Magnesium Metabolism in 4-Year-Old to 8-Year-Old Children


  • Steven A Abrams,

    Corresponding author
    1. Children's Nutrition Research Center, United States Department of Agriculture/Agricultural Research Service (USDA/ARS), Houston, TX, USA
    2. Department of Pediatrics, Texas Children's Hospital, Baylor College of Medicine, Houston, TX, USA
    • Address correspondence to: Steven A Abrams, MD, USDA/ARS Children's Nutrition Research Center, 1100 Bates St, Houston, TX 77030, USA. E-mail:

    Search for more papers by this author
  • Zhensheng Chen,

    1. Children's Nutrition Research Center, United States Department of Agriculture/Agricultural Research Service (USDA/ARS), Houston, TX, USA
    2. Department of Pediatrics, Texas Children's Hospital, Baylor College of Medicine, Houston, TX, USA
    Search for more papers by this author
  • Keli M Hawthorne

    1. Children's Nutrition Research Center, United States Department of Agriculture/Agricultural Research Service (USDA/ARS), Houston, TX, USA
    2. Department of Pediatrics, Texas Children's Hospital, Baylor College of Medicine, Houston, TX, USA
    Search for more papers by this author


Magnesium (Mg) is a key factor in bone health, but few studies have evaluated Mg intake or absorption and their relationship with bone mineral content (BMC) or bone mineral density (BMD) in children. We measured Mg intake, absorption, and urinary excretion in a group of children 4 to 8 years of age. Mg absorption was determined using a dual-tracer stable isotope technique, with 25Mg given intravenously and 26Mg given orally. We found a small, but significantly greater Mg absorption efficiency (percentage absorption) in males than females (67% ± 12% versus 60% ± 8%, p = 0.02) but no difference in estimated net Mg retention (average of 37 mg/d in both males and females). Relating dietary Mg intake to estimated Mg retention showed that an intake of 133 mg/d, slightly above the current estimated average requirement (EAR) of 110 mg/d, led to a net average retention of 10 mg/d, the likely minimum growth-related need for this age group. Covariate analysis showed that Mg intake and total Mg absorption, but not calcium intake or total absorption were significantly associated with both total body BMC and BMD. These results suggest that usual Mg intakes in small children in the United States meet dietary requirements in most but not all children. Within the usual range of children's diets in the United States, dietary Mg intake and absorption may be important, relatively unrecognized factors in bone health. © 2014 American Society for Bone and Mineral Research.


Osteoporosis remains a major public health problem in the United States. It is hypothesized that osteoporosis in the elderly has its basis in achieving an inadequate peak bone mass during adolescence. Although increasing calcium and vitamin D intake have been shown to have some effect in increasing bone mineral content (BMC) and bone mineral density (BMD) during adolescence, it is likely that other nutritional factors remain limiting in optimizing peak bone mass.[1, 2]

One of the most likely candidates for a nutritional cofactor in maximizing bone mineral outcomes is magnesium (Mg). Mg is a critical mineral involved in regulating bone development as a component of the bone mineral matrix and in regulating hormonal status.[3] Mg deficiency impairs parathyroid hormone function leading to failure to mineralize bone and increased bone demineralization in all age groups.[4, 5]

However, data related to Mg intake and metabolism and their effects in children and adolescents are very limited as a result of the challenges of assessing Mg intake and absorption. Modern stable isotope studies are expensive and only a handful of centers have the technical capacity to perform and analyze them. Most estimates for dietary Mg requirements were made using interpolation of data from infants and adolescents.[6-8]

We conducted stable isotope studies of Mg absorption as well as measurements of BMC and BMD in a group of 4-year-old to 8-year-old children who subsequently participated in a controlled trial of vitamin D supplementation.[9] We hypothesized that usual Mg intakes in the United States would meet estimated growth requirements and that both calcium and Mg intake would be significantly related to total body BMC and BMD in this age group. Although peak bone growth occurs at a later age, younger children are an important group to ensure all forms of nutritional adequacy, including bone health.

Subjects and Methods

Subjects and research studies

Subjects for the study were selected to approximately match the ethnic distribution of the greater Houston, TX, USA area. To be enrolled, subjects had to be healthy, and not using any medications or multivitamins/multiminerals. Written informed consent was obtained from a parent or legal guardian for each subject. The Institutional Review Board of Baylor College of Medicine and Affiliated Hospitals approved the protocol.

Subjects were enrolled who were 4.0 to 8.9 years of age at the time of starting the study. Subjects were not enrolled who had a body mass index (BMI) Z-score > 2.0. Although a lower limit of BMI Z-score was not set, only 1 subject had a BMI Z-score < –2.0.

Calcium and vitamin D outcomes for this study have been reported, as related to the subsequent controlled trial of vitamin D supplementation.[9]

Study procedures

This study was conducted at the General Clinical Research Center (GCRC) of Texas Children's Hospital in Houston, TX, USA. To confirm eligibility, usual dietary nutrient intake was determined from two 24-hour telephone dietary recalls and a 3-day weighed diet.

Studies of total BMC were performed by dual-energy X-ray absorptiometry (DXA) using a Hologic Delphi Model (Hologic Inc., Waltham, MA, USA). The DXA instrument undergoes regularly scheduled quality control testing for phantom reproducibility and signal uniformity.

Dietary methods

At the screening visit, the study dietitian asked subjects what they usually ate on a normal day via two 24-hour dietary recalls, and food preferences were obtained. Inpatient menus for the overnight study visit were based on the usual daily calcium intake. All foods and beverages during the inpatient and outpatient visits were preweighed and postweighed to accurately determine intake. Subjects were provided food scales and were instructed to keep weighed food records for 3 days at home. To reflect the marketplace changes in dietary food contents during the study, dietary intake data were collected using Nutrition Data System for Research software, developed by the Nutrition Coordinating Center (NCC), University of Minnesota, Minneapolis, MN, USA.

Mg absorption and analytical methods

Mg absorption efficiency (fractional absorption of magnesium, commonly expressed as the percentage of absorption) was measured using a dual-tracer stable isotope technique that we have described in detail.[8-10] A stable isotope of Mg (6 mg of 25Mg) was given intravenously over about 2 minutes. Subsequently, the children were given a standard breakfast. Toward the end of breakfast, the subjects were given a stable isotope of Mg (12 mg of 26Mg) that had been mixed with 120 mL of orange juice fortified with calcium and vitamin D. Calcium isotope studies were performed simultaneously.[9]

Beginning with breakfast, a complete 72-hour urine collection was obtained (first 24 hours inpatient, then home collection). Urine samples were prepared for inductively coupled plasma mass spectrometry analysis as described.[10, 11]

Fractional absorption of Mg was calculated as the relative recovered oral versus intravenous isotope in a 72-hour urine specimen collected starting at the time of the tracer administration.[7, 8] Total Mg absorption was calculated as the product of intake and fractional absorption. An estimate of endogenous fecal Mg of 20 mg/d was used based on our studies[11] to determine approximate net Mg retention. We have demonstrated that endogenous fecal excretion is not an important regulatory point for Mg metabolism and thus an overall estimate in this age group of about 20 mg was reasonable.[11, 12]

Sample size determination and statistical analysis

Sample size for this study was based on the sample needed for the controlled trial of vitamin D supplementation and calcium absorption, of which this study was a part.[9] Although 63 subjects were enrolled in that study, limitations in the availability of Mg isotopes was such that we were only able to complete these studies in the first 50 enrolled subjects of that trial. All data were analyzed using SPSS (Version 16.0; SPSS, Inc. Chicago, IL, USA).

Study data are reported as Mean ± SD except as specifically noted. Relationships among outcome variables, primarily Mg intake, absorption, and BMC and BMD were done by analysis of covariance (ANCOVA), with weight, height, BMI, gender, and calcium intake used in the analysis (as noted). Other dietary factors including protein and total energy intake were also included in the analyses.


Subject characteristics

A total of 50 subjects had Mg absorption studies completed. For 1 of those subjects, a DXA scan was not obtained as a result of technical issues. There were 22 boys and 28 girls and their anthropometry by gender is shown in Table 1. For the 50 subjects who received Mg absorption studies in this cohort, mean calcium intake was 852 ± 201 mg/d and was not significantly different between males and females (p = 0.21).

Table 1. Study population
 Age (years)Weight (kg)Height (cm)BMI Z-score
  • BMI = body mass index.
  • aSignificance of differences between males and females.
Males (n = 22)6.9 ± 1.523.6 ± 4.1121.4 ± 8.80.1 ± 0.7
Females (n = 28)6.4 ± 1.321.8 ± 4.7118.4 ± 10.1–0.1 ± 0.8
Overall (n = 50)6.6 ± 1.422.6 ± 4.4119.7 ± 9.60.0 ± 0.8
Gender p valuea0.

Mg absorption and retention

Results for Mg absorption and retention for the study population by gender are shown in Table 2. Overall, for the entire population, mean Mg intake was 177 ± 36 mg/day, absorption fraction was 63 ± 10% and total Mg absorption was 111 ± 31 mg/day. Serum Mg averaged 2.1 ± 0.1 mg/dL. Total urinary Mg for the whole group was 54 ± 26 mg/day. Mg retention was 37 ± 32 mg/day using an estimate of endogenous fecal Mg losses as described above.

Table 2. Mg Absorption and Balance Data for Study Population
 Mg intake (mg/d)Mg absorption (%)Total Mg absorption (mg/d)Urinary Mg (mg/d)Estimated Mg retention (mg/d)
  • aSignificance of differences between males and females.
Male (n = 22)184 ± 3167 ± 12123 ± 3367 ± 2637 ± 26
Female (n = 28)170 ± 3860 ± 8101 ± 2544 ± 2237 ± 36
Gender p valuea0.150.0160.010.0020.98

The relationship between total Mg absorption and urinary excretion was significant as shown in Fig. 1, r = 0.38, p = 0.006. The relationship between Mg intake and estimated retention is shown in Fig. 2. Because intake is part of the equation for retention, caution should be used in interpreting this relationship statistically. However, it can be helpful to evaluate the intake at which growth requirements for Mg are met. In this case, the relationship between intake and absorption is described by retention = 0.62, intake = –72.4, r = 0.69. Solving this for a net retention of 10 mg/d gave a Mg intake of 133 mg/d and for a net retention of 20 mg/d it gave a Mg intake of 152 mg/d.

Figure 1.

Relationship between total Mg absorption and urinary Mg excretion: y = 0.32x + 18, r = 0.38, p < 0.05.

Figure 2.

Relationship between Mg intake and estimated Mg retention: y = 0.62x – 72.3, r = 0.69.

Relationship between mineral intake and absorption and BMC and BMD

The linear relationship between total body BMC and Mg absorption is shown in Fig. 3. As shown in Table 3, Mg intake and absorption, but not gender or calcium intake or absorption, were significantly related to total body BMC and BMD in 4-year-old to 8-year-old children after adjusting for age, weight, and height. When Mg intake and absorption fraction were included together as independent covariates, intake (p < 0.01), but not the absorption fraction (p = 0.19), was significantly associated with total body BMC. Urinary Mg was not a significant predictor of BMC (p = 0.51).

Figure 3.

Linear relationship between total Mg absorption and BMC: y = 2.79x + 368, r = 0.50, p < 0.01. BMD = bone mineral content.

Table 3. Relationships to Total Body BMC and BMD Using Models Including Either Calcium and Mg Absorption or Intake as Covariates
  1. Values are p values for covariate analysis along with the beta for the relationship and the SE of the beta value. n = 49. Replacing weight with BMI led to minimal changes in values shown. BMC was total g of bone mineral; BMD is in g/cm.
  2. BMC = bone mineral content; BMD = bone mineral density; SE = standard error; BMI = body mass index.
(A) Intake
Age (years)0.01231.
Weight (kg)<0.00123.65.40.0700.0070.004
Mg intake (mg/d)0.0031.10.30.0090.0010.00024
Calcium intake (mg/d)0.13–<0.00010.00004
Height (cm)0.631.
(B) Absorption
Age (years)0.1018.520.90.050.0170.008
Weight (kg)<0.0128.
Mg absorption (mg/d)0.0020.980.310.080.000420.00022
Calcium absorption (mg/d)0.10––0.00020.0007
Height (cm)0.790.772.940.09–0.0040.002

Other anthropometric and dietary variables were considered as possible confounders in this relationship. When potassium intake was added as a covariate, there was no significant relationship with BMC or BMD (p = 0.71 and p = 0.92, respectively) and the relationship between BMD and BMC with Mg intake remained significant.

For energy intake, evaluating this relationship with the same covariates as in Table 3 related to intake also revealed no significant relationship for BMC or BMD (p = 0.20 and p = 0.21, respectively), and the relationships between BMD and BMC with Mg intake remained significant. When fat mass (g) was included in the covariate analysis, BMC was significantly related to Mg intake (p = 0.013), but not to fat mass itself (p > 0.19).

Similarly, evaluating the relationship of protein intake with the same covariates as in Table 3 (including height and weight) also revealed no significant relationship for protein intake with BMC or BMD (p = 0.31 and p = 0.61, respectively), and the relationships between BMD and BMC with Mg intake remained significant and were not substantially affected when protein intake was included in the analysis. This was also true when adding both energy intake and protein intake to the analysis together. In this combined model, the relationship between Mg intake and BMC and BMD showed p = 0. 018 and p = 0.04, respectively.

When height and weight were removed from the model, then energy intake (p = 0.008) and protein intake (p = 0.07) were related to total BMC but were not as closely related with BMD (p = 0.88 for protein and p = 0.08 for energy).


We found several unique characteristics of Mg metabolism in small children. We found that Mg, but not calcium intake or absorption, was significantly correlated to BMC and BMD. As a result of their interrelationship, it is difficult to identify which component of total Mg absorption, Mg intake or Mg absorption efficiency (fractional absorption), is the more important factor in determining bone mineral status, although it appears that Mg intake was more important than Mg absorption efficiency. Like calcium, Mg absorption efficiency is inversely proportional to total dietary Mg intake.[11] This, along with inherent limitations in using a single tracer dose to represent all dietary intake over the course of time, may have led to a failure to identify absorption as the primary determinant of bone mineral status. Regardless, the finding of a relationship between Mg in the diet, and its absorption and bone mineral status in children is unique and may lead to further efforts to evaluate these relationships with larger population datasets. However, few datasets include Mg absorption, and Mg intake is difficult to assess using standard methods. Our use of weighed diets increased the likelihood of accurately assessing Mg intake and thus identifying these relationships.

We further showed that there was a close relationship between the total amount of Mg absorbed and the urinary Mg excretion. As endogenous fecal Mg excretion is minimal, this suggests regulation of “excess” absorption via renal excretion, similar to that seen for calcium. This process is likely needed because of the limited ability of the body to use excess Mg for bone development and the need to maintain a narrow range of serum Mg for cardiac function as well as parathyroid hormone regulation.

Of importance, we found that apparently an estimated Mg balance (retention) is positive in children in this age group on usual diets and that Mg retention is slightly above the amount believed needed for growth and metabolic needs.[6] Dietary intakes of Mg about 130 to 150 mg/d, below the mean of our study and population mean data,[6] lead to net retentions of Mg that are positive and supportive of bone and tissue accretion of Mg (+10–20 mg/d). It should be noted, however, that the years studied, 4 to 8 years of age, are not peak growth years, and intake of Mg likely needs to be greater during adolescence to meet pubertal growth needs. We also recognize that we did not specifically measure endogenous secretory Mg losses in these children. However, we have shown in several populations of children that these are minimal and not a major source of regulation of metabolism.[11, 12]

There have been a few previous studies of Mg absorption and metabolism in pediatric populations after the first year of life. A study of Mg absorption and kinetics using stable isotopes was performed in 5 adolescent girls.[13] This study looked at the effects of high versus low calcium intakes on Mg balance and found no effect of different calcium intake levels.

This was followed by studies conducted at our center involving young adolescents.[14] In these studies, we found a close relationship between whole-body fat-free mass and Mg kinetics.[7, 12] We were able to confirm the need for a full 72-hour urine collection time period to accurately assess Mg absorption. This finding was further confirmed in adults.[15]

Only one controlled trial evaluated the effects of Mg supplementation in young adolescent females on bone mineral outcomes. Carpenter and colleagues[16] randomized 10-year-old to 14-year-old adolescents to 300 mg/d of Mg (n = 23) or placebo (n = 27) for 12 months. They found that hip BMC was significantly increased in the supplement group compared to the placebo group.

A study of 52 children at 8 years of age showed that only energy intake, protein intake, and magnesium intake were significantly correlated with BMD and those values, along with calcium and zinc intake, were significantly correlated with BMC. It is not certain why we did not find an effect of energy or protein independently in our model, but it is likely because we included height and weight in our model; these parameters were not included in the models used in that report.[17]

Our results provide further support that Mg should be more strongly considered as a key factor in bone health. Calcium intake, when not severely deficient, is inconclusively related to bone mineral status in children, and supplementation with 1000 IU/d of vitamin D in this population did not affect calcium absorption or bone outcomes.[9]

Mg in the diet can be generally found in foods that can be advocated for in children, including whole grains, dairy, and nuts. It is possible that some of our outcomes related more generally to healthy diets in children, but this is not apparent from nutrient intake or BMI data and is unlikely to be the primary reason for the relationship of BMC and BMD with Mg intake and absorption. Although genetic factors are also undoubtedly critical in determining bone health, and severe deficiencies of calcium and vitamin D are limiting in some circumstances; our findings suggest that, within the usual range of diets in the United States, Mg intake may be important in optimizing bone mineral development in children. The mechanisms of this effect are not defined and require further evaluation, but may relate to hormonal factors, including parathyroid hormone secretion.


All authors state that they have no conflicts of interest.


This work is a publication of the U.S. Department of Agriculture (USDA)/Agricultural Research Service (ARS) Children's Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, and Texas Children's Hospital, Houston, TX, USA. This project was funded in part with federal funds from the USDA/ARS under Cooperative Agreement number 58-6250-0-008, NCRR General Clinical Research for Children Grant number RR00188. The contents of this publication do not necessarily reflect the views or policies of the USDA, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. government. We acknowledge the support and contributions of Texas Children's Hospital Investigational Pharmacists Jennifer Lynds and Tara McCartney, the Texas Children's Hospital General Clinical Research Center Staff, and students Melissa Mohammed, Michelle Taub, and Jennifer Haden.

Authors' roles: Study design: SA, KH, ZC. Study conduct and data collection: SA, KH, ZC. Data analysis: SA, KH, ZC. Data interpretation: SA, KH. Drafting manuscript: SA. Revising manuscript content and approving final version of manuscript: all authors. SA takes responsibility for the integrity of the data analysis.