Total Energy Intake and Childhood Overweight

Total Energy Intake and Childhood Overweight: Observational Studies

Longitudinal Studies

There are a total of 14 longitudinal studies were identified from 12.study cohorts of children and adolescents.

Two studies found a significant, positive association between total energy intake and adiposity (Deheeger M, Akrout M., et al 1996, Alexy U, Sichert-Hellert W et al, 1999). Both of these studies earned a neutral rating
Two studies (Robertson SM, Cullen KW, et al 1999, Berkey CS, Rockett HRH, et al 2000) concluded with mixed results. Robertson SM, Cullen KW, et al 1999 found a significant positive association between total energy in the year prior to the adiposity take-off [LDR1] and adiposity, and a non-significant association between the average intake of total energy over the study duration and adiposity. This study earned a neutral rating. Berkey CS, Rockett HRH, et al 2000 found a positive association from year 1 to year 2 for girls but not for boys.
Nine studies found no significant association between total energy intake and adiposity. Eight of these studies earned a positive rating, one earned a neutral rating.
One study (Gazzaniga JM & Burns TL. 1993) found a negative association between total energy intake and percentage body fat. This small study (n=48) received a positive quality rating.

US Nationally Representative Cross-sectional Studies

Only one US nationally representative cross-sectional study was identified (Troiano RP, Briefel RR et al 2000, positive quality rating) that examined the association between total energy intake and adiposity in children ages 2 to 19 years. This study found no significant association between total energy intake and adiposity.

Other Cross-sectional Studies

28 non-US nationally representative cross-sectional studies were identified. Two of these were cross-sectional results reported from 2 of the longitudinal studies discussed above.

Only 3 of the 28 studies (Rodriguez-Artalejo F, Garces C, et al, 2002, Gillis LJ, Kennedy LC, et al, 2002, Gillis LJ, Bar-Or O, 2003) found significant positive associations between total energy intake and adiposity in children. All three of these studies earned positive ratings.
Three of the 28 studies concluded with mixed results:
- One found no significant association in boys, but a significant inverse association in girls (Lluch A, Herbeth B, et al, 2000).
- Another study (Stewart KJ, Seemans C, et al, 1999) found an inverse association among both sexes when the sum of skinfold thickness measures were used as the measure of adiposity, and no significant association when BMI was used as the measure of adiposity;
Seventeen of the 28 studies found no significant association between total energy and adiposity. Seven of the 16 studies that found no significant association earned positive quality ratings. The remaining nine studies earned neutral ratings.
Seven of the 28 studies found significant inverse associations, suggesting that overweight and obese children were consuming less total energy than lean and non-overweight children. Four of the 7 studies that found significant inverse associations earned positive quality ratings.. The other three earned a neutral rating.

Finally, one of the 28 studies (Johnson-Down L, OLoughlin J, et al, 1997) concluded that it was not appropriate to study the relationship between diet and BMI, given the substantial underreporting that was detected among the heavier children. This study earned a + quality rating.

Summary

A total of 43 observational studies examined either the longitudinal or cross-sectional association between total energy intake and some measure of adiposity in children. These studies varied substantially in sample size, with the largest including 10,769 children (Troiano RP, Briefel RR et al 2000) and the smallest including 13 (Manos TM, Gutin B, et al, 1993), and they also varied in terms of geographical location (25 were conducted in the US and 18 outside the US). Two of these studies (Obarzanek E, Schreiber GB, et al. 1994, Manos TM, Gutin B, et al, 1993) examined this association only in girls, while the other 34 studies examined this association among both boys and girls. A majority of these studies were conducted among non-Hispanic white populations.

There is little consistency between studies in the methods used for dietary intake. Dietary intake methods include 3- to 5-day food records, multiple 24-hour recalls, food frequency questionnaires, and diet histories. Such assessment tools are limited in their capacity to measure actual energy intake. Common methodological limitations include the underreporting of total intake, and in children there is a unique limitation when intake is reported by a parent or caregiver who may inaccurately report children’s intake. Thus, current dietary assessment tools may not accurately identify children’s true energy intake and may not detect the small differences in energy intake that can lead to gains in adiposity over time, above and beyond that required for normal growth.

Total Energy Intake and Childhood Overweight Overview Table

View table in new window

Study and Group Name of Study	Quality Rating of Article	Statistical Analysis	Control Variables	Duration of Study Sample Size Sex of Population Studied Age Race Location of Study	Measurements	Total Energy Intake Associations
Cohort Designs
Berkey CS, Rockett HRH, et al 2000 , girls Growing Up Today Study (offspring of Nurses; Health Study II participants)	Positive	Linear regression analysis	Race/ethnic group, Baseline BMI, Annual change in height, Menstrual history in girls, Tanner stage & Age. 1 year change in BMI was adjusted for time lag between the 2 returned questionnaires; fat & fiber intakes were energy-adjusted. All analysis performed separately by gender.	1 year 6149 Girls Child (6-12 years), Adolescent (13 years to 18 years) African American, Hispanic, Asian, Caucasian, Native American, Other USA	FFQ	Increase in total energy intake (yr 1 to yr 2): Positive association
Berkey CS, Rockett HRH, et al 2000 , boys Growing Up Today Study (offspring of Nurses’ Health Study II participants)	Positive	Linear regression analysis	Race/ethnic group, Baseline BMI, Annual change in height, Menstrual history in girls, Tanner stage & Age. 1 year change in BMI was adjusted for time lag between the 2 returned questionnaires; fat & fiber intakes were energy-adjusted. All analysis performed separately by gender.	1 year 4620 Boys Child (6-12 years), Adolescent (13 years to 18 years) African American, Hispanic, Asian, Caucasian, Native American, Other USA	FFQ	Increase in total energy intake (yr 1 to yr 2): No association
Newby PK, Peterson KE et al, 2003	Positive	Linear regression	Sex, Baseline age, Weight, Total energy intake, Change in height, Birthweight, Maternal education, Race and ethnicity, Residence, Poverty level	1 year 1379 Boys Pre-school child (2-5 years) African American, Hispanic, Asian, Caucasian, Native American North Dakota, USA	BMI, dietary intake	Total energy to weight change: No association
Scaglioni S, Agostoni C, et al, 2000	Positive	Student t-tests, Non-parametric Wilcoxon & Mann-Whitney tests& Multiple logistic analysis.	Gender, Weight & Length at birth & at 1 y of age, Parental age.	4 years 147 Boys, Girls Infant (0-2 years), Pre-school child (2-5 years) Caucasian Milan, Italy	Overweight status at age 5, measured ht & wt (Overweight: BMI > 90th%). FFQ followed-up with 24-hr recall to confirm serving sizes	Total energy intake: No association
Klesges RC, Klesges LM, et al 1995	Positive	Multiple regression analysis.	Baseline BMI, Gender, Age, Family risk, Gender by familial risk interaction, Baseline % kcal as fat, aerobic activity, change (y 2-3) in % kcal as fat, Change (y 2-3) in leisure activity.	2-years 146 Boys, Girls Pre-school child (2-5 years) Caucasian, Other Memphis, TN, US	Change in BMI, measured ht & wt, FFQ completed collaboratively by parent & child	Total Energy baseline level & change from yr 2 - 3: No association
Davison KK, Birch LL 2001 IJO	Positive	Hierarchical regression	Girls & Parents total energy intake was adjusted for body weight, Family income, Parent education status, Girls BMI at age 5 years	2 years 142 Girls Pre-school child (2-5 years), Child (6-12 years) Caucasian Pennsylvania, USA	Measured ht & wt difference between age 5y & 7y, mothers provided 3, 24-hr recalls of their daughters food intake, mothers completed the shortened version of the Childrens Physical Activity Scale (CPA), Semi-quantitative food frequency questionnaire	Total energy: No association
Maffeis C, Talamini G, et al 1998 long	Positive	Multiple regression analysis & Multivariate logistic regression analysis.	Age, Gender, Puberty development (clinically assessed), Parents’ BMI.	4 years 112 Boys, Girls Child (6-12 years) Caucasian Italy	Obesity status at baseline, measured ht & wt (Obesity = rel BMI > 120%; rel BMI = BMI/BMI at 50th % for age & gender). Diet History obtained through interview with mothers & children	Energy intake: No association
Rolland-Cachera MF, et al 1995	Positive	Correlation coefficients	Baseline energy intake (at age 2 years), Baseline BMI (at age 2 years), Social class (fathers occupation), Parental BMI (self-reported by parents).	7 years 112 Boys, Girls Infant (0-2 years), Pre-school child (2-5 years), Child (6-12 years) Not Specified Paris, France	BMI at age 8, measured ht & wt & Subscapular & Tricep skinfolds at age 8 Dietary History method with info collected from childs mother & verified with 24-hr recall	Energy intake: No association
Skinner JD, Bounds W et al 2004	Positive	forward selection stepwise regression analysis, anova, correlation analyses	None specified	7 years 70 Boys, Girls Infant (0-2 years), Pre-school child (2-5 years), Child (6-12 years) Caucasian Tennessee, USA	Height (or length), weight, 24 hour recall	Total energy intake: No association
Deheeger M, Akrout M., et al 1996	Neutral	ANOVA & Correlation tests	Not specified	7 years 112 Boys, Girls Infant (0-2 years), Pre-school child (2-5 years), Child (6-12 years) Not Specified Paris, France	BMI at age 8 y, measured ht & wt (3 tertiles of the BMI distribution: lean, medium & fat), Diet History confirmed with 24-hr Recall	Daily Total Energy Increase between the ages 4 & 6 y: Positive association
Robertson SM, Cullen KW, et al 1999 Studies of Child Activity and Nutrition (SCAN)	Neutral	ANOVA	Matched on age, ethnicity & gender for the same year of data collection and 3 were randomly selected as control subjects per case	3 years 48 Boys, Girls Pre-school child (2-5 years), Child (6-12 years) African American, Hispanic, Caucasian Texas, US	Adiposity Take-off, Sum of 7 skinfolds increased by 1.5 standard deviations or more in any year of the study.	Total energy: No association
Bogaert N, Steinbeck KS et al 2003	Neutral	Unpaired t-tests, analysis of covariance, correlations	History of breastfeeding, meal pattern, food consumed away from home, parental macronutrient intake (including fatty acids), childrens lipid subfractions	1 year 41 Boys, Girls Child (6-12 years) Not Specified Australia	Body mass index, Body composition (bioelectrical impedance analysis), Body fat (skinfold test)	Energy intake: No association
Case Control
Tanasescu M, et al 2000	Neutral	Multiple logistic regression	Gender, Mother’s BMI, Mother’s marital status, TV viewing, Fruit juice intake (others tested, but not included in final model)	Not specified 53 Boys, Girls Child (6-12 years) Hispanic Hartford, CT, USA	Obesity (BMI >85th percentile) Dairy (71-item FFQ classified into 11 food groups change over previous year); calcium (4-stage 24-hour recall), Fruit juice not clear if 100% fruit juice only (FFQ change over previous year), Soda, Fruit (excluding juice) vegetable	Total energy: No association
Bandini LG, Vu D, et al 1999	Neutral	Students t-test	Gender, Underreporting	43 Boys, Girls Adolescent (13 years to 18 years) Not Specified Boston, MA, USA	Obesity (% body fat), Sweetened soda (14-day diet record) Fruit (excluding juice), vegetable (excluding potatoes) (14-day diet record)	Total Energy: Negative association
Manos TM, Gutin B, et al, 1993	Neutral	Student t-test	none	13 Girls Child (6-12 years) African American Georgia, US	Obesity status (BMI > 85th%) 5-consecutive days of 24-hr Recalls completed by girls & their mothers	Energy intake: No association
Longitudinal Study
Wang Y, Ge K, Popkin BM, 2003 China Health and Nutrition Survey	Positive	Analysis of variance (4 groups: overwt to overwt; overwt to non-overwt; non-overwt to overwt; non-overwt to non-overwt), Cochran-Mantel-Haenszel test	Age, Gender	2 years 95 Boys, Girls Child (6-12 years), Adolescent (13 years to 18 years) Asian China	Overweight (BMI>25) Fruit + vegetables (3-day diet record)	Total energy intake and BMI: No association
Gazzaniga JM & Burns TL. 1993 Muscatine Coronary Risk Factors Project	Positive	Pearson correlation coefficients & Multiple-regression analysis.	Gender, REE, Energy expended for physical activity, Age, Body weight.	48 Boys, Girls Child (6-12 years) Not Specified Iowa, US	% Body Fat estimated by taking average of two skinfold thickness measurements, triceps & subscapular & subjects sex, age, ht & wt, 3-consecutive 24-h Recalls	Total daily energy: Negative association with % body fat
Alexy U, Sichert-Hellert W et al, 1999 Dortmund Nutritional and Anthropometri-cal Longitudinal Designed Study (DONALD), 1990-1997	Neutral	Pearson correlation coefficients to test interactions between fruit juice intake and anthropometric indices; Wilcoxon rank sum test to test differences between groups.	Not specified.	3 years 205 Boys, Girls Pre-school child (2-5 years) Not Specified Germany	BMI, Weight gain 100% fruit juice >12 fl oz/day (3-day weighed diet record)	Intake of energy and BMI: Positive association (r=0.18, P<0.05)
Eck LH, Klesges RC, et al. 1992	Neutral	ANOVA	Cohort separated into high & low-risk groups based on parental overweight.	1 year 187 Boys, Girls Pre-school child (2-5 years) Caucasian TN, US	Weight change over a 1-year period, measured height & weight, FFQ modified for use with children completed collaboratively by parents and child	Total energy: No association
Nationally Representative Cross-Sectional Study
Troiano RP, Briefel RR et al 2000 NHANES III	Positive	Not described	Age, Gender, Ethnicity	10371 Boys, Girls Pre-school child (2-5 years), Child (6-12 years), Adolescent (13 years to 18 years) African American, Hispanic, Caucasian USA	Overweight (not specified), Soda calories (24-hour recall)	Total Energy: No association
Other Cross-Sectional Study
Obarzanek E, Schreiber GB, et al. 1994 National Heart, Lung, and Blood Institute Growth & Health Study (NGHS)	Positive	Multivariate regression analysis	Race/ethnicity, Age, Income, Level of education	2147 Girls Child (6-12 years) African American, Caucasian Berkeley, CA, Cincinnati, OH & Rockville, MD	BMI measured ht & wt & Sum of Skinfold measures 3-consecutive day Food Records recorded by girls	Total energy intake: No association
Kelishadi R, Pour MH et al 2003	Positive	Multiple linear regression		2000 Boys, Girls Child (6-12 years), Adolescent (13 years to 18 years) Other Isafahan and Arak, Iran	BMI, Fruits and vegetables (3-d diet record and FFQ)	Total energy intake: No association
Bao W, Nicklas TA, et al, . 1996 Bogalusa Heart Study	Positive	Spearman correlation	Race, Gender	1419 Boys, Girls Child (6-12 years) African American, Caucasian Louisiana, US		Total energy intake (with fat free mass): Positive association; Total energy intake (with body fat): No association; Total energy intake (with weight): weak Positive association
Garaulet M, Martinez A., et al 2000	Positive	ANOVA & Pearsons simple correlation	Gender	1331 Boys Adolescent (13 years to 18 years) Hispanic Spain	BMI, Measure height & weight, (Overweight: BMI > 23; Normal-weight: BMI < 23), 7-consecutive days Food Records	Total energy intake: Negative association (though overweight tended to under report--so association is questionable)
Rodriguez-Artalejo F, Garces C, et al, 2002	Positive	Spearman correlation coefficient	Birth weight. Children were selected through random cluster-sampling of schools, stratified by sex & socio-economic level	1112 Boys, Girls Child (6-12 years) Not Specified Spain	Ponderal index, BMI, Overweight = BMI > 17.6; Obesity = BMI > 20.1, measured ht & wt FFQ completed by childrens mothers	Total energy: Positive association
Guillaume M, Lapidus L, 1998	Positive	ANOVA & Multiple regression analysis.	Age, Gender.	955 Boys, Girls Child (6-12 years) Not Specified Province de Luxembourge, Belgium	BMI, measured ht & wt. 3-day Food Record completed by children with assistance of teachers & parents	Total energy: no association (boys and girls)
Lluch A, Herbeth B, et al, 2000 Stanislas Family Study	Positive	ANOVA & Pearsons correlation analysis.	Gender & Age	560 Boys, Girls Adolescent (13 years to 18 years) Not Specified France	Relative weight, measured ht & wt, (calculated using Lorentzs index (ideal weight), taking into account gender), 3-day Food Record completed by childs mother with childs help	Energy intake and overweight: Negative association (girls) Energy intake and overweight: No association (boys)
Maffeis C, Provera S et al 2000	Positive	Multiple regression analysis, Kruskal-Wallis test	Gender, Parents BMI, Energy intake/BMR ratio (index of validity for self-report diet), Fat intake (many tested for inclusion in final model)	530 Boys, Girls Child (6-12 years) Asian, Caucasian, Other Italy	% body fat BMI Relative body wt status: overweight 110-120%, obese >120% % total energy intake from snacks (self-defined; diet history)	Energy intake/BMR: Negative association
Johnson-Down L, OLoughlin J, et al, 1997 Canadian Heart Health Initiative project called Coeur en sante St. Louis du Parc	Positive	Multivariate logistic regression analyses.	None	498 Boys, Girls Child (6-12 years) African American, Asian, Caucasian Montreal, Canada	BMI % measured ht & wt; Overweight > 85th %, 24-hr Recall)	Authors concluded it was not appropriate to study the relationship of diet and BMI given the underreporting among the heavy children
Stewart KJ, Seemans C, et al, 1999 FRESH (Food Reeducation for Elementary School Health)	Positive	Pearsons product moment correlation	Because there was no substantial differences between males & females & among the racial groups, their data were combined	468 Boys, Girls Child (6-12 years) African American, Caucasian, Other Baltimore, Maryland, US	BMI, measured ht & wt & Sum of triceps & subscapular skinfolds sites combined 24-hr Diet Food Logs recorded by children	Total daily energy intake: Negative association
Gillis LJ, Bar-Or O, 2003	Positive	T- tests (Weight status), Stepwise multiple regression (% body fat)	Age, Gender, SES, Energy intake, Percentage of Recommended Nutrient Intake for age and gender, fat in grams and as a percentage of Health and Welfare Canada Guidelines for age, fat as a percentage of total energy, sugar in grams and as a percentage of total energy, intake from each good group and as a percentage of total energy [above used only in analysis of % body fat]	181 Boys Child (6-12 years), Adolescent (13 years to 18 years) Caucasian Canada	% Body fat Cheese (dietary history) Sweetened beverages (Diet history)	Total energy: Positive association
Gillis LJ, Kennedy LC, et al, 2002	Positive	Linear regression analysis.	Gender, Age. Non-obese subjects were recruited by stratifying age, gender and socioeconomic background.	181 Boys, Girls Child (6-12 years), Adolescent (13 years to 18 years) Not Specified Ontario, Canada		Total calories: Positive association; Energy as a percent of required kcal for age and sex: Positive association
Dennison BA, Rockwell HL et al, 1997	Positive	General linear models	Age, Gender, Maternal height, Child age-gender interaction	168 Boys, Girls Pre-school child (2-5 years) Caucasian, Other Schoherie county, rural upstate NY, US	Obesity (BMI or ponderal index >90th percentile), 100% fruit juice >12 fl oz/d (7-consecutive days of Dietary Records)	Total energy: No association
McGloin AF, Livingstone MBE, 2002	Positive	ANOVA & Regression analysis.	Gender	114 Boys, Girls Child (6-12 years) Caucasian, Other Northern Ireland, UK	Obesity status (BMI > 95th% according to British growth standards) low-risk (LR) (child with two lean parents) or high-risk (children with at least one biological parent with BMI > 29.5), measured ht & wt & Body fatness (difference bt body weight & lean mass) 7-day Weighed Dietary Records	Total energy intake: No association
Maffeis C, Pinelli L, et al, 1996	Positive	Pearson product moment linear correlation coefficient & ANCOVA.	RMR (indirect calorimetry).	82 Boys, Girls Child (6-12 years) Caucasian Verona, Italy	BMI, measured ht & wt Obesity: BMI > 97th % of reference values for age & sex. Diet History completed by mothers & children	Total energy: Negative association
Atkin Lisa-Marie, Davies PS 2000 Feasibility Study for National Diet and Nutrition Survey of Children	Positive	Pearson’s correlation coefficient, ANOVA & Multiple regression analysis.	Gender	77 Boys, Girls Pre-school child (2-5 years) Not Specified UK	% Body Fat (calculated from fat mass & body weight, body composition assessed by measurement of total body water using oxygen-18 dilution method), 4-day Weighed Food Records	Total Energy: No association
Albertson AM, Anderson GH et al 2003	Neutral	Analysis of variance, logistic regression, Wald chi-square test	Age	603 Boys, Girls Pre-school child (2-5 years), Child (6-12 years) Caucasian, Other USA		Energy intake and BMI: No association
Kelley C, Krummel D et al 2004	Neutral	Student’s t-test, chi-square analysis	None	297 Boys, Girls Child (6-12 years) Caucasian West Virginia, USA	BMI, Dietary intake (food frequency questionnaire)	Total energy: No association
Tucker LA, Seljaas GT, 1997	Neutral	Regression analysis & Partial correlation analysis	Gender, Total energy intake, Physical fitness (run/walk test), Parental body mass (self-reported ht & wt).	253 Boys, Girls Child (6-12 years) Not Specified Utah, US	Body fat % (formula using skinfold measurements). FFQ completed collaboratively by parents & children	Energy intake with % body fat: Negative association
Hanley AJG, Harris SB et al, 2000	Neutral	Multiple logistic regression	Age, Gender	242 Boys, Girls Pre-school child (2-5 years), Child (6-12 years), Adolescent (13 years to 18 years) Native American Sandy Lake First Nation, Canada	Overweight (BMI > 85th percentile), Vegetable (34 item FFQ into 7 groups change over previous 3 mo), 24-hr Recall)	Total energy: No association
Gray and Smith 2003 Healthy U Project	Neutral	1-way ANOVA and pearsons correlations for normally distributed data Mann-Whitney or Kurskal-Wallis test and Spearman Correlation for non-normally distributed data	None	155 Boys, Girls Child (6-12 years), Adolescent (13 years to 18 years) Native American	Fitness (Presidential Fitness: Flexibility-v-sit; endurance-1/4, 1/2, 1 mile run; agility-shuttle run; ab strength-partial curl ups; upper body strength-push ups; balance-one leg stand; BMI (height and weight); skinfolds (4 site); mid arm circumference	Energy intake and BMI: No association
Arluk SL, Branch JD, 2003	Neutral	Chi-squared & Logistic regression.	none	98 Boys, Girls Child (6-12 years) African American, Hispanic, Asian, Caucasian, Other Virginia, US	BMI %, self-reported ht & wt; Obesity > 95th %.	Total energy and BMI: No association
Ortega RM, Requejo AM et al, 1995	Neutral	T-test for comparison of cases vs. controls; ANOCOVA for association with BMI	Gender	64 Boys, Girls Adolescent (13 years to 18 years) Not Specified Madrid, Spain	Overweight (BMI >23) Dairy (5-day weighed food record and FFQ) Fruit, vegetable (5-day weighed food record)	Total energy intake: No association
Koivisto U, Fellenius J, et al, 1994	Neutral	Test for Correlation	Age, Gender	50 Boys, Girls Pre-school child (2-5 years), Child (6-12 years) Not Specified Sweden	Weight-length index (Overweight > 109)	Total Energy: No association
Rocandio AM, Ansotegui L, 2001	Neutral	Chi-squared tests & Student t-tests.	none	32 Boys, Girls Child (6-12 years) Not Specified Spain	Weight-for-height, measured ht & wt (overweight > 90th % compared with 1985 Spanish reference values) 7-day Weighed Food Records	Energy intake/day: Negative association
Francis CC, Bope AA, et al 1999	Neutral	Matched pair analysis and Wilcoxon signed rank tests	Children were matched for age, gender, weight	24 Boys, Girls Child (6-12 years) Caucasian Texas, USA	% Abdominal fat & Fat-free mass, (Dual-energy x-ray absorptiometry), 7-day Food Diaries complied by mothers	Total Energy: No association