[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
January 2008

Family Dietary Coaching to Improve Nutritional Intakes and Body Weight ControlA Randomized Controlled Trial

Author Affiliations

Author Affiliations: Nutri-Health, Rueil-Malmaison, France (Mr Paineau and Drs Zourabichvili and Bornet); Department of Internal Medicine, Poissy St-Germain-en-Laye Hospital, St-Germain-en-Laye, France (Dr Beaufils); Department of Nutrition and Functional Investigations, INSERM U695, Bichat-Claude Bernard Hospital, Paris, France (Dr Boulier); Department of Nutrition, EA3502, Hôtel-Dieu, Paris, France (Dr Chwalow); INRA-CORELA, Ivry-sur-Seine, France (Dr Combris); INSERM E211, Tours Hospital, Tours, France (Dr Couet); Department of pediatrics, Children's Hospital, Toulouse, France (Drs Jouret and Tauber); Agence Française de Sécurité Sanitaire des Aliments, Maisons-Alfort, France (Dr Lafay); CRNH Rhône-Alpes, Lyon 1 University, Hospices Civils de Lyon, INSERM U870, INRA U1235, Lyon, France (Dr Laville); French Ministry of Research, Paris, France (Dr Mahe); Department of pediatrics, Necker Hospital, Paris, France (Dr Ricour); Department of Nutrition, EA2694 Lille 2 University, Lille, France (Dr Romon); Louis Pasteur University, Medical Faculty, EA 1801, Strasbourg, France (Dr Simon); Department of Endocrinology, CRNH Ile-de-France, Jean Verdier Hospital, Bondy, France (Dr Valensi); and Quanta Medical, Rueil-Malmaison, France (Ms Chapalain and Dr Zourabichvili). Dr Cassuto is in private practice in Paris, France.

Arch Pediatr Adolesc Med. 2008;162(1):34-43. doi:10.1001/archpediatrics.2007.2

Objective  To test the hypothesis that family dietary coaching would improve nutritional intakes and weight control in free-living (noninstitutionalized) children and parents.

Design  Randomized controlled trial.

Setting  Fifty-four elementary schools in Paris, France.

Participants  One thousand thirteen children (mean age, 7.7 years) and 1013 parents (mean age, 40.5 years).

Intervention  Families were randomly assigned to group A (advised to reduce fat and to increase complex carbohydrate intake), group B (advised to reduce both fat and sugar and to increase complex carbohydrate intake), or a control group (given no advice). Groups A and B received monthly phone counseling and Internet-based monitoring for 8 months.

Outcome Measures  Changes in nutritional intake, body mass index (calculated as weight in kilograms divided by height in meters squared), fat mass, physical activity, blood indicators, and quality of life.

Results  Compared with controls, participants in the intervention groups achieved their nutritional targets for fat intake and to a smaller extent for sugar and complex carbohydrate intake, leading to a decrease in energy intake (children, P < .001; parents, P = .02). Mean changes in body mass index were similar among children (group A, + 0.05, 95% confidence interval [CI], − 0.06 to 0.16; group B, + 0.10, 95% CI, − 0.03 to 0.23; control group, + 0.13, 95% CI, 0.04-0.22; P = .45), but differed in parents (group A, + 0.13, 95% CI, − 0.01 to 0.27; group B, − 0.02, 95% CI, − 0.14 to 0.11; control group, + 0.24, 95% CI, 0.13-0.34; P = .001), with a significant difference between group B and the control group (P = .01).

Conclusions  Family dietary coaching improves nutritional intake in free-living children and parents, with beneficial effects on weight control in parents.

Trial Registration  clinicaltrials.gov Identifier: NCT00456911

Public health strategies to prevent obesity include nutritional recommendations to limit intake of fats and sugars and to increase intake of complex carbohydrates.1 Although these recommendations are widely used, insufficient evidence is available to support their feasibility, sustainability, or even their efficacy. If low-fat diets have often been associated with weight loss,24 long-term compliance with marked fat reductions is questionable, as fats enhance palatability.5,6 A lower consumption of sugar-sweetened soft drinks may help to prevent overweight,7,8 but high-carbohydrate diets have also been associated with better weight control.9,10 A significant increase in complex carbohydrates is difficult to sustain because it leads to major changes in food habits.11

Overall, available intervention studies on weight control are often short-term and concern small cohorts.12 They mostly investigate the effects of concurrent changes in diet, physical activity, and behavior, thus making causal inference between nutritional changes and clinical benefits impossible. In this context, well-designed studies targeting only dietary modifications are needed to better understand the acceptable levels of nutritional changes and their impacts on health.

As general messages have shown little efficacy in changing dietary habits,13 these studies should be based on specific methods, such as individualized coaching,14 Internet-based monitoring,15 and home food delivery.8 A family approach as well as actions in schools for children16 may also induce sustainable dietary changes.17 In the Etude Longitudinale Prospective Alimentation et Santé (ELPAS) study, we hypothesize that family dietary coaching for one school year will allow a nutritional shift toward following recommendations and improve weight control in free-living (noninstitutionalized) children and parents.


One thousand thirteen families were included in this 10-month, parallel, randomized intervention trial. In each family, one second- or third-grade pupil (aged 7-9 years) and one of his or her parents participated. This children's age was chosen for 2 reasons: (1) this age corresponds to the period following the adiposity rebound, when overweight and obesity may appear,18 and (2) children of this age are receptive to dietary interventions,19 because they have passed the phase of dietary neophobia.20 Volunteers were recruited from 54 elementary schools in Paris, France, from March 2005 through June 2005. A mailing was performed in July 2005 to complete the recruitment with families from nonparticipating schools. All families were informed of the general nature of the intervention (a study on food and health) but were unaware of the primary hypothesis, eg, that nutritional changes would affect body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared). Inclusion criteria were being a second- or third-grade pupil, having an affiliation with the French health care system, and providing written informed consent. Because our study was designed to test nutritional targets for the general population and to avoid discrimination between children, no inclusion criteria were based on pathologic, ethnic, or social/educative indicators. Ethical approval was given for the study by the ethics committee of Poissy St-Germain-en-Laye Hospital, St-Germain-en-Laye, France.


Families were randomly assigned to 1 of the following groups (Figure 1):

Figure 1.
Study design and participant flow. BMI indicates body mass index; asterisk, categories in which exclusions are known are presented, more than 1 reason could be given for exclusion.

Study design and participant flow. BMI indicates body mass index; asterisk, categories in which exclusions are known are presented, more than 1 reason could be given for exclusion.

  1. Group A (297 families) received advice on how to reduce dietary fats (< 35% of total energy intake) and how to increase complex carbohydrates (> 50% of total energy intake).

  2. Group B (298 families) received advice on how to reduce both dietary fats (< 35% of total energy intake) and sugars (− 25% of initial crude intake) and how to increase complex carbohydrates (> 50% of total energy intake).

  3. The control group (418 families) received no dietary advice.

Groups A and B were given explanations on how to increase complex carbohydrate intake to maintain an isocaloric diet. In accordance with current nutritional recommendations, advice for fats targeted saturated fatty acids, and advice for sugars targeted extrinsic sugars.

We performed a school-based randomization to avoid contact between volunteers from different groups. Schools were stratified by district, status (school participating or not in the study), and number of participants in each school to ensure that all 3 groups would be homogeneous with regards to social/educative characteristics and recruitment methods. Randomization was performed according to a computer-generated randomization list (PROC SURVEYSELECT; SAS Institute Inc, Cary, North Carolina).


Computer-based interventions have proved efficient in previous trials.21,22 Therefore, to enhance participant motivation and monitoring, we developed a Web site dedicated to the study (http://www.elpas.fr). About half of the families, those who did not have a computer and/or Internet connection, were given adequate equipment at inclusion. The ELPAS Web site gave personal access to the study's self-administered questionnaires (diet, physical activity, meal preparation, and quality of life) along with updated information, an individual and interactive agenda, an email address, and various other functions. The online food questionnaire included 3425 food items and 700 portion-size pictures (NutriXpert; Medical Expert System, Paris, France); before initiation of the study, the families were instructed how to use the computer program. They then performed 3-day dietary records (2 weekdays including Wednesday, and one weekend day) monthly from September 2005 to June 2006. Energy and macronutrient intakes were calculated using computerized national food composition tables23 completed with additional food items. Adult underreporters were identified using the Schofield equation24 with the Goldberg cutoff limits25 (1.04 × basal metabolic rate). At baseline, these underreporters were called by the dieticians who checked for omission and errors.

All families completed the baseline evaluation September 2005 through October 2005. From November 2005 through June 2006, families in the intervention groups received monthly telephone counseling by a trained dietician. The dietary intervention thus occurred for 8 months. Telephone calls, which were about 30 minutes long, were dedicated to analyzing food habits of the participants according to their last food records and determining pragmatic advice to reach their specific dietary targets. Advice included reduction in portions and/or frequency of consumption as well as food substitutions and meal preparation modifications. Prior to the study, dieticians were instructed by experts on how to optimize dietary advice by developing an original approach based on analysis of the subject's social/educative and psychological characteristics. In addition to telephone counseling, children and parents received monthly newsletters. Families were also invited to a series of events (eg, conferences, museum visits), and 3 lessons on nutritional education were programmed in participating schools. To test for a causal inference between nutritional changes and clinical outcomes, no specific recommendations were made regarding exercise.

Participants in the control group received general information about nutrition, but no individualized advice, to maintain motivation and to avoid a high dropout level. They were followed at the same intervals as participants in the intervention groups and were asked to record their diets in an identical fashion.


All anthropometric measurements were performed by trained staff, blinded to the experimental design, at baseline (September-October 2005) and at the end of the intervention (May-June 2006) in both children and parents. Height without shoes was measured to the nearest 0.1 cm using a portable stadiometer and weight in underwear was measured to the nearest 0.1 kg on a portable electronic scale. Body mass index was then calculated.26,27 In children, the age-adjusted z score for BMI was calculated with the following equation28: z = [(BMI/M)L – 1]/LS, where L, M, and S, respectively, represent power transformation, median, and coefficient of variation. These values were collected from the appropriate table corresponding to the sex and age of the child.29 We defined overweight children as having BMIs greater than or equal to the age-specific and sex-specific international cutoff points,30 and classified overweight adults on the basis of World Health Organization references.31 Chest, waist, hip, and knee circumferences were measured to the nearest 0.1 cm according to established guidelines32 using measuring tape. Fat mass (FM) was determined by means of a single-frequency (50-kHz) leg-to-leg bioelectrical impedance analyzer (Tefal, Rumilly, France), as described elsewhere.33 Prior to the study, we validated this equipment in 55 French children using deuterium dilution and air-displacement plethysmography; we obtained a good correlation for FM measurement between bioelectrical impedance analyzer and the reference method (adjusted R2 = 0.8) (D.P., unpublished data, 2003-2004). Blood pressure and heart rate were measured using a fully automatic blood pressure monitor.

Fasting blood samples were collected in parents only, at baseline and at the end of intervention. Blood for glucose and insulin analysis was collected in tubes containing iced ethylenediaminetetraacetic acid prepared with sodium fluoride. Serum was separated and aliquots were frozen at − 80°C in a deep-freeze system equipped with a central security alarm. Serum glucose, insulin, triglycerides, total cholesterol, and high-density lipoprotein cholesterol levels were analyzed with an automated instrument using standard kits. Low-density lipoprotein cholesterol and insulin resistance index were calculated according to methods by Friedewald et al34 and Matthews et al,35 respectively. Aliquots were kept at − 80°C for further analysis.

Overall physical activity was assessed at baseline and at the end of intervention using previously validated physical activity questionnaires (Modifiable Activity Questionnaire in children36,37 and the Multinational Monitoring of Trends and Determinants in Cardiovascular Disease Optional Study of Physical Activity in parents36,38). The children's questionnaire was administered by trained staff, and the parent's questionnaire was self-administered. In parents only, food-related quality of life was assessed at baseline, at the halfway point of the intervention, and at the end of the intervention using a specific self-administered questionnaire developed and validated for our study. This questionnaire was created from several existing scales3942 and underwent linguistic and psychometric validation using the Rasch model.43 Socioeconomic indicators were collected at the end of the intervention using a self-administered questionnaire.44


The sample size calculation for this study was based on the previously reported changes in BMI among French children45 and adults.46 We compared each intervention group with the control group to test 2 public health strategies to improve weight control. With a 2-sided .05 significance level (α = .05) and BMI change as the primary outcome measure, studying 236 children (162 adults) in each intervention group and 334 children (229 adults) in the control group makes it possible to detect a significant difference at 80% power. To account for an expected 20% dropout rate, we decided to recruit 295 families in each intervention group and 420 families in the control group (total 1010 families). We intentionally allocated a higher number of families in the control group to maintain the global type 1 error (α) at the .05 level.

Statistical analyses were all conducted using the SAS statistical program, version 8.2 (SAS Institute). All analyses were completed on an intent-to-treat basis, excluding only subjects with no value for BMI neither at baseline nor at the end of intervention (Figure 1). We used adequate procedures for adjustment of a type 1 error, such as the Dunnet method for variance and covariance analyses.47 Missing data for BMI were imputed using the mean value in the whole cohort.

Baseline comparability of intervention and control groups was assessed using analysis of variance with one factor (group) for continuous variables and with χ2 or Fisher exact tests for categorical variables. Differences between groups in changes from baseline to end of intervention (final value – initial value) were investigated using analysis of covariance with the baseline value as cofactor (PROC GLM; SAS Institute). When analysis of covariance indicated significant differences between the intervention groups and the control group (P < .05), comparisons were made between each intervention group and the control group. Nonparametric analyses of variance and covariance by ranks were performed when data showed abnormal distribution (Kolmogorov-Smirnov test) and/or nonhomogeneity of variances (Levene test). Baseline characteristics are reported as mean (SD). Changes throughout the study are reported as mean with 95% confidence intervals (CIs).


Participant characteristics at baseline are presented in Table 1. More than 80% of the parents were women. At baseline, there were no significant differences between groups for anthropometric indicators. A difference was found for age in children between group B and controls (respectively, 7.83 [0.64] vs 7.69 [0.64] years, P = .01). The baseline prevalence of overweight (including obesity) was 18% in children and 33% in parents, with no differences between groups. These rates agree with available data from French observational studies.48,49 Of the baseline sample, 84.8% (859 families) completed the study, indicating a dropout rate of 15.2%, with no significant difference in the percentage of dropout between groups (P = .46). The main reason for dropout was lack of time to complete the dietary records and lack of motivation. Most dropouts occurred in the first 4 months of the study. Those who did and did not complete the study did not differ for sex or initial BMI. In parents, dropouts were statistically younger than study completers, but the difference was very limited (mean age, 39.4 vs 40.6 years, P = .02). After imputation, complete data for BMI were available for 949 children and 947 parents.

Table 1. 
Baseline Characteristics of the 1013 Study Participantsa
Baseline Characteristics of the 1013 Study Participantsa

In parents, a difference was found for percentages of underreporters at the end of intervention between group B and controls (respectively, 41% vs 27%; P < .001). We decided not to exclude underreporters from nutritional analysis because (1) dieticians found that low energy intake reporting was often explained by undereating in the study's population and (2) the dietary interventions may result in a decrease in energy intake. These 2 reasons would wrongly classify participants as underreporters.


Changes in nutritional intakes from baseline to the end of intervention are presented in Table 2 (children) and Table 3 (parents); changes in nutrition were expressed as crude intake (grams) and percentage of total energy intake. Similar results were obtained in parents when excluding underreporters (Table 4). The nutritional target was achieved for fats in the intervention groups. The intake of sugars decreased only in group B and the intake of complex carbohydrates increased only in group A. The interventions led to total carbohydrate intakes exceeding 50% of total energy intake in children and about 46% in parents. Compared with controls, total energy intake decreased in children (groups A and B) and in parents (group B).

Table 2. 
Children's Changes in Anthropometric and Dietary Characteristics
Children's Changes in Anthropometric and Dietary Characteristics
Table 3. 
Parents' Changes in Anthropometric, Dietary, and Biologic Characteristics
Parents' Changes in Anthropometric, Dietary, and Biologic Characteristics
Table 4. 
Parents' Changes in Dietary Intake, Excluding Underreporters
Parents' Changes in Dietary Intake, Excluding Underreporters

Changes in anthropometric measures are presented in Table 2 (children) and Table 3 (parents). As expected for children of this age, all anthropometric indicators increased in both intervention and control groups during the course of the school year. No significant differences were found between groups. When comparing the BMI z scores, we found a trend toward negative changes in all 3 groups (group A, − 0.13; group B, − 0.09; control group, − 0.06; P = .19).

In parents, BMI and FM increased in the control group during the intervention period (BMI, P < .001; FM, P = .04). Dietary interventions had a beneficial impact on these indicators, with significant decreases compared with controls in group B only (BMI, P = .01; FM, P = .04). Similar results were found for hip circumference (group B vs controls, P = .005). All groups showed the same changes in cardiovascular and blood indicators during the study.


At baseline, participants reported moderate activities, with no differences between groups. In children, changes in physical activity throughout the study did not differ between groups, either for daily screen viewing (group A vs controls, P = .47; group B vs controls, P = .10) or for activities in clubs (group A vs controls, P = .71; group B vs controls, P = .85). Identical results were found in parents (group A vs controls, P = .19; group B vs controls, P = .54). In parents, food-related quality of life did not change differently between groups throughout the study (P = .94), irrespective of the dimension of quality of life (eating pleasure, P = .65; social relations, P = .94; psychological dimension, P = .73; physical capacity, P = .30).


Our primary aims were to investigate to what extent family dietary coaching would cause nutritional changes and improve weight control in free-living children and parents. To our knowledge, this is the first time that an intervention study based on family dietary coaching and targeting only dietary changes has been carried out on a large cohort of free-living families for one school year. Groups were comparable at baseline on almost all indicators; the difference observed for age in children was too limited to have any effect on the outcome measurements. Our results are therefore probably not because of differences in the groups that were unrelated to the intervention. Moreover, because no difference in the evolution of physical activity levels was found between groups, a causal inference may be made between nutritional changes and clinical evolution.

The ELPAS adult cohort is a group of educated, motivated, mostly female volunteers, and therefore it is not representative of the French population as a whole. Higher social/professional categories accounted for about half of the parents, which is about 5 times greater than in the general French population. As a result, the efficacy of family dietary coaching in this population might not have been observed in less educated participants. However, it is relevant to test family dietary coaching in this population, which is theoretically more compliant to nutrition education, before implementing adapted methods in other groups.

Nutritional changes mainly occurred in the first 3 months of the intervention and were then maintained until the end of intervention, suggesting that such changes were sustainable in the context of the study (Figure 2). Percentages of adult underreporters agreed with previous findings in adults.50,51 Nutritional targets were achieved for fats and to a smaller extent for sugars. However, advice on sugars was based only on extrinsic sugars, limiting the opportunities to decrease them. Based on studies in normal to moderately overweight men, Drummond and Kirk52,53 found that free-living populations may find it hard to maintain concurrent reductions in fats and sugars, mainly for palatability reasons. However, in our study, we obtained concurrent reductions in fats and sugars in group B. A trend toward a decrease in the intake of sugars was even observed in group A (− 5 g/d in children, P = .08; − 7 g/d in parents, P = .07), though this group received advice only on fats. We also observed that the increase in complex carbohydrates was insufficient to counterbalance the energy deficit due to decrease in fats and sugars, especially in group B. The lower compliance regarding complex carbohydrates in group B compared with group A may be because of (1) a specific decrease in consumption of food containing sugars and complex carbohydrates in group B, (2) some difficulty in achieving 3 vs 2 nutritional targets, or (3) a more restrictive attitude toward food in group B. In parents, the decrease in energy intake was associated with an increase in body weight in 2 groups (group A and control group), suggesting that final energy intakes were still higher than requirements and/or that underreporting increased throughout the study. Similar observations were made in children in groups A and B. Overall, family dietary coaching leads to an improvement of the macronutrient repartition in the diet, with final values close to current recommendations. Further analysis will be performed to assess the effects of the dietary advice on micronutrient intake and food habits.

Figure 2.
Changes in nutritional intake throughout the intervention. Group A was assigned to a low-fat, high–complex carbohydrate diet. Group B was assigned to a low-fat, low-sugar, high–complex carbohydrate diet. The control group was assigned to a usual diet. CC indicates complex carbohydrates; F, fats; and SS, simple sugars.

Changes in nutritional intake throughout the intervention. Group A was assigned to a low-fat, high–complex carbohydrate diet. Group B was assigned to a low-fat, low-sugar, high–complex carbohydrate diet. The control group was assigned to a usual diet. CC indicates complex carbohydrates; F, fats; and SS, simple sugars.

In children, the interventions had no effect on the clinical indicators. This result was expected in growing children, as previously studied approaches focusing on diet alone in this population were found ineffective54 except when they targeted soft drinks.7,8 However, we initially hypothesized that a longer intervention based on family dietary coaching would induce clinical benefits, even in free-living, mostly healthy children. Given that effective studies in nonobese children aged 6 to 10 years include a combination of actions toward screen viewing, activity, and diet, our study confirms that primary prevention of childhood obesity should not be limited to dietary intervention, though nutritional education may improve dietary habits in the long run.

In parents, participants in the control group showed an increase in clinical indicators (BMI, FM, and body weight), as expected for this population.55 These indicators were improved in the intervention groups, especially in group B (decrease in BMI and FM compared with controls), showing that family dietary coaching was effective in improving weight control in this population. The changes in clinical indicators and the decrease in fat intake are positive outcomes of the intervention with regards to prevention of obesity and cardiovascular diseases. Regarding blood indicators, no differences were found between groups in our study, as expected in subjects with normal values for blood indicators at baseline.

Although the sample size was insufficient to formally test for effects within subgroups, it was desirable to further characterize the effects of the intervention on participants with varying initial BMI with an exploratory analysis (Table 5). Intervention and control groups were compared within strata defined by BMI status (normal weight or overweight). In overweight children, BMI was stabilized throughout the study, irrespective of their group. This finding is interesting, because BMIs of 27 overweight children not participating in the study but attending the same schools increased during the school year (+ 0.36, 95% CI, 0.08-0.65, P = .004). This result might indicate that participating in an educational program may improve BMI in overweight children, whatever the intensity of dietary coaching (general information in the control group vs family dietary coaching in the intervention groups). Previous studies suggested that interventions may be effective in overweight children but not in children with a healthy BMI.56 In parents, we found that nonoverweight participants in group B had the smallest mean increase in BMI, while overweight participants in both intervention groups showed a mean decrease in BMI throughout the study, as already reported by Astrup et al.57

Table 5. 
Changes in Body Mass Index (BMI)a in Families Assigned to Dietary Coaching
Changes in Body Mass Index (BMI)a in Families Assigned to Dietary Coaching

From a public health perspective, general messages seem to have little efficacy in inducing nutritional changes in the general population.13,14 In our study, we used family dietary coaching to modulate nutritional intakes toward recommendations, with beneficial effects on weight control in parents. This approach makes it possible to account for individual characteristics that influence food habits, such as socioeconomic status and education. Family dietary coaching has an individual cost of around 1 €/d/person (US $1.42/d/person), which should be compared with the cost implications of obesity for health care and society. In our study, we found that compliance with advice on complex carbohydrates was lower than compliance with advice on fats and sugars, suggesting that current recommendations on complex carbohydrates are hard to achieve in the general population. No differences were found between groups for food-related quality of life, showing that interventions were well accepted.

Further studies should be conducted to better assess the sustainability of the dietary changes. The few available long-term studies found a decreasing efficacy of nutritional interventions over time,15,58 especially when several dietary goals are defined, such as a reduction in both fats and sugars.53 Two studies in children showed maintenance of the favorable changes observed 1 to 4 years after the intervention had ended.59,60

Overall, family dietary coaching may be a promising approach to induce sustainable nutritional changes in a free-living population. Further studies with more sociodemographically diverse samples are needed to evaluate the generalizability of these findings.

Correspondence: Damien L. Paineau, MS, Nutri-Health, Immeuble Ampère, 8 rue Eugène et Armand Peugeot, 92566 Rueil-Malmaison Cedex, France (d.paineau@nutri-health.eu).

Accepted for Publication: July 18, 2007.

Author Contributions: Mr Paineau and Dr Bornet had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Paineau, Beaufils, Boulier, Cassuto, Chwalow, Combris, Couet, Jouret, Lafay, Laville, Mahe, Ricour, Romon, Simon, Tauber, Valensi, Chapalain, Zourabichvili, and Bornet. Acquisition of data: Paineau, Boulier, Cassuto, Chwalow, Combris, Couet, Jouret, Simon, Tauber, Valensi, and Bornet. Analysis and interpretation of data: Paineau, Beaufils, Boulier, Cassuto, Chwalow, Combris, Couet, Jouret, Lafay, Laville, Mahe, Ricour, Romon, Simon, Tauber, Valensi, Chapalain, Zourabichvili, and Bornet. Drafting of the manuscript: Paineau. Critical revision of the manuscript for important intellectual content: Paineau, Beaufils, Boulier, Cassuto, Chwalow, Combris, Couet, Jouret, Lafay, Laville, Mahe, Ricour, Romon, Simon, Tauber, Valensi, Chapalain, Zourabichvili, and Bornet. Statistical analysis: Paineau, Chapalain, Zourabichvili, and Bornet. Obtained funding: Paineau and Bornet. Study supervision: Paineau, Beaufils, Boulier, Cassuto, Chwalow, Combris, Couet, Jouret, Lafay, Laville, Mahe, Ricour, Romon, Simon, Tauber, Valensi, Chapalain, and Bornet.

Financial Disclosure: Dr Cassuto served as an independent consultant for the Centre d'Etudes et de Documentation du Sucre.

Funding/Support: Funding was provided by the French Ministry of Research (2002 Réseau Alimentation Référence Europe 31), and by the ELPAS study's private partners (Avenance Enseignement, the Centre d'Etudes et de Documentation du Sucre, and the Louis Bonduelle Foundation). The private partners did not participate in conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the manuscript. The Centre d'Etudes et de Documentation du Sucre participated in the study design.

Additional Information: All statistical analyses were performed by statisticians from the Quanta Medical Group (quality certification ISO 9001-2000) in compliance with statistical procedures determined before the study.

Additional Contributions: We thank the families, schools, and administrators (Rectorat de Paris) who participated in this project. Michel Vidailhet, MD, Brabois Hospital, Vandoeuvre les Nancy, France, helped with supervising the study and drafting the manuscript. Donald White, BA, assisted in correcting the English text.

World Health Organization, Global Strategy on Diet, Physical Activity and Health.  Paris, France World Health Organization2004;
Howard  BVManson  JEStefanick  ML  et al.  Low-fat dietary pattern and weight change over 7 years: the Women's Health Initiative Dietary Modification Trial. JAMA 2006;295 (1) 39- 49
Bray  GAPaeratakul  SPopkin  BM Dietary fat and obesity: a review of animal, clinical and epidemiological studies. Physiol Behav 2004;83 (4) 549- 555
Saris  WHAstrup  APrentice  AM  et al.  Randomized controlled trial of changes in dietary carbohydrate/fat ratio and simple vs complex carbohydrates on body weight and blood lipids: the CARMEN study, The Carbohydrate Ratio Management in European National diets. Int J Obes Relat Metab Disord 2000;24 (10) 1310- 1318
Drummond  SKirk  Tde Looy  A Are dietary recommendations for dietary fat reduction achievable? Int J Food Sci Nutr 1996;47 (3) 221- 226
Varela-Moreiras  G Controlling obesity: what should be changed? Int J Vitam Nutr Res 2006;76 (4) 262- 268
James  JKerr  D Prevention of childhood obesity by reducing soft drinks. Int J Obes (London) 2005;29 ((suppl 2)) S54- S57
Ebbeling  CBFeldman  HAOsganian  SKChomitz  VREllenbogen  SJLudwig  DS Effects of decreasing sugar-sweetened beverage consumption on body weight in adolescents: a randomized, controlled pilot study. pediatrics 2006;117 (3) 673- 680
Maillard  GCharles  MALafay  L  et al.  Macronutrient energy intake and adiposity in non obese prepubertal children aged 5-11 y (the Fleurbaix Laventie Ville Sante Study). Int J Obes Relat Metab Disord 2000;24 (12) 1608- 1617
Astrup  A Carbohydrates as macronutrients in relation to protein and fat for body weight control. Int J Obes 2006;30S4- S9Article
Kantor  LSVariyam  JNAllshouse  JEPutnam  JJLin  B-H Choose a variety of grains daily, especially whole grains: a challenge for consumers. J Nutr 2001;131 ((2S-1)) 473S- 486S
Gibson  LJPeto  JWarren  JMDos Santos Silva  I Lack of evidence on diets for obesity for children: a systematic review. Int J Epidemiol 2006;35 (6) 1544- 1552
Wammes  BBreedveld  BLooman  CBrug  J The impact of a national mass media campaign in The Netherlands on the prevention of weight gain. Public Health Nutr 2005;8 (8) 1250- 1257
Ash  SReeves  MBauer  J  et al.  A randomised control trial comparing lifestyle groups, individual counselling and written information in the management of weight and health outcomes over 12 months. Int J Obes (Lond) 2006;30 (10) 1557- 1564
Williamson  DAWalden  HMWhite  MA  et al.  Two-year internet-based randomized controlled trial for weight loss in African-American girls. Obesity (Silver Spring) 2006;14 (7) 1231- 1243
Sharma  M School-based interventions for childhood and adolescent obesity. Obes Rev 2006;7 (3) 261- 269
Edwards  CNicholls  DCroker  HVan Zyl  SViner  RWardle  J Family-based behavioural treatment of obesity: acceptability and effectiveness in the UK. Eur J Clin Nutr 2006;60 (5) 587- 592
Rolland-Cachera  MFDeheeger  MBellisle  FSempe  MGuilloud-Bataille  MPatois  E Adiposity rebound in children: a simple indicator for predicting obesity. Am J Clin Nutr 1984;39 (1) 129- 135
Rinderknecht  KSmith  C Social cognitive theory in an after-school nutrition intervention for urban Native American youth. J Nutr Educ Behav 2004;36 (6) 298- 304
Loewen  RPliner  P The Food Situations Questionnaire: a measure of children's willingness to try novel foods in stimulating and non-stimulating situations. Appetite 2000;35 (3) 239- 250
Wylie-Rosett  JSwencionis  CGinsberg  M  et al.  Computerized weight loss intervention optimizes staff time: the clinical and cost results of a controlled clinical trial conducted in a managed care setting. J Am Diet Assoc 2001;101 (10) 1155- 1162
Haerens  LDeforche  BMaes  LStevens  VCardon  GDe Bourdeaudhuij  I Body mass effects of a physical activity and healthy food intervention in middle schools. Obesity (Silver Spring) 2006;14 (5) 847- 854
SUVIMAX, Table de Composition des Aliments.  Paris, France Economica2005;
Schofield  WN Predicting basal metabolic rate, new standards and review of previous work. Hum Nutr Clin Nutr 1985;39 ((suppl 1)) 5- 41
Goldberg  GRBlack  AEJebb  SA  et al.  Critical evaluation of energy intake data using fundamental principles of energy physiology, 1: derivation of cut-off limits to identify under-recording. Eur J Clin Nutr 1991;45 (12) 569- 581
Dietz  WHRobinson  TN Use of the body mass index (BMI) as a measure of overweight in children and adolescents. J pediatr 1998;132 (2) 191- 193
Kraemer  HCBerkowitz  RIHammer  LD Methodological difficulties in studies of obesity, I: measurement issues. Ann Behav Med 1990;12 (3) 112- 118Article
Cole  TJ Using the LMS method to measure skewness in the NCHS and Dutch National height standards. Ann Hum Biol 1989;16 (5) 407- 419
Rolland-Cachera  MFCole  TJSempe  MTichet  JRossignol  CCharraud  A Body Mass Index variations: centiles from birth to 87 years. Eur J Clin Nutr 1991;45 (1) 13- 21
Cole  TJBellizzi  MCFlegal  KMDietz  WH Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320 (7244) 1240- 1243
 Obesity, preventing and managing the global epidemic: report of a WHO consultation. World Health Organ Tech Rep Ser 2000;894i- xii, 1-253
Lohman  TGRoche  AFMartorell  R Anthropometric Standardization Reference Manual.  Champaign, IL Human Kinetics Publishers1988;
Boulier  AChumlea  WCDe Lorenzo  A  et al.  Body composition estimation using leg-to-leg bioelectrical impedance: a six-site international crossvalidation study. International Journal of Body Composition Research 2005;3 (1) 31- 39
Friedewald  WTLevy  RIFredrickson  DS Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18 (6) 499- 502
Matthews  DRHosker  JPRudenski  ASNaylor  BATreacher  DFTurner  RC Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28 (7) 412- 419
Pereira  MAFitzGerald  SJGregg  EW  et al.  A Collection of Physical Activity Questionnaires for health-related research. Med Sci Sports Exerc 1997;29 ((suppl 6)) S1- S205
Aaron  DJKriska  AMDearwater  SRCauley  JAMetz  KFLaPorte  RE Reproducibility and validity of an epidemiologic questionnaire to assess past year physical activity in adolescents. Am J Epidemiol 1995;142 (2) 191- 201
Roeykens  JRogers  RMeeusen  RMagnus  LBorms  Jde Meirleir  K Validity and reliability in a Flemish population of the WHO-MONICA Optional Study of Physical Activity Questionnaire. Med Sci Sports Exerc 1998;30 (7) 1071- 1075
Rozin  PFischler  CImada  SSarubin  AWrzesniewski  A Attitudes to food and the role of food in life in the U.S.A., Japan, Flemish Belgium and France: possible implications for the diet-health debate. Appetite 1999;33 (2) 163- 180
Ware  JEJSherbourne  CD The MOS 36-item short-form health survey (SF-36), I: conceptual framework and item selection. Med Care 1992;30 (6) 473- 483
Meadows  KSteen  NMcColl  E  et al.  The Diabetes Health Profile (DHP), a new instrument for assessing the psychosocial profile of insulin requiring patients: development and psychometric evaluation. Qual Life Res 1996;5 (2) 242- 254
Meadows  KAAbrams  CSandbaek  A Adaptation of the Diabetes Health Profile (DHP-1) for use with patients with Type 2 diabetes mellitus: psychometric evaluation and cross-cultural comparison. Diabet Med 2000;17 (8) 572- 580
Paineau  DBaudoin  CGrairia  M  et al.  Development and validation of a food-related quality-of-life scale in the French population. Cah Nutr Diét In press
Institut National de la Statistique et des Études Économiques, Enquête permanente sur les conditions de vie des ménages (EPCV).  Paris, France Institut National de la Statistique et des Études Économiques2001;
Rolland-Cachera  MFCastetbon  KArnault  N  et al.  Body mass index in 7-9-y-old French children: frequency of obesity, overweight and thinness. Int J Obes Relat Metab Disord 2002;26 (12) 1610- 1616
Maillard  GCharles  MAThibult  N  et al.  Trends in the prevalence of obesity in the French adult population between 1980 and 1991. Int J Obes Relat Metab Disord 1999;23 (4) 389- 394
Dunnet  CW A multiple comparison procedure for comparing several treatments with a control. J Am Stat Assoc 1955;501096- 1121Article
Vincelet  CGalli  JGrémy  I Surpoids et obésité en Ile-de-France. Observatoire régional de santé d’Ile-de-France, Union régionale des caisses d’assurance maladie d’Ile-de-France. 2006;
ObÉpi-Roche, 4ème enquête épidémiologique nationale sur l’obésité et le surpoids en France [press release].  September19 2006;
Pryer  JAVrijheid  MNichols  RKiggins  MElliott  P Who are the ‘low energy reporters' in the dietary and nutritional survey of British adults? Int J Epidemiol 1997;26 (1) 146- 154
Briefel  RRSempos  CMcDowell  MChien  SAlaimo  K Dietary methods research in the third National Health and Nutrition Examination Survey: underreporting of energy intake. Am J Clin Nutr 1997;65 (4) ((suppl 4)) 1203S- 1209S
Drummond  SKirk  T The effect of different types of dietary advice on body composition in a group of Scottish men. J Hum Nutr Diet 1998;11473- 485Article
Drummond  SKirk  T Assessment of advice to reduce dietary fat and non-milk extrinsic sugar in a free-living male population. Public Health Nutr 1999;2 (2) 187- 197
Summerbell  CDWaters  EEdmunds  LDKelly  SBrown  TCampbell  KJ Interventions for preventing obesity in children. Cochrane Database Syst Rev 2005; (3) CD001871
Korkeila  MRissanen  AKaprio  JSorensen  TIKoskenvuo  M Weight-loss attempts and risk of major weight gain: a prospective study in Finnish adults. Am J Clin Nutr 1999;70 (6) 965- 975
Datar  ASturm  R Physical education in elementary school and body mass index: evidence from the early childhood longitudinal study. Am J Public Health 2004;94 (9) 1501- 1506
Astrup  ARyan  LGrunwald  GK  et al.  The role of dietary fat in body fatness: evidence from a preliminary meta-analysis of ad libitum low-fat dietary intervention studies. Br J Nutr 2000;83S25- S32
Borg  PFogelholm  MKukkonen-Harjula  K Food selection and eating behaviour during weight maintenance intervention and 2-y follow-up in obese men. Int J Obes Relat Metab Disord 2004;28 (12) 1548- 1554
Nemet  DBarkan  SEpstein  YFriedland  OKowen  GEliakim  A Short- and long-term beneficial effects of a combined dietary-behavioral-physical activity intervention for the treatment of childhood obesity. pediatrics 2005;115 (4) e443- e449
Manios  YKafatos  APreventive Medicine and Nutrition Clinic University of Crete Research Team, Health and nutrition education in primary schools in Crete: 10 years follow-up of serum lipids, physical activity and macronutrient intake. Br J Nutr 2006;95 (3) 568- 575