabflityValue

-0,029 +0.015 +1.503 +0.020 +0,281 +0.005

0,003 0,003 0.319 0.006 0.145 0.003

81.9 32.5 22.2 12.9 3.79 2.99

<0,001 <0.001 <0.001 <0.001 0.052 0.084

ratio of urine sodium excretion to potassium excretion as potential predictors. Fitting the model in a stepwise fashion, four of these variables were found to be significant, with GFR entering the model first as the strongest individual predictor and potassium excretion entering last. Male gender and potassium excretion were borderline significant; other variables were highly significant. The coefficient for GFR is negative, reflecting the fact that low GFR values are associated with a higher rate of hypertension (Fig 1). The coefficients for black race (1.503) and male gender (0.281), after taking exponentials, translate to odds ratios of 4.5 and 1.32, respectively. The effect of race on blood pressure classification is not due to black patients having a lower GFR, since mean GFR was higher in the black patients than in the white patients classified as having hypertension (43.2 + 20.6 mL/min/1.73 m 2 v 36.8 + 18.8 mL/min/1.73 m2; P < 0.001). Black patients show a higher mean GFR than white patients in the nonhypertensive group as well (53.6 2 29.2 mL/mirdl.73 m 2 v 49.9 _+ 24.6 mL/ mirdl.73 me), although the difference was not significant (P = 0.66). To evaluate the possible effects of seeming misclassification, the 52 subjects classified as nonhypertensive with elevated blood pressure at the first baseline visit were characterized separately. Their mean GFR (46 _+ 22.1 mL/mird 1.73 m 2) and BMI (26.2 ,+ 4.3 kg/m 2) were not statistically different from these parameters in the nonhypertensive group, whose blood pressure levels were within the normal range (GFR, 51 _+ 25.5 mL/mirdl.73 m2; BMI, 25.2 _+ 4.2 kg/m2), although there was a significant difference in age (51 _+ 15 years v 46 _+ 13 years; P = 0.03). All factors in the multiple logistic regression model refit without these 52 patients were still highly significant (GFR, BMI, black race, and age [P < 0.001] and male gender [P = 0.006]).

DISCUSSION

The prevalence of hypertension in 1,795 patients with chronic renal disease admitted to baseline in the MDRD Study was 83%, slightly less than the prevalence of 89% reported by the Northern Italian Cooperative Study Group in 406 patients with chronic renal failure. ~6 This difference is probably explained by the fact that we included patients with renal disease and normal or near-normal GFR. Using this large MDRD Study cohort, we investigated some factors that might determine the prevalence of hypertension in chronic renal disease. Using 12q-iothalamate clearance to measure renal function, we found a striking relationship between GFR and the prevalence of hypertension (Fig 1). When the GFR was normal or near-normal (approximately 83 mL/mirdl.73 m2), only 65% of the patients had hypertension, while more than 95% were hypertensive when the GFR was approximately 12 mL/min/1.73 m 2. These results are in general agreement with previous reports using serum creatinine or creatinine clearance to evaluate renal function. In the largest such study (290 patients), Vendemia et al reported that the prevalence of hypertension was 40% with normal renal function (serum creatinine < 1.0 mg/dL), 60% with mild renal impairment (serum creatinine between 1.1 and 3.0 mg/dL), 79% with severe renal failure (serum creatinine > 7.0 rag/ dL), and 88% with end-stage renal disease. 4 An increasing prevalence of hypertension with increasing serum creatinine was also found by Orofino et al in 288 patients with biopsy-proven chronic glomerulonephritis. 7 With regard to the type of renal disease, we confirmed the report of Vendemia et al that hypertension prevalence is higher in patients with glomerular disease than in those with tubulointerstitial disease, independent of GFR. 4 Among pa-

HYPERTENSION IN CHRONIC RENAL DISEASE

tients with glomerulonephritis, Vendemia et al also showed the highest prevalence of hypertension in focal glomerulosclerosis and the lowest in patients with Berger's disease until the serum creatinine was greater than 7 mg/dL, when the prevalence reached almost 100% regardless of glomerulonephritis type. Other investigators have shown that the prevalence of hypertension varies with the type of glomerulonephritis independent of GFR, with the highest in focal glomerular sclerosis and membranoproliferative glomerulonephritis and the lowest in minimalchange disease. 2'6'v'17We did not examine the relationship between GFR and hypertension prevalence in the various diagnostic categories, nor did we attempt to subclassify diagnostic types of glomerulonephritis since renal biopsy was not obtained in most patients. It also should be noted that GFR was measured with patients taking their usual antihypertensive medications. The effect these medications might have had on the observed associations is unknown. There are several interesting and novel findings in the present study with regard to factors other than renal function and type of disease that predict the prevalence of hypertension in chronic renal disease. Perhaps the most important one from a therapeutic point of view is the difference in BMI between hypertensive and nonhypertensive patients and the striking linear relationship between BMI and the prevalence of hypertension (Fig 2), an effect independent of GFR (Fig 3). Although BMI is a major determinant of blood pressure in both nonhypertensive populations 18'19 and in patients with essential hypertension, 2°-22 the present study is the first to suggest that it also is a factor in determining blood pressure in patients with parenchymal renal disease. Thus, weight reduction might be as helpful in controlling hypertension in patients with renal parenchyreal disease as in essential hypertension. 2325 Other new findings in the present study are the effect of race, age, and gender on hypertension prevalence in renal disease. The effect of race is not explained by differences in GFR, nor by the increased prevalence of nephrosclerosis in the black patients. Regarding the latter, the prevalence of hypertension was higher in black patients without hypertensive nephrosclerosis than in white patients without that diagnosis (90.4% v 78.6%). This finding, when considered together with the well-known increased prevalence of es-

819

sential hypertension in African-Americans, 26 suggests that black patients are more likely to develop hypertension than white patients, regardless of etiology. The higher prevalence of hypertension with older age and male gender in patients with chronic renal disease (Table 6) also has not been previously described. Although the urine Na/K was significantly higher in hypertensive than in nonhypertensive women (Table 5), neither sodium excretion nor urine Na/K was higher in the hypertensive group compared with the nonhypertensive group. These observations are in general agreement with studies of blood pressure determinants in general populations. Thus, in the INTERSALT Study, urine sodium and potassium, as well as the Na/K, predicted blood pressure levels across populations but not necessarily within populations. 27 Furthermore, both within and across populations, urine Na/K was a stronger predictor of blood pressure in women than in men. 28 Although sodium intake was not different, three factors that increase sodium sensitivity of blood pressure in essential hypertension did predict hypertension prevalence in this cohort (increased BMI, black race, and increasing age),29-32 It may be important that factors which affect sodium sensitivity of blood pressure also predict the prevalence of hypertension in chronic renal disease, but the mechanisms that underlie these associations cannot be ascertained from observational, epidemiologic studies. The relationship we and others have observed between GFR and prevalence of hypertension could be due in part to an effect of GFR on sodium sensitivity of blood pressure. This possibility is supported by several observations. Sodium sensitivity is probably more common in patients with endstage renal disease than in those with essential hypertension. 6'33'34 In addition, it has been estimated that 90% of hypertensive patients with end-stage renal disease are sodium sensitive6'33 compared with approximately 50% of patients with essential hypertension.34 Koomans et al found that the increment in blood pressure for any given degree of volume expansion was greater in patients with creatinine clearance less than 22 mL/min than in those with creatinine clearance greater than 32 mL/min.35 A possible explanation for this finding was the subsequent demonstration by Koomans et a136 that patients with chronic renal disease deposit a significantly greater percentage of an acute saline load

820

intravascularly than control subjects. This phenomenon, which could be due to changes in compliance of the interstitial space, 37 needs further study as one possible explanation for the increasing prevalence of hypertension as GFR declines. Sixty-five percent of patients with chronic renal disease and normal GFR had hypertension (Fig 1). This observation, which is similar to that of Vendemia et al,4 suggests that some change in renal function leads to hypertension development before GFR declines. Only a few studies have been performed on patients with renal parenchymal hypertension and normal GFR, mostly in patients with polycystic kidney disease. 38'39 These studies showed, not surprisingly, increased plasma volume and a shift in the pressure diuresis curve. The importance of hypertension development in renal disease is emphasized by the observation that hypertensive patients were more likely to show progression of renal failure prior to entering baseline of the MDRD Study. This finding is similar to that in two previous reports 7'4° and suggests either that progressive renal failure is a determinant of hypertension development or that the spontaneous development of hypertension is a significant determinant of whether a given kidney disease progresses (excluding hypertensive nephrosclerosis). In either case, the bulk of the evidence suggests that once hypertension develops, it becomes a determinant of progression rate. 11'4°44 It should be noted that type I and insulin-dependent type II diabetic patients were excluded from participation in the MDRD Study and only a small number of non-insulin-dependent type II diabetic patients were included. The results of this study, therefore, may not be extrapolated to patients with diabetic nephropathy. The cohort reported here, selected from a larger population of 2,507 persons formally screened by 15 MDRD Study clinics, were recruited nationally from a wide variety of sources, including physician referrals, mass media advertising, a toll-free number, and review of medical records. 9 Accordingly, the data presented in Table 3 provide an approximation of the status of hypertension treatment and control in the nondiabetic chronic renal disease population in the United States at the time of screening for the MDRD Study (January 1988 through March 1991). 9 As might be expected, the percentage of hypertensive renal disease patients treated for hypertension (91%) and controlled (54%) is

VARDAMAN M, BUCKALEW

higher than that of hypertensive persons in the general population (49% and 21%, respectively).10 However, if controlling hypertension to less than 140/90 mm Hg slows progression of renal failure, 11'40-44 these data suggest that there is considerable potential for reducing the incidence of end-stage renal disease in the United States by increased emphasis on hypertension control in the population with established renal disease. REFERENCES 1. Kaplan NM: Hypertension in the population at large, in Kaplan NM: Clinical Hypertension (ed 6). Baltimore, MD, Williams & Wilkins, 1994, pp 1-22 2. Smith MC, Dunn MJ: Hypertension in renal parenchyreal disease, in Laragh JH, Brenner BM (eds): Hypertension: Pathophysiology, Diagnosis, and Management. New York, NY, Raven, 1990, pp 1583-1599 3. Kaplan NM: Renal parenchymal hypertension, in Kaplan NM: Clinical Hypertension (ed 6). Baltimore, MD, Williams & Wilkins, 1994, pp 299-317 4. Vendemia F, Fornasieri A, Velis O, Baroni M, Scarduelli B, D'Amico G: Different prevalence rates of hypertension in various reno-parenchymal diseases, in Blaufox MD, Bianchi C (eds): Secondary Forms of Hypertension: CmTent Diagnosis and Management. New York, NY, Grune & Stratton, 1981, pp 89-94 5. Blythe WB: Natural history of hypertension in renal parenchymal disease. Am J Kidney Dis 5:A50-A56, 1985 (suppl) 6. Danielsen H, Kornerup HJ, Olsen S, Posborg V: Arterial hypertension in chronic glomerulonephritis. An analysis of 310 cases. Clin Nephrol 19:284-287, 1983 7. Orofino L, Quereda C, Lamas S, Orte L, Gonzalo A, Mampaso F, Ortufio J: Hypertension in primary chronic glomerulonephritis: Analysis of 288 biopsied patients. Nephron 45:22-26, 1987 8. Modification of Diet in Renal Disease Study Group (prepared by Beck GJ, Berg RL, Coggins CH, Gassman JJ, Hunsicker LG, Schluchter MD, Williams GW): Design and statistical issues of the Modification of Diet in Renal Disease Trial. Controlled Clin Trials 12:566-586, 1991 9. Kusek JW, Coyne T, de Velasco A, Drabik MJ, Finlay RA, Gassman JJ, Kiefer S, Powers SN, Steinman TI: Recruitment experience in the full-scale phase of the Modification of Diet in Renal Disease Study. Controlled Clin Trials 14:538-557, 1993 10. Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure: The fifth report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC-V). Arch Intern Med 153:154-183, 1993 11. Modification of Diet in Renal Disease Study Group (prepared by Anderson A, Caggiula AW, Klahr S, Kusek JW, Levey AS, Williams GW): The Modification of Diet in Renal Disease Study: Design, methods, and results from the feasibility study. Am J Kidney Dis 20:18-33, 1992 12. Frisancho AR: Anthropometric Standards for the As-

HYPERTENSION IN CHRONIC RENAL DISEASE sessment of Growth and Nutritional Status. Ann Arbor, MI. University of Michigan Press, 1990 13. Durnin JVGA, Womersley J: Body fat assessed from total body density and its estimation from skinfold thickness: Measurements of 481 men and women aged from 16 to 72 years. Br J Nutr 32:77-92, 1974 14. Greene T, Bourgoignie JJ, Habwe V, Kusek JW, Sneetselaar LG, Soucie JM, Yamamoto ME: Baseline characteristics in the Modification of Diet in Renal Disease Study. J Am Soc Nephrol 3:1819-1834, 1993 15. Hosmer DW, Lemeshow S: Applied Logistic Regression. New York, NY, Wiley, 1989 16. Alberti D, Locatelli F, Graziani G, Buccianti G, Redaelli B, Giangrande A, Marcelfi D, and the No,hem Italian Cooperative Study Group: Hypertension and chronic renal insufficiency: The experience of the Northern Italian Cooperative Study Group. Am J Kidney Dis 21:124-130, 1993 (snppl 2) 17. Guidi E, Magni M, di Belgiojoso GB, Minetti L, Bianchi G: Blood pressure in patients with four different primary glomerulopathies. Clin Exp Hypertens A6:1357-1366, 1984 18. Johnson AL, Comoni JC, Cassel JC, Tyroler HA, Heyden S, Haines CG: Influence of race, sex and weight on blood pressure behavior in young adults. Am J Cardiol 35:523-530, 1975 19. Voors AW, Webber LS, Frerichs RR, Berenson GS: Body height and body mass as determinants of basal blood pressure in children--The Bogalusa Heart Study. Am J Epidemiol 106:101-108, 1977 20. McNamara PM: Relationship of adiposity to blood pressure and the development of hypertension. The Framingham Study. Ann Intern Med 67:48-59, 1967 21. Stamler R, Stamler J, Riedlinger WF, Algera G, Roberts R: Weight and blood pressure. JAMA 240:1607-1610, 1978 22. Dyer A, Elliott P: The INTERSALT Study: Relations of body mass index to blood pressure. J Hum Hypertens 3:299-308, 1989 23. Reisin E, Abel R, Modan M, Silverberg DS, Eliahou HE, Modan B: Effect of weight loss without salt restriction on the reduction of blood pressure in overweight hypertensive patients. N Engl J Med 298:1-6, 1978 24. Stamler R, Stamler J, Gosch FC, Civinelli J, Fishman J, McKeever P, McDonald A, Dyer AR: Primary prevention of hypertension by nutritional-hygienic means. Final report of a randomized, controlled trial. JAMA 262:1801-1807, 1989 25. Prineas R: Clinical interaction of salt and weight change on blood pressure level. Hypertension 17:I-143-I-149, 1991 (suppl I) 26. Tyroler HA: Socioeconomic status, age, and sex in the prevalence and prognosis of hypertension in blacks and whites, in Laragh JH, Brenner BM (eds): Hypertension: Pathophysiology, Diagnosis, and Management. New York, NY, Raven, 1990, pp 159-174 27. Elliott P: The INTERSALT Study: An addition to the evidence of salt and blood pressure, and some implications. J Hum Hypertens 3:289-298, 1989 28. Elliott P, Dyer A, Stamler R: The INTERSALT Study: Results for 24 hour sodium and potassium, by age and sex. J Hum Hypertens 3:323-330, 1989

821 29. Rocchini AP, Key J, Bondie D, Chico R, Moorehead C, Katch V, Martin M: The effect of weight loss on the sensitivity of blood pressure to sodium in obese adolescents. N Engl J Med 321:580-585, 1989 30. Weinberger MH, Miller JZ, Luft FC, Grim CE, Fineberg NS: Definitions and characteristics of sodium sensitivity and blood pressure resistance. Hypertension 8:11-127-II-134, 1986 31. Weinberger MH, Fineberg NS: Sodium and volume sensitivity of blood pressure: Age and pressure change over time. Hypertension 18:67-71, 1991 32. Luft FC, Miller JZ, Grim CE, Fineberg NS, Christian JC, Daugherty SA, Weinberger MH: Salt sensitivity and resistance of blood pressure: Age and race as factors in physiological response. Hypertension 17:I-102-I-108, 1991 33. Kaplan MD: Renal parenchymal hypertension, in Kaplan NM (ed): Clinical Hypertension (ed 4). Baltimore, MD, Williams & Wilkins, 1986, pp 292-316 34. Hollenberg NK, Williams GH: Abnormal renal function, sodium-volume homeostasis, and renin system behavior in normal-renin essential hypertension, in Laragh JH, Brenner BM (eds): Hypertension: Pathophysiology, Diagnosis, and Management. New York, NY, Raven, 1990, pp 1349-1370 35. Koomans HA, Roos JC, Boer P, Geyskes GG, Mees EJD: Salt sensitivity of blood pressure in chronic renal failure. Hypertension 4:190-197, 1982 36. Koomans HA, Geers AB, Boer P, Roos JC, Mees EJD: A study on the distribution of body fluids after rapid saline expansion in normal subjects and in patients with renal insufficiency: Preferential intravascular deposition in renal failure. Clin Sci 64:153-160, 1983 37. Lucas J, Floyer MA: Renal control of changes in the compliance of the interstitial space: A factor in the aetiology of renoprival hypertension. Clin Sci 44:397-416, 1973 38. Valvo E, Gammaro L, Tessitore N, Panzetta G, Lupo A, Loschiavo C, Oldrizzi L, Fabris A, Rugiu C, Ortalda V, Maschio G: Hypertension of polycystic kidney disease: Mechanism and hemodynamic alterations. Am J Nephrol 5:176-181, 1985 39. Schmid M, Mann JFE, Stein G, Herter M, Nussberger J, Klingbeil A, Ritz E: Natriuresis-pressure relationship in polycystic kidney disease. J Hypertens 8:277-283, 1990 40. Hannedouche T, Chauveau P, Kalou F, Albouze G, Lacour B, Jungers P: Factors affecting progression in advanced chronic renal failure. Clin Nephrol 39:312-320, 1993 41. Bergstrom J, Alvestrand A, Bucht H, Gutierrez A: Progression of chronic renal failure in man is retarded with more frequent clinical follow-ups and better blood pressure control. Clin Nephrol 25:1-6, 1986 42. Brazy PC, Stead WW, Fitzwilliam JF: Progression of renal insufficiency: Role of blood pressure. Kidney Int 35:670-674, 1989 43. Brazy PC, Fitzwilliam JF: Progressive renal disease: Role of race and antihypertensive medications. Kidney Int 37:1113-1119, 1990 44. Oldrizzi L, Rugiu C, De Biase V, Maschio G: The place of hypertension among the risk factors for renal function in chronic renal failure. Am J Kidney Dis 21:119-123, 1993