![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
*
Division of Human and Animal Nutrition, TNO Nutrition and Food Research Institute, 3700 AJ Zeist, The Netherlands,
Department of Epidemiology, National Institute for Food and Nutrition, 02-903 Warsaw, Poland,
**
ORS PACA INSERM and CIC APHM-INSERM, 13006 Marseille, France,
Osteoporosis Clinic, Aarhus Amtssygehus, 8000 Aarhus, Denmark,
Centro Sanitario, University of Calabria, 87030 Arcavacata di Rende, Italy,
Department of Anatomy, University of Oulu, 90220 Oulu, Finland,
#
Deaconess Institute of Oulu, 90100 Oulu, Finland, and
§
CAREPS, 38043 Grenoble, France
2To whom correspondence should be addressed.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
KEY WORDS: • dietary calcium • iron status • ferritin • transferrin saturation • cross-sectional • girls • women
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Differences in bioavailability
of dietary iron may be one of the factors explaining regional
differences (Galan et al. 1991
). Less than 10% of daily
iron intake has to be absorbed to compensate for daily losses (up to
20% for menstruating and pregnant women). Therefore, factors
determining absorption may be more relevant to iron status than the
absolute intake of iron (Passmore and Eastwood 1986
).
Also, the form in which iron is present in the diet is important. Heme
iron is absorbed more efficiently than nonheme iron and is thought to
be less affected by other dietary constituents (Fairbanks 1994
).
Studies of Hallberg et al. (1991
,
Hallberg et al. 1992a
,
Hallberg et al. 1992b
) showed that calcium
inhibits iron absorption when taken during the same meal. This
inhibitory effect was observed for both heme and nonheme iron. The
inhibitory effect of calcium seems to be limited to the meals in which
both calcium and iron are consumed (Cook et al. 1991
,
Dawson et al. 1986
, Deehr et al. 1990
,
Galan et al. 1991
, Gleerup et al. 1993
,
Hallberg et al. 1992a
). Further, a dose-dependent
relationship was found in which calcium intake in excess of 300 mg/d
did not lead to further inhibition of iron absorption (Hallberg et al. 1991
).
The risk of iron deficiency is elevated in rapidly growing persons such
as teenagers and in women during their reproductive years. Thus far,
studies have focused mainly on the influence of supplementary calcium
intake on iron absorption (Hallberg et al. 1991
,
Hallberg et al. 1992a
,
Hallberg et al. 1992b
). To study the
influence of nonsupplement calcium intake on iron in a group of persons
most at risk of iron deficiency, we investigated the association
between calcium intake and iron status in the CALEUR study, a large
cross-sectional study among girls and women in six European
countries. (Kardinaal et al. 1999
)
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). In each of the participating countries, girls aged
11–15 y and young adult women aged 20–23 y with high and low calcium
intake were selected from random population samples of about 750 girls
and 375 women. All subjects were of Caucasian origin. In Finland
(Oulu), Denmark (Aarhus), Poland (Warsaw) and The Netherlands (Zeist),
random samples from the local population registries were obtained and
subjects invited to participate; response rates varied from 25.4 to
51.5% for girls and from 28.6 to 62% for women. In France, girls and
women were recruited via general practitioners and gynecologists in two
geographical areas, Rhone-Alps and Marseille; in Italy, girls were
recruited from all eight secondary schools in the town of Rende, and
women from the University of Calabria (response rates were 100%). Of
those selected, the participation rate varied between 64 and 84% for
girls and between 52 and 91% for women: data were collected for 1,116
girls and 526 women. The study was approved by local
medical-ethical committees, and all participants (and their
parents, if required) gave their informed consent. Subjects were
excluded when indicating a chronic disease in general, diseases related
to bone or calcium metabolism in particular, use of corticosteroids,
participating in sports more than 7 h/wk, current or prior pregnancy,
menstruation at irregular intervals (for the women only), vegetarianism
or macrobiotism or a prescribed diet other than an
energy-restricted diet. Dietary intake.
To estimate calcium intake in a comparable way in the six countries, a 3-d food record method was used. The subjects were asked to record everything they consumed during a consecutive Wednesday, Thursday and Friday, the week before their visit to the institute. Time of day, food, quantity and recipes of composite dishes were recorded. The parent responsible for meal preparation was asked to assist in completing the food records. At the visit to the institute, the food records were checked by a dietitian for completeness; household measures were verified by comparison with standard measures. Daily consumption of food products in grams were converted to nutrient intake using local food composition tables. Mean intakes of calcium, iron and energy were calculated as the average over 3 d.
Height and weight were measured with the subjects wearing light clothing and no shoes. Subjects completed a self-administered questionnaire on menstrual function and use of oral contraceptives, smoking habits, alcohol use, time spent outdoors, height, weight and education of parents and physical activity. Physical activity was determined for the previous month and covered activities at school, at work and in leisure time (sports and household activities). For the girls, the questionnaire comprised 58 items, and for the women 88. The questionnaire was checked in an interview setting.
A 10-mL blood sample was drawn from nonfasting subjects in a 10-mL tube
with clot activator and cooled to 4°C. Serum was prepared within
2 h by centrifugation at 3000 x g for 10 min.
Serum was stored at -20°C. Serum ferritin was measured as the main
marker for iron status. Serum ferritin levels reflect stored iron.
Serum iron and the main iron-binding protein, transferrin, reflect
the iron in transit and were used to calculate the transferrin
saturation {(transferrin saturation (%) = iron (g/L)/[transferrin
(g/L) x 1.41)]} (Wick 1996
). Transferrin saturation
can be used as a short-term marker of iron concentration
(Wick 1996
). Serum iron, transferrin and ferritin were
measured with the Hitachi 911 (Boehringer Mannheim, Mannheim, Germany).
Ferritin and transferrin were analyzed immunoturbidimetrically with a
ferritin test kit (No. 1661400; Boehringer Mannheim) and a transferrin
test kit (No. 1360752; Boehringer Mannheim). Serum iron was analyzed
after separation of Fe3+ from transferrin. After reduction,
Fe2+ forms colored complexes with FerroZine(TM) (Hach
Chemical Co., Ames, IA). The coefficient of variation for the analysis
of serum ferritin was 8.1 and 3.9% at mean values of 30 and 60 µg/L
in the quality control samples, respectively. The variation coefficient
for analysis of transferrin was 2.5% at a mean level of 3.3 g/L and
for analysis of serum iron the variation coefficient was 1.8% at a
mean value of 18 µmol/L in the quality control sample.
Blood samples were available for 1,083 girls and 525 women. In a selection of samples with extremely high or low ferritin levels (ferritin <3.0 or >10.0 µg/L), duplicate measurement was performed. When the second measurement was comparable to the first measurement, the first value was used. Samples of four subjects (three girls and one woman) were excluded from statistical analyses because of instability of the sample, leaving data for 1,080 girls and 524 women for analysis.
Statistical analyses.
Mean and SD of serum iron, serum ferritin, serum transferrin and transferrin saturation and of dietary intake levels were calculated. Pearson correlations between potential confounding factors and calcium intake and parameters for iron status were calculated. As potential confounders, age, height, weight, menses, smoking, tea and coffee consumption, alcohol consumption, energy intake, protein intake and vitamin C intake were considered. Variables associated with calcium intake and serum ferritin levels or transferrin saturation were included as a covariable in the statistical models with serum ferritin or transferrin saturation as independent variables. Categorical variables were put into the model as dummy variables. Though the covariables were not significantly associated with all the parameters of iron status, we choose to use a fixed set of covariables in the statistical models. With analysis of covariance, mean levels of the serum iron parameters were calculated per quartile of calcium intake, adjusted for the covariables. Linear regression analysis was used to calculate the effect of calcium intake on serum levels overall and per country, unadjusted and adjusted for covariables. The influence of the simultaneous intake of iron and calcium during a meal was measured by dividing calcium intake into calcium ingested simultaneously with iron and the remaining calcium. Simultaneous intake of iron and calcium was defined as a moment of the day when at least 20% of the total daily iron and at least 20% of the total daily calcium was consumed. For statistical analysis the BMDP statistical package was used (BMDP version 7.0; BMDP Statistical Software Inc., Los Angeles, CA). P values <0.05 were considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The ranges for (serum iron, serum ferritin, serum transferrin and the
transferrin saturation) the different countries were comparable
(Table 1
).For the pooled data, serum iron ranged from 2.9 to 46.8 mmol/L for
girls and from 2.1 to 45.1 mmol/L for women. Ranges for serum ferritin
and serum transferrin were 2.3–161.1 µg/L and 2.0–5.3 g/L,
respectively, for girls and 1.1–191.6 µg/L and 2.1–5.5 g/L,
respectively, for women. Transferrin saturation ranged from 3.4 to
60.3% for girls and from 3.0 to 63.3% for women, with the lowest
levels in Italy and the highest in Denmark. The lowest levels of serum
iron and serum ferritin were seen in Finland, both for girls and women.
Low-serum iron and ferritin levels were also seen in Italian women.
Latent iron deficiency (ferritin <12 µg/L) was found in 4.3% of the
girls and 7.4% of the women (Fig. 1
).
|
|
).Intakes in The Netherlands, Finland and Denmark were all relatively
high. The lowest iron intakes were in Italy, whereas girls and women in
Poland had the highest iron intakes.
|
In Table 3
the mean levels of serum iron, ferritin, transferrin and transferrin
saturation per quartile of calcium intake are presented. The means were
adjusted for a fixed set of covariables, age, country, tea, protein,
vitamin C and iron consumption and for menses (in girls only). Adjusted
means differed significantly between quartiles of calcium intake for
transferrin levels in girls and serum iron and transferrin saturation
in women.
|
).Both for girls and for women a significant inverse association was
found between calcium intake and serum ferritin level (r
= -0.09 and -0.07 for girls and women, respectively,
P < 0.05). After adjustment was made for the set of
covariables, the adjusted linear regression coefficient for the pooled
data was -0.57 ± 0.20 for girls. This means that for every
100-mg/d increase of calcium intake the ferritin level is reduced by
-0.57 µg/L. For an average ferritin level of 34.5 µg/L, this means
a reduction in ferritin level of 1.6%. For women a linear regression
coefficient of -1.37 ± 0.46 was found, i.e., a reduction in
ferritin concentration of 3.3% for every 100-mg/d increase in calcium
intake at an average ferritin level of 40.2 µg/L. Calculation of the
association for each country separately revealed a consistent inverse
association between calcium intake and serum ferritin, even though in
none of the countries this association reached significance. When
transferrin saturation as a measure of short-term iron
concentration was put in the model as the dependent variable, we found
a significant adjusted inverse regression coefficient for the girls of
-0.18 ± 0.08 per 100-mg/d increase of calcium. The regression
coefficient calculated for the separate countries showed only a
significant inverse association in Denmark. For women, no association
was observed between calcium intake and transferrin saturation.
|
). Our linear regression model, adjusted for iron intake,
age, menarche, country, protein intake and vitamin C intake, did not
reveal a difference in inhibition between calcium taken simultaneously
with iron and the remaining calcium. The two linear models in which the
calcium variables were separated neither revealed a significant
association with calcium taken with iron nor with the remaining
calcium. The contribution of the two calcium sources independent of the
other source was assessed by putting both calcium variables in one
regression model. In the adjusted model with serum ferritin as the
dependent variable, the linear regression coefficients for
calcium-consumed simultaneously with iron and for the remaining
calcium were -0.56 ± 0.25 and -0.58 ± 0.22, respectively,
in girls and -1.34 ± 0.56 and -1.37 ± 0.50, respectively,
in women (P < 0.05 for all values). The dose dependence of the association between calcium intake and serum ferritin levels was checked over strata of calcium intake levels. A division was made between intake levels of <300, 300–600, 600–900, 900–1200, and >1200 mg/d. No indication of a threshold effect was seen. None of the linear regression coefficients in the strata showed a significant association between calcium intake and serum ferritin levels. In all strata an inverse association was observed except for a positive linear regression coefficient in women with a calcium intake lower than 300 mg. No trend could be detected.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The study was conducted in six countries throughout Europe, which contributed to a wide variety in intakes of both calcium and iron. In each country, two groups of subjects were recruited, a group of girls and a group of young women. To ensure comparability of the data between groups and among countries, all procedures were similar. Because the study populations are not necessarily representative for the countries, prudence is called for when using these data as a measure of intake for a specific country. Also, blood was collected according to the same protocol, and all blood samples were analyzed centrally in one laboratory.
Dietary assessment took place on three consecutive days. From the Dutch
National Food Consumption Survey, it was calculated that the day-to-day
variation in calcium intake in The Netherlands was small (Van Erp-Baart 1996
). A 3-d dietary record could thus give a
representative estimate of usual calcium intake. Day-to-day variation
in iron is expected to be more pronounced and consequently the 3-d
record is not likely to accurately reflect habitual individual intake.
On the other hand, measurement of serum iron parameters variable is
also exposed to biological and analytical variances. The consequence of
these factors is that whenever an association really exists, the
association is either not found, or it seems to be weaker than in
reality.
In this study dietary information was collected at different periods over the year for the separate countries. Therefore, comparison of intakes between countries can give a biased view. Blood collection took place in the same period as the dietary assessment was made. Because blood levels reflect intake, seasonality plays only a minor role in the association between calcium intake levels and iron status.
Several studies show an inhibitory effect of calcium intake on iron
absorption (Deehr et al. 1990
, Cook et al. 1991
, Galan et al. 1991
, Hallberg et al. 1992a
). In these intervention studies, calcium was provided via
a supplement or added to the meal (Hallberg et al. 1991
,
Hallberg et al. 1992a
,
Hallberg et al. 1992b
). Some studies provide
extra yogurt or milk during the meal (Gleerup et al. 1993
, Gleerup et al. 1995
). An
effect of calcium on iron absorption was reported to be most pronounced
when calcium was provided during the same meal in which iron was
consumed. Gleerup et al. (1993)
showed that calcium
given 2 or 4 h before a meal had no inhibitory effect on iron
absorption. Some studies, however, did not find an inhibitory effect of
milk on iron absorption (Turnlund et al. 1990
,
Tidehag et al. 1995
), and in a study among lactating
Gambian women, no effect of calcium supplementation on serum ferritin
level was observed (Yan et al. 1996
).
A far as we know, our study is the first to look at normal dietary
intake levels instead of supplementation with calcium. Contrary to most
studies which evaluate iron absorption, our interest was the influence
of calcium intake on iron status. This study, therefore, may yield a
better assessment of the long-term effect of calcium on iron
absorption. Our study did not reveal an effect of time of calcium
consumption on iron status. It may well be that the effect of
simultaneous consumption of calcium and iron is readily reflected in
iron absorption, whereas the iron status of blood is a more
long-term variable on which this effect is less clear. In our study
we did not distinguish between heme and nonheme iron. However, because
the inhibitory effect of calcium was observed in both forms of iron
(Gleerup et al. 1995
, Hallberg et al. 1991
), we assume that the distinction is of less importance. On
the other hand, it is therefore not possible to assess iron
availability from food as proposed by Tseng et al. (1997)
.
The mechanism of inhibition of iron absorption by calcium is not yet
clear. A complex formation with iron and phytate was suggested but, on
the other hand, a complex formation may enhance rather than inhibit
calcium absorption by forming a calcium phytate complex. As both heme
and nonheme iron absorption is inhibited, Hallberg and
colleagues (1992a)
argued that the mechanism must involve
inhibition of iron extrusion from the enterocyte. Recent studies
suggest that calcium competes for iron-binding sites on the
intestinal iron-binding protein mobilferrin.
It can be argued whether the finding that higher calcium intake is
associated with reduced serum ferritin levels is of biological
relevance. The normal absorption of iron from the diet is estimated to
be 10% (about 1 mg/d). To maintain the iron balance, several
mechanisms are involved. In case of an iron deficiency, the uptake of
iron may be increased by up to 20–30% (Wick et al. 1996
). Several studies showed that iron absorption is strongly
and inversely associated with serum ferritin level (Hultén et al. 1995
). Therefore, reducing calcium intake does not
necessarily lead to increased serum ferritin levels. Because inadequate
calcium intake may lead to other serious health problems such as
osteoporosis (Heaney 1993
), it seems inappropriate to
advise strongly against the consumption of calcium to prevent iron
deficiency.
From our results we conclude that dietary calcium intake is weakly, inversely associated with the iron status of blood in girls and young women, irrespective of whether calcium was ingested simultaneously with iron.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
3 Current address: DMV international, 5460 BA
Veghel, The Netherlands.
Manuscript received June 10, 1998. Initial review completed September 24, 1998. Revision accepted January 20, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1.
Cook J. D., Dassenko S. A., Whittaker P. Calcium supplementation: effect on iron absorption. Am. J. Clin. Nutr. 1991;53:106-111
2.
Dawson-Hughes B., Seligson F. H., Hughes V. A. Effects of calcium carbonate and hydroxyapatite on zinc and iron retention in postmenopausal women. Am. J. Clin. Nutr. 1986;44:83-88
3.
Deehr M. S., Dallal G. E., Smith K. T., Taulbee J. D., Dawson-Hughes B. Effects of different calcium sources on iron absorption in postmenopausal women. Am. J. Clin. Nutr. 1990;51:95-99
4. Fairbanks V. F. Iron in medicine and nutrition. Shils M. E. Olson J. A. Shike M. eds. Modern nutrition in health and disease 1994 Lea & Febiger Philadelphia, A Waverly Company.
5. Galan P., Cherouvrier F., Preziosi P., Hercberg S. Effects of the increasing consumption of dairy products on iron absorption. Eur. J. Clin. Nutr. 1991;45:553-559[Medline]
6. Gleerup A., Rossander-Hultén L., Hallberg L. Duration of the inhibitory effect of calcium on non-haem iron absorption in man. Eur. J. Clin. Nutr. 1993;47:875-879[Medline]
7.
Gleerup A., Rossander-Hultén L., Gramatkovski E., Hallberg L. Iron absorption from the whole diet: comparison of the effect of two different distributions of daily calcium intake. Am. J. Clin. Nutr. 1995;61:97-104
8.
Hallberg L., Brune M., Erlandsson M., Sandberg A.-S., Rossander-Hultén L. Calcium: effect of different amounts on nonheme and heme-iron absorption in humans. Am. J. Clin. Nutr. 1991;53:112-119
9. Hallberg L., Rossander-Hultén L., Brune M., Gleerup A. Inhibition of haem-iron absorption in man by calcium. Br. J. Nutr. 1992;69:533-540
10. Hallberg L., Rossander-Hultén L., Brune M., Gleerup A. Calcium and iron absorption: mechanism of action and nutritional importance. Eur. J. Clin. Nutr. 1992;46:317-327[Medline]
11. Heaney R. P. Bone mass, nutrition and other lifestyle factors. Am. J. Med. 1993;95(5a suppl):29S-33S[Medline]
12. Hultén L., Gramatkovski E., Gleerup A., Hallberg L. Iron absorption from the whole diet. Relation to meal composition, iron requirements and iron stores. Eur. J. Clin. Nutr. 1995;49:794-808[Medline]
13. Kardinaal A. F. M., Ando S., Charles P., Charzewska J., Rotily M., Väänänen K., Van Erp-Baart A. M. J., Heikkinen J., Thomsen J., Maggiolini M., Deloraine A., Chabros E., Juvin R., Schaafsma G. Dietary calcium and bone density in adolescent girls and young women in Europe. JBMR 1999;April
14. Passmore R., Eastwood M. A. Human Nutrition and Dietetics 8th edn. 1986 Churchill Livingstone Hong Kong.
15.
Tidehag P., Sandberg A.-S., Hallmans G., Wing K., Türk M., Holm S., Grahn E. Effect of milk and fermented milk on iron absorption in ileostomy subjects. Am. J. Clin. Nutr. 1995;62:1234-1238
16.
Tseng M., Chakraborty H., Robinson D. T., Mendez M., Kohlmeier L. Adjustment of iron intake for dietary enhancers and inhibitors in population studies: bioavailable iron in rural and urban residing Russian women and children. J. Nutr. 1997;127:1456-1468
17.
Turnlund J. R., Smith R. G., Kretsch M. J., Keyes W. R., Shah A. G. Milk's effect on the bioavailability of iron from cereal-based diets in young women by use of in vitro and in vivo methods. Am. J. Clin. Nutr. 1990;52:373-378
18. Van Erp-Baart, A. M. J. (1996) Calciuminneming en de bijdrage van voedingsmiddelen aan de calciuminneming van 10-16 jarige en 20-26 jarige Nederlandse vrouwen. TNO-report V96.349. (in Dutch).
19. Wick M., Pinggera W., Lehmann P. Iron metabolism diagnosis and therapy of anemies 3th edn. 1996 Springer Verlag Wien-New York.
20. Yan L., Prentice A., Dibba B., Jarjou L. M. A., Stirling D. M. The effect of long-term calcium supplementation on indices of iron, zinc and magnesium status in lactating Gambian women. Br. J. Nutr. 1996;76:821-831[Medline]
This article has been cited by other articles:
|
|
 |
|
  L. Grinder-Pedersen, K. Bukhave, M. Jensen, L. Hojgaard, and M. Hansen Calcium from milk or calcium-fortified foods does not inhibit nonheme-iron absorption from a whole diet consumed over a 4-d period Am. J. Clinical Nutrition, August 1, 2004; 80(2): 404 - 409. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
