点击显示 收起
1 From the Calcium Research Unit, the Department of Applied Chemistry and Microbiology, University of Helsinki, and the Helsinki University Central Hospital.
2 Address reprint requests to CJE Lamberg-Allardt, Department of Applied Chemistry and Microbiology, Calcium Research Unit, University of Helsinki, PO Box 27, FIN-00014 Helsinki, Finland. E-mail: christel.lamberg-allardt{at}helsinki.fi.
ABSTRACT
Background: Calcium supplements are widely used to prevent osteoporosis. However, little is known about the metabolic effects of different dosages and of the timing of the dosages.
Objective: The aim was to study the effects of the timing of the dose (study 1), the effects of the size of the dose (study 2), and the effects of small repetitive doses (study 3) of calcium on calcium and bone metabolism in women.
Design: The investigation was conducted in 3 parts, each with 10 participants. In study 1, calcium loads (0 and 25 mg/kg body wt) were taken at 0900 and 2100. In study 2, calcium loads of 0, 250, and 1000 mg were taken at 0900. In study 3, calcium loads of 0 and 200 mg were taken 4 times/d. Markers of calcium and bone metabolism were followed.
Results: There was no significant difference in the response of serum parathyroid hormone (PTH) to the calcium load taken at 0900 and that at 2100. There was a significant dose-response effect of the calcium load on serum ionized calcium (P = 0.00005) and serum PTH (P = 0.0003). Small calcium doses (200 mg) taken 4 times/d kept the PTH secrection at a lower level than during the control day (P = 0.016). None of the doses caused significant changes in the markers of bone formation and resorption measured.
Conclusions: The calcium loads had no significant effect on the markers of bone formation and resorption measured, although even small calcium doses decreased serum PTH and increased serum ionized calcium concentrations rapidly. The effect was similar whether calcium was taken in the morning or in the evening.
Key Words: Parathyroid hormone • calcium intake • bone metabolism • timing of supplementation • women
INTRODUCTION
The discussion concerning dietary calcium has mainly focused on recommended intakes. However, because dietary calcium also has direct metabolic effects [eg, calcium decreases parathyroid hormone (PTH) secretion], it is important to determine the optimal time and method of dosing. For instance, PTH (1)—which is an important regulator of calcium metabolism—and markers of bone formation (2) and bone resorption (3) follow a diurnal rhythm. Therefore, the effects of calcium may depend on the time of the day that calcium is ingested. The size of the calcium load may also be important. The acute suppressive effect of calcium doses >400 mg (4, 5) on serum PTH secretion is well established and the duration of the effect was shown to be 8–10 h (6). Although a higher percentage of calcium is absorbed from small than from large calcium loads (7), it is not known whether the total amount absorbed from a calcium load comparable with the amount of calcium typically ingested in a Western diet, ie, 200–400 mg, has an effect on calcium and bone metabolism.
The aim of the present investigation was to determine, in a controlled setting, the acute effects of oral calcium loads on PTH secretion and on markers of bone resorption and formation. Special care was taken to ensure that the timing of the blood and urine samples and the composition and timing of the meals were identical during the control and experimental sessions. The emphasis of the investigation was to determine possible differences and benefits of different dosages of calcium and the timing of the dosages on calcium and bone metabolism. In addition, because dietary calcium is usually ingested in small doses many times during the day, we also studied the effect of repetitive small doses of calcium on calcium and bone metabolism. Because it is known that there are differences in the absorbabilities of different calcium salts, we chose a calcium supplement widely used in Europe that was shown to be as effective as calcium carbonate (8), a calcium salt commonly used in the United States.
SUBJECTS AND METHODS
Subjects
Thirty women aged 21–34 y participated in the investigation (Table 1). The subjects were healthy, had no disease known to affect bone or mineral metabolism, and were not taking any medications besides oral contraceptives. The study protocol was approved by the University of Helsinki Ethics Committee and each subject gave her informed consent.
View this table:
TABLE 1. Basic characteristics of the study subjects1
Study design
The subjects maintained 4-d food records, from which habitual calcium, phosphorus, and energy intakes were estimated. The investigation was conducted in 3 parts, each with 10 participants, to study the effect of the timing of the calcium dose (study 1), the effect of the size of the calcium dose (study 2), and the effect of small repetitive doses (study 3) on calcium and bone metabolism. The meals served to the participants were identical in every session within the study. No extra meals or snacks were allowed, but water was provided ad libitum. The subjects fasted overnight before coming to the research unit for all sessions in studies 2 and 3 and for the morning sessions in study 1. The subjects remained at the unit until the session for that day or night was completed.
The order of the sessions was randomized. Calcium was administered as a single dose of calcium (1 g Ca2+) salt (Calcium Sandoz: 6810 mg calcium lactate gluconate, 300 mg calcium carbonate, 1350 mg acid citrate anhydrous, and 40 mg aspartame; Laboratoires Sandoz, Sarl, France) in water with a small standardized meal. Because calcium (9) and phosphorus (10) were shown to affect serum PTH concentrations, the calcium and phosphorus intakes from foods were kept as low as possible. All of the meals combined were calculated to provide 375 mg Ca and 878 mg P daily in study 1, 100 mg Ca and 365 mg P daily in study 2, and 200 mg Ca and 803 mg P daily in study 3 (11).
The study design is depicted in Figure 1. All subjects participated in control and experimental sessions on separate days. In study 1, the subjects took part in 2 morning (beginning at 0900) and 2 night (beginning at 2100) sessions. Subjects were instructed to consume dairy products providing 500 mg Ca during the daytime but to fast 4 h before coming to the research unit for the night sessions. Calcium doses were 0 or 25 mg/kg body wt. The meals were similar at all sessions, except that only the first meal was served during the night sessions, which started at 2100. The night sessions were identical to the day sessions, with only 3 exceptions: the night sessions started at 2100, no meals were served after the small evening meal at 2100, and the subjects stayed in bed between 2300 and 0700. The subjects were awoken for the collection of blood and urine samples.
FIGURE 1. . Designs of studies 1, 2, and 3. Arrows indicate the time at which the meals and calcium loads were given and when blood and urine samples were collected. In study 1, the 2 arrows in parentheses indicate that these 2 meals were eaten during the day sessions but not during the night sessions.
In study 2, the subjects took part in 3 sessions: a control session (water), a session in which they received a 250-mg (6.25 mmol) Ca load with water, and a session in which they received a 1000-mg (25 mmol) Ca load with water. Other than the differences in the calcium load, the study days were identical (Figure 1). In study 3, the subjects took part in 2 sessions starting at 0900. In addition to the meals, the subjects were given either water or 800 mg Ca divided into four 200-mg doses.
Strontium absorption test
An oral strontium absorption test (12), which measures the calcium absorption capacity, was performed 1 wk after the last study session in all 10 subjects of study 2. The subjects ingested 2.5 mmol stable strontium chloride in water with a light standardized meal at 0900; a blood sample was taken 4 h later.
Laboratory methods
Serum was separated from blood and serum and urine samples were stored at -20°C until analyzed. All samples from the same person were analyzed in the same assay in a randomized order. Serum intact PTH concentrations were measured with an immunoradiometric assay (Allegro intact PTH assay kits; Nichols Institute, San Juan Capistrano, CA). The concentration of serum osteocalcin was measured with an Allegro osteocalcin kit (Nichols Institute), which recognizes the intact form of osteocalcin. Serum concentrations of the carboxy terminal propeptide of collagen type I procollagen (PICP), a marker of bone collagen formation, and cross-linked telopeptide of type I collagen (ICTP), a marker of bone resorption, were measured with immunoradiometric assay kits from Orion Diagnostica (Espoo, Finland). The bone-specific alkaline phosphatase activity (B-ALP) was measured with an enzyme immunoassay (Metra Biosystems, Palo Alto, CA). The concentration of urine deoxypyridinoline (DPD) was measured with an enzyme immunoassay that preferentially recognizes the free form of DPD (Metra Biosystems, Palo Alto, CA). DPD was corrected for creatinine excretion (DPD/Cr). In the immunoanalyses, the intraassay CVs were 2–8.5% and the interassay CVs were 3–8% in our laboratory. Serum total calcium, serum phosphate, urinary calcium, urinary phosphate, and urinary creatinine concentrations were measured by using routine laboratory methods, and the CVs were between 1.5% and 2.5%. Serum ionized calcium concentrations were measured from anaerobically handled serum samples with an ion selective analyzer (Microlyte, Kone Inc, Finland) within 90 min of the sample collection. Serum strontium concentrations were measured by atomic absorption spectrophotometry (model 1100B; Perkin-Elmer, Norwalk, CT) at 460.7 nm by using lanthanum chloride:hydrochloric acid diluent to remove interference with phosphate.
Statistical analysis
The data are expressed as means ± SEMs. Logarithmic transformation was used to normalize nonnormal distributions. Repeated-measures analysis of variance (ANOVA) was used to compare the results of the study sessions. If the sphericity assumption was rejected, the Hyunh-Feldt adjustment was used (13). If a significant difference was found with repeated-measures ANOVA, post hoc testing was conducted by using contrast analysis with Bonferroni correction. The comparison of single time points was made by using paired t tests with Bonferroni correction. Correlations were calculated by using Pearson's correlation analysis. Analyses were made with BMDP statistical software (SPSS Inc, Chicago) in a VAX/VMS minicomputer system (Digital Equipment Corp).
RESULTS
Study 1: effect of the timing of the calcium dose
Serum ionized calcium concentrations increased in response to the calcium load; however, the time of the day when the calcium load was given (0900 or 2100) did not significantly affect the response (Figure 2). The total postload urinary excretion of calcium did not differ significantly between day (1.98 ± 0.22 mmol) and night (2.06 ± 0.27 mmol) by repeated-measures ANOVA. Compared with the control session, the serum phosphate concentration increased in response to the oral calcium load and the response was not significantly different between day and night (Table 2). Urinary phosphate excretion did not decrease significantly after the calcium load. The timing of the dose (0900 or 2100) did not significantly affect the response curves of serum PTH after the calcium loads, although PTH decreased significantly in response to the calcium load (Figure 2). The calcium load had no significant effect on the markers of bone resorption (serum ICTP and urinary DPD/Cr) or bone formation (serum PICP and B-ALP activity) (Table 3). There was a significant variation over time in serum ICTP, urinary DPD/Cr, serum PICP, and B-ALP activity.
FIGURE 2. . Study 1: effect of the timing of the calcium load on changes (
View this table:
TABLE 2. Serum phosphate and urinary phosphate excretion in study 1 (effect of the timing of the calcium dose)1
View this table:
TABLE 3. Markers of bone resorption and bone formation at night in study 1 (effect of the timing of the calcium dose)1
Study 2: effect of the size of the dose
There was a significant dose-dependent effect of the 250- and 1000-mg Ca loads on serum ionized calcium: P = 0.013 and P = 0.0000 (contrast analysis with Bonferroni correction), respectively, compared with the control session (Figure 3). There was a significant difference in the response curves of serum ionized calcium (P = 0.0001; contrast analysis with Bonferroni correction) after the 2 calcium loads. Urinary calcium excretion increased significantly after both calcium loads (dose x time effect, P = 0.014; repeated-measures ANOVA) relative to values on the control day. Postload urinary calcium excretion was 1.47 ± 0.43 mmol on the control day and 1.79 ± 0.51 and 2.39 ± 0.84 mmol after the 250-mg and 1000-mg loads, respectively (250 compared with 1000 mg Ca: P = 0.005; paired t test with Bonferroni correction). There was a significant increase in serum phosphate after both calcium loads compared with the control session (Table 4). Compared with the control session, urinary phosphate excretion decreased after the calcium load.
FIGURE 3. . Study 2: effect of the size of the calcium dose on changes (
View this table:
TABLE 4. Serum phosphate and urinary phosphate excretion in study 2 (effect of the size of the calcium dose)1
The serum PTH concentration declined significantly compared with the control session after both calcium loads (Figure 3), more profoundly after the 1000-mg load (P = 0.0025; contrast analysis with Bonferroni correction) than after the 250-mg load (P = 0.00005; contrast analysis with Bonferroni correction). In addition, there was a significant difference between the PTH response curves after the 2 calcium loads (P = 0.0014; contrast analysis with Bonferroni correction). At 6 h postload, the serum PTH concentration was still significantly decreased compared with that during the control session after the 1000-mg load (P = 0.002) but not after the 250-mg load. The calcium load had no significant effect on serum ICTP, serum PICP, or the urinary excretion of DPD/Cr (data not shown). Serum B-ALP activity was not measured in study 2.
Study 3: effect of 4 repetitive calcium doses
Serum ionized calcium increased significantly compared with the control session after 4 repetitive doses of 200 mg Ca (Figure 4). The postload urinary excretion of calcium was higher than that during the control session (2.9 ± 0.4 compared with 1.9 ± 0.3 mmol/d, P = 0.005; repeated-measures ANOVA). Serum PTH decreased significantly compared with the control session beginning after the first of the 4 doses (35 ± 4 compared with 22 ± 3 ng/L 2 h postload; P = 0.009) and remained significantly different after the subsequent 3 doses. The postload urinary excretion of DPD/Cr and serum B-ALP activity did not differ significantly from that during the control session after the repetitive doses of calcium (data not shown).
FIGURE 4. . Study 3: effect of the repetition of calcium doses on changes (
Strontium absorption and PTH response to an oral calcium load
Strontium absorption (8.0 ± 1.3%) correlated negatively with the change in the serum PTH concentration after the 1000-mg Ca load (r = -0.83, P = 0.003; Pearson correlation analysis), indicating that the greater the calcium absorption capacity, the greater the decrease in PTH.
DISCUSSION
It is well known that an oral calcium load of 1 g decreases serum PTH concentrations and increases serum ionized calcium concentrations and urinary calcium excretion (9, 8, 14). Because the typical calcium load of a normal meal is <500 mg, we studied the effect of small calcium loads on calcium metabolism. The present study extends the previous findings by showing that there was a response in serum PTH concentrations even after a calcium load as small as 200 mg. In addition, serum PTH concentrations followed a diurnal rhythm after repetitive doses of calcium, but at lower concentrations than during the control session. Because it was shown that variations in serum PTH concentrations are important to bone (15), a repetitive dosage regimen could be especially beneficial to bone. On the other hand, a single high dose could be more effective because it causes a greater decrease in serum PTH.
The markers of calcium metabolism follow a diurnal rhythm (16, 1, 3), a finding that was confirmed in the present study. Because PTH concentrations peak at night (1), it is postulated that the timing of the calcium dose could be important in inhibiting bone resorption, which is greater at night than during the day (2). In the present study the diurnal variation was taken into account by including a control session in which subjects did not receive calcium supplementation. In addition, special care was taken to ensure that the sampling times during different study sessions were identical. The response curves of serum PTH were, however, similar whether the calcium load was taken at 0900 or at 2100. Bluhmsohn et al (6) stated that calcium taken in the evening reverses the nocturnal increase in serum PTH concentrations, whereas calcium taken in the morning does not. Serum PTH concentrations returned to baseline 8–10 h after a single oral dose of 1000 mg Ca in the present study and in other studies (6, 17). Thus, calcium taken in the morning no longer has an effect on serum PTH concentrations by evening. However, McKane et al (18) showed that the secretory capacity of the parathyroid gland decreases with long-term, high calcium intakes (2.4 g/d). In addition, we showed that serum PTH concentrations in the morning are lower with habitually high calcium intakes (1.3 g/d) than with low intakes, even within a normal calcium intake range (19).
The PTH response to an oral calcium load is used as an indicator of bioavailability of calcium from calcium supplements (5, 20). In the present study there was a dose-response effect on the changes in serum ionized calcium and serum PTH concentrations after the 250- and 1000-mg Ca loads. This extends the previous finding of a dose-response effect of 500- and 1000-mg Ca loads on changes in serum ionized calcium and serum PTH concentrations (21). Increases in serum ionized calcium were similar after higher loads of 1000 and 2000 mg Ca (4), which may have been due to the saturation of the active mechanism of calcium absorption (22).
The correlation between the time of maximal PTH suppression and the strontium absorption capacity may suggest that PTH suppression depends at least partially on the intestinal calcium absorption capacity of the individual. Although urinary calcium excretion does not equal the amount of calcium absorbed, it has been used as a marker of calcium absorption (23). In the present study, urinary calcium excretion was also dose dependent, confirming a previous finding (24). However, urinary calcium excretion as a percentage of the calcium load was smaller after the higher (1000 mg) calcium load, indicating that a smaller percentage of calcium was absorbed from the higher load. This finding confirms the results of Heaney et al (7).
In the present study, an acute oral calcium load had no significant effect on the markers of bone resorption (serum ICTP and urinary DPD/Cr) in the 24-h study period; furthermore, neither the dosage nor the timing of the dose had a significant effect. Our results confirm the results of a previous study in which a single calcium load (1000 mg) ingested in the evening had no significant effect on markers of bone resorption (serum ICTP and the urinary excretion of pyridinoline cross-links) during the subsequent 11 h (17). A recent study indicated that free DPD might not be as sensitive a marker as is total DPD (25). Some studies showed a decrease in DPD/Cr after an oral calcium load (14, 26, 27); however, one of these studies did not include a control session (27) and one did not use randomization (26). Horowitz et al (14) reported a decline in the urinary excretion of hydroxyproline, DPD, and pyridinoline relative to values during a control session after an oral calcium load of 1000 mg in men. This discrepancy may have been due to the marker used, ie, free DPD rather than total DPD (25). More likely, however, is that the design of the study conducted by Horowitz et al was not comparable with ours. In that study, the subjects were in a fasting state during both the control and the calcium-load sessions. Fasting per se could increase bone resorption because of, for example, the lack of some gastrointestinal hormones, such as vasoactive intestinal peptide. In fact, there is evidence that fasting increases bone resorption (28) and that vasoactive intestinal peptide, which increases postprandially, decreases osteoclastic resorption (29). Thus, the suppressive effect of calcium on bone resorption in the study by Horowitz et al may have resulted because bone resorption is higher than normal in a fasting state and may be inhibited by a meal. Nevertheless, it was shown previously that supplementation with 1000 mg Ca for 14 consecutive days decreases the urinary excretion of bone-resorption markers (6). This inhibiting effect of calcium on bone resorption was shown to occur as early as 3 d after supplementation began (30).
Osteoblasts have receptors for PTH (31) and PTH infusion decreases markers of bone formation within 24 h (32). Furthermore, we showed that an oral phosphate load acutely affects markers of bone formation (10). Moreover, there is evidence that extracellular calcium concentrations could affect osteoblasts per se (33). For these reasons we hypothesized that an oral calcium load could have an effect on markers of bone formation. However, the oral calcium load had no effect on any of the markers of bone formation (serum PICP, B-ALP, and osteocalcin) analyzed in this study. To our knowledge there are only a few studies in which markers of bone formation were studied after a calcium load. In one of these studies, calcium supplementation for 2 wk did not significantly affect markers of bone formation, osteocalcin, or B-ALP (34). However, Akesson et al (35) showed an increase in serum PICP after 7 d of calcium supplementation following calcium depletion for 22 d.
On the basis of the results of the present study, we conclude that there is no significant difference in the acute effects of calcium supplementation on serum PTH concentrations regardless of whether the doses are taken during the day or night. Thus, considering the acute response, there is no need to specifically recommend that calcium supplements be taken at night, at least not by women before menopause. However, there is a possibility that chronic supplementation could induce an adaptation process in favor of one of the regimens.
The dose clearly has an effect on PTH concentrations: the higher the calcium dose, the lower the PTH concentration. In addition, if a high dose of calcium is distributed in small doses over an entire day, PTH concentrations averaged over the entire day are low but do still fluctuate throughout the day. However, we do not know which is better for bone health: larger fluctuations in serum PTH concentrations after one large dose of calcium or low PTH concentrations during the day as a result of many small doses of calcium. Because we did not see any effect of the timing of the dose or the amount of the dose of calcium on the bone markers studied, we cannot make any definite conclusions about which regimen is best for bone health. This must be evaluated further in long-term studies.
REFERENCES