Relation Between Low Calcium Intake, Parathyroid Hormone, and Blood Pressure
Abstract—In a population health survey in 1995, serum parathyroid hormone (PTH) was measured in 1113 subjects, aged 30 to 79 years, and was found to be elevated (>6.9 pmol/L) in 118 subjects. In 1998, this group and 131 subjects with normal PTH levels were invited for reexamination, and 82 and 90 subjects from each respective group attended the follow-up. At the follow-up, 72 subjects had elevated and 100 had normal serum PTH levels. Those with elevated serum PTH levels (8 subjects with hyperparathyroidism were excluded) had significantly lower serum calcium levels and intake of calcium than those with normal PTH (2.24±0.09 and 2.29±0.10 mmol/L [mean±SD] and 400.3±227.3 and 592.1±459.6 mg/d, respectively; P<0.01). Serum levels or intake of vitamin D did not differ between the 2 groups. Subjects with elevated PTH in both 1995 and 1998 had significantly lower bone mineral content and bone mineral density in the lumbar spine than did those with persistently normal PTH levels (P<0.05). In the females, but not in the males, the systolic and diastolic blood pressures were significantly higher in those with elevated serum PTH (158.0±27.5 versus 141.5±19.2 mm Hg and 90.5±13.6 versus 82.6±8.6 mm Hg, respectively; P<0.01). This difference was even more pronounced when those with persistently elevated PTH were considered separately. In conclusion, reduced intake of calcium is frequently associated with high levels of serum PTH. This is associated with moderately reduced bone mineral content and bone mineral density in the lumbar spine. In women, high levels of serum PTH are also associated with markedly increased blood pressure.
The serum calcium level is tightly regulated by the parathyroid hormone (PTH). Thus, a small decrease in serum calcium elicits a prompt increase in the secretion of PTH, which in turn mobilizes calcium from the skeleton and increases the renal tubular reabsorption of calcium. Furthermore, PTH stimulates the hydroxylation of 25-hydroxyvitamin D to the more biologically potent 1,25-dihydroxyvitamin D in the kidneys,1 which again leads to an increased absorption of calcium from the intestine. This negative-feedback system keeps the serum calcium within fairly narrow limits.1 Therefore, an increased PTH level may be caused not only by primary hyperparathyroidism but also by factors tending to reduce the serum calcium level, like vitamin D deficiency.2 3
In Tromsø, northern Norway, there have been 4 large health surveys since 1974.4 In the last survey, which took place in 1994 to 1995, serum PTH was measured in a subgroup of 1113 subjects and was found to be elevated in 118. The purpose of the present study was to reexamine the subjects with an elevated serum PTH and to look for possible causes of the PTH elevation, in particular the intakes and serum levels of calcium and vitamin D.
Furthermore, because a high serum PTH level has been associated with hypertension5 6 7 8 and because calcium, vitamin D, and PTH participate in bone metabolism, we also compared blood pressure (BP), bone mineral content (BMC), and bone mineral density (BMD) between subjects with elevated and normal PTH levels.
Subjects and Study Protocol
The Tromsø study is a general health survey in the Tromsø area, northern Norway.4 In the fourth survey, which took place in 1994 and 1995, 27 180 men and women participated. All subjects who were examined between March 21 and May 19, 1995, and who were between the ages of 30 and 79 years (a total of 1113 subjects) had blood samples drawn for the measurement of PTH. One hundred eighteen of these subjects had PTH values >6.9 pmol/L. In 1998, these subjects and an age- and gender-matched group (drawn from the cohort that had a PTH measurement <7.0 pmol/L in 1995) were invited to participate in a follow-up study.
The follow-up study was performed at the Clinical Research Unit at the University Hospital of Tromsø. After a general health examination, a medical history was taken, and present medication was recorded. All subjects filled out a questionnaire on dietary habits that included questions on daily use of calcium and vitamin D tablets. By use of a Norwegian food table,9 the daily intakes of calcium and vitamin D were calculated.
Height, weight, and BP were measured as previously described.10 BMC and BMD were measured with a Lunar DPX-L Dual-Energy X-ray Absorptiometer (software version 1.3, Lunar Radiation Corp).
Blood samples were analyzed for serum calcium and creatinine with the use of a Hitachi 917 with reagents from Boehringer-Mannheim. PTH was measured by an Immulite intact PTH assay (Diagnostic Products Corp). 25-Hydroxyvitamin D3 was measured as previously described.11 The interassay coefficients of variation for these 4 assays were 2.0%, 2.5%, 6% to 8%, and 7% to 12%, respectively.
Comparisons between the groups with elevated and normal PTH levels regarding intake and serum levels of calcium and vitamin D, as well as body mass index (BMI), were performed with linear regression, with the parameter in question as a dependent variable, gender and PTH group (elevated/normal) as fixed factors, and age as a covariable. Comparisons regarding BP, BMC, and BMD were performed similarly, with BMI added as a covariable. The same analyses were repeated in males and females separately. In addition, a gender-specific multiple linear regression model was used to assess independent predictors of serum PTH concentration, systolic and diastolic BP, BMC, and BMD. Age, BMI, serum calcium and 25-hydroxyvitamin D3, and intake of calcium and vitamin D were included as independent variables in the model. The regression coefficient is given as the standardized β coefficient (the coefficient of the independent variables when all variables are expressed in standardized [z score] form). Comparison of BP between groups according to combinations of high or low serum PTH levels in 1995 and 1998 was performed by ANOVA with least significant difference as a post hoc test. All tests were 2-sided, and a value of P<0.05 was considered statistically significant. Unless otherwise stated, the data are presented as mean±SD. Statistical analyses were performed with the SPSS version 9.0 (SPSS Inc).
The study was approved by the regional ethics committee, and all subjects gave written informed consent to participate.
Of the initial 118 subjects with serum PTH levels >6.9 pmol/L in 1995, 1 had died, and 2 had moved from the Tromsø area. The hospital records of the remaining subjects were reviewed, and 10 were considered too sick or otherwise unfit to participate. Three subjects were aged >80 years and were not invited to participate for that reason. One person had undergone an operation for hyperparathyroidism, thus leaving 101 subjects that were invited to the follow-up study. Of these, 10 did not respond to the invitation, and 9 were unwilling to participate. Of the 82 subjects finally examined, 26 (31.7%) subjects had normal serum PTH levels at follow-up, and 56 (68.3%) subjects had elevated PTH levels at follow-up.
From the cohort with serum PTH <7.0 pmol/L in 1995, 131 gender- and age-matched subjects were invited to a follow-up examination. Ninety subjects were willing to participate and completed the examinations. Of these, 74 (82.2%) subjects had normal serum PTH levels at follow-up, and 16 (17.8%) subjects had serum PTH levels >6.9 pmol/L at follow-up.
Thus, at follow up, PTH levels were elevated in a total of 72 subjects and normal in 100 subjects.
Possible Causes of Elevated PTH Levels at Follow-Up
Among the 72 subjects with elevated serum PTH levels at follow-up, 1 male and 7 females were considered to have hyperparathyroidism (serum PTH >6.9 pmol/L together with serum calcium >2.55 mmol/L). Their mean±SD age was 70.0±7.8 years; systolic and diastolic BP, 156.0±9.5 and 90.5±11.9 mm Hg, respectively; and BMC at the lumbar spine, femoral neck, Ward’s triangle, and trochanter, 57.5±14.6, 4.1±0.7, 2.0±0.6, and 10.2±3.3 g, respectively. These 8 subjects are not included in the following analyses.
Mean serum calcium was significantly (P<0.01) lower in the group with high serum PTH levels, whereas mean serum 25-hydroxyvitamin D3 did not differ significantly between the 2 groups (Table 1⇓). The intake of calcium was significantly (P<0.01) lower in those with elevated PTH levels (Tables 1⇓ and 2⇓). In 8 (12.5%) of those with elevated serum PTH, the intake was <200 mg/d, and in 36 (56.3%) of those with elevated serum PTH, the intake was <400 mg/d. The corresponding numbers in those with normal PTH levels were 9 (9.0%) and 39 (39.0%), respectively.
The 2 groups did not differ significantly regarding vitamin D intake (Table 1⇑). None had a serum creatinine >150 μmol/L, and none had gastric or bowel resection, celiac disease, or other known causes of malabsorption. None were using anticonvulsant drug therapy or other medication known to affect the PTH levels.
In the multiple linear regression model, the intake of calcium was significantly (P<0.01) associated with serum PTH in the females but not in the males (Table 3⇓).
Both systolic and diastolic BPs were significantly (P<0.05) higher in the group with elevated PTH levels (Table 1⇑). However, a significant difference in BP was seen in females only (P<0.01, Table 2⇑). In the females, slightly more subjects with high serum PTH levels were on BP medication than those with normal serum PTH (29.0% and 24.1%, respectively). After exclusion of those on BP medication, the difference in BP was still significant (P<0.01, Table 2⇑). The difference in BP was even more pronounced when those who had elevated serum PTH levels in both 1995 and 1998 were considered separately from those who had normal serum PTH levels on both occasions (Table 4⇓).
In the multiple linear regression model, only serum PTH reached statistical significance (P<0.001) as an individual predictor for systolic and diastolic BP. However, this was seen in females only (Table 3⇑).
Because BMI was higher in the group with elevated serum PTH, the difference in BP between females with high and normal serum PTH levels at follow-up is shown stratified for BMI quartiles (BMI ≤23.0, 23.1 to 25.7, 25.8 to 28.6, and >28.6 kg/m2) in Figure 1⇓.
BMC and BMD
BMC and BMD were similar in the 2 serum PTH groups (Table 1⇑). In the multiple linear regression model, BMI was a significant (P<0.001) predictor of BMC and BMD in the females but not in the males. For the females, this was seen at all 4 places of measurement (standardized β coefficients from 0.44 to 0.59). None of the other variables reached statistical significance in this respect. However, when those with elevated serum PTH levels on both occasions (and thus those with presumably the most profound calcium intake deficit) were considered separately from those with persistently normal serum PTH levels, the BMC at the lumbar spine was significantly reduced (64.5±17.2 versus 67.4±18.4 g, P<0.02), as was the BMD at the lumbar spine and trochanter (1.065±0.190 versus 1.099±0.203 g/cm2 and 0.816±0.149 versus 0.837±0.159 g/cm2, respectively; P<0.05).
BMC at the lumbar spine and the trochanter for these 2 PTH subgroups is shown stratified for BMI quartiles in Figure 2⇓.
In the present study, 68% of those with an elevated serum PTH level in 1995 had an increased PTH level when reexamined 3 years later. Because the blood samples were not drawn with subjects in the fasting state, the PTH secretion may have been affected by the mineral content in food recently ingested. This may partly explain why an initially elevated PTH level was found normalized in some of the subjects.
As could be expected, some of those with elevated serum PTH levels at the follow-up were found to have hyperparathyroidism. However, the lack of other obvious causes for the increased serum PTH levels, like diseases associated with malabsorption, was remarkable.
From studies on selected subjects, vitamin D deficiency and a significant inverse relation between serum 25-hydroxyvitamin D and PTH have frequently been found in the elderly.2 3 However, in the present study, there was no significant difference in serum level or intake of vitamin D between those with high or normal serum PTH levels. Therefore, lack of vitamin D can hardly be the main explanation for the increased serum PTH levels in the present study, even though the average vitamin D intake was slightly below the recommendation of 10 μg (400 IU) per day for those aged 51 to 70 years and 15 μg (600 IU) per day for those aged ≥71 years.12
On the other hand, the serum calcium level and the intake of calcium were remarkably low in the group with high serum PTH. Admittedly, the food questionnaire used by us cannot measure calcium intake as accurately as a dietary interview. However, the questionnaire did include all dairy products, which in a Norwegian diet are the main calcium sources. Furthermore, the calcium intake was much lower than that found in the control group and also much lower than that recommended.13 This difference in calcium intake is the most likely explanation for the low serum calcium levels and the compensatory increased serum PTH. The multiple linear regression analysis supports this view, in view of the fact that there was a strong negative association between intake of calcium and serum PTH.
However, it must be emphasized that in spite of the difference in calcium intake, there was a considerable overlap between the 2 groups. Thus, in addition to the low calcium intake, additional individual factors must also be present, such as a slightly reduced ability to absorb calcium or an increased calcium excretion, possibly because of reduced vitamin D sensitivity.
One would expect to see the most apparent effects of a low calcium intake on the skeleton. Thus, a markedly reduced calcium intake could lead to a negative calcium balance,14 loss of calcium from the skeleton, and the development of osteoporosis. This was also found in the present study but only when those with persistently elevated PTH levels were compared with those with persistently normal serum PTH levels. Those with persistently elevated serum PTH were those with the lowest calcium intake and, presumably, those with the most profound calcium deficit. Because BMC and BMD increase with increasing BMI and because BMI was considerable higher in those with elevated PTH levels, the difference between the 2 groups was more evident when stratified for BMI.
However, the most interesting observation in the present study was the large difference in BP between females with high and normal serum PTH levels. The difference was ≈20 mm Hg for systolic BP and 10 mm Hg for diastolic BP and was even larger after excluding those on medication for hypertension. In spite of this large difference in BP, it was somewhat surprising that only slightly more subjects with high serum PTH levels were on BP medication compared with those with normal serum PTH levels. As shown in the stratified analysis, the difference in BP could not be ascribed to differences in BMI values, which were higher in the group with elevated serum PTH. Furthermore, in the multiple linear regression analysis, serum PTH in women was strongly associated with both systolic and diastolic BPs. Finally, the most elevated BPs were seen in those with persistently elevated serum PTH, a subgroup that also had the highest serum PTH levels as well as the lowest calcium intake.
Although an association between BP and serum levels of PTH has been described before,5 6 8 those studies were not population-based and did not find the gender difference that we observed. On the contrary, in the study by Young et al,7 males but not females with hypertension had elevated PTH levels. This is hard to reconcile with the present findings and is most likely the result of different study populations.
For a long time, there has been a discussion of whether a reduced intake of calcium is associated with hypertension.15 16 17 This has been demonstrated in several18 19 20 but not all epidemiological studies,21 and the effect in intervention studies has been marginal.22 In this respect, our group with persistently elevated serum PTH is unique, because it represents subjects who most likely have been on a very low- calcium diet for a number of years. Although the group obviously is a selected one, it still represents a significant proportion of the hypertensive population. Thus, of the original 1113 subjects examined in 1995, 15.9% had a systolic BP >160 mm Hg, and of these, 16.9% had a PTH level >6.9 pmol/L.
In addition to throwing light on pathophysiological relations that otherwise would not be evident, this group would also be ideal for testing whether an adequate calcium intake is of importance for maintaining a normal BP. If that is the case, the effect of calcium supplementation in this group should be profound.
We cannot say whether the effect on the BP was caused by the high PTH level or by the reduced calcium intake. This is also difficult to determine because the 2 events are physiologically linked together, with minor changes in serum calcium eliciting large changes in serum PTH.23 Furthermore, an association is, of course, no proof of a causal relation, and we cannot rule out that the high BP in the females and the low BMC and BMD in those with persistently elevated serum PTH levels were caused by factors not measured by us. One possible candidate in this respect could be parathyroid hypertensive factor.24 Parathyroid hypertensive factor is suppressed by high dietary calcium intake, and parathyroid hypertensive factor–like activity levels have been found to be elevated in hypertensive individuals.24 25
In spite of the above uncertainty, we find it fair to conclude that if an elevated serum PTH level is found in an otherwise healthy female and the serum calcium is not above or in the upper normal range, one likely explanation could be a reduced calcium intake. A controlled clinical trial to see whether calcium supplementation will reduce the BP in a group of patients with elevated PTH levels is clearly indicated.
This study was supported by a grant from the Norwegian Research Council and the University Hospital of Tromsø. The assistance by the staff, and Annika Gustafsson in particular, at the Clinical Research Unit at the University Hospital of Tromsø is gratefully acknowledged.
- Received September 7, 1999.
- Revision received September 28, 1999.
- Accepted December 21, 1999.
Resnick LM, Müller FB, Laragh JH. Calcium-regulating hormones in essential hypertension: relation to plasma renin activity and sodium metabolism. Ann Intern Med. 1986;105:649–654.
Brickman AS, Nyby MD, von Hungen K, Eggena P, Tuck ML. Calcitropic hormones, platelet calcium, and blood pressure in essential hypertension. Hypertension. 1990;16:515–522.
Statens ernæringsråd. Den Store Matvaretabellen. Oslo, Norway: Universitetsforlaget, Oslo; 1995.
Bønaa KH, Arnesen E. Association between heart rate and atherogenic blood lipid fractions in a population: the Tromsø study. Circulation. 1992;86:394–405.
Standing committee on the Scientific Evaluation of Dietary Reference Intakes, Institute of Medicine. Dietary reference intakes: calcium, phosphorus, magnesium, vitamin D, and fluoride. Washington, DC: National Academy Press; 1997.
Heaney RP, Recker RR, Saville PD. Calcium balance and calcium requirements in middle-aged women. Am J Clin Nutr. 1977;30:1603–1611.
Cappuccio FP, Elliot P, Allender PS, Pryer J, Follman DA, Cutler JA. Epidemiological association between dietary calcium intake and blood pressure: a meta-analysis of published data. Am J Epidemiol. 1995;142:933–945.
Joffres MR, Reed DM, Yano K. Relationship of magnesium intake and other dietary factors to blood pressure: the Honolulu heart study. Am J Clin Nutr. 1987;45:469–475.
Witteman JCM, Willet WC, Stampfer MJ, Colditz GA, Sacks FM, Speizer FE, Rosner B, Hennekens CH. A prospective study of nutritional factors and hypertension among US women. Circulation. 1989;80:1320–1327.