Abstract We analyzed the hypotensive mechanisms of a thiazide-type diuretic, mefruside, on the basis of the pressure-natriuresis relationship. We performed a 5-week study in eight patients with essential hypertension who were given a high sodium diet (15 to 18 g NaCl per day) during the 1st and 5th weeks, a severely sodium-restricted diet (1 to 3 g/d) during the 2nd week, and a mildly sodium-restricted diet (5 to 7 g/d) during the 3rd and 4th weeks. Mefruside (25 mg/d) was administered during the 4th and 5th weeks. Urinary sodium excretion rate and mean arterial pressure were measured at the end of each week, and the pressure-natriuresis relationship was drawn by plotting urinary sodium excretion rate on the ordinate and mean arterial pressure on the abscissa before and after mefruside treatment. Before treatment, the pressure-natriuresis relationship was linear, and mean arterial pressure was changed as a consequence of sodium intake alteration (1st week, 117±9 mm Hg; 2nd week, 105±7; 3rd week, 109±9). After treatment, however, the change in mean arterial pressure was very small (4th week, 102±8 mm Hg; 5th week, 104±7). Mefruside steepened the slope of the relationship (20.8±10.5 versus 143±85 [mmol/d]/mm Hg, P<.005) without significantly shifting the x intercept (104±6 versus 101±9 mm Hg, P=NS) of the relationship. The increase in the slope was greater in patients whose slope had been depressed and blood pressure was sodium sensitive before mefruside treatment. The hypotensive effect of mefruside during a high sodium diet correlated positively with both the hypotensive effect of sodium restriction (r=.84, P<.01) and the increase in the slope by mefruside (r=.83, P<.02). Thus, mefruside lowers blood pressure especially in patients with high sodium sensitivity mainly by making blood pressure sodium insensitive through its diuretic action. Strict sodium restriction seems unnecessary when diuretics are administered for blood pressure control.
- hypertension, essential
- sodium, dietary
- pressure-natriuresis relationship
- antihypertensive therapy
- hypertension, sodium sensitive
Diuretics have been most widely used as the first choice of antihypertensive drugs. In general, the pressure-natriuresis relationship is shifted rightward along the pressure axis toward a higher blood pressure (BP) and the slope is depressed, resulting in sodium-sensitive hypertension, with increased severity1 2 3 4 in essential hypertension. If the hypotensive mechanism of diuretics is essentially the same as that of sodium restriction, the hypotensive effect of diuretics must be stronger in the advanced stage of hypertension in patients with a depressed slope of the pressure-natriuresis relationship and therefore high sodium sensitivity of BP. In this study, we examined the effect of a thiazide-type diuretic, mefruside, on the pressure-natriuresis relationship and analyzed the relationships between the effects of mefruside and sodium restriction as well as between the effects of mefruside and sodium sensitivity.
Eight inpatients (four men, four women; 44 to 67 years old; mean, 56±8 years) with essential hypertension, who had given their informed consent, were studied in the Nephrology Unit of the National Cardiovascular Center Hospital. Their BPs were above 160/95 mm Hg in an outpatient clinic in two patients without medication and in six patients taking antihypertensive drugs. Administration of antihypertensive drugs was discontinued at least 2 weeks before hospitalization. During the initial hospitalization, a routine examination was performed to exclude secondary hypertension and evaluate the severity of the hypertensive state. No patient showed evidence of a detectable cause for hypertension or malignant hypertension.
After the initial hospitalization lasting more than 1 week, the patients were studied during five stages, each lasting 7 days: a high sodium diet containing 15 to 18 g NaCl per day (stages I and V), a low sodium diet containing 1 to 3 g NaCl per day (stage II), and a medium sodium diet containing 5 to 7 g NaCl per day (stages III and IV). During stages I through III, patients were given no antihypertensive drugs. During stages IV and V, they were given 25 mg mefruside once a day. Mefruside, one of the most commonly used antihypertensive diuretics in Japan, is a thiazide-type diuretic that has a sulfonamide group and a structural similarity with loop diuretics.5 Urinary sodium excretion rate (UNaV) and creatinine excretion rate were measured on the last 3 days of each stage. Serum sodium, potassium, uric acid, and creatinine concentrations were measured on the last day of each stage. Creatinine clearance was calculated from serum creatinine and the average values of the urinary creatinine excretion rate. BP was measured with patients in the supine position with an automatic sphygmomanometer (model BP-103, Nippon Kohrin Co) every hour from 6 am to 9 pm on the last day of each stage, and the average reading of 16 measurements was adopted. Mean arterial pressure (MAP) was calculated as one third of the pulse pressure plus the diastolic pressure.
Determination of the Pressure-Natriuresis Curve
A pressure-natriuresis curve was drawn by linking data points in each patient obtained in a steady state of sodium balance under three different amounts of sodium intake before mefruside administration and under two different amounts of sodium intake after mefruside administration. MAP (in millimeters of mercury) and UNaV (in millimoles per day) were plotted on the x and y axes, respectively.6 7 8 To compare the major characteristics of the curve, that is, the shift of the curve along the arterial pressure axis and steepness of the curve, before and after mefruside administration, we calculated the extrapolated x intercept, A (millimeters of mercury), and the slope, B (millimoles per day per millimeter of mercury), as follows4 9 10 :
where H and L are the data obtained in a steady state of sodium balance under relatively high and relatively low sodium diets, respectively. The reciprocal of the slope of the curve, 1/B (millimeters of mercury per millimole), corresponds to the salt sensitivity index.11 12 13
Results are expressed as mean±SD, and the significance of the effects of salt restriction and mefruside was tested by a two-way classification ANOVA and ANCOVA with repeated measures. The comparison of the extrapolated x intercept and slope of the pressure-natriuresis curve between before and after mefruside administration was done by Student’s t test for paired samples. The correlation coefficient was obtained by the least squares method. The reciprocal of the slope (1/B, millimeters of mercury per millimole) was used for analysis of the slope (B, millimoles per day per millimeter of mercury), because dietary sodium intake was altered primarily and a consequent secondary change in MAP was observed.4 9
Body Weight, UNaV, and BP
Table 1⇓ summarizes body weight, UNaV, and BP in the eight patients with essential hypertension during each experimental stage. When sodium intake was altered, UNaV changed, reflecting the amount of sodium intake in each stage and indicating that a steady state of sodium balance had been achieved. Two-way ANOVA showed that systolic BP, diastolic BP, and MAP were all lowered significantly by both sodium restriction and mefruside. Body weight decreased significantly by mefruside when stage V was compared with stage I and when stage IV was compared with stage III.
Pressure-Natriuresis Relationship Before Mefruside Administration
The pressure-natriuresis curve (arterial pressure–urinary sodium output relationship) before mefruside administration was obtained by plotting UNaV in stages I, II, and III on the ordinate as a function of MAP on the abscissa (Fig 1⇓). The extrapolated x intercept, A, and slope, B, of the pressure-natriuresis curve were compared between two phases: a decreasing phase from stage I to II (A: 104±6 mm Hg; B: 20.8±10.5 [mmol/d]/mm Hg) and an increasing phase from stage II to III (A: 104±7 mm Hg; B: 23.3±13.0 [mmol/d]/mm Hg). The x intercept and slope were not significantly different between these two phases. In addition, there was a significant positive overall linear relationship between MAP and the corresponding UNaV during stages I through III all together (r=.617, P<.01, n=24). These results indicate that the pressure-natriuresis curve was virtually linear. Therefore, the pressure-natriuresis curve from stage I to II was adopted for the curve before mefruside administration.
Effect of Mefruside on the Pressure-Natriuresis Curve
The pressure-natriuresis curve after mefruside administration was obtained by linking two data points of stages IV and V. The extrapolated x intercept, A, was not different before and after mefruside administration (104±6 versus 101±9 mm Hg, P=NS), whereas the slope, B, was increased after mefruside administration (20.8±10.5 versus 143±85 [mmol/d]/mm Hg, P<.005). These results indicated that the hypotensive action of mefruside was based on the increased slope of the pressure-natriuresis curve but was not based on the leftward shift of the curve along the BP axis. In addition, there was a significant negative relationship between the increase in the slope by mefruside and the steepness of the slope before mefruside administration (r=−.81, P<.02), showing that the hypotensive effect of mefruside by steepening the slope of the curve was greater in patients with a depressed slope and therefore higher sodium sensitivity.
Difference in Hypotensive Mechanisms of Mefruside Under Different Amounts of Sodium Intake
The hypotensive effect of mefruside during two different amounts of sodium intake was analyzed in relation to the hypotensive effect of sodium restriction from stage I to II. The hypotensive effect of mefruside during high sodium from stage I to V was positively correlated with that of sodium restriction (r=.84, P<.01, Fig 2⇓), whereas that during medium sodium from stage III to IV was not (r=.50, P=NS). Therefore, the hypotensive effect of mefruside during the two different amounts of sodium intake was analyzed in relation to the changes by mefruside in both the extrapolated x intercept, A, and slope, B, of the pressure-natriuresis curve. The hypotensive effect during high sodium from stage I to V was significantly correlated with the increase in the slope (r=.83, P<.02, Fig 3⇓) but was not correlated with the change in the x intercept (r=−.06, P=NS). On the other hand, the effect during medium sodium from stage III to IV had a tendency to correlate with the decrease in the x intercept (r=.67, P<.1) but not with the change in the slope (r=.20, P=NS). These results suggest that the hypotensive mechanisms of mefruside differ between a high and medium sodium intake. Mefruside may lower BP by increasing the slope of the pressure-natriuresis curve during high sodium and by shifting the curve leftward toward a lower BP level along the BP axis during relatively low sodium.
Serum Chemistry and Creatinine Clearance
Serum sodium and creatinine concentrations did not change significantly among the five stages (Table 2⇓). Serum potassium concentration was elevated by sodium restriction and was lowered by mefruside. Serum uric acid concentration was significantly elevated by both sodium restriction and mefruside. On the other hand, creatinine clearance was reduced only in stage IV compared with the other stages by the combined effect.
In our experimental design, the pressure-natriuresis curve (arterial pressure–urinary sodium output relationship) was derived by linking two data points in each patient and plotting the steady-state MAP and UNaV on the x and y axes, respectively. Although it is ideal to generate data over a larger continuum to assess this curve more precisely, the arterial pressure–natriuresis relationship has usually been found to be very near to linear within the experiment range of sodium intake used in animals6 14 15 and within the wide range of sodium intake in normal humans.16 17 On these bases, Parfrey and colleagues2 3 analyzed the pressure-natriuresis curve in essential hypertension assuming its linearity. In fact, our present study showed that this relationship was statistically linear for 1 to 18 g NaCl per day of sodium in individual patients with essential hypertension. Although the amount of sodium was kept constant only for a week in this study, on the basis of multiple studies in the past, 4 to 5 days has been found to be sufficient for sodium balance and for arterial pressure to reach a steady state during a fixed sodium intake.18 Thus, we believe that it is possible to determine by the present simple protocol the major characteristics of the pressure-natriuresis curve, namely, the shift of the curve along the arterial pressure axis and the steepness of the curve.
Recently, we proposed that hypertension can be ascribed to one of three major renal mechanisms12 13 : (1) increased preglomerular vascular resistance from heart to glomeruli, mainly due to afferent arteriolar resistance; (2) decreased whole-kidney ultrafiltration coefficient, due to decreased filtration surface area per glomerulus, decreased hydraulic permeability of the glomerular filtration barrier, and/or decreased number of glomeruli; and (3) increased rate of tubular sodium reabsorption. The first mechanism produces non–sodium-sensitive hypertension, and the latter two produce sodium-sensitive hypertension. The hypotensive mechanism of diuretics may be the reverse of the third mechanism above. When renal tubular sodium reabsorption is inhibited by diuretics, sodium balance becomes negative if glomerular filtration rate and tubular sodium load remain normal, resulting in a fall in body fluid volume and BP. As systemic BP is lowered, glomerular capillary hydraulic pressure and glomerular filtration rate are also reduced. The resultant decrease in tubular sodium load makes it possible for sodium balance to be maintained despite inhibited tubular sodium reabsorption. In this way, when tubular sodium reabsorption is inhibited, sodium balance can be maintained only with the reduction in tubular sodium load because of the fall in systemic and glomerular capillary pressures. In other words, by inhibiting tubular sodium reabsorption, diuretics multiply urinary sodium output created at any level of glomerular filtration rate and tubular sodium load as well as at any level of systemic and glomerular capillary pressures.6 The above relationship between sodium balance and BP can be illustrated clearly by the effect of a thiazide-type diuretic, mefruside, on the pressure-natriuresis curve.6 Mefruside steepened the slope of the pressure-natriuresis curve, especially in patients whose slope was depressed before mefruside administration (Fig 1⇑). Since the reciprocal of the slope of the pressure-natriuresis curve reflects the sodium sensitivity of BP,11 12 13 mefruside lowers BP especially in patients with high sodium sensitivity by reducing their sodium sensitivity. Mefruside made it possible for sodium balance to be maintained under lowered systemic BP based on steepening the slope of the pressure-natriuresis curve by inhibiting tubular sodium reabsorption.
The hypotensive effect of mefruside was greater during high sodium than low sodium, and BP became relatively insensitive to the amount of sodium after mefruside treatment. It is reported that serum potassium is further reduced by high sodium intake with diuretics, since tubular sodium load and sodium-potassium exchange at distal tubules are enhanced by increased sodium intake.19 This finding and the fact that the hypotensive effect of diuretics is diminished by an excessive amount of sodium intake20 usually encourage clinicians to restrict sodium intake during diuretic administration. However, the present study has indicated that this effect was almost negligible. Instead, uric acid was significantly elevated by mefruside in combination with sodium restriction. Taken together, these data suggest that strict sodium restriction may be unnecessary during diuretic administration.
Our data also suggest that the hypotensive mechanisms of mefruside may be different under different amounts of sodium intake. It is certain that the hypotensive effect of mefruside during high sodium was based on its diuretic action, because the effect was correlated with both the hypotensive effect of sodium restriction (Fig 2⇑) and the increase in the slope of the pressure-natriuresis curve (Fig 3⇑). Since the slope is believed to be determined by glomerulotubular balance of sodium between the glomerular ultrafiltration coefficient and tubular sodium reabsorption,12 13 the steepened slope by mefruside can be attributed to the inhibition of tubular sodium reabsorption, as discussed above. On the other hand, the hypotensive effect of mefruside during relatively low sodium was correlated with neither the hypotensive effect of sodium restriction nor the increase in the slope. The hypotensive effect had a tendency to correlate with the shift of the pressure-natriuresis curve toward a lower BP level along the BP axis. Since the x intercept of the curve seemed determined mainly by the preglomerular vascular resistance from heart to glomeruli,12 13 the effect during low sodium may be based on the reduction in preglomerular resistance, probably because of the vasodilating action of diuretics. Diuretics initially lower cardiac output and increase total peripheral vascular resistance. However, by 1 week, “reverse whole body autoregulation” is occurring, and peripheral resistance decreases while cardiac output returns to normal.6 8 21 Thus, after 7 days, diuretics are acting as vasodilators. The very high sodium diet may prevent or retard this reverse autoregulatory process. Mefruside, which is structurally similar to loop diuretics, may also have a weak direct vasodilating action in addition to the above indirect action.
We have previously proposed classifying antihypertensive drugs into three groups based on their effects on the pressure-natriuresis relationship.4 22 The first group, consisting of vasodilators such as calcium antagonists, shifts the pressure-natriuresis curve to the left without affecting slope. The hypotensive effect of this group can be obtained independently of the amount of salt intake, and the hypotensive effects of this group and sodium restriction are thought to be additive. The second group, consisting of β-blockers and angiotensin-converting enzyme inhibitors, shifts the pressure-natriuresis curve leftward, with a decrease in slope. The hypotensive effect is augmented under sodium restriction and weakened under sodium overload and is thought to be synergistic with sodium restriction. The third group of antihypertensive drugs, consisting of diuretics, can increase the slope of the pressure-natriuresis curve, as described in this study. Since the hypotensive effect of diuretics is greater during sodium overload than during sodium restriction, the hypotensive effects of diuretics and sodium restriction are thought to be antagonistic. In addition, diuretics will potentiate the hypotensive effects of the β-blocker propranolol23 24 and the angiotensin-converting enzyme inhibitor captopril25 26 by increasing the slope that has been depressed by the two drugs. This classification of antihypertensive drugs may be useful in clinical practice for selecting the most effective drug or combination of drugs in the treatment of hypertensive patients with various pressure-natriuresis curves and sodium sensitivities.
It is already well known that the pressure-natriuresis curve is shifted to the right and the slope is depressed with increased severity of hypertension.1 2 3 4 Parfrey and colleagues2 3 reported that children born to hypertensive parents showed a pressure-natriuresis curve shifted to the right even before the onset of hypertension compared with children born to normotensive parents. Thus it is clear that as essential hypertension progresses from borderline to mild and severe forms, the pressure-natriuresis curve will be shifted farther to the right and the slope depressed. Therefore, our data suggest that the hypotensive effect of diuretics is less effective in patients with borderline and mild essential hypertension, who have a steep slope and low sodium sensitivity, than in those with severe hypertension, who have a depressed slope and high sodium sensitivity.
This work was supported by Research Grants for Cardiovascular Diseases (C-1994-6 and C-1995-3) and for “Progressive Renal Disease” from “Specially Selected Diseases by the Ministry of Health and Welfare Research Project,” from the Ministry of Health and Welfare of Japan.
Reprint requests to Fumio Saito, MD, Department of Cardiovascular Medicine, Surugadai Hospital, Nihon University School of Medicine, 1-8-13 Kanda-Suruga-dai, Chiyoda-ku, Tokyo 101, Japan.
- Received October 24, 1995.
- Revision received November 9, 1995.
- Accepted January 15, 1996.
Omvik P. Pressure natriuresis in hypertension. Acta Med Scand. 1981;646(suppl):77-89.
Parfrey PS, Markandu ND, Roulstone JE, Jones BE, Johnes JC, Macgregor GA. Relation between arterial pressure, dietary sodium intake and renin system in essential hypertension. Br Med J. 1981;283:94-97.
Parfrey PS. Salt in essential hypertension. In: Sleight P, Freis ED, eds. Hypertension. Butterworth’s International Medical Reviews, Cardiology 1. London, UK: Butterworth; 1982:322-339.
Guyton AC. Arterial Pressure and Hypertension, Circulatory Physiology III. Philadelphia, Pa: WB Saunders; 1980.
Guyton AC. Renal function curve: a key to understanding the pathogenesis of hypertension. Hypertension. 1987;10:1-6.
Guyton AC. Dominant role of the kidneys and accessory role of whole-body autoregulation in the pathogenesis of hypertension. Am J Hypertens. 1989;2:575-585.
Kimura G, Saito F, Kojima S, Yoshimi H, Abe H, Kawano Y, Yoshida K, Ashida T, Kawamura M, Kuramochi M, Ito K, Omae T. Renal function curve in patients with secondary forms of hypertension. Hypertension. 1987;10:11-15.
Kimura G, Imanishi M, Sanai T, Kawano Y, Kojima S, Yoshida K, Abe H, Ashida T, Yoshimi H, Kawamura M, Kuramochi M, Omae T. Intrarenal hemodynamics in patients with essential hypertension. Circ Res. 1991;69:421-428.
Koomans HA, Roos JC, Boer P, Geyskes GG, Dorhout Mees EJ. Salt sensitivity of blood pressure in chronic renal failure: evidence for renal control of body fluid distribution in man. Hypertension. 1982;4:190-197.
Kimura G, Brenner BM. The renal basis for salt sensitivity in hypertension. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis and Management. New York, NY: Raven Press Publishers; 1995:1569-1588.
Norman RA Jr, Enobakhare JA, DeClue JW, Douglas BH, Guyton AC. Arterial pressure-urinary output relationship in hypertensive rats. Am J Physiol. 1978;234:R98-R103.
Sugai T, Nakagawa Y, Takeda K, Imai S. Arterial pressure-urinary output relationship in DOCA-saline hypertensive rats. Am J Physiol. 1983;245:R633-R636.
Luft FC, Rankin LI, Bloch R, Weyman AE, Willis LR, Murray RH, Grim CE, Weinberger MH. Cardiovascular and humoral responses to extremes of sodium intake in normal black and white men. Circulation. 1979;60:697-706.
Roos JC, Koomans HA, Dorhout Mees EJ, Delawi IMK. Renal sodium handling in normal humans subjected to low, normal, and extremely high sodium supplies. Am J Physiol. 1985;249:F941-F947.
Reineck HJ, Stein JH, Seldin DW. Integrated responses of the kidney to alterations in extracellular fluid volume. In: Seldin DW, Giebisch G, eds. The Kidney: Physiology and Pathophysiology. New York, NY: Raven Press Publishers; 1985:1137-1161.
Winer BM. The antihypertensive actions of benzothiadiazines. Circulation. 1961;23:211-218.
Kimura G, Abe H, Ashida T, Kawano Y, Yoshimi H, Imanishi M, Kawamura M, Kojima S, Yoshida K, Kuramochi M, Omae T. Hypotensive mechanisms of antihypertensives based on the water tank model. Am J Hypertens. 1989;2:532-536.
Owens CJ, Brackett NC. Role of sodium intake in the antihypertensive effect of propranolol. South Med J. 1978;71:43-46.
Geyskes GG, Boer P, Vos J, Leenen FHH, Mees EJ. Effects of salt depletion and propranolol on blood pressure and plasma renin activity in various forms of hypertension. Circ Res. 1975;36(suppl I):I-248-I-256.
Brunner HR, Gavras H, Waeber B. Enhancement by diuretics of the antihypertensive action of long-term angiotensin converting enzyme blockade. Clin Exp Hypertens. 1980;2:639-657.