| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles |
Medical Research Council Dunn Nutrition Unit, Cambridge, United Kingdom and Keneba, The Gambia; and Institute of Food Research, Norwich, United Kingdom
Address all correspondence and requests for reprints to: Dr. Ann Prentice, Micronutrient Research, Medical Research Council Dunn Nutrition Unit, Downham’s Lane, Milton Road, Cambridge, United Kingdom CB4 1XJ. E-mail: ann.prentice{at}mrc-dunn.cam.ac.uk
Abstract
The effect of 18 months of lactation on indexes of calcium and bone metabolism was studied in 60 Gambian women accustomed to a very low calcium intake. Half the women consumed a calcium supplement from 10 days postpartum for 52 weeks (supplement, 714 mg Ca/day; total Ca intake, 992 ± 114 mg/day), and half consumed placebo (total Ca intake, 288 ± 128 mg/day). Fasting blood and 24-h urine samples were collected at 1.5, 13, 52, and 78 weeks of lactation and analyzed for calciotropic hormones (intact PTH, 1,25-dihydroxyvitamin D, and calcitonin), bone turnover markers (osteocalcin, bone alkaline phosphatase, and urinary deoxypyridinoline), and plasma minerals (calcium and phosphate). The first months of lactation were associated with increased bone turnover and plasma phosphate, and decreased PTH and 1,25-dihydroxyvitamin D. These effects diminished by 52 weeks, although breast milk volumes remained high. The Gambians had higher PTH, 1,25-dihydroxyvitamin D, and bone formation than British women with a greater customary calcium intake. None of the biochemical indexes was affected by calcium supplementation, with the possible exception of bone alkaline phosphatase (-29% at 52 weeks; P = 0.015). These data demonstrate that lactation-associated changes in calcium and bone metabolism are physiological and are independent of dietary calcium supply in women with very low calcium intakes.
THE CALCIUM required for breast milk production can represent a substantial proportion of daily calcium intake for many lactating women (1). Lactation-associated bone loss and changes in biochemical markers of calcium and bone metabolism have been reported, which may suggest that some lactating women consume insufficient dietary calcium for their needs. A number of recent studies have found no association between maternal calcium intake and changes in either bone mineral status or biochemical indicators of calcium metabolism (2, 3, 4, 5, 6, 7, 8, 9). However, these investigations were conducted in women with a moderate to high calcium intake, often above recommended levels, and there have been no studies of women with very low calcium intakes. In addition, there is no information for women who breastfeed for periods longer than 6–9 months.
Gambian women, like many others in rural areas of Africa, have high requirements for calcium over many years as a result of sequential periods of pregnancy and extended lactation. It is customary for these women to breastfeed each infant for approximately 2 yr, and breast milk production remains high throughout this period (10). However, calcium intakes of women in rural areas of The Gambia are very low, averaging around 300 mg/day, and are not increased during lactation (11, 12). To examine the potential benefits of an increase in calcium intake for lactating Gambian mothers, we recently conducted a randomized, double blind, placebo-controlled supplementation study (12). Women were supplemented from 10 days postpartum for the first year of lactation with sufficient calcium to more than triple their intake and to bring it close to recommended levels. We have previously reported that breast milk calcium secretion and changes in forearm bone mineral status during lactation were not affected by the increase in calcium intake, but, as expected, the supplement increased urinary calcium and decreased urinary phosphate excretion (12). The Gambian women had a high calcium absorption efficiency (13), and there was evidence of significant renal conservation of calcium, particularly at 3 and 12 months of lactation (12). We present here the effects of 18 months of lactation on calciotropic hormones, bone turnover markers, and other indexes of calcium and bone metabolism and the biochemical response to the calcium supplement.
Subjects and Methods
Subjects and study design
Sixty women residing in the rural Gambian villages of Keneba and Manduar, West Kiang, took part in the study. They were 16–41 yr old, and parity was 1–13. The group represented all healthy women who gave birth in the 2 villages during 1 calendar yr. Women with twins were excluded. The study was approved by the Medical Research Council Gambia ethics committee, and informed consent was obtained from all participants.
The women visited the Dunn Nutrition Unit’s Keneba Clinic for baseline measurements 1.5 weeks after delivery, immediately after the end of their traditional confinement period (mean ± SD, 9 ± 1 days). Each subject was assigned, double blind, to receive either a calcium supplement (chewable calcium carbonate; Calcichew, Shire Pharmaceuticals, Hampshire, UK) or a placebo (dextrose; Dextro-energy, CPC-UK, Surrey, UK) for the following 12 months. Randomization was by permuted block of four to minimize potential bias due to seasonal effects. The calcium supplement provided 1000 mg calcium and was consumed on 5 days/week. Overall compliance was 100%, resulting in an increase in calcium intake of 714 mg Ca/day (17.9 mmol/day) in the calcium-supplemented group during the supplementation period. Further details of the randomization procedure and study design have been reported previously (12).
Data and sample collection
The effects of the calcium supplement on biochemical indexes of calcium and bone metabolism were assessed in blood and urine samples collected before the start of supplementation (1.5 weeks postpartum), during supplementation at 13 and 52 weeks (92 ± 6 and 366 ± 5 days postpartum), and 6 months after withdrawing the supplement (78 weeks, 547 ± 5 days postpartum). Retrospective information indicated that three women were in the early stages of pregnancy at 78 weeks. The results for these mothers at this time point were excluded from data analysis.
Blood was collected by venipuncture between 0800–1000 h after an overnight fast. Samples were collected into precooled syringes, dispensed immediately into cooled tubes containing either lithium heparin or potassium ethylenediamine tetraacetate as anticoagulant, mixed thoroughly, and spun in a refrigerated centrifuge within 45 min. The plasma was immediately separated and frozen at -20 C. Samples were transported to the Cambridge laboratory on dry ice and stored at -20 C before analysis.
Urine was collected in the subject’s home over the 24-h period following the clinic visit. The women were provided with a urine collection kit consisting of acid-washed urine bottles, a funnel, and an insulated box containing cooled freezer packs. A fieldworker visited the subject at the start and end of the collection and at various times during the day to remove urine bottles to the clinic refrigerator and to bring fresh freezer packs. At the end of the collection, urine fractions were pooled and mixed, and total volume was recorded. Titratable acidity and urinary pH were measured, and an acidified aliquot was stored for the measurement of calcium, phosphate, and creatinine (12). A separate, unacidified aliquot was taken for the measurement of free deoxypyridinoline. This was stored at -20 C and transported to the Cambridge laboratory on dry ice. A creatinine clearance rate of less than 60 mL/min·1.73 m2 was taken as an indication of incomplete urine collection (12). Creatinine clearances below this threshold were measured for 0, 2, 4, and 12 women at 1.5, 13, 52, and 78 weeks, respectively. Urinary data from these collections were excluded from further consideration.
Other measurements were performed during the study, including single photon absorptiometry, anthropometry, and estimates of dietary intake, calcium absorption efficiency, breast milk volume, breast milk mineral content, and vitamin D, iron, zinc, and magnesium status. These have been described in detail previously, as have the results of the urine analysis other than those for deoxypyridinoline (12, 13, 14, 15, 16).
Subject characteristics
All women in this region breastfeed their infants on demand for approximately 2 yr, introducing their infants to complementary weaning foods from about 3–4 months of age. The mothers in this study were no exception; at 1.5 weeks all were exclusively breastfeeding except for occasional drinks of water, at 13 weeks 25% had introduced the infant to small quantities of other foods, and at 52 weeks all were providing a mixed diet in addition to breast milk. All women were still breastfeeding at 78 weeks, except 1 who had recently become pregnant.
Full details about the supplementation program, the methodologies used,
and the characteristics of the participants have been reported
previously (12, 13, 14, 16). Published results relevant to the presentation
of the biochemical data include the following. There were no
differences in baseline characteristics between women in the
calcium-supplemented and placebo groups (Table 1
). Calcium intake from foods was similar
in the two groups, resulting in total intakes that were 3 times higher
in the calcium-supplemented group during the intervention phase (Table 1). At 13 weeks, there was evidence of lactation-associated decreases
in forearm bone mineral content that was not altered by calcium
supplementation (12). In addition, calcium supplementation did not
affect breast milk volume or breast milk calcium concentration, but
increased urinary calcium output by 6.6% of the ingested dose and
decreased phosphate excretion (Table 1) (12). In the lactating Gambian
women as a whole, breast milk volumes averaged 0.87 ± 0.20 L/day
at 13 weeks and 0.76 ± 0.21 L/day at 52 weeks (14). There was no
evidence of vitamin D deficiency [25-hydroxyvitamin D concentration at
13 weeks, 25.9 ± 6.4 µg/L (64.6 ± 16.0 nmol/L)] (15) or
of recent infection (all had plasma 
1-antichymotrypsin
concentration at 13 weeks <0.64 g/L) (16), but many subjects had
biochemical indications of poor iron and zinc status (16).
|
Blood and urine were collected from 11 premenopausal Gambian women living in the same villages who were neither pregnant nor lactating (NPNL). Samples were collected on 1 occasion from each woman using materials and methods identical to those used for the lactating women. These women were aged 26–38 yr (mean, 31 ± 4 yr), parity was 0–8 (mean, 4 ± 3), and they had a mean plasma 25-hydroxyvitamin D concentration of 32.4 ± 9.1 µg/L (80.9 ± 22.7 nmol/L). The reasons for these women being neither pregnant nor lactating at the time of the study were not investigated, but may have included reduced fertility in either the woman or her husband, or the husband’s absence from the village for study or employment. Only 1 of the NPNL women had never conceived. It was not possible to obtain NPNL data from the subjects involved in the lactation study because most became pregnant before they stopped breastfeeding.
Samples were also collected from 15 British women at 13 weeks of lactation (88 ± 8 days) living in Norwich, Norfolk (12), and from 10 NPNL premenopausal British women living in Cambridge (17). Both sets of women were recruited by advertisement from the general population to provide samples for assay alongside the Gambian samples. The lactating women were 22–41 yr old (mean, 29 ± 5 years), parity was 1–4 (mean, 2 ± 1), weight was 64.6 ± 11.0 kg, and height was 1.64 ± 0.59 m. Calcium intake was 1168 ± 32 mg/day (29.2 ± 7.9 mmol/day), the breast milk calcium concentration was 261 ± 38 mg/L (6.53 ± 0.95 mmol/L), and the breast milk volume was 0.78 ± 0.24 L/day. Of these women, 60% were exclusively breastfeeding; the others had introduced small quantities of other foods. The samples from the lactating mothers were collected in the summer months (July to August), and their mean plasma 25-hydroxyvitamin D was 27.1 ± 7.7 µg/L (67.6 ± 19.2 nmol/L). The NPNL British women were 21–46 yr of age (mean, 29 ± 9 yr), parity was 0–4 (0.7 ± 1.4), weight was 64.6 ± 11.4 kg, height was 1.65 ± 6.1 m, and calcium intake was 968 ± 244 mg/day (24.2 ± 6.1 mmol/day). Samples from these women were collected throughout the year, and their mean plasma 25-hydroxyvitamin D was 19.4 ± 10.5 µg/L (48.5 ± 26.3 nmol/L) (17). Seven subjects in the British lactating group and 4 British NPNL women were practicing chemical contraception (most used an oral contraceptive, 1 lactating subject had received a postpartum injection). Contraceptive use had no measurable effect on any analyte in this study, and the data have been combined.
Biochemical assays
Plasma osteocalcin, calcitonin, and intact PTH were measured by RIA (Incstar, Stillwater, MN). Plasma 1,25-dihydroxyvitamin D was analyzed using RIA after immunoextraction (IDS, Boldon, Tyne and Wear, UK). Calcium, inorganic phosphate, creatinine, and albumin were measured by colorimetry on a centrifugal analyzer as follows: calcium, methylthymol blue reaction in the presence of 8-hydroxyquinoline; phosphate, direct phosphomolybdate reaction without deproteinization; creatinine, kinetic buffered Jaffé reaction without deproteinization; and albumin, specific immunonephelometry (Roche Products, Welwyn Garden City, UK; Atlantic Antibodies, Incstar). Plasma calcium was normalized to an albumin concentration of 36 g/L by adding or subtracting 4 mg/L (0.1 mmol/L) calcium for every 4 g/L that the albumin concentration was above or below 36 g/L. Plasma bone-specific alkaline phosphatase activity was measured on the centrifugal analyzer using a lectin precipitation procedure that also quantified total alkaline phosphatase activity (Boehringer Mannheim, Mannheim, Germany: optimized Deutsche Gesellschaft für Klinische Chemie (DGKC) method, 37 C). Urinary free deoxypyridoline was analyzed using competitive enzyme immunoassay (Metra Biosystems, Wheatley, Oxfordshire, UK).
Plasma obtained from blood anticoagulated with lithium heparin was used for the majority of analytes, PTH was measured in plasma prepared with potassium ethylenediamine tetraacetate. PTH, osteocalcin, calcitonin, bone alkaline phosphatase, and minerals were assayed in plasma at the first thawing, which was carefully controlled to maintain the plasma at a low temperature. The accuracy and precision of the biochemical assays was assured using in-house precision controls and the following materials: Lyphochek Normal Urine Control and Immunoassay Control (Bio-Rad, Anaheim, CA) and Roche Serum Control N (Roche Products). For financial reasons, 1,25-dihydroxyvitamin D and urinary deoxypyridinoline were measured in only a subset of lactating Gambian women, selected at random across the year of recruitment to ensure similar numbers in the calcium-supplemented and placebo groups.
Bone alkaline phosphatase activity at 1.5 weeks was recorded as zero (±10 U/L) in 20% of samples and was low in many more, although total alkaline phosphatase activity was elevated at this time point. This suggested interference with the lectin precipitation assay by factors present in the plasma at this time, possibly residual placental alkaline phosphatase. Results for this analyte at 1.5 weeks have been excluded.
Data analysis
Statistical analysis was performed using ANOVA, regression, and linear model software on DataDesk 4.1 (Data Descriptions, Ithaca, NY, 1993). The Data Desk ANOVA is suitable for use with unbalanced longitudinal datasets; missing data points for an individual are not interpolated, whereas other data are retained. Bone alkaline phosphatase and total alkaline phosphatase were converted to natural logarithms to correct a marked positive skew in the distribution; the other analytes had a distribution that approximated normality.
The effect of the supplement on each analyte was examined in two ways. Firstly, conditional regression was performed at each stage of lactation separately, with the value for the analyte at 1.5 weeks and supplementation group (calcium/placebo coded as 1/0) as independent variables. This effectively examines the effect of supplement on the change in analyte from baseline after adjusting for the initial value, thereby avoiding regression toward the mean. Secondly, ANOVA (repeated measures, nested form) was performed, with subject, stage of lactation (1.5, 13, 52, and 78 weeks), supplementation group (calcium/placebo), and a stage x group interaction term as discrete, independent variables. This method was also used, after removing the terms involving the supplementation group, to explore time trends and the effects of season within individuals. Four seasons were defined: January to March (postharvest, cold season), April to June (postharvest, hot season), July to September (early rainy season), and October to December (late rainy season). Comparisons between Gambian and British women were conducted by ANOVA; the Mann-Whitney nonparametric test gave similar results. The conservative Scheffe’s post-hoc test was used to examine the significance of differences between pairs of time points, between supplementation groups, between seasons, and between countries to minimize problems associated with multiple comparisons. Percent differences were calculated as 100 x (difference/mean) (18). For those analytes expressed in natural logarithms, the difference between loge values multiplied by 100 equals the percent difference (18).
Results
Response to calcium supplementation
The indexes of calcium and bone metabolism were similar in the two
groups of Gambian women at 1.5 weeks, before the start of
supplementation (Figs. 1-3). Calcium
supplementation had no significant effect on calciotropic hormone
concentrations (Fig. 1), on bone turnover as measured by plasma
osteocalcin and urinary deoxypyridinoline (Fig. 2), or on plasma calcium and phosphate
concentrations (Fig. 3) at any stage of
lactation. The supplemented group had lower plasma bone alkaline
phosphatase activity than the placebo group at 52 weeks (-29%;
P = 0.015), but not at any other time. This difference
was not reflected in total alkaline phosphatase activity [mean ±
SE at 52 weeks (loge): supplemented, 5.10
± 0.06 U/L; placebo, 5.21 ± 0.18 U/L; P = 0.20]
or in urinary deoxypyridinoline excretion (Fig. 2). The mean
osteocalcin concentration was slightly lower in the
calcium-supplemented group at 52 weeks (Fig. 2), but this was not
significant either before or after correcting for the concentration at
1.5 weeks (P = 0.29 and 0.77, respectively). Similarly,
although there was a trend toward a lower PTH concentration in the
calcium-supplemented group (Fig. 1), this was present before the start
of supplementation, and there was no evidence of a difference between
the groups at any stage of lactation after correcting for initial value
(P > 0.70). There was no indication of any difference
in response to the calcium supplement among teenage mothers (n =
10), those under 25 yr of age (n = 22), or those who were nursing
their first or second baby (n = 14).
|
|
|
The data for the Gambian women at each stage of lactation are
given in Table 2 for supplement and
placebo groups combined. Data from the small group of NPNL women are
provided for comparison. Table 3 details
the statistical analysis investigating the effects of stage of
lactation on analyte levels within individuals. Three distinct patterns
across time were apparent. 1) Plasma PTH, 1,25-dihydroxyvitamin D, and
albumin concentrations were lower in early lactation and increased
during lactation. Bone alkaline phosphatase activity also increased
from 13–52 weeks. Although this appeared to be more pronounced for
women in the placebo group (Fig. 2
), there was no significant
interaction between stage of lactation and supplementation group
(P = 0.15). 2) Urinary deoxypyridinoline excretion and
plasma levels of calcitonin, calcium corrected for albumin, and total
alkaline phosphatase activity were higher at 1.5 weeks than at later
stages of lactation. 3) Plasma osteocalcin, calcium before albumin
correction, and phosphate concentrations showed a marked increase from
1.5–13 weeks, with a decrease in later lactation. Osteocalcin levels
remained higher at 52 and 78 weeks than at 1.5 weeks, whereas plasma
calcium and phosphate values were close to those in early lactation.
The lactating women had higher plasma phosphate concentrations than the
NPNL women at all stages of lactation (P < 0.001).
|
|
No significant relationships were observed between any of the calciotropic hormones or bone turnover markers and the urinary excretion of calcium or phosphorus at any stage of lactation. In addition, the change in urinary excretion from 1.5–13 weeks of lactation was not associated with the change in or mean value of any of these analytes over the same period regardless of whether the calcium-supplemented and placebo groups were considered separately or combined. A positive relationship was observed between urinary phosphate excretion and plasma phosphate concentration at 1.5 weeks (P = 0.012), but this was not evident in later lactation in either group. Urinary calcium output was inversely correlated with plasma calcium at 52 weeks in the calcium-supplemented group (P = 0.001), but not in the placebo group and not at any other stage of lactation.
Neither breast milk calcium nor phosphorus at 13 weeks, expressed as concentration or daily output, was related to any of the analytes or to their change from 1.5 weeks. Similarly, there was no association between the decrease in bone mineral status observed from 1.5–13 weeks of lactation and the change in or mean concentration of any of the biochemical indexes.
Effects of age and season
Osteocalcin, bone alkaline phosphatase, and total alkaline
phosphatase were negatively correlated with age at all stages of
lactation (e.g. at 13 weeks, P 
0.001),
indicating greater bone formation in the younger women. Replacing age
with parity produced similar results. The increase in osteocalcin from
1.5–13 weeks was also inversely related to age (after adjusting for
the initial value, P = 0.008), demonstrating that the
response to lactation was greater in the younger women. Age or parity
had no effect on the changes in bone alkaline phosphatase or total
alkaline phosphatase activities during lactation. No other analyte was
significantly correlated with age or parity.
No seasonal variation was discernible in any of the calciotropic
hormones or bone turnover markers. Significant seasonal effects in
plasma calcium and albumin concentrations were observed within
individuals after adjusting for stage of lactation (P
0.001). The highest levels of these analytes were observed during
the period from January to June (the two postharvest quarters), and the
lowest was seen from July to December (rainy season). The magnitude of
the seasonal difference averaged 11% for both analytes.
Differences between Gambian and British women
Gambian women at 13 weeks of lactation had significantly greater
concentrations of PTH, 1,25-dihydroxyvitamin D, osteocalcin, and total
alkaline phosphatase activity than the group of British women at the
same stage of lactation (Table 4), with a
trend in the same direction for bone alkaline phosphatase. Differences
of similar magnitude were seen for these analytes between Gambian and
British NPNL women, although these were only statistically significant
for 1,25-dihydroxyvitamin D and bone alkaline phosphatase. Albumin was
lower in both lactating and NPNL Gambians than in British women. There
were no country differences in calcitonin concentration in either
lactating or NPNL women. Lactating Gambian women, but not NPNL women,
had higher plasma phosphate concentrations and lower urinary
deoxypyridinoline excretion than their British counterparts. These
country effects were still apparent after adjustment for age.
|
This is the first study to examine longitudinal changes in calcium and bone metabolism in women who lactated for more than 18 months and in women who were accustomed to a very low calcium intake (300 mg/day). Highly significant changes were observed during the 78 weeks of lactation in all analytes measured, and similar patterns were obtained for women consuming their usual calcium intake and for those consuming a calcium supplement.
The calcium supplement, which effectively tripled calcium intake between 1.5–52 weeks of lactation, had no discernible effect on calciotropic hormone levels, plasma minerals, or bone turnover. There was one exception; the calcium-supplemented group had lower bone alkaline phosphatase activity than the placebo group at 52 weeks. However, the significance of this difference was weak and may have been a chance finding given the large number of statistical comparisons made in this study. This lack of an effect of the supplement at the biochemical level was unexpected, as Australian lactating women responded rapidly to an oral calcium load by increasing plasma calcium concentration and urinary calcium output, and decreasing PTH concentration and urinary outputs of phosphate and hydroxyproline (19). However, in our study, the effect of calcium supplementation was measured in 24-h urine samples and fasting blood collected 3 and 12 months after the supplement had been introduced, by which time alterations in urinary output were obvious, and other adjustments may have had time to take effect. Two previous studies of lactating women with habitual diets considerably higher than that of the Gambians also detected no effects of calcium supplementation on calciotropic hormones or bone turnover markers after 3–6 months (5, 20). In addition, the biochemical results of the Gambian study parallel the lack of effect of the supplement on the breast milk calcium content and bone mineral status (12).
For both groups of lactating Gambian women, the PTH concentration at 1.5 weeks was below that in NPNL women in the same community and increased during lactation to a concentration at 78 weeks that was slightly above NPNL levels. 1,25-Dihydroxyvitamin D behaved similarly. Calcitonin was raised in early lactation and decreased to NPNL levels by 13 weeks. These data provide no support for the concept of lactation as a period of physiological hyperparathyroidism (21). Elevated concentrations of PTH and 1,25-dihydroxyvitamin D have been noted in association with breastfeeding in some studies, but not all, and generally only in the later stages of lactation or after weaning (5, 20, 21, 22, 23). This suggests that the increase in these hormones is connected with the decline in breast milk production and the cessation of lactation. A possible association with rising estrogen levels and the return of menses has also been reported (20, 24).
There was no indication that any of the three calciotropic hormones was directly associated with the changes in forearm bone mineral status or with the decrease in daily urinary calcium excretion at 13 and 52 weeks of lactation in the Gambian women (12). This latter finding is consistent with an Australian study in which renal calcium conservation, as measured in fasting urine samples, was not associated with an increase in PTH or 1,25-dihydroxyvitamin D (23). Renal phosphate conservation was also observed in the Australian women, as indicated by increases in renal phosphate threshold and serum phosphate concentration (23). Elevated plasma phosphate was a striking feature of lactation in the Gambian women, particularly at 13 weeks, suggesting that their renal phosphate threshold was similarly increased, although this was accompanied by only a small, nonsignificant, decrease in daily urinary phosphate output (12).
Substantial changes in bone turnover were observed in the first months of lactation among the Gambian women. Urinary deoxypyridinoline excretion, a marker of bone resorption, was high at 1.5 weeks and decreased as lactation progressed. Osteocalcin, a marker of bone formation, was at NPNL levels at 1.5 weeks, but was elevated at 13 weeks, decreasing again in later lactation, although the concentrations remained above those at 1.5 weeks. Bone alkaline phosphatase activity was also raised during lactation. These findings are consistent with the reports of increased bone turnover at 3–6 months of lactation in several other studies (4, 5, 8, 25, 26).
The longitudinal Gambian data suggest that there may be a time delay, particularly in early lactation, between the rise in bone resorption and the elevation in bone formation. If this were the case, the rapid, temporary increase in the number of new bone resorption cavities would result in the net release of calcium, other minerals, and matrix components from the skeleton for a short period. Restoration of this bone would occur after a few weeks as part of the normal remodeling cycle, once osteoblast recruitment and activity had increased. Such a pattern would explain the loss and subsequent regain of bone mineral during lactation observed in this and other studies (2, 3, 4, 5, 6, 7, 8, 9, 12) and has parallels with the effects of the bone remodeling transient on bone mineral status after the administration of drugs and other factors that have an incremental effect on bone turnover (27, 28). Interestingly, as this study has shown, provision of a calcium supplement to lactating women appears to have no effect on bone turnover or bone mineral status. This is in contrast to studies of children and older women, in whom calcium supplementation is associated with an increase in bone mineral and a decrease in bone turnover (28, 29).
These findings suggest that the alteration in calcium and bone metabolism that accompanies human lactation represents a physiological, homeorhetic response (30) that is independent of dietary calcium intake. It can be speculated that such a response may have evolved to guarantee breast milk production in the first days and weeks after delivery in the face of uncertain dietary supply. The hormonal mechanism driving this response is, as yet, unknown, but the evidence suggests that the classical calciotropic hormones do not play a key role. Lactation is accompanied by alterations in the secretion of many hormones, such as PRL and estradiol, which are likely to have effects on bone and the kidney, and are known to act as homeorhetic regulators in other situations (30). There is also evidence that PTH-related peptide may have an important function in this respect, especially in early lactation (25, 31). This hormone is produced by the active mammary gland, is secreted in large amounts into breast milk, and is found in the maternal circulation during lactation (25, 31). A mechanism by which a factor derived in the mammary gland is produced and released into the circulation in response to suckling would explain our recent finding that lactational bone loss is related to breast milk output (6).
The Gambian women, both those who were lactating and those who were NPNL, had considerably higher concentrations of PTH, 1,25-dihydroxyvitamin D, and plasma markers of bone formation than their British counterparts. Their vitamin D status, as measured by 25-hydroxyvitamin D, was in the normal range. Calcitonin levels were similar in the two countries despite an earlier study that suggested that Gambians have higher values (32). The high concentrations of PTH, 1,25-dihydroxyvitamin D, and bone formation markers in the Gambian subjects are unlikely to reflect racial differences, because Gambians born and raised in West Africa but resident in the United Kingdom have concentrations of these analytes similar to those in Britons (17). There is evidence that African-Americans also have elevated PTH and 1,25-dihydroxyvitamin D concentrations compared to Caucasians (33, 34), but this has generally been associated with lower bone turnover (33, 35). Although the most plausible explanation for the higher circulating levels of these hormones in the Gambian subjects is that they are related to diet and other environmental factors, the supplementation study indicates that low calcium intake, at least during lactation, is not implicated.
In conclusion, this study has shown that lactation is associated with marked increases in bone turnover and plasma phosphate, without corresponding alterations in circulating PTH and 1,25-dihydroxyvitamin D levels. These effects occur in the first weeks after delivery and, for women who breastfeed for many months, diminish as lactation progresses even when, as in the Gambian subjects, breast milk volume remains high. The changes appear to be physiological, possibly part of an orchestrated homeorhetic response to accommodate breast milk production, and are not related to calcium supply, as in this group of women who were consuming a very low calcium diet, a substantial increase in calcium intake did not alter the biochemical changes materially or affect breast milk calcium secretion or bone mineral status.
Acknowledgments
We are grateful to all those in The Gambia and Britain who assisted with various aspects of the study, including the medical supervision (Terry Aspray, Anne Dale, Simon Hoare, and Kayode Erinoso), sample collection and transportation (Steve Austin, Mustapha Ceesay, Sainabou Darboe, Warren Lee, Yankuba Sawo, and S. Gabrielle Wharf), organization of the randomization of supplements (Odile Dewit), and provision of samples from nonpregnant, nonlactating subjects (Terry Aspray, Bakary Dibba, and Ann Laskey).
Footnotes
1 This work was supported by the Commission of the European
Communities under the Science and Technology for Development Program
"Medicine, Health and Nutrition in the Tropics,"
TS2*0228.UK(JR). 
2 Visiting scholar at the Dunn Nutrition Unit from the University of
Witwatersrand, South Africa.
Received July 29, 1997.
Revised October 27, 1997.
Accepted January 12, 1998.
References
This article has been cited by other articles:
 |
|
 |
|
  F. F Bezerra, L. M. Mendonca, E. C Lobato, K. O O'Brien, and C. M Donangelo Bone mass is recovered from lactation to postweaning in adolescent mothers with low calcium intakes Am. J. Clinical Nutrition, November 1, 2004; 80(5): 1322 - 1326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. L Vargas Zapata, C. M Donangelo, L. R Woodhouse, S. A Abrams, E M. Spencer, and J. C King Calcium homeostasis during pregnancy and lactation in Brazilian women with low calcium intakes: a longitudinal study Am. J. Clinical Nutrition, August 1, 2004; 80(2): 417 - 422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Prentice Micronutrients and the Bone Mineral Content of the Mother, Fetus and Newborn J. Nutr., May 1, 2003; 133(5): 1693S - 1699. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. F. Bezerra, F. P. Laboissiere, J. C. King, and C. M. Donangelo Pregnancy and Lactation Affect Markers of Calcium and Bone Metabolism Differently in Adolescent and Adult Women with Low Calcium Intakes J. Nutr., August 1, 2002; 132(8): 2183 - 2187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. G. Vajda, B. M. Bowman, and S. C. Miller Cancellous and Cortical Bone Mechanical Properties and Tissue Dynamics During Pregnancy, Lactation, and Postlactation in the Rat Biol Reprod, September 1, 2001; 65(3): 689 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. C. Nordin Calcium requirement is a sliding scale Am. J. Clinical Nutrition, June 1, 2000; 71(6): 1381 - 1383. [Full Text] [PDF]
|
|
|
|
|
|
A. Prentice Maternal calcium metabolism and bone mineral status Am. J. Clinical Nutrition, May 1, 2000; 71(5): 1312S - 1316. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. J. Kalkwarf, B. L. Specker, and M. Ho Effects of Calcium Supplementation on Calcium Homeostasis and Bone Turnover in Lactating Women J. Clin. Endocrinol. Metab., February 1, 1999; 84(2): 464 - 470. [Abstract] [Full Text]
|
|
|
|
|
|
S. A. Abrams Bone Turnover during Lactation--Can Calcium Supplementation Make a Difference? J. Clin. Endocrinol. Metab., April 1, 1998; 83(4): 1056 - 1058. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
