Examination of the Relationship of Dairy Product Consumption and Dietary Calcium with Body Mass Index Percentile in Children

Aim: The purpose of this study was to assess the relationships of dairy product and calcium intake with BMI percentile in children. Methods: In this cross-sectional analysis, the relationships of dairy product (cup equivalents/day) and dietary calcium (mg/day) (measured by 24 hour recall) consumption with BMI percentile [derived from Centers for Disease Control and Prevention (CDC) sex-specific BMI-for-age growth charts] were assessed in 101 healthy U.S. children (ages 8-13). Results: Subject characteristics included (mean + SD): dairy product consumption (2.8 + 1.4 cup equivalents/day), dietary calcium (1039.4 + 431.4 mg/day), and BMI percentile (70.1 + 29.7). Dairy product intake and dietary calcium were inversely related to BMI percentile (p= 0.005). In subjects <85 th BMI percentile (n= 50 normal weight and n= 3 underweight), there was an inverse association of dairy product intake with BMI percentile (p= 0.001) and calcium intake with BMI percentile (p< 0.001). However, no association of dairy product consumption or calcium intake with BMI percentile was observed in overweight (85 th percentile BMI < 95 th percentile, n= 19) or obese subjects (BMI > 95 th percentile, n= 29). Removal of underweight subjects from the analysis only minimally impacted the results. Differences in dietary intake (MyPyramid food groups, calories, macronutrients, micronutrients) among normal, overweight, and obese subgroups were examined. No significant differences were detected between groups, and removing underweight subjects did not change the findings. Conclusions: These results provide the basis for further investigation into a potential role of dairy and calcium in achieving a healthy body weight in children.


INTRODUCTION
The prevalence of overweight and obese children and adolescents in the United States has increased dramatically over the past three decades [1].Of concern, overweight and obese children are at an increased risk for developing serious illnesses (e.g., type 2 diabetes, metabolic syndrome, and hypertension) and suffering negative psychological consequences (e.g., low self-esteem and depression) [2,3].The economic consequences are also vastly detrimental; in the past three decades obesity related health care costs have tripled among overweight children [4].
Understanding the role of dietary factors in weight management is critical and has attracted much interest.Studies indicate that dairy and/or calcium intake may play a beneficial role in healthy body weight management in adults and children [5][6][7].Results from some studies suggest that high calcium and/or dairy intake may reduce body fat accumulation or enhance fat mass loss during energy restriction [8][9][10][11][12].For example, in a two year intervention in 54 young, healthy women (ages 18 to 31 years) calcium intake adjusted for calories was negatively associated with changes in body weight and fat mass, specifically in those women whose energy intake was at or below the overall group mean of 1876 kcal/day [9].However, conflicting results have been produced from other studies [13][14][15][16][17][18][19][20][21][22][23][24][25][26].In a review by Barr, intervention trials that were designed to study bone status, without including individual calorie intake in the analysis of the trials, showed no weight or body fat relationship to calcium or dairy product intake [13].Thus, the evidence in this area of inquiry is conflicting and the magnitude of the effect is uncertain.Determining the impact of dairy product and/or calcium consumption on body weight regulation is critical in the design of dietary recommendations for achieving and maintaining optimal body weight.
The purpose of this study was to assess the relationships of dairy product intake (cup equivalents/day) and calcium consumption (mg/day) with BMI percentile in children prior to enrollment in an after-school nutrition education program.The central hypothesis of this study was that dairy products and/or calcium intake play a role in modulating a healthy body weight in children.

METHODS
One hundred and one healthy U.S. children, ages 8 to 13 years, were assessed prior to enrollment in an after-school nutrition program.One extreme height outlier (2007 cm; assumed to be a data entry error) was removed from the analysis.The Ohio State University Institutional Review Board approved the study; parents of all subjects provided informed consent.
A validated method for obtaining 24 hour recalls with assisted food records was administered to assess subjects' total diet and food group consumption [27].Children were educated on how to keep a simple food diary, and instructed to write small descriptions of the foods they ate for the following 24 hours in a distributed food journal.On the following day a trained dietetic intern administered a 24 hour recall, using the United States Department of Agriculture (USDA) Five-Step Multiple Pass Method.During each interview, interns used the food records to help children recall the foods eaten.Children were first prompted to make a short list all the foods eaten during the previous day, while the intern transcribed these foods onto a standardized form.Children were asked to be as specific as possible about the foods and beverages they reported.Food records were then entered into the dietary analysis software Nutricalc 2.1 (version 2.1, The McGraw-Hill Companies, Higher Education, New York, New York) by a trained dietitian.
Height was measured to the nearest 0.1 cm with a portable stadiometer (Seca 214) and body weight was measured to the nearest 0.1 kg with an electronic digital scale (Tanita HD 317) with subjects in light clothing and no shoes.BMI percentile was derived from the sex-specific for-age growth chart for children and adolescents available from Centers for Disease Control and Prevention (CDC) [28].
This study was a cross-sectional secondary analysis of baseline data by design.Descriptive statistics are reported as mean + SD.Distributions of several baseline characteristics were compared across weight status subgroups (underweight, normal, overweight, and obese) using Mann-Whitney tests.The relationships of amounts of dairy product consumed (cup equivalent/day) and dietary calcium (mg/day) with BMI percentile were assessed using linear regression models.Caloric intake (kcal/day) and age were considered as potential covariates.Gender and weight status subgroups were incorporated into the analysis as potential moderators.Multivariate analysis of variance (MANOVA) was employed to assess differences in dietary intake including (1) daily servings of MyPyramid food groups [grains (ounce equivalent), vegetables (cup equivalent), fruits (cup equivalent), dairy (cup equivalent), meat and beans (ounce equivalent)]; (2) calories and macronutrients [calories (kcal/day), protein (g/day), carbohydrates (g/day), fiber (g/day), sugar (g/day), total fat (g/day), saturated fat (g/day)]; and (3) micronutrients [vitamin A (IU/day), vitamin C (mg/day), calcium (mg/day), and iron (mg/day)] among weight status subgroups.All models were fitted twice, once using all subjects and once excluding underweight subjects (n=3) with BMIs below the 5 th percentile.Data were analyzed using Statistical Package for the Social Sciences (SPSS), version 19.0, Chicago, IL.Statistical significance was determined at an alpha level of 0.05.

RESULTS
Participating subjects were 101 U.S. children, ages 8 to 13 years old, both boys (n= 58) and girls (n= 43), and largely classified as either Caucasian (n= 51) or African American (n= 42).Baseline characteristics of the subjects are presented in Table 1 both in aggregate and divided by weight status subgroup.Subjects' mean + SD weight and BMI percentile were 39.6 + 13.5 kg and 70.1 + 29.7, respectively.The mean + SD dairy and calcium intakes of the subjects were 2.8 + 1.4 cup equivalent/day and 1039.4 + 431.4 mg/day, respectively.Mann-Whitney tests for baseline differences among weight status subgroups found significant differences between underweight subjects and each other subgroup in dairy consumption (p= 0.044 against normal, p= 0.014 against overweight, and p= 0.027 against obese), dietary calcium (p= 0.020, p= 0.001, and p= 0.017), and daily calorie intake (p= 0.017, p= 0.021, and p= 0.006).For each of these differences, underweight subjects had higher mean values than subjects in other groups.The only other significant differences were in weight, where each group differed from each other group (p 0.001) except for underweight vs. normal (p= 0.115), and in height, where normal subjects had a lower average value than obese subjects (p= 0.001).
Regression models were fitted to the data to assess the relationships of amounts of dairy product consumed (cup equivalent/day) and dietary calcium (mg/day) with BMI percentile (Table 2).Higher levels of dairy product intake and dietary calcium were significantly associated with lower BMI percentile ( = -5.71%/cupper day, p= 0.005; = -0.019%/mgper day, p= 0.005, respectively).The strength of these associations was lessened slightly when underweight subjects (n= 3) were removed (dairy intake: = -3.81%/cupper day, p= 0.061; dietary calcium: = -0.013%/mgper day, p= 0.048).Dairy product intake and dietary calcium were highly correlated (r= 0.935), so using them simultaneously as independent variables resulted in inflation of standard errors, rendering both effects nonsignificant.Consequently, models with both variables included simultaneously were not considered further.
Age and caloric intake were assessed as potential covariates in the regression models of BMI percentile on dairy product intake and dietary calcium.Neither age nor caloric intake was significant in either model (p= 0.914 and p= 0.838 respectively for dairy intake, and p= 0.852 and p= 0.939 respectively for dietary calcium), but the effects of dairy intake and dietary calcium on BMI percentile remained significant and consistent with the previous models ( = -5.51%/cupper day, p= 0.018; = -0.020%/mgper day, p= 0.017, respectively).Removing underweight subjects, both age and caloric intake were still non-significant in both models, and results were similar to the original regression models with underweight subjects removed (dairy intake: = -4.34%/cupper day, p= 0.055; dietary calcium: = -0.017%/mgper day, p= 0.033).
Gender and weight status subgroup were entered into the regression models both as main effects and interacted with the dairy/calcium variable to assess moderating effects (Table 2).Neither the main effect of gender nor the interaction of gender with the dairy/calcium variable was found to be significant in either of the models (p= 0.793 and p= 0.783 for the main effects in the dairy intake and dietary calcium models, respectively; p= 0.547 and p= 0.605 for the interaction effects in those models).These findings did not change substantially when underweight subjects were removed.Weight status subgroup had significant main effects in both models, and the interaction term was significant in the model with dietary calcium as an independent variable (p= 0.049).For the model including dairy intake, weight status subgroup, and their interaction as independent variables, the interaction approached significance (p= 0.077).In the underweight/normal group higher levels of dairy product intake were significantly associated with lower BMI percentile ( = -5.74%/cupper day, p= 0.001), but no such association was found in the overweight or obese groups (p= 0.984 and p= 0.952, respectively).Similar patterns were observed in the model with dietary calcium.In the normal group, higher levels of dietary calcium were significantly associated with lower BMI percentile ( = -0.021%/mgper day, p< 0.001), but no such association was found in the overweight or obese groups (p= 0.933 and p= 0.917, respectively).Removal of underweight subjects from the analysis had only minor impacts on the results.The interaction of weight status subgroup with dairy intake became nonsignificant (p= 0.281), although tests of effects in each group still found a significant association between dairy intake and BMI percentile in the normal weight group (p= 0.021) and no significant association in the overweight or obese groups.Similarly, the interaction of weight status subgroup with dietary calcium became non-significant (p= 0.221), although tests of effects in each group still found a significant association between dietary calcium and BMI percentile in the normal weight group (p= 0.017) and no significant association in the The inverse relationship of dairy intake and dietary calcium with BMI percentile remained when they were controlled for potential covariates, age and caloric intake. b The inverse relationship of dairy intake and dietary calcium with BMI percentile in the normal/underweight subgroup remained when they were controlled for potential covariates, age and caloric intake.
overweight or obese groups.All of the models incorporating moderating effects of gender and weight status subgroup were run with age and caloric intake as covariates.Inclusion of these terms in the model changed p-values slightly, but all findings of significance were unchanged.
A MANOVA model examined differences between the weight status subgroups in dietary intake measured as daily servings of MyPyramid food groups as well as calories, macronutrients, and micronutrients (Table 3).No significant differences (Pillai's Trace F= 1.379, df= 32 and 168, p= 0.100) were detected in these dietary variables between groups.None of the individual tests of differences between groups were found to be significant.Removing underweight subjects from the analysis had no impact on the MANOVA findings.

DISCUSSION
In addition to the well-established role of calcium in bone and mineral metabolism, recent studies suggest that high calcium intake may lead to reduced body fat accumulation or enhance fat mass loss during energy restriction [8][9][10][11][12].Some [5][6][7][8][9][10][11][12][29][30][31][32][33][34][35][36][37], but not all [13][14][15][16][17][18][19][20][21][22][23][24][25][26], evidence support an inverse relationship between calcium and/or dairy product intake and body fat mass.The majority of the work conducted in this area of inquiry has focused on the adult population; while only a relatively limited number of studies have investigated the relationship in children and adolescents [14].Due to the conflicting published data, limited longitudinal, human clinical trials, and the alarming increase in rates of childhood obesity and the strong role of dietary factors, there is a critical need to better understand the impact of a high calcium and/or dairy diet on body weight regulation, particularly in children.
In the current study, we demonstrated that dairy intake and dietary calcium were inversely related to BMI percentile using a sample of healthy U.S. children, aged 8 to 13 years old.When relationships were assessed in weight status subgroups (normal, overweight, and obese), the inverse association between dairy and calcium with BMI percentile was observed only in normal subjects.Results from the MANOVA test indicated no significant differences in dietary intake among the subgroups.
Results from the current study bear some similarity to other evidence published in the peer-reviewed literature.For example, in a cross-sectional baseline analysis of children aged, 9 to 14 years old, conducted by Rockett et al., it was shown that overweight subjects consumed fewer dairy products than non-overweight children [38].In a study published by Novotny et al., similar results were produced [6].In that analysis, which included Asian and Caucasian early adolescents aged 9 to 14 years old, calcium intake, age, and physical activity inversely correlated with iliac skinfold  [21].In another study of young girls, Lorenzen et al. demonstrated that there was no impact of calcium supplementation on height, body weight, body fatness, or lean mass over a year period [20].And in a recently published review, it was demonstrated that milk and milk products do not negatively affect body weight and composition in children and adolescents [41].
One factor that makes the interpretation of studies difficult in the critical analysis of calcium consumption and body composition is that the primary source of calcium is dairy products.Independent from its calcium content, dairy protein (a macronutrient having high satiety properties) [39] and other dairy components (e.g., caseinomacropeptide or CMP) [40] have been implicated themselves in regulating body composition and food intake or affecting mediators of food intake.In the current study, we demonstrated that both calcium and dairy (separately) were inversely correlated with BMI percentile.
There are several strengths and limitations to this study that need to be addressed.One such strength is the use of free-living subjects.While a study conducted in a controlled environment might yield more accurate results (i.e., capturing and controlling for daily variations in influential lifestyle factors), one might argue that the results of this study better reflect what might occur in response to a public health recommendation to increase dairy calcium aimed at children.The method employed for dietary data collection (24-hour recall, using the USDA Five-Step Multiple Pass Method) is another strength of the current study since it has enhanced recall with reduced respondent burden.In addition to these strengths, there are limitations worth noting.First, a cause and effect relationship can't be established due to the observational nature of this study.The cross-sectional (observation made at single time point) versus longitudinal (multiple observations made over time) design of this study presents an additional limitation.The lack of physical activity data is another study weakness, which introduces a potential unmeasured confounding variable to the current study.Unfortunately, this limitation could not be overcome, since this study was a secondary analysis of an existing dataset.
In summary, we demonstrated an inverse relationship between calcium/dairy intake and BMI percentile in preadolescent children.Interestingly, when weight status subgroups were explored, the relationship was lost in overweight and obese children, but maintained in normal weight children.Results from this study provide evidence for a potential beneficial role of dairy and calcium consumption in achieving a healthy body weight in preadolescent children and the basis for further examination of the relationship between dietary calcium/dairy and body weight and fat mass.Additional longitudinal and/or experimental research (i.e., a randomized controlled trial) is needed to further explore the relationship of calcium/dairy product intake with changes in body weight and fat mass in children and adolescents.