The Effects of a Combined Low-Sodium, High-Potassium, High-Calcium Diet on Blood Pressure in Patients with Mild Hypertension

Home , Potassium

ORIGINAL ARTICLE The effects of a combined low-sodium, high-potassium, high-calcium diet on blood pressure in patients with mild hypertension

E Grossman, A Vald, E Peleg, B Sela and T Rosenthal Chorley Hypertension Unit, Chaim Sheba Medical Center, Tel Hashomer, and Tel Aviv University Sackler School of Medicine, Israel

Nutritional sodium, potassium and calcium are con- (P Ͻ 0.05). Eighteen patients did not comply with the sidered to be important regulators of blood pressure diet. Systolic BP (SBP) only slightly decreased, from 142 (BP). The present study evaluates the effects of com- mm Hg (95% CI, 137–146 mm Hg) to 138 mm Hg (95% bined low-sodium (LS), high-potassium (HK), high-cal- CI, 133–142 mm Hg); (P = 0.11). The change in SBP was cium (HCa) diet on BP in patients with mild essential related to the change in urinary sodium excretion hypertension. Thirty-six patients (26 M, 10 F), 24–67 (R = 0.46; P = 0.006). After 1 month of the HK diet, uri- years of age (mean 46 ± 8), participated in the study. nary potassium excretion increased by only Patients were divided into three groups and given a diet 5 mmols/day (P = NS). BP was unaffected by HK and consisting of three 1-month segments, which they fol- HCa diet. At the end of the study, urinary sodium lowed in different order. Group 1 (n = 11) received LS excretion decreased from 183 mmols/day (95% CI, 155– diet followed by the addition of HCa and then HK. The 211 mmols/day) to 148 mmols/day (95% CI, 131– order in Group 2 (n = 12) was HK-LS-HCa; and in 165 mmols/day); (P Ͻ 0.05), urinary potassium excretion Group 3 (n = 13) it was HCa-HK-LS. The third month of slightly increased from 75 mmols/day (95% CI, 68–82 the study all patients were eating a combined LS, HK mmols/day) to 85 mmols/day (95% CI, 76–94 and HCa diet. Urinary electrolytes were measured to mmols/day); (P = 0.09), and urinary calcium excretion conﬁrm compliance with the diets. After 1 month of the remained unchanged. BP did not decrease. It is con- LS diet urinary sodium excretion decreased signiﬁ- cluded that only the LS diet may be advantageous in cantly by 25 mmols/day (95% CI, 1–48 mmols/day); patients with mild essential hypertension.

Keywords: sodium; potassium; calcium; diet; blood pressure

Introduction Patients and methods Nutritional sodium, potassium and calcium are con- Patients sidered to be important regulators of blood pressure Thirty-six patients (26 M, 10 F), from our out-patient (BP). Some studies have shown that sodium restriction ± 1–5 clinic, 24–67 years of age (mean 46 8 years) par- lowers BP while others have found no effect. Pot- ticipated in the study. The relevant clinical data at assium and calcium supplementation was also con- 6–12 enrolment in the study are given in Table 1. sidered beneﬁcial in lowering BP, although the All participants had essential hypertension and ﬁnding on the value of manipulating these electro- = 13–18 were either on no drug therapy (n 20), or treated lytes in hypertension are controversial. Each with a ␤-blocker (n = 7), or a calcium channel dietary manipulation has been evaluated as a sole blocker (n = 9) as a single agent throughout the intervention in these various studies, but the combi- study. Patients were included only if they stated nation of all three has seldom been studied. their willingness to maintain the diet. Excluded The present study was undertaken to assess from the study were patients with secondary hyper- whether a combined low-sodium (LS), high-potas- tension, ischaemic heart disease, congestive heart sium (HK), high-calcium (HCa) diet is achievable, failure, renal failure (creatinine Ͼ1.5 mg/dL) and and whether it affects BP in patients with mild treated diabetes mellitus. essential hypertension. Study design Patients were divided into three groups and given a diet which they followed in different order, as illus- Correspondence: Dr Ehud Grossman, Hypertension Unit, Chaim = Sheba Medical Center, Tel Hashomer 52621, Israel trated in Figure 1. Group 1 (n 11) began with the Received 29 November 1996; revised and accepted 5 May 1997 LS diet followed by the addition of HCa and HK Sodium, potassium and calcium in hypertension E Grossman et al 790 Table 1 Patients characteristics at time of enrolment in the study (mean Ϯ s.d.)

Group 1 Group 2 Group 3

Number 11 12 13 Height (cm) 168 ± 10 174 ± 9 175 ± 5 Weight (kg) 76 ± 10 89 ± 13 85 ± 12 BMI (kg/m2) 26.8 ± 2.9 28.6 ± 3.9 28.1 ± 3.8 Waist/Hip 0.94 ± 0.12 1.04 ± 0.08 1.03 ± 0.05 MAP (mm Hg) 103 ± 4 112 ± 11 112 ± 13 HR (beats/min) 76 ± 14 73 ± 11 79 ± 11 Urinary sodium (mmols/day) 143 ± 58 194 ± 96 219 ± 78 Urinary sodium (mmols/gr 138 ± 77 151 ± 71 180 ± 50 creatinine) Urinary potassium (mmols/day) 76 ± 22 75 ± 29 77 ± 12 Urinary potassium (mmols/gr 75 ± 35 56 ± 12 68 ± 24 creatinine) Urinary calcium (mg/day) 128 ± 63 185 ± 106 121 ± 83 Energy intake (Kcal) 2260 ± 765 2467 ± 1130 2600 ± 509 Carbohydrate intake (g/day) 282 ± 92 308 ± 120 330 ± 85 Fat intake (g/day) 81 ± 46 97 ± 59 92 ± 35 Protein intake (g/day) 114 ± 31 105 ± 44 115 ± 27 Calcium intake (mg/day) 1115 ± 129 985 ± 175 710 ± 94

BMI = body mass index; MAP = mean arterial pressure; HR = heart rate.

diets. Group 2 (n = 12) began with HK followed by (Version 6 Copyright 1990, N-Sgeared Company, LS and HCa diets. Group 3 (n = 13) started on the USA). HCa diet followed by HK and LS diets. The third BP, heart rate, body weight, waist and hip circum- month of the study all patients were eating a com- ference and 24 h urinary electrolytes were measured bined LS, HK and HCa diet. at baseline and monthly for 3 months. BP was meas- Recommendations for the diets were: 50 mmols ured with a sphygmomanometer three times in the sodium per day for the LS diet, supplementation of sitting position after 5 min rest; phases 1 and 5 of 30–60 mmols potassium per day for the HK diet, and the Korotkoff sounds were used to determine the 1000 mg calcium per day for the HCa diet. systolic and diastolic BPs (SBP/DBP), respectively. LS diet was attained by cooking without adding Sodium and potassium in urine were measured with salt, and using low-sodium substitutes to the regular a flame photometer, and calcium was measured diet, such as fresh vegetables and fish instead of spectrophotometrically on a monarch 2000 analyser olives, sour cucumbers and tuna. The HK diet was with the indicator methylthymol blue reagent of attained by increasing the consumption of veg- bioMe´rieux (France). To confirm complete urine col- etables and legumes and by moderately increasing lection the values were also corrected for creatin- the consumption of fresh fruits. Suggestions for food ine excretion. preparation were also given, in particular, subjects were encouraged to steam rather than boil legumes Data analysis and vegetables. A compensatory moderate reduction ± in the intake of animal protein (meat) and total fat Values are presented as mean s.d. Analysis of vari- (olive oil) was suggested, however, to keep the total ance (ANOVA) for repeated measurements (Statview + energy and macronutrient intake unchanged. HCa 512 , Brainpower Inc, Calabasas, CA, USA) was diet was attained by increasing the consumption of used to compare values at baseline with those at dairy products, such as milk and cheese, sesame, each diet manipulation. Pearson’s correlation was and fish. Patients were asked to alter only electrolyte used to test the relationship between changes in BP intake and to avoid changes in other dietary habits and urinary electrolytes, and between the estimated and lifestyle during the study. A nutritional coun- electrolyte intake and the 24-h urinary excretions. A selor was provided to help subjects maintain the P value of less than 0.05 defined statistical signifi- prescribed diet. To monitor compliance patients cance. were seen frequently by the dietitian who calculated the electrolyte intake. Dietary intake was assessed at Results each clinic visit using a recall questionnaire that Nutrient analysis was filled by the patient under the supervision of the dietitian. Foods were then coded for computer Energy intake remained unchanged during the study analysis by using a Nutritionist 3 computer program (baseline value, 2424 kcal/day [95% CI, 2147– Sodium, potassium and calcium in hypertension E Grossman et al 791 (Figure 2). Among the 18 who did comply, 11 exhibited a decrease in SBP (Figure 2) while seven showed no change or even an increase in this para- meter. These patients were probably salt-resistant subjects. The change in SBP was related to the change in urinary sodium excretion (Figure 2, R = 0.46, P = 0.006). Eliminating the seven salt-resistant subjects from the calculation further strengthened this correlation (Figure 2, R = 0.64, P = 0.0003). The rate of compliance and the rate of response to LS diet were not affected by age, sex, BMI, and waist/hip ratio. Salt sensitivity was unrelated to age. Urinary sodium excretion increased over time in patients who started the study on LS diet (Group 1, Figure 3).

The effects of a high potassium diet After 1 month of HK diet, urinary potassium excretion increased by only 5 mmols/day from 82 mmols/day (95% CI, 71–92 mmols/day) to 87 mmols/day (95% CI, 76–97 mmols/day); (P = 0.27). BP was unaffected, and changes in neither systolic nor diastolic BP were related to the change in urinary potassium excretion.

Figure 1 Study design; LS = low sodium diet, HK = high potassium diet, HCa = high calcium diet.

2701 kcal/day]; and at the end of the study 2326 kcal/day [95% CI, 2135–2517 kcal/day]; P = NS). Estimated sodium and potassium intake sig- niﬁcantly correlated with the 24-h urinary excretions (R = 0.43 for sodium, and R = 0.4 for potassium, P Ͻ 0.001 for both). Therefore, for these electrolytes urinary excretions are presented. Esti- mated calcium intake far exceeded and only slightly correlated with the 24-h urinary excretion (R = 0.23, P Ͻ 0.01), and therefore, for calcium the estimated intakes are also used.

The effects of a low sodium diet After 1 month of the LS diet, urinary sodium excretion decreased significantly by 25 mmols/day (95% CI, 1– 48 mmols/day), from 170 mmols/day (95% CI, 150–190 mmols/day) to 145 mmols/day (95% CI, 120–170 mmols/day); (P Ͻ 0.05). SBP decreased only slightly and not significantly from 142 mm Hg (95% CI, 137–146 mm Hg) to 138 Figure 2 (a) Relation between the change in urinary sodium mm Hg (95% CI, 133–142 mm Hg); (P = 0.11), and excretion and the change in systolic blood pressure (⌬SBP) for the entire group (R = 0.46; P = 0.006). (b) Relation between the DBP remained unchanged. change in urinary sodium excretion and the change in systolic Analyses of individual results showed that 18 blood pressure (⌬SBP) after eliminating the salt-resistant patients patients (50%) did not comply with the diet (R = 0.64; P = 0.0003). Sodium, potassium and calcium in hypertension E Grossman et al 792 month, demonstrating how difficult dietary sodium restriction is to maintain in free living individuals. Poor dietary compliance has been described by other investigators who attempted to decrease sodium intake by altering food habits.1,24,25 Most people find the LS diet tasteless and unacceptable, and the poor compliance can be attributed in part to lack of obvious feedback such as weight reduction. These findings raise the question of whether the LS diet is a practical therapeutic strategy for long-term BP management. The adherence to the HK diet was even lower than the adherence to the LS diet, as urinary potassium excretion increased by only 10 mmols/day after 3 months. While the HK diet can be obtained by supplementation of potassium chloride tablets, the present study was designed to evaluate whether altering food habits rather than drug therapy can lower Figure 3 Urinary sodium excretion over time in patients who BP in hypertensive patients. started the study on a low sodium (LS) diet. The adherence to HCa diet is difficult to assess, since urinary calcium excretion does not reflect calcium intake.26 Moreover, LS and HK intake may The effects of a high calcium diet decrease urinary calcium excretion.9,26,27 Since our Nutrient analysis showed a significant increase in patients were all on a LS or HK diet, the lack of calcium intake after 1 month of HCa diet from 933 increase in urinary calcium excretion does not mg/day (95% CI, 770–1097 mg/day) to 1029 mg/day necessarily mean lack of compliance with the HCa (95% CI, 930–1126 mg/day); (P Ͻ 0.05). However, diet. In fact, estimated calcium intake increased sig- urinary calcium excretion was unchanged during nificantly during the study. this period despite adherence to the diet. BP was We failed to show a significant decrease in BP unaffected and changes in systolic or diastolic BP with the combined diet. In fact, the HK and HCa diet were unrelated to the change in calcium intake or had no effect whatsoever on BP. A LS diet is urinary calcium excretion. regarded as effective in lowering BP and several studies have provided evidence of the relationship between sodium intake and BP.28–32 Studies evaluat- The effects of combined LS, HK and HCa diet ing the role of decreased dietary sodium as mono- After 3 months of dietary intervention, urinary therapy for hypertension have yielded conflicting sodium excretion decreased significantly by results.1–5 MacGregor et al33 showed a significant fall 34.6 mmols/day (95% CI, 1.4–67.8 mmols/day); in BP with the LS diet, but his patients used salt- (P Ͻ 0.05), urinary potassium increased slightly free products to maintain the reduction in salt (P = 0.09), urinary sodium/potassium ratio intake. The lack of decrease in BP in the present decreased significantly (P Ͻ 0.05). Estimated cal- study may be due to an insufficient decrease in diet- cium intake increased by 281 mg/day (95% CI, 172– ary sodium intake. Sodium restriction is more bene- 390 mg/day); (P Ͻ 0.01) while urinary calcium ficial in women and in older hypertensive excretion remained unchanged (Table 2). Body patients;3,29,34 our group included mainly middle- weight, systolic, diastolic and mean arterial pressure aged men. Moreover, it is known that the BP remained the same (Table 2). Change in mean response to sodium restriction is heterogeneous and arterial pressure was unrelated to change in urinary only salt-sensitive patients will respond. Our study sodium or potassium excretion, but was related to included salt-resistant patients, making the mean the change in urinary sodium/potassium ratio decrease in BP in the entire group not significant. (Figure 4, R = 0.34; P = 0.05). Despite this lack of a decrease in BP in the entire group, there was a correlation between the change Discussion in SBP and the change in urinary sodium excretion, supporting the contribution of sodium intake to BP. Reduction of sodium intake is recommended as one Our findings may also explain the inconsistent find- of the interventions to lower BP,19 and several stud- ings on the effect of sodium restriction on BP ies have shown that HK and HCa diet may also described by other investigators.1–5 lower BP in animal models and in hypertensive Potassium supplementation can lower BP by acce- patients.6–12,20–23 The present study evaluated the lerating diuresis and natriuresis mainly in older effect of each diet alone and their combined effect hypertensive patients.6,7,13 In the present study, on on BP. Despite careful selection of patients who mainly middle-aged men, the increase in urinary expressed readiness to follow the diet, and close potassium was moderate and probably insufficient monitoring of compliance only 50% complied with to lower BP. In addition, patients ate a LS diet, and the LS diet, and even fewer with the HK diet. In fact, the hypotensive response to potassium supplemen- urinary sodium excretion increased with time in tation is mild when sodium intake is kept low.13 subjects who complied with the diet in the first A combination of the LS and HK diet led to a sig- Sodium, potassium and calcium in hypertension E Grossman et al 793 Table 2 Urinary electrolyte and haemodynamic values at baseline and after 3 months (mean ± s.d.)

Baseline End of study Change (95% CI)

(to −1.4 67.8−) 34.6− ءUrinary sodium 183 ± 76 148 ± 53 (mmols/day) (to −0.9 58.8−) 28.7− ءUrinary sodium 157 ± 67 118 ± 47 (mmols/gr creatinine) (to 22.9 0.1) 11.5 ءءUrinary 75 ± 23 85 ± 29 potassium(mmols/day) Urinary potassium 66 ± 26 71 ± 27 7.1 (−7.1 to 21.3) (mmols/gr creatinine) Urinary calcium 141 ± 95 130 ± 88 −3.1 (−42 to 36) (mg/day) Calcium intake 933 ± 496 1233 ± 377 281 (172 to 390) (mg/day) (to −0.31 0.99−) 0.65− ءءءUrinary Na/K ratio 2.52 ± 1.06 1.87 ± 0.89 Body weight (kg) 84 ± 11 83 ± 11 −1.0 (−2.0 to 0.0) SBP (mm Hg) 142 ± 17 139 ± 12 −2.5 (−6.5 to 1.5) DBP (mm Hg) 93 ± 11 92 ± 7 −0.9 (−4 to 2.2)

CI = conﬁdence interval; SBP = systolic blood pressure; DBP = diastolic blood pressure. .P Ͻ 0.01 vs baselineءءء ;P = 0.09 vs baselineءء ;P Ͻ 0.05 vs baselineء

increase calcium intake. Weinberger et al36 demon- strated congruity between sodium and calcium responsiveness of BP in normal and hypertensive humans, showing that only salt-sensitive patients exhibited a significant decrease in BP in response to calcium supplementation. Since our study included salt-resistant patients, the BP response in the entire group was not significant. Our results suggest that the HCa diet may not lower BP in sodium-depleted, middle-aged, hypertensive patients. In conclusion, dietary changes are difficult to ach- ieve in the long run, even with intense counselling and concentrated effort in patients willing to change life habits. Mild sodium restriction is more attain- able than a HK diet, but compliance decreases over time. The response to dietary intervention is heterogeneous and sodium restriction may be effective only in salt-sensitive patients. Thus, only the LS diet should be tried in patients with mild essential hypertension. Figure 4 Relationship between the change in urinary sodium/potassium ratio and the change in mean arterial pressure (⌬MAP). References 1 Davis BR et al. Lack of effectiveness of a low- nificant decrease in the urinary sodium/potassium sodium/high-potassium diet in reducing antihypertensive medication requirements in overweight persons ratio. The change in mean arterial pressure was with mild hypertension. TAIM Research Group. Trial related to the change in this ratio, supporting the of Antihypertensive Interventions and Management. relationship of sodium and potassium intake to BP. Am J Hypertens 1994; 7: 926–932. The HCa diet did not affect the BP. Previous stud- 2 Elliott P. Nutritional factors in blood pressure. J Hum ies, mainly in animal models of hypertension, Hypertens 1994; 8: 595–601. showed that calcium supplementation decreases or 3 Geleijnse JM et al. Reduction in blood pressure with a prevents the development of hypertension in low sodium, high potassium, high magnesium salt in sodium-loaded animals.11,21,35 Our patients were older subjects with mild to moderate hypertension. Br relatively sodium-depleted and therefore may not Med J 1994; 309 (6952): 436–440. have responded to calcium supplementation. A 4 Law MR, Frost CD, Wald NJ. By how much does dietary salt reduction lower blood pressure? III – Analysis beneficial effect of calcium supplementation on BP 10–12,36 of data from trials of salt reduction. Br Med J 1991; 302 has been reported in hypertensive patients, (6780): 819–824. 10,36 but some of these studies used calcium tablets 5 Weder AB, Egan BM. Potential deleterious impact of and others were done in elderly patients,10 while we dietary salt restriction on cardiovascular risk factors. studied middle-aged patients and used diet alone to Klin Wochenschr 1991; 69 (Suppl 25): 45–50. Sodium, potassium and calcium in hypertension E Grossman et al 794 6 Mano M et al. Potassium accelerates urinary sodium 22 Benedetti RG, Wise KJ, Massey LK. The hemodynamic excretion during salt loading without stimulating atrial effect of dietary calcium supplementation on rat reno- natriuretic polypeptide secretion. Clin Exp Pharmacol vascular hypertension. Basic Res Cardiol 1993; 88: Physiol 1992; 19: 795–801. 60–71. 7 Smith SR, Klotman PE, Svetkey LP. Potassium chlor- 23 Kishimoto H et al. Chronic parathyroid hormone ide lowers blood pressure and causes natriuresis in administration reverses the antihypertensive effect of older patients with hypertension. J Am Soc Nephrol calcium loading in young spontaneously hypertensive 1992; 2: 1302–1309. rats. Am J Hypertens 1993; 6: 234–240. 8 Krishna GG, Miller E, Kapoor S. Increased blood pres- 24 Kumanyika S. Behavioral aspects of intervention stra- sure during potassium depletion in normotensive men. tegies to reduce dietary sodium. Hypertension 1991; N Engl J Med 1989; 320: 1177–1182. 17 (Suppl I): I190–I195. 9 Krishna GG, Kapoor SC. Potassium depletion exacer- 25 Wylie Rosett J et al. Trial of antihypertensive inter- bates essential hypertension. Ann Intern Med 1991; vention and management: greater efficacy with weight 115 (2): 77–83. reduction than with a sodium-potassium intervention. 10 Takagi Y et al. Calcium treatment of essential hyper- J Am Diet Assoc 1993; 93: 408–415. tension in elderly patients evaluated by 24 h monitor- 26 He J et al. Effect of dietary electrolytes upon calcium ing. Am J Hypertens 1991; 4: 836–839. excretion: the Yi People Study. J Hypertens 1992; 10 11 Saito K et al. Calcium supplementation in salt-depen- (7): 671–676. dent hypertension. Contrib Nephrol 1991; 90: 25–35. 27 Luft FC et al. Salt sensitivity and resistance of blood 12 McCarron DA. Epidemiological evidence and clinical pressure. Age and race as factors in physiological trials of dietary calcium’s effect on blood pressure. responses. Hypertension 1991; 17 (Suppl 1): I102– Contrib Nephrol 1991; 90: 2–10. I108. 13 Krishna GG. Role of potassium in the pathogenesis of 28 Muntzel M, Drueke T. A comprehensive reivew of the hypertension. Am J Med Sci 1994; 307 (Suppl 1): salt and blood pressure relationship. Am J Hypertens S21–S25. 1992; 5: 1S–42S. 14 Luft FC et al. The effect of dietary interventions to 29 Law MR, Frost CD, Wald NJ. By how much does diet- reduce blood pressure in normal humans. J Am Coll ary salt reduction lower blood pressure? I – Analysis Nutr 1989; 8 (6): 495–503. of observational data among populations. Br Med J 15 Tanji JL et al. Dietary calcium supplementation as a 1991; 302 (6780): 811–815. treatment for mild hypertension. J Am Board Fam 30 Intersalt Cooperative Research Group: Intersalt: An Pract 1991; 4: 145–150. international study of electrolyte and blood pressure. 16 Simon JA, Browner WS, Tao JL, Hulley SB. Calcium Results for 24-hour urinary sodium and potassium intake and blood pressure in elderly women. Am J Epi- excretion. Br Med J 1988; 297: 319–328. demiol 1992; 136: 1241–1247. 31 Elliot P. Observational studies of salt and blood pres- 17 Kynast Gales SA, Massey LK. Effects of dietary cal- sure. Hypertension 1991; 17 (Suppl I): I3–I8. cium from dairy products on ambulatory blood pres- 32 Stamler R. Implications of the INTERSALT Study. sure in hypertensive men. J Am Diet Assoc 1992; 92: Hypertension 1991; 17 (Suppl I): I16–I20. 1497–1501. 33 MacGregor GA et al. Double-blind study of three 18 Cutler JA, Brittain E. Calcium and blood pressure. An sodium intakes and long-term effects of sodium epidemiologic perspective. Am J Hypertens 1990; 3: restriction in essential hypertension. Lancet 1989; 2: 137S–146S. 1244–1247. 19 Joint National Committee on Detection, Evaluation, 34 Nestel PJ et al. Enhanced blood pressure response to and Treatment of High Blood Pressure: The fifth dietary salt in elderly women, especially those with report. Arch Intern Med 1993; 153: 154–184. small waist: hip ratio. J Hypertens 1993; 11: 1387– 20 Makynen H, Arvola P, Vapaatalo H, Porsti I. High cal- 1394. cium diet effectively opposes the development of 35 Ando K et al. Antihypertensive effect of dietary cal- deoxycorticosterone-salt hypertension in rats. Am J cium loading in angiotensin II-salt rats. Am J Physiol Hypertens 1994; 7: 520–528. 1991; 261 (5 Pt 2): R1070–R1074. 21 Oparil S et al. Dietary Ca2+ prevents NaCl-sensitive 36 Weinberger MH, Wagner UL, Fineberg NS. The blood hypertension in spontaneously hypertensive rats via pressure effects of calcium supplementation in sympatholytic and renal effects. Am J Clin Nutr 1991; humans of known sodium responsiveness. Am J Hyp- 54 (Suppl 1): 227S–236S. ertens 1993; 6: 799–805.