The presence of early-stage atherosclerosis, such as impairment of endothelial function, intima thickening and fatty streaks, has been documented extensively in children and young adults, suggesting that the pathogenesis of cardiovascular disease starts in childhood.1^–3

Low-grade inflammation has a role in the impairment of endothelial function4 and atherosclerosis,5 6 and has been associated with increased risk for cardiovascular disease and diabetes.7 8 In adults, the levels of C-reactive protein (CRP), an acute-phase protein that is the most sensitive marker of systemic inflammation,9 predict the development of cardiovascular disease.10

Magnesium, the second most abundant intracellular cation,11 is an essential cofactor of enzymatic pathways involved in energetic metabolism and modulation of glucose transport across cell membranes.12 13 In adults, hypomagnesemia is associated with an increased risk for metabolic syndrome,14 type 2 diabetes,15 high blood pressure 16 and atherogenic lipid profile.16 17

Recently, we documented a significant inverse relationship between serum magnesium and CRP,18 which supports the hypothesis that hypomagnesemia is a risk factor for development of low-grade chronic inflammatory syndrome19 20 and cardiovascular disease.

In children, epidemiological studies have shown that elevation of CRP is linked to obesity,21 22 but there are no reports of its association with low serum magnesium levels, and to the best of our knowledge there are no reports about the interaction between hypomagnesemia and inflammation. Thus, the aim of this study was to determine the relationship between serum magnesium and high-sensitivity CRP (hsCRP) levels in healthy children.

METHODS

After the approval of the protocol by the Mexican Social Security Institute Research Committee, and after obtaining informed consent from the children and their parents, a community-based cross-sectional population study was carried out.

The study population was determined using two-stage cluster sampling In the first stage, a random sample of elementary schools from Durango, in northern Mexico, were obtained; in the second stage, a random sample of children aged 9 to 13 year olds were invited to participate.

All the participants were required to be in good health. For this purpose, a standardised interview and clinical examination were conducted, and detailed information on medical history was collected. Disorders related to increased CRP and magnesium depletion such as inflammatory disease, alcohol consumption, cigarette smoking, cardiovascular disease, chronic disorders of the joints and connective tissues, infectious diseases, diarrhoea, diabetes, high blood pressure, surgical stress, diuretic therapy and reduced renal function were exclusion criteria. In addition, children receiving anti-inflammatory drugs and/or magnesium supplementation were not included. The Tanner stage, which defines stages of pubertal development, was determined;23 only children at Tanner stage 1 or 2 were included.

Eligible children were allocated to three groups according to their body mass index (BMI): (a) normal weight; (b) overweight and (c) obese.

Measurements

Weight and height were measured between 8:00 and 9:00 h, with children in a standing position, wearing light clothing and no shoes. BMI was calculated as weight (in kg) divided by height (in m) squared. Waist circumference was measured to the nearest cm with a flexible steel tape measure. The following anatomical landmarks were used: laterally, midway between the lowest portion of the rib cage and iliac crest; and anteriorly, midway between the xiphoid process of the sternum and the umbilicus.24 Total body fat was measured by bioelectric impedance (Omron BF 300, Vernon Hills, Ill, USA).

Blood pressure was measured using a mercury sphygmomanometer, with the children seated and their arms bared and supported at heart level, after at least 5 min of rest, and using an appropriate cuff size.25 An average of three readings separated by 2 min was used.

Assays

A venous whole-blood sample was collected after 8–10 h of fasting. hsCRP was measured by automated microparticle enzyme immunoassay (IMx, Abbot Laboratories, Minneapolis, MN, USA). The detection limit of hsCRP was 0.05 mg/dl, with an intra–interassay coefficient of variation of 4.1% and 5.8%, respectively. Serum magnesium concentrations were measured using a colorimetric method; the intra- and interassay variations were 1.0% and 2.5%.

Triglycerides were measured enzymatically (Data Pro Plus random access clinical analyzer Arlington, TX, USA), and the HDL-cholesterol fraction was obtained after precipitation by a phosphotungstic reagent. The intra- and interassay coefficients of variation were 1.7% and 3.1% for triglycerides, and 1.3% and 2.6% for HDL-cholesterol. Serum glucose was measured using the glucose–oxidase method. The intraassay and interassay coefficients of variation were 1.1% and 1.5%, respectively. Insulin levels were measured by microparticle enzyme immunoassay (Abbot Axsym System, Alameda, CA, USA), with intra- and interassay variation coefficients of 4.5 and 6.9, respectively.

Definitions

Obese was defined as a BMI equal or greater than the 95th percentile, overweight by a BMI between the 85th and 95th percentiles, and normal weight by a BMI lower than the 85th percentile.26

Decreased serum magnesium levels were defined as a magnesium concentration lower than 1.8 mg/dl,16 and serum hsCRP concentration ⩾3 mg/l defined the elevated hsCRP levels.27

Statistical analysis

Two-tailed unpaired student’s t test was used for comparison of normally distributed variables (Mann–Whitney U test for skewed data). One-way ANOVA test was used for estimating differences between the groups. Pearson’s analysis was performed to examine the correlation between measures for obesity, hsCRP and serum magnesium levels. For this purpose, skewed numerical data were transformed to the Logn.

Multivariate logistic regression adjusted by body-fat percentage was performed to compute the relationship between hsCRP and serum magnesium levels. A 95% confidence interval was considered. Data were analysed using the statistical package SPSS for Windows 12.0.

RESULTS

A total of 488 children (age range 10 to 13 years) were included in the study, 248 (50.8%) boys and 240 (49.2%) girls; of these, 255 (52.2%), 113 (23.1%) and 120 (24.6%) were classed as normal weight, overweight and obese, respectively.

The characteristics of participants are shown in table 1. Girls had higher BMI, waist circumference and body fat than boys. For both boys and girls, systolic and diastolic blood pressure as well as glucose, insulin and triglycerides showed a gradual increase, and HDL-cholesterol a gradual decrease, from normal weight to obese groups. Serum magnesium and hsCRP levels did not differ significantly between boys and girls.

One hundred and nine children (22.3%) exhibited high hsCRP levels (40 obese, 29 overweight and 40 normal weight); and 101 (20.7%) had low serum magnesium levels (31 obese, 30 overweight, and 40 normal weight). Among children with high hsCRP levels, 95 (87.1%) had hypomagnesemia; of these 72.5% (29/40), 89.6% (26/29) and 100% (40/40) were in the groups for obese, overweight and normal weight, respectively. For boys the median hsCRP level was 0.60 mg/l (range 0.05–10.8) and the mean (SD) serum magnesium levels 2.01 (0.48) mg/dl. For girls, the median hsCRP levels was 0.62 (range 0.05–10.0) and the mean serum magnesium levels 1.97 (0.49) mg/dl.

The prevalence of elevated hsCRP and low serum magnesium levels was 19.5% (95/488). Children with elevated hsCRP and hypomagnesemia showed the highest fasting glucose, insulin and triglycerides and the lowest HDL-cholesterol levels (table 2).

Table 3 shows the Pearson’s correlations among measures for obesity, serum magnesium and hsCRP levels. For both girls and boys there were a strong negative correlation between serum magnesium and hsCRP levels, and a positive correlation between BMI, waist circumference and body-fat percentage with hsCRP, but not with serum magnesium levels. Among measures for obesity, BMI and waist circumference were strongly correlated (p<0.01).

The principal finding of this study is that the reduction of serum magnesium levels corresponds to increased hsCRP, a trend that is statistically significant, r =  –0.708, p<0.001 (fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Relationship between serum magnesium levels and high-sensitivity C-reactive protein (hsCRP) in Mexican children aged 10 to 13 years, n = 488. There is a significant inverse relationship between magnesium and hsCRP (r = –0.708, p<0.001).

In the multivariate analysis, body-fat percentage and serum magnesium levels were associated with hsCRP levels. After adjustment for body-fat percentage, serum magnesium remained strongly associated with hsCRP (table 4).

DISCUSSION

Results of this study show a significant association between low serum magnesium and elevated hsCRP levels. In adults, we reported that low serum magnesium levels are related to elevated hsCRP concentration28 and more recently we have shown that magnesium intake is inversely associated with systemic inflammation.29 30 Our results extend these findings to children, in whom there are no previous data on the association between subclinical inflammation and hypomagnesemia.

In this study, the prevalence of elevated hsCRP was 22.3%. Although there are not specific reports on prevalence of low-grade chronic inflammation and little is known about CRP concentration in children,31 previously reports indicate that the range of CRP varies from 4.0 to 7.7%.32 33 However, the 85th percentile of CRP concentration among Mexican–American individuals from 3–19 years who participated in the National Health and Nutrition Examination Survey 1999–2000 was 3.3 mg/l,31 a figure that represents a prevalence of low-grade chronic inflammation greater than 15.0%. In the same way, the median CRP that we reported (0.62 and 0.60 mg/l for boys and girls) was similar to the National Health and Nutrition Examination Survey 1999–2000 reports (0.60 and 0.66 mg/l for boys and girls).34 CRP concentration varies by age, ethnicity and adiposity, with the highest concentrations among Mexican-American children,31 34 variables that could explain differences in prevalence reported. Also, data about prevalence of hypomagnesemia in healthy children are scarce; in this study the prevlance was 20.7%, similar to that reported by Ahsan et al (21.7%).35

Although data for children are not so well developed as they were in adults, since inflammation and hypomagnesemia per se have important roles in many disease processes, the interaction of between elevated hsCRP and low serum magnesium levels in children suggests the presence of increased risk of development of metabolic and cardiovascular disease in childhood, a finding that agrees with previous reports showing that cardiovascular disease often starts in youth.2 3 On this point, a strong relationship between elevated hsCRP concentration and both insulin resistance36 and metabolic syndrome,37 as well as between magnesium deficiency and insulin resistance,38 has been reported among children and adolescents.

Obesity is a well-known risk factor associated with elevated hsCRP concentration;11 33 37 furthermore, we hypothesized that low serum magnesium levels could also be a risk factor for elevation of hsCRP.28 The results of this study support our hypothesis that there is a relationship between hypomagnesemia and inflammation. A strong association between hypomagnesemia and low-grade inflammation was found. This association remained significant after adjustment for measures of obesity. In addition, children with inflammation and hypomagnesemia, although less fat, had significantly higher hsCRP levels than children with inflammation and normo-magnesemia, and all normal-weight children with inflammation exhibited hypomagnesemia.

The release of substance P, one of the earliest events in the chronic inflammation response,39 is linked to hypomagnesemia,40 41 which provides a biologically plausible background for our hypothesis. However, since cross-sectional studies imply methodological conflict in the demonstration of the cause–effect temporal exposition, we are not certain of the role of magnesium in pathophysiology of inflammation. Further research is required in this field.

In obese children and adults, magnesium deficiency is associated with insulin resistance,38 42 suggesting that hypomagnesemia plays an important role in the pathophysiology of metabolic disease. In healthy individuals, magnesium deficiency is commonly related to intakes below the recommended dietary allowance (RDA), and as consequence of profound changes in diet, it has been estimated that at least half the US population fails to meet the RDA for magnesium,43 increasing their risk for cardiovascular disease.29 However, surprisingly, there is a lack of information on the relationship between hypomagnesemia and inflammation in childhood and its potential interaction for increasing disease risk. Our results shows that children who have both elevated hsCRP and low serum magnesium levels exhibit the highest fasting glucose, insulin and triglycerides levels as well as the lowest HDL-cholesterol levels, suggesting that hypomagnesemia and low-grade inflammation are interactive risk factors for metabolic glucose disturbances and atherosclerosis.

Several limitations of this study ought to be mentioned. First, the cross-sectional design of the study makes it difficult to draw inferences about causation; it cannot yet be established whether hypomagnesemia is a risk factor for inflammation or merely an associated epiphenomenon in children. Second, because we only included children within a limited age range, our conclusions cannot be applied to other populations. Third, although sampling strategy was based on a population random sample, but because all the participants were Mexicans, the extent to which the results can be generalised to other ethnic groups is uncertain. Thus, further research, including a widre population and a prospective design, is needed.

In conclusion, to our knowledge this study is the first reporting the association between low serum magnesium and elevated hsCRP levels in healthy children. Children who exhibited both elevated CRP levels and hypomagnesemia display a atherogenic lipid profile and adverse changes in glucose metabolism.