journal of Internal Medicine 1990: 228: 477-482
Milk: does it affect blood pressure? A controlled intervention study E. C. H. VAN BERESTEIJN, M. VAN SCHAIK & G . SCHAAFSMA* From the Department of Nutrition. The Netherlands Institute for Dairy Research (NIZO). Ede. The Netherlands
Abstract. Van Beresteijn ECH, van Schaik M, Schaafsma G (Department of Nutrition, The Netherlands Institute for Dairy Research (NIZO), Ede, The Netherlands). Milk: does it affect blood pressure ? A controlled intervention study. Journal 01Znternal Medicine 1990; 228: 4 7 7 4 8 2 . In a double-blind trial, the effect on blood pressure of supplementation of normal milk (1180 mg Ca2+,1650 mg K+ and 110 mg Mg2+d-l) vs. 'mineral-poor' milk (95 mg Ca2+, 580 mg K+ and 10 mg Mg2+d-l) was studied. Young healthy normotensive female students consumed one of the two supplements while on a low calcium diet (< 500 mg Ca2+d-') for a period of 6 weeks. In both the normal milk- and 'mineral-poor ' milk-supplemented groups systolic blood pressure decreased slightly. However, this decrease was persistently greater in the milk-supplemented group. The individual mean systolic blood pressure change during normal milk treatment ( - 4.1 %) was significantly greater (P = 0.03) than that during 'mineral-poor' milk treatment ( - 1.3%).An effect of normal milk supplementation on diastolic blood pressure could not be demonstrated. The results of the present study indicate a small hypotensive effect of milk consumption, which is attributable to its content of essential minerals. Keywords: blood pressure, calcium. magnesium, milk, normotensive women, potassium.
Introduction Recent interest in the relationship between calcium and blood pressure originated from epidemiological studies which suggested the existence of an inverse relationship between dietary calcium intake and blood pressure. These studies have recently been reviewed [l]. It is known that, in Western countries, about 70-75% of total dietary calcium is derived from dairy products [2]. Clinical trials [3-71 provided conflicting results with regard to the effect of oral calcium on blood pressure. One-sided attention to the effects of calcium may not be justified, since dairy products are an important source not only of calcium but also of other essential minerals, such as potassium and magnesium. It has been shown that both minerals may be hypotensive [8-141. It may therefore be questioned whether these minerals in dairy products * Present address: "NO-CIVO Institutes. Department of Nutrition, Zeist. The Netherlands.
have contributed to the effect on blood pressure observed in the epidemiological studies cited. Consequently, a double-blind placebo-controlled trial was designed in order to test the inverse relationship between dairy minerals and blood pressure in normotensive subjects. The results are presented in this paper.
Study population and methods The study group consisted of 60 normotensive female students of the Training College for Dietitians. They were 19-23 years of age, 50-80 kg in body weight, and were not receiving any medical treatment at the time of recruitment, except for six participants who were using oral contraceptives. All participants gave their written consent. During the course of the study seven participants dropped out for personal reasons. As in a previous study [7]. the students were assigned to two groups according to a randomized block design that accounted for habitual calcium 477
478
E. C. H. V A N BERESTEIJN et al.
intake, oral contraceptive use and body mass index. During the experimental period both groups adopted a low calcium diet ( < 500 mg Ca2+d-'). One group received supplements of normal dairy products and the other received specially produced ' mineral-poor ' dairy products. The study design was approved by the Ethical Review Committee of the Agricultural University, Wageningen. The study was conducted over a period of 7 weeks. Baseline data on blood pressure, body height, body weight and 24-h urinary excretion of calcium, phosphorus, potassium, magnesium and sodium were collected during the first week of the experiment, while subjects were on their usual diet. The habitual intake of dairy products was estimated by means of a specially developed questionnaire (TNO-CIVO Institutes, Department of Nutrition, Zeist, The Netherlands). Calcium intake from dairy products was calculated using a computerized food composition table [15], and represented about 70-75% of total calcium intake [2]. The duration of the experimental period was 6 weeks. The normal product was semi-skimmed milk obtained from the dairy industry. The ' mineral-poor ' milk was developed and manufactured at the Netherlands Institute for Dairy Research. The detailed composition of the milk is shown in Table 1. The amino acid composition of the two types of milk was comparable. Both types were sterilized under standard conditions and poured into tetrapacks of Table 1. Composition of the experimental milk per 100 g product ~
Normal milk Fat (8)' Protein (g) Casein ( g ) t Demineralized whey powder (g)$ Lactose (g) Ca ( m d § p (md§ K (md§ Mg (md§ Na (mg)§
H,O (8)
1.5 3.4 -
4.7 118 99 165 11
45 89
'Mineral-poor' milk
1.5 2.6 5.4
9.5 25
58 0.5 40 90
* Water-free butterfat for mineral-poor milk only. t Sodium and potassium caseinate (spray-dried. 1 :2 w/w)(DMV Milk Industries, Veghel. The Netherlands). $ Deminal 90. 13.9%protein and 81.6%lactose. (NoordNederland. Beilen. The Netherlands). By analysis.
2 5 0 ml. Every week yoghurt was prepared from both types of milk according to standard procedures, except that the incubation time was longer for the 'mineral-poor' milk as a result of delayed growth of the yoghurt culture. The subjects were offered a total of 11d-' according to their personal preference for milk or yoghurt. Blood pressure was measured five times at every weekly session, after a standardized rest period of 5 min in the supine position. Measurements were made on the right arm using a n automatic device (Copal digital sphygmomanometer, model UA 2 3 1. Adquipment Medical, Kotterdam. The Netherlands). The mean value of the five repeated measurements was taken to represent the weekly blood pressure value. Blood pressure was measured twice during the baseline period and during the sixth week of the experimental period, and once a week during the intermediate period. For each subject, the blood pressure measurements obtained during the baseline period and during week 6 of the experimental period were averaged, and the two mean values were regarded as the initial and final values, respectively. Blood pressure values were determined at each time point. The individual mean blood pressure was calculated as the mean of the six blood pressure measurements obtained during treatment. Body height was measured. Body weight was recorded once a week and the energy intake was adjusted as necessary to prevent changes in body weight. Body mass index (kg m-') was calculated. In order to control patient compliance with the calcium-restricted diet and milk supplements, and to obtain information about the changes in calcium, sodium, potassium and magnesium intake, each subject collected 24-h urine samples, one during the baseline period and two during the experimental period (weeks 2 and 6). Urine samples were analysed for total volume and for concentrations of creatinine, calcium, phosphorus, sodium, potassium and magnesium, using standard laboratory methods. Urinary excretion of each variable was expressed as its ratio to creatinine. Statis tics On the basis of the blood pressure variation obtained in a previous study [7], it could be calculated that, with 30 participants in both the normal milk and the 'mineral-poor' milk group, a treatment effect of 3 mmHg could be detected with a type I error (a)of
MILK AND BLOOD PRESSURE
5% and a statistical power of 50%. A two-way unbalanced analysis of variance using a randomized block design [16] did not reveal any effect of classification, according to body mass index and habitual calcium intake, on blood pressure by treatment. As an inverse relationship was expected, onesided t-tests for unpaired observations were used to compare the effects of normal milk and 'mineralpoor' milk treatment on blood pressure at various time intervals and on individual mean blood pressure changes during treatment. A paired t-test (one-sided) was used to assess changes within groups. In order to allow for differences in initial blood pressure, the change in blood pressure was expressed as a percentage of the initial value. This allows a more clinical interpretation of the results.
479
time in the normal and 'mineral-poor' milk groups. In both groups systolic blood pressure decreased slightly during the experimental period. The decrease was persistently greater in the normal milk-supplemented group. The difference between group means was significant at week 3 (P c 0.05), and tended to be different (P c 0.10) at most other time points. The individual mean blood pressure change (-4.1 +0.9%, corresponding to -5.1 1.3 mmHg, mean +SEM) during normal milk treatment was significantly greater (P = 0.03) than that during 'mineralpoor ' milk treatment ( - 1.3 1.1%, corresponding to -2.2k1.3 mmHg, meanfSEM). No effect of normal milk supplementation on diastolic blood pressure was observed.
*
Urinary parameters
Results Baseline characteristics
Baseline data are shown according to treatment group in Table 2 and Figs l(a) and (b). No significant differences were found between groups in the variables collected during the baseline period. Body weight
Table 2 shows that final body weight did not differ significantly between groups. No significant weight changes were observed during the study. Blood pressure
Figure 2 shows the mean percentage changes in systolic and diastolic blood pressure as a function of
The urinary excretion of calcium, phosphorus, sodium, magnesium and potassium during the experimental period is shown in Figs l(a) and (b). Excretion is expressed as the ratio of each variable to creatinine. Compared to baseline values, calcium excretion decreased significantly in the ' mineralpoor' milk group (€' < 0.05), but did not change in the normal milk group. The differences between groups were significant (P c 0.01) (Fig. la). Phosphorus excretion decreased in the ' mineralpoor ' milk group and increased in the normal milk group. Neither change was significant, but the two groups did differ significantly (P c 0.05) (Fig. la). Magnesium excretion decreased in the 'mineralpoor' milk group (P c 0.01), but remained unchanged in the normal milk group. The differences between the two groups were significant (P < 0.05) (Fig. lb). Potassium excretion increased significantly
Table 2. Baseline and final values for 'mineral-poor' milk (n = 25) and normal milk (n = 28) groups
Age (years) Body weight (kg) Body mass index (kg m-') Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Dietary calcium intake (mg d-l)* Dietary potassium intake (mg d--l)* Dietary magnesium intake (mg d-')*
'Mineral-poor' milk group
Normal milk group
Baseline value
Baseline value
20k1.2 61.2 f8.7 20.5f2.1 113.0f 12.2 63.3 f 7.2 953 f294 85 3 k 3 0 8 69k28
Final value 61.7k8.3 20.7f 2.0 110.7+10.0 62.9 k 8.2
Final value
20 & 0.9 60.7k8.0 20.8k2.5 114.0f 10.1 63.2 f 8.2 906f318 754+352 6 0 +2 7
6 1 . 0 k 7.9 20.9 f2.5 109.0f 7.6 63.0 f 7.4
Mean values fSD are shown. From dairy products only (range Ca2+351-1877 mg d-', range K+ 288-2088 mg d-l, range Mg2+18-160 mg d - I ) . 30
I M B 228
480
E. C. H. V A N B E R E S T E I J N et al.
0.25 0
,
,
,
0
2
6
2.01
2
0
6 Time (weeks)
Fig. I(a). Urinary excretion of calcium and phosphorus expressed as ratios of the latter to creatinine. vs. time. Mean valueskSEM are = normal milk group: (A) = ‘mineral-poor’ milk group. Significance level between groups: * P < 0.05: ** P c 0.01. shown. (A)
0
2
0
6
2
0
6
6
2
Time (weeks)
Fig. l(b). Urinary excretion of sodium. magnesium and potassium expressed as ratios of the latter to creatinine vs. time. Mean values +SEM are shown. (A)= normal milk group: (A) = ‘mineral-poor’ milk group. Significance level between groups: * P < 0.05: **P c 0.01: ***P < 0.001.
.(b) f 2
-2 0
a
a
-6 I
0
1
2
3
I
I
4
5
6
0
1
2
3
4
1
5
6
Time (weeks)
Fig. 2. Mean percentage changes in (a) systolic and (b) diastolic blood pressure from basal values vs. time. Bar represents 1 S M . (A) = normal milk group: (A) = ‘mineral-poor’ milk group. Significance level between groups: * P < 0.05.
in the normal milk group (P < 0.05), and decreased significantly in the ‘mineral-poor’ milk group (P < 0.01). The two groups differed significantly (P < 0.001) (Fig. lb). Sodium excretion did not change significantly during the experimental period, and there were no differences in excretion between the two groups (Fig. lb).
Discussion Milk and milk products are important sources of essential minerals and trace elements. In this population the contribution of dairy products to the USA Recommended Daily Allowances [17] of calcium (800 mg d-’), magnesium (300 mg d-l) and potass-
MILK A N D BLOOD PRESSURE
ium (range 1525-4575 mg d-l) was 112%, 21% and 52% of the lower limit, respectively. For reasons of taste, necessary because of the double-blind nature of the study, it was not possible to develop a 'mineral-poor' milk totally depleted of minerals. Sodium in particular appeared to be an essential component of taste. In this experiment the contribution of sodium from normal milk was small compared to total sodium intake. Therefore the sodium level in the ' mineral-poor ' milk was not changed, but the concentrations of calcium, phosphorus, magnesium and potassium were substantially reduced (Table 1). The results of this study showed that, in young healthy normotensive females, supplementing the diet with normal milk for 6 weeks resulted in a slight reduction in systolic blood pressure (2.8%, corresponding to approximately 3 mmHg), which persisted throughout the experimental period. Although the change was small, this effect may be important for several reasons. The participants in this study were healthy young normotensive subjects, who are certainly less susceptible to diet-induced reduction of blood pressure than hypertensive individuals. Thus large decreases in blood pressure were not expected. The persistent slight reduction in blood pressure might therefore be considered to represent evidence of normal milk consumption as a determinant of blood pressure. The risk of cardiovascular mortality and morbidity increases steadily with increasing blood pressure [18]. Rose [19] has estimated that all the life-saving benefits of current hypertensive treatment might be equalled by a downward shift of the total blood pressure distribution in the population by only 2-3 mmHg. The reduction in blood pressure was observed after only 1 week of intervention, suggesting that blood pressure is a dynamic indicator of current electrolyte status after milk consumption. To our knowledge this is the h s t reported study of the effect of the essential mineral content of milk on blood pressure. A tentative comparison of our observations with the results of trials on calcium supplementation alone [3-71 suggests that the effect of milk on blood pressure is greater and more rapid than that of calcium alone, which may be due to its essential mineral content. On the basis of the present study it could not be established whether the effect was caused by individual minerals, other than calcium, or by a combination of minerals, as interaction between these has been shown to occur, In two double-blind trials [ l o , 111 the effect of
481
potassium supplementation (2.5 g d-') on blood pressure was studied in normo- and hypertensive subjects. In both studies, diastolic blood pressure fell significantly, by 2.4 mmHg and 4.0 mmHg, respectively. Systolic blood pressure showed a significant decrease of 7.0 mmHg in hypertensive subjects and a non-significant decrease of 1.1mmHg in normotensive individuals. Furthermore, there is some evidence of an inverse relationship between the sodium/potassium ratio and blood pressure [20]. Decreasing this ratio from 1.7 to 0.5 reduced the systolic blood pressure (by 3.3 mmHg) significantly in young people with mildly elevated blood pressure [21]. Both systolic and diastolic blood pressure were significantly reduced (by 12/8 mmHg) in hypertensive patients who required diuretic therapy after oral magnesium supplementation (365 mg d-', [14]). No effect of oral magnesium supplementation (365 mg d-l) was observed in untreated hypertensive patients [ 2 2 ] .In the present study the differences in potassium intake (1000 mg d-l) and magnesium intake (100 mg d-'), introduced by the supplementation of both milks, were considerable, although less than those in previous clinical trials. However, the results demonstrated that normal milk did have a small blood-pressure-lowering effect. The difference in calcium intake (1100 mg d-l) was comparable to that reported for previous clinical trials [3-51. Patient compliance with the dietary instructions and milk supplements was good, as conhmed by the urinary excretion of all minerals. Dairy products are rich in calcium, phosphorus, magnesium and potassium ; restriction of their intake reduces their urinary excretion. However, normal milk supplementation prevents this (Figs l a and lb). For a given change in dietary calcium, urinary excretion of calcium will change by approximately 6% of the dietary change [23, 241. The changes in urinary excretion of calcium observed in the present study are in agreement with this finding. Urinary excretion of phosphorus, sodium, potassium and magnesium reflects their intake because intestinal absorption of these minerals is less well regulated than that of calcium. We can offer no explanation for the fluctuations in diastolic blood pressure measurements, observed in both groups or for the decrease in urinary sodium excretion observed after 2 weeks. These effects should probably be considered as having been affected by the time of measurement. 30-2
482
E. C. H. V A N BERESTEIJN et al.
In conclusion, the results of the present study indicate that, owing to its content of essential minerals, milk has a slight antihypertensive action, which requires further confirmation.
Acknowledgements We wish to thank T. Lossonszy. M. Dijkmeier, T. Dijkstra and the second-year students (1986/1987) of the Training College for Dietitians, The Hague, for their co-operation and participation. We are indebted to R. Dekker, who assisted with blood pressure measurements, J. van der Heiden-Winkeldermaat, who estimated habitual intake of dairy products, and R. Visser, who carried out the urine analyses. We thank Dr. J. de Wit and M. Adams for development of the placebo milk. The placebo milk and yoghurts were produced in the Technology Hall, NIZO. We thank DOMO, Arnhem for their co-operation in sterilization and packaging the milk into tetrapacks.
References 1 McCarron DA, Morris CD. Epidemiological evidence associating dietary calcium and calcium metabolism with blood pressure, Am ] Nephrol 1986: 6 (Suppl. 1): 3-9. 2 Marston RM, Welsh SO. Nutrient content of United States food supply. National Food Review 1982: NFR-25: 7-1 3. US Department of Agriculture, Economic Research Service. November. 3 Belizan JM, Villar J. Pineda 0 et al. Reduction of blood pressure with calcium supplementation in young adults. Am Med ASSOC1983; 249: 1161-5. 4 McCarron DA, Morris CD. Blood pressure response to oral calcium in persons with mild to moderate hypertension. Ann Intern Med 1985: 103: 825-31. 5 Strazullo P, Siani A, Guglielmi S et al. Controlled trial of longterm oral calcium supplementation in essential hypertension. Hypertension 1986; 8: 1084-8. 6 Johnson NE. Smith EL. Freudenheim JL. Effects on blood pressure of calcium supplementation in women. Am J Clin Nutr 1985: 42: 12-7. 7 Beresteijn ECH van. Schaafsma G. de Waard H. Oral calcium and blood pressure: a controlled intervention trial. A m J Clin Nutr 1986: 44: 883-8. 8 Langford HC. Dietary potassium and hypertension : epidemiologic data. Ann Intern Med 1983: 98: 770-2.
9 Khaw KT. Barrett-Conner E. Dietary potassium and blood pressure in a population. Am / Clin Nutr 1984: 39: 963-8. 10 Khaw KT. Thom S. Randomised double-blind cross-over trial of potassium on blood pressure in normal subjects. Lancet 1982 : ii: 1127-9. 11 MacCregor GA. Smith SJ, Markandu ND, Banks RA. Sagnella GA. Moderate potassium supplementation in essential hypertension. Lancet 1982 : ii: 567-70. 12 Peterson B, Schnell M. Christiansen C, Transbel I. Serum and erythrocyte magnesium in normal elderly Danish people. Acta Med Scand 1977: 2 0 1 : 3 1 4 . 13 McCarron DA. Calcium and magnesium nutrition in human hypertension. Ann lntern Med 1983: 98: 800-5. 14 Dyckner T. Wester PO. Effect of magnesium on blood pressure. Br M e d ] 1983: 286: 1847-9. 1 5 Nederlandse Voedingsstoflen Bestand (NEVO). 's Gravenhage : The Netherlands Bureau for Nutrition Education, 1986-1 987. 16 Snedecor GW, Cochran WG. Statistical Methods'. Iowa: Iowa State University Press, 1973 : 299-338. 17 Truswell AS, Irwin T. Beaton GH. Recommended dietary intake around the world. Nutr Abstr Rew 1983; 53: 993. 18 Kannel WB. Sorlie P. Hypertension in Framingham. In: Paul 0.ed. Epidemiology and Control of Hypertension. New York: Stratton Intercontinental Medical Book Corporation. 1975 : 5 5 3-8 7. 19 Rose G. Strategy of prevention : lessons from cardiovascular disease. Br Med I I981 : 282: 1847-51. 20 Dahl LK.Leitt G. Heine M. Influence of dietary potassium and sodium/potassium molar ratios on the development of salt hypertension ] Exp Mcd 1972: 136: 318-30. 21 Grobbee DE, Hofman A. Roelandt JT, Boomsma F. Schalekamp MA, Valkenburg HA. Sodium restriction and potassium supplementation in young people with mildly elevated blood pressure. J Hypertension 1987: 5: 13 5-9. 22 Cappuccio FP. Markandu ND. Beynon CW. Shore AC. Sampson B. MacGregor GA. Lack of effect of oral magnesium on high blood pressure: a double blind study. Br Med / 1985 : 291: 235-8. 23 Heaney PR, Saville PD. Recker RR. Calcium absorption as a function ofcalcium intake. ] Irrb Clin Med 1975: 8 8 : 881-95. 24 Lemann J, Adams ND. Gray RW. Urinary calcium excretion in human beings. N E n g l ] Men 1979; 301: 535-41. Received 31 October 1989, accepted 18 January 1990.
Correspondence: Dr ir. E. C. H. van Beresteijn, Netherlands Institute for Dairy Research (NIZO). P.O. Box 20. 6710 BA Me. The Netherlands.
Milk: does it affect blood pressure? A controlled intervention study E. C. H. VAN BERESTEIJN, M. VAN SCHAIK & G . SCHAAFSMA* From the Department of Nutrition. The Netherlands Institute for Dairy Research (NIZO). Ede. The Netherlands
Abstract. Van Beresteijn ECH, van Schaik M, Schaafsma G (Department of Nutrition, The Netherlands Institute for Dairy Research (NIZO), Ede, The Netherlands). Milk: does it affect blood pressure ? A controlled intervention study. Journal 01Znternal Medicine 1990; 228: 4 7 7 4 8 2 . In a double-blind trial, the effect on blood pressure of supplementation of normal milk (1180 mg Ca2+,1650 mg K+ and 110 mg Mg2+d-l) vs. 'mineral-poor' milk (95 mg Ca2+, 580 mg K+ and 10 mg Mg2+d-l) was studied. Young healthy normotensive female students consumed one of the two supplements while on a low calcium diet (< 500 mg Ca2+d-') for a period of 6 weeks. In both the normal milk- and 'mineral-poor ' milk-supplemented groups systolic blood pressure decreased slightly. However, this decrease was persistently greater in the milk-supplemented group. The individual mean systolic blood pressure change during normal milk treatment ( - 4.1 %) was significantly greater (P = 0.03) than that during 'mineral-poor' milk treatment ( - 1.3%).An effect of normal milk supplementation on diastolic blood pressure could not be demonstrated. The results of the present study indicate a small hypotensive effect of milk consumption, which is attributable to its content of essential minerals. Keywords: blood pressure, calcium. magnesium, milk, normotensive women, potassium.
Introduction Recent interest in the relationship between calcium and blood pressure originated from epidemiological studies which suggested the existence of an inverse relationship between dietary calcium intake and blood pressure. These studies have recently been reviewed [l]. It is known that, in Western countries, about 70-75% of total dietary calcium is derived from dairy products [2]. Clinical trials [3-71 provided conflicting results with regard to the effect of oral calcium on blood pressure. One-sided attention to the effects of calcium may not be justified, since dairy products are an important source not only of calcium but also of other essential minerals, such as potassium and magnesium. It has been shown that both minerals may be hypotensive [8-141. It may therefore be questioned whether these minerals in dairy products * Present address: "NO-CIVO Institutes. Department of Nutrition, Zeist. The Netherlands.
have contributed to the effect on blood pressure observed in the epidemiological studies cited. Consequently, a double-blind placebo-controlled trial was designed in order to test the inverse relationship between dairy minerals and blood pressure in normotensive subjects. The results are presented in this paper.
Study population and methods The study group consisted of 60 normotensive female students of the Training College for Dietitians. They were 19-23 years of age, 50-80 kg in body weight, and were not receiving any medical treatment at the time of recruitment, except for six participants who were using oral contraceptives. All participants gave their written consent. During the course of the study seven participants dropped out for personal reasons. As in a previous study [7]. the students were assigned to two groups according to a randomized block design that accounted for habitual calcium 477
478
E. C. H. V A N BERESTEIJN et al.
intake, oral contraceptive use and body mass index. During the experimental period both groups adopted a low calcium diet ( < 500 mg Ca2+d-'). One group received supplements of normal dairy products and the other received specially produced ' mineral-poor ' dairy products. The study design was approved by the Ethical Review Committee of the Agricultural University, Wageningen. The study was conducted over a period of 7 weeks. Baseline data on blood pressure, body height, body weight and 24-h urinary excretion of calcium, phosphorus, potassium, magnesium and sodium were collected during the first week of the experiment, while subjects were on their usual diet. The habitual intake of dairy products was estimated by means of a specially developed questionnaire (TNO-CIVO Institutes, Department of Nutrition, Zeist, The Netherlands). Calcium intake from dairy products was calculated using a computerized food composition table [15], and represented about 70-75% of total calcium intake [2]. The duration of the experimental period was 6 weeks. The normal product was semi-skimmed milk obtained from the dairy industry. The ' mineral-poor ' milk was developed and manufactured at the Netherlands Institute for Dairy Research. The detailed composition of the milk is shown in Table 1. The amino acid composition of the two types of milk was comparable. Both types were sterilized under standard conditions and poured into tetrapacks of Table 1. Composition of the experimental milk per 100 g product ~
Normal milk Fat (8)' Protein (g) Casein ( g ) t Demineralized whey powder (g)$ Lactose (g) Ca ( m d § p (md§ K (md§ Mg (md§ Na (mg)§
H,O (8)
1.5 3.4 -
4.7 118 99 165 11
45 89
'Mineral-poor' milk
1.5 2.6 5.4
9.5 25
58 0.5 40 90
* Water-free butterfat for mineral-poor milk only. t Sodium and potassium caseinate (spray-dried. 1 :2 w/w)(DMV Milk Industries, Veghel. The Netherlands). $ Deminal 90. 13.9%protein and 81.6%lactose. (NoordNederland. Beilen. The Netherlands). By analysis.
2 5 0 ml. Every week yoghurt was prepared from both types of milk according to standard procedures, except that the incubation time was longer for the 'mineral-poor' milk as a result of delayed growth of the yoghurt culture. The subjects were offered a total of 11d-' according to their personal preference for milk or yoghurt. Blood pressure was measured five times at every weekly session, after a standardized rest period of 5 min in the supine position. Measurements were made on the right arm using a n automatic device (Copal digital sphygmomanometer, model UA 2 3 1. Adquipment Medical, Kotterdam. The Netherlands). The mean value of the five repeated measurements was taken to represent the weekly blood pressure value. Blood pressure was measured twice during the baseline period and during the sixth week of the experimental period, and once a week during the intermediate period. For each subject, the blood pressure measurements obtained during the baseline period and during week 6 of the experimental period were averaged, and the two mean values were regarded as the initial and final values, respectively. Blood pressure values were determined at each time point. The individual mean blood pressure was calculated as the mean of the six blood pressure measurements obtained during treatment. Body height was measured. Body weight was recorded once a week and the energy intake was adjusted as necessary to prevent changes in body weight. Body mass index (kg m-') was calculated. In order to control patient compliance with the calcium-restricted diet and milk supplements, and to obtain information about the changes in calcium, sodium, potassium and magnesium intake, each subject collected 24-h urine samples, one during the baseline period and two during the experimental period (weeks 2 and 6). Urine samples were analysed for total volume and for concentrations of creatinine, calcium, phosphorus, sodium, potassium and magnesium, using standard laboratory methods. Urinary excretion of each variable was expressed as its ratio to creatinine. Statis tics On the basis of the blood pressure variation obtained in a previous study [7], it could be calculated that, with 30 participants in both the normal milk and the 'mineral-poor' milk group, a treatment effect of 3 mmHg could be detected with a type I error (a)of
MILK AND BLOOD PRESSURE
5% and a statistical power of 50%. A two-way unbalanced analysis of variance using a randomized block design [16] did not reveal any effect of classification, according to body mass index and habitual calcium intake, on blood pressure by treatment. As an inverse relationship was expected, onesided t-tests for unpaired observations were used to compare the effects of normal milk and 'mineralpoor' milk treatment on blood pressure at various time intervals and on individual mean blood pressure changes during treatment. A paired t-test (one-sided) was used to assess changes within groups. In order to allow for differences in initial blood pressure, the change in blood pressure was expressed as a percentage of the initial value. This allows a more clinical interpretation of the results.
479
time in the normal and 'mineral-poor' milk groups. In both groups systolic blood pressure decreased slightly during the experimental period. The decrease was persistently greater in the normal milk-supplemented group. The difference between group means was significant at week 3 (P c 0.05), and tended to be different (P c 0.10) at most other time points. The individual mean blood pressure change (-4.1 +0.9%, corresponding to -5.1 1.3 mmHg, mean +SEM) during normal milk treatment was significantly greater (P = 0.03) than that during 'mineralpoor ' milk treatment ( - 1.3 1.1%, corresponding to -2.2k1.3 mmHg, meanfSEM). No effect of normal milk supplementation on diastolic blood pressure was observed.
*
Urinary parameters
Results Baseline characteristics
Baseline data are shown according to treatment group in Table 2 and Figs l(a) and (b). No significant differences were found between groups in the variables collected during the baseline period. Body weight
Table 2 shows that final body weight did not differ significantly between groups. No significant weight changes were observed during the study. Blood pressure
Figure 2 shows the mean percentage changes in systolic and diastolic blood pressure as a function of
The urinary excretion of calcium, phosphorus, sodium, magnesium and potassium during the experimental period is shown in Figs l(a) and (b). Excretion is expressed as the ratio of each variable to creatinine. Compared to baseline values, calcium excretion decreased significantly in the ' mineralpoor' milk group (€' < 0.05), but did not change in the normal milk group. The differences between groups were significant (P c 0.01) (Fig. la). Phosphorus excretion decreased in the ' mineralpoor ' milk group and increased in the normal milk group. Neither change was significant, but the two groups did differ significantly (P c 0.05) (Fig. la). Magnesium excretion decreased in the 'mineralpoor' milk group (P c 0.01), but remained unchanged in the normal milk group. The differences between the two groups were significant (P < 0.05) (Fig. lb). Potassium excretion increased significantly
Table 2. Baseline and final values for 'mineral-poor' milk (n = 25) and normal milk (n = 28) groups
Age (years) Body weight (kg) Body mass index (kg m-') Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Dietary calcium intake (mg d-l)* Dietary potassium intake (mg d--l)* Dietary magnesium intake (mg d-')*
'Mineral-poor' milk group
Normal milk group
Baseline value
Baseline value
20k1.2 61.2 f8.7 20.5f2.1 113.0f 12.2 63.3 f 7.2 953 f294 85 3 k 3 0 8 69k28
Final value 61.7k8.3 20.7f 2.0 110.7+10.0 62.9 k 8.2
Final value
20 & 0.9 60.7k8.0 20.8k2.5 114.0f 10.1 63.2 f 8.2 906f318 754+352 6 0 +2 7
6 1 . 0 k 7.9 20.9 f2.5 109.0f 7.6 63.0 f 7.4
Mean values fSD are shown. From dairy products only (range Ca2+351-1877 mg d-', range K+ 288-2088 mg d-l, range Mg2+18-160 mg d - I ) . 30
I M B 228
480
E. C. H. V A N B E R E S T E I J N et al.
0.25 0
,
,
,
0
2
6
2.01
2
0
6 Time (weeks)
Fig. I(a). Urinary excretion of calcium and phosphorus expressed as ratios of the latter to creatinine. vs. time. Mean valueskSEM are = normal milk group: (A) = ‘mineral-poor’ milk group. Significance level between groups: * P < 0.05: ** P c 0.01. shown. (A)
0
2
0
6
2
0
6
6
2
Time (weeks)
Fig. l(b). Urinary excretion of sodium. magnesium and potassium expressed as ratios of the latter to creatinine vs. time. Mean values +SEM are shown. (A)= normal milk group: (A) = ‘mineral-poor’ milk group. Significance level between groups: * P < 0.05: **P c 0.01: ***P < 0.001.
.(b) f 2
-2 0
a
a
-6 I
0
1
2
3
I
I
4
5
6
0
1
2
3
4
1
5
6
Time (weeks)
Fig. 2. Mean percentage changes in (a) systolic and (b) diastolic blood pressure from basal values vs. time. Bar represents 1 S M . (A) = normal milk group: (A) = ‘mineral-poor’ milk group. Significance level between groups: * P < 0.05.
in the normal milk group (P < 0.05), and decreased significantly in the ‘mineral-poor’ milk group (P < 0.01). The two groups differed significantly (P < 0.001) (Fig. lb). Sodium excretion did not change significantly during the experimental period, and there were no differences in excretion between the two groups (Fig. lb).
Discussion Milk and milk products are important sources of essential minerals and trace elements. In this population the contribution of dairy products to the USA Recommended Daily Allowances [17] of calcium (800 mg d-’), magnesium (300 mg d-l) and potass-
MILK A N D BLOOD PRESSURE
ium (range 1525-4575 mg d-l) was 112%, 21% and 52% of the lower limit, respectively. For reasons of taste, necessary because of the double-blind nature of the study, it was not possible to develop a 'mineral-poor' milk totally depleted of minerals. Sodium in particular appeared to be an essential component of taste. In this experiment the contribution of sodium from normal milk was small compared to total sodium intake. Therefore the sodium level in the ' mineral-poor ' milk was not changed, but the concentrations of calcium, phosphorus, magnesium and potassium were substantially reduced (Table 1). The results of this study showed that, in young healthy normotensive females, supplementing the diet with normal milk for 6 weeks resulted in a slight reduction in systolic blood pressure (2.8%, corresponding to approximately 3 mmHg), which persisted throughout the experimental period. Although the change was small, this effect may be important for several reasons. The participants in this study were healthy young normotensive subjects, who are certainly less susceptible to diet-induced reduction of blood pressure than hypertensive individuals. Thus large decreases in blood pressure were not expected. The persistent slight reduction in blood pressure might therefore be considered to represent evidence of normal milk consumption as a determinant of blood pressure. The risk of cardiovascular mortality and morbidity increases steadily with increasing blood pressure [18]. Rose [19] has estimated that all the life-saving benefits of current hypertensive treatment might be equalled by a downward shift of the total blood pressure distribution in the population by only 2-3 mmHg. The reduction in blood pressure was observed after only 1 week of intervention, suggesting that blood pressure is a dynamic indicator of current electrolyte status after milk consumption. To our knowledge this is the h s t reported study of the effect of the essential mineral content of milk on blood pressure. A tentative comparison of our observations with the results of trials on calcium supplementation alone [3-71 suggests that the effect of milk on blood pressure is greater and more rapid than that of calcium alone, which may be due to its essential mineral content. On the basis of the present study it could not be established whether the effect was caused by individual minerals, other than calcium, or by a combination of minerals, as interaction between these has been shown to occur, In two double-blind trials [ l o , 111 the effect of
481
potassium supplementation (2.5 g d-') on blood pressure was studied in normo- and hypertensive subjects. In both studies, diastolic blood pressure fell significantly, by 2.4 mmHg and 4.0 mmHg, respectively. Systolic blood pressure showed a significant decrease of 7.0 mmHg in hypertensive subjects and a non-significant decrease of 1.1mmHg in normotensive individuals. Furthermore, there is some evidence of an inverse relationship between the sodium/potassium ratio and blood pressure [20]. Decreasing this ratio from 1.7 to 0.5 reduced the systolic blood pressure (by 3.3 mmHg) significantly in young people with mildly elevated blood pressure [21]. Both systolic and diastolic blood pressure were significantly reduced (by 12/8 mmHg) in hypertensive patients who required diuretic therapy after oral magnesium supplementation (365 mg d-', [14]). No effect of oral magnesium supplementation (365 mg d-l) was observed in untreated hypertensive patients [ 2 2 ] .In the present study the differences in potassium intake (1000 mg d-l) and magnesium intake (100 mg d-'), introduced by the supplementation of both milks, were considerable, although less than those in previous clinical trials. However, the results demonstrated that normal milk did have a small blood-pressure-lowering effect. The difference in calcium intake (1100 mg d-l) was comparable to that reported for previous clinical trials [3-51. Patient compliance with the dietary instructions and milk supplements was good, as conhmed by the urinary excretion of all minerals. Dairy products are rich in calcium, phosphorus, magnesium and potassium ; restriction of their intake reduces their urinary excretion. However, normal milk supplementation prevents this (Figs l a and lb). For a given change in dietary calcium, urinary excretion of calcium will change by approximately 6% of the dietary change [23, 241. The changes in urinary excretion of calcium observed in the present study are in agreement with this finding. Urinary excretion of phosphorus, sodium, potassium and magnesium reflects their intake because intestinal absorption of these minerals is less well regulated than that of calcium. We can offer no explanation for the fluctuations in diastolic blood pressure measurements, observed in both groups or for the decrease in urinary sodium excretion observed after 2 weeks. These effects should probably be considered as having been affected by the time of measurement. 30-2
482
E. C. H. V A N BERESTEIJN et al.
In conclusion, the results of the present study indicate that, owing to its content of essential minerals, milk has a slight antihypertensive action, which requires further confirmation.
Acknowledgements We wish to thank T. Lossonszy. M. Dijkmeier, T. Dijkstra and the second-year students (1986/1987) of the Training College for Dietitians, The Hague, for their co-operation and participation. We are indebted to R. Dekker, who assisted with blood pressure measurements, J. van der Heiden-Winkeldermaat, who estimated habitual intake of dairy products, and R. Visser, who carried out the urine analyses. We thank Dr. J. de Wit and M. Adams for development of the placebo milk. The placebo milk and yoghurts were produced in the Technology Hall, NIZO. We thank DOMO, Arnhem for their co-operation in sterilization and packaging the milk into tetrapacks.
References 1 McCarron DA, Morris CD. Epidemiological evidence associating dietary calcium and calcium metabolism with blood pressure, Am ] Nephrol 1986: 6 (Suppl. 1): 3-9. 2 Marston RM, Welsh SO. Nutrient content of United States food supply. National Food Review 1982: NFR-25: 7-1 3. US Department of Agriculture, Economic Research Service. November. 3 Belizan JM, Villar J. Pineda 0 et al. Reduction of blood pressure with calcium supplementation in young adults. Am Med ASSOC1983; 249: 1161-5. 4 McCarron DA, Morris CD. Blood pressure response to oral calcium in persons with mild to moderate hypertension. Ann Intern Med 1985: 103: 825-31. 5 Strazullo P, Siani A, Guglielmi S et al. Controlled trial of longterm oral calcium supplementation in essential hypertension. Hypertension 1986; 8: 1084-8. 6 Johnson NE. Smith EL. Freudenheim JL. Effects on blood pressure of calcium supplementation in women. Am J Clin Nutr 1985: 42: 12-7. 7 Beresteijn ECH van. Schaafsma G. de Waard H. Oral calcium and blood pressure: a controlled intervention trial. A m J Clin Nutr 1986: 44: 883-8. 8 Langford HC. Dietary potassium and hypertension : epidemiologic data. Ann Intern Med 1983: 98: 770-2.
9 Khaw KT. Barrett-Conner E. Dietary potassium and blood pressure in a population. Am / Clin Nutr 1984: 39: 963-8. 10 Khaw KT. Thom S. Randomised double-blind cross-over trial of potassium on blood pressure in normal subjects. Lancet 1982 : ii: 1127-9. 11 MacCregor GA. Smith SJ, Markandu ND, Banks RA. Sagnella GA. Moderate potassium supplementation in essential hypertension. Lancet 1982 : ii: 567-70. 12 Peterson B, Schnell M. Christiansen C, Transbel I. Serum and erythrocyte magnesium in normal elderly Danish people. Acta Med Scand 1977: 2 0 1 : 3 1 4 . 13 McCarron DA. Calcium and magnesium nutrition in human hypertension. Ann lntern Med 1983: 98: 800-5. 14 Dyckner T. Wester PO. Effect of magnesium on blood pressure. Br M e d ] 1983: 286: 1847-9. 1 5 Nederlandse Voedingsstoflen Bestand (NEVO). 's Gravenhage : The Netherlands Bureau for Nutrition Education, 1986-1 987. 16 Snedecor GW, Cochran WG. Statistical Methods'. Iowa: Iowa State University Press, 1973 : 299-338. 17 Truswell AS, Irwin T. Beaton GH. Recommended dietary intake around the world. Nutr Abstr Rew 1983; 53: 993. 18 Kannel WB. Sorlie P. Hypertension in Framingham. In: Paul 0.ed. Epidemiology and Control of Hypertension. New York: Stratton Intercontinental Medical Book Corporation. 1975 : 5 5 3-8 7. 19 Rose G. Strategy of prevention : lessons from cardiovascular disease. Br Med I I981 : 282: 1847-51. 20 Dahl LK.Leitt G. Heine M. Influence of dietary potassium and sodium/potassium molar ratios on the development of salt hypertension ] Exp Mcd 1972: 136: 318-30. 21 Grobbee DE, Hofman A. Roelandt JT, Boomsma F. Schalekamp MA, Valkenburg HA. Sodium restriction and potassium supplementation in young people with mildly elevated blood pressure. J Hypertension 1987: 5: 13 5-9. 22 Cappuccio FP. Markandu ND. Beynon CW. Shore AC. Sampson B. MacGregor GA. Lack of effect of oral magnesium on high blood pressure: a double blind study. Br Med / 1985 : 291: 235-8. 23 Heaney PR, Saville PD. Recker RR. Calcium absorption as a function ofcalcium intake. ] Irrb Clin Med 1975: 8 8 : 881-95. 24 Lemann J, Adams ND. Gray RW. Urinary calcium excretion in human beings. N E n g l ] Men 1979; 301: 535-41. Received 31 October 1989, accepted 18 January 1990.
Correspondence: Dr ir. E. C. H. van Beresteijn, Netherlands Institute for Dairy Research (NIZO). P.O. Box 20. 6710 BA Me. The Netherlands.
