The prevalence of overweight in children has markedly increased over the past few decades in France, as in all Western countries. We sought to describe the yearly prevalence of childhood overweight from 1996 to 2006 and to assess whether a shift in trends could be observed dating from the time the Nutrition and Health National Program (PNNS) was set up in France in 2001, in particular according to gender, age and family economic status.
We used annual overweight prevalence of standardized 6- to 15-year-old populations (total=26 600) with weight and height measured at health examination centers in the central/western part of France between 1996 and 2006. Regression slopes of overweight prevalence were evaluated between 1996 and 2006, and specifically between 1996 and 2001, and 2001 and 2006. The annual prevalence and estimated slopes were compared in subgroups, taking into account gender, age and economic status of the family.
The prevalence increased between 1996 (11.5%) and 1998 (14.8%) and was stable between 1998 and 2006 (15.2%). According to linear regression, the overall trend in prevalence of overweight children between 1996 and 2006 was stable (slope=0.19, P=0.08). Similarly, the prevalence of overweight increased between 1996 and 1998 in boys and girls, in 6–10 year olds, in 11–15 year olds and in non-disadvantaged children, and remained stable thereafter. The prevalence of overweight in the disadvantaged group increased between 1996 (12.8%) and 2001 (18.9%) (slope=1.16, P=0.004) and was stable between 2001 and 2006 (18.2%) (slope=0.09, P=0.78).
The results of this study reveal a stable prevalence of overweight since 1998 in most groups studied, and since 2001 in the disadvantaged group.
The prevalence of childhood overweight has increased worldwide throughout the past several decades.1, 2, 3 Overweight is now the most common childhood disorder in Europe, affecting one child out of six in 2002, and, in some parts of Europe, one child out of three.4 Recent stabilization or a decrease in the prevalence of overweight and obese children has been observed in several countries, including France.5, 6, 7, 8, 9, 10 An effect of public health initiatives on the recent stabilization of overweight prevalence in children is hypothesized.5, 11
To manage the burden of overweight populations, governments were advised to set up national programs4 and international policies.12 Many countries have begun health campaigns to prevent overweight. In France in 2001, the government established a Nutrition and Health National Program (‘Programme National Nutrition Santé,’ PNNS) to enhance the overall health status of the population through nutrition.13 One of its priority objectives has been to stop the increase in childhood overweight and obesity. To attain this objective, the measures that have been implemented include the dissemination of national food-based guidelines for the general population and specific population groups such as children and adolescents. These measures have been oriented toward school settings, and include improvement in school meals and removal of vending machines.13
The impact of such public health programs is particularly difficult to evaluate.14 An assessment of overweight and obesity prevalence in the population before and after the launch of public health plans would be useful. However, in France6, 8 and other countries5, 9, 15, 16 repeated cross-sectional studies investigating trends over a period of several years have been limited to comparisons between prevalences at 5- or 10-year intervals. The use of annual overweight prevalence based on continuous data collection enables more accurate assessment of changes in trends and the possibility of determining whether these changes occurred as a result of implementation of public health measures. However, such data are rarely available at the national level owing to the high cost and difficulties in data collection.
In this study, we described trends in overweight using annual prevalence rates from 1996 to 2006 in 6- to 15-year-old children measured at health examination centers in the central/western region of France. We hypothesized a possible shift in prevalence trends from the time the PNNS was launched in 2001. We also sought to estimate whether overweight prevalence trends were variable according to gender, age and family economic status.
Subjects and data collection
The sample consisted of 6- to 15-year-old children attending one of the health examination centers in the central/western part of France (parts of 3 regions: Centre, Pays de la Loire and Normandie) from January 1996 to December 2006. All individuals affiliated with the French national health insurance system (corresponding to about 85% of the French population) are able to benefit from a free medical and laboratory check-up every 5 years. Most of the subjects were spontaneously present at the center for a free check-up. The population is generally aware of this possibility. Practical information is publicly available in city halls and places where the disadvantaged can be easily reached. About one-fourth of the examinations carried out in the health examination centers are reserved for the unemployed, who receive income support or who are 16–25 years of age and in an employment integration plan. Children with at least one parent who fulfilled these criteria were qualified as belonging to a ‘disadvantaged family.’
Social and economic characteristics were collected through standardized questionnaires. Weight and height were measured by trained nurses, with participants wearing only underwear, according to standardized procedures.17 Data collection by the examination centers was approved by the Comité National Informatique et Liberté (CNIL number 26674).
Continuous collection of data involved 26 671 individuals spread over the 11 years of the study. Seventy-one children with incomplete information on age, weight and height were excluded from the analysis. The remaining sample included 26 600 children aged 6–15 years, from 1996 to 2006. During the 11 years of the study, the annual sample size ranged from 2000 to 3500 children (Table 1).
Body mass index was calculated as weight in kilograms divided by the square of height in meters and was rounded off to the nearest tenth decimal. Overweight was defined according to International Obesity Task Force (IOTF) cutoffs for each year of age.18 The IOTF cutoff values for overweight are based on percentiles passing through body mass index of 25 kg/m2 at the age of 18 years; thus, overweight includes obese children.
Prevalence of overweight in 6- to 15-year-old central/western French children was calculated for each year of the study. To adjust for potential differences in age and gender distribution over time, direct standardization was performed according to age and gender distribution from the 1999 national census.19 As no data were available at the national level for economically disadvantaged status, as defined above, internal standardization of economic status was performed considering the mean rates (26%) of disadvantaged children in the present sample during the period 1996–2006. We calculated standard deviations of standardized prevalence rates.20 Using linear regression models, we estimated slopes (95% confidence interval (CI)) of overweight prevalence trends during 1996–2006. To test our hypothesis on the impact of the PNNS, trends were analyzed for 1996–2001 (preceding the launch of PNNS) and 2001–2006 (following the launch of the PNNS). We compared slopes (95% CI) of the evolution of prevalence of overweight between 1996–2001 and 2001–2006 using bilateral Z-tests.
Thereafter, all analyses were performed separately according to gender (boys and girls), age category (6–10 years and 11–15 years) and economically disadvantaged status (yes and no). We compared prevalence rates in boys versus girls, in 6–10 year olds versus 11–15 year olds and in economically disadvantaged versus non-disadvantaged children.20 In addition, we compared linear regression slopes (95% CI) in each subgroup, using bilateral Z-tests, during the entire period and each subperiod. Standardization was performed for each of these subgroups. For the three subgroups, standardization was performed for the two other stratification variables mentioned above. For example, gender subgroups were standardized according to age and economic status.
All statistical analyses were carried out using standard procedures (SAS version 8.02; SAS Institute, Cary, NC, USA). Statistical significance was judged at α< 0.05.
A sample description is presented in Table 1. Gender distribution was stable throughout the study period, with boys and girls being equally represented. The proportion of 6- to 10-year-old children ranged from 41.6 to 47.0%, whereas the proportion of economically disadvantaged children ranged from 18.6 to 35.6% across the years.
Trends in overweight prevalence in the entire sample
The prevalence of childhood overweight (Table 2) increased between 1996 and 1998 (11.5–14.8%) and was followed by stabilization starting in 1999 and lasting until the end of the study (Figure 1). Linear regression slopes showed a significant increase in the predicted prevalence of overweight between 1996 and 2001 and a stable prevalence between 2001 and 2006 (Table 3).
Trends in overweight prevalence in the subgroups
For each year of the 1996–2006 period, except for 1999, the prevalence of overweight was comparable in boys and girls (Table 2). An increase in the prevalence of overweight was observed from 1996 to 1998 in boys (10.7–14.1%) and from 1996 to 1999 in girls (12.4–16.3%). Following those dates, the prevalence was generally stable in both groups until the end of the study. On the basis of linear regression analyses, no significant increase in the prevalence of overweight was observed in either gender over the 1996–2006 period (Table 3). During the 1996–2001 period, there was an increase in the prevalence of overweight that was statistically significant in boys only. During the 2001–2006 period, the prevalence of overweight was stable in both genders. Comparison of linear regression slopes between 1996–2001 and 2001–2006 indicated a significant difference in girls only (Table 3).
For each year throughout the entire period of the study (1996–2006), the prevalence of overweight was similar in 6- to 10- and 11- to 15-year-old children (Table 2). An increase in the prevalence of overweight was observed from 1996 to 1998 in 6- to 10-year-old children (10.5–15.6%) and from 1996 to 1999 in 11- to 15-year-old children (12.8–15.2%). Following these periods, the prevalence was stable in both groups until the end of the study. According to linear regression analyses, we showed a significant increase in the prevalence of overweight in 11- to 15-year-old children, but no increase in 6- to 10-year-old children during the 1996–2006 period (Table 3). During the 1996–2001 period, there was a significant increase in the prevalence of overweight in 6- to 10-year-old children. During the 2001–2006 period, the prevalence of overweight was stable in both age groups. It can be noted that during this period the regression slope was negative for the 6–10 year olds and positive for the 11- to 15-year-old children, with a significant difference between age groups (P=0.008). Comparison of linear regression slopes between 1996–2001 and 2001–2006 indicated a highly significant difference in 6- to 10-year-old children only.
The annual prevalence of overweight was significantly higher in disadvantaged children than in non-disadvantaged ones for 4 years during the 1996–2006 period: 2000, 2001, 2004 and 2005 (Table 2). As shown in Figure 1, childhood overweight prevalence trends in the non-disadvantaged group matched those of the whole sample, with an increase between 1996 and 1998 (11.2–14.9%) and a stable prevalence between 1998 and 2006. However, overweight prevalence in the economically disadvantaged group increased until 2001 (12.8–18.9%) and remained generally stabilized later on. According to linear regression analyses, there was a significant increase in the prevalence of overweight in disadvantaged children, but no increase in non-disadvantaged children during the 1996–2006 period (Table 3). During the 1996–2001 period, there was a significant increase in the prevalence of overweight in disadvantaged children. During the 2001–2006 period, the prevalence of overweight was stable in both groups. Comparison of linear regression slopes between 1996–2001 and 2001–2006 indicated a significant difference in disadvantaged children only.
Trends in overweight prevalence in the entire sample
Our study is the first to describe trends in annual overweight prevalence in a large sample of 6- to 15-year-old French children using annual data and to evaluate these trends before and after the launch of the French PNNS. In this gender-, age- and deprivation-status-standardized population of children living in the central/western part of France, the prevalence of overweight was shown to increase between 1996 and 1998 and to subsequently stabilize until the end of the study period, in 2006 (Figure 1). The overall trend over the 1996–2006 period was not significant. Only a few studies, based on repeated cross-sectional data collection, had previously evaluated trends in overweight prevalence in France. The prevalence of overweight showed an increase in 4- to 17-year-old children between 1980 and 1990.21 Following an increase in the prevalence of childhood overweight in the 1990s, more recent data suggested stabilization during the 2000s. National surveys showed global stability of overweight in 7–9 year olds between 2000 and 2007,8 and in 3–17 year olds between 1998–1999 and 2006–2007.6
Most countries have encountered an increased frequency of overweight and obesity in children in the past few decades.1, 2, 3 However, as is the case in France, recent data suggest stabilization of the prevalence of overweight in several countries and even a decrease in some groups. Data in Sweden showed a decreasing frequency in 10–year-old girls between 2000–2001 and 2004–2005 in Gothenburg City, and trends were stable in boys of the same age.9 Another study in Stockholm County in the same age group of children10 showed stable trends in both genders between 1999 and 2003. A national cross-sectional study carried out in 2002 and 2007 in Switzerland indicated a decrease in overweight in 6- to 13-year-old children.5 Finally, in the United States, the prevalence of high body mass index for age was no different between 2003–2004 and 2005–2006, and there were no significant trends between 1999 and 2006, as estimated by logistic regression.7
A possible link between public health campaigns and the recent stabilization of overweight and obesity prevalence has been suggested in Switzerland and the United States.5, 11 In Sweden, no large-scale obesity preventive actions had been carried out in the study area or in the country, but in the past few years, an awareness has grown about the importance of promoting healthy food habits and physical activity among children.9 To evaluate the potential effect of the PNNS on overweight risk in French children, overweight prevalence trends were estimated using linear regression analysis during the 1996–2001 and 2001–2006 periods, that is, before and after the launch of the PNNS. According to our results, the overall prevalence of overweight would have increased during the 1996–2001 period and remained stable during the 2001–2006 period. However, annual prevalence estimations indicate that stabilization had already begun, at least in this French region, during 1998–1999, that is, before 2001. Continuous data collection enabled us to more accurately identify when the shift in trends occurred than when using the 5- or 10-year interval cross-sectional data.
It is particularly difficult to evaluate a national public health program as most people are expected to be exposed to it. An assessment of overweight and obesity prevalence in the population before and after the launch of public health plans is therefore useful. In particular, continuous collection of anthropometry is complementary to repeated cross-sectional analyses. However, continuous collection of measured anthropometry is not easy on a large scale even in accessible populations such as children. From these results, it is not possible to conclude whether the stabilization in overweight prevalence in children resulted from the PNNS, which was set up in 2001. However, the stabilization beginning in 1998–1999 coincided with increasing consciousness of childhood overweight issues in France. As a consequence, the first governmental report on nutrition and health pointed out in 1998 the problem of obesity as a priority (S Hercberg and A Tallec, unpublished). Childhood obesity was specifically emphasized and largely relayed by the media to the population and health professionals. It can be hypothesized that this increasing awareness of the childhood obesity issue initiated stabilization that may have been partly maintained owing to the PNNS.
Moreover, it must be noted that our sample involved subjects volunteering for a free check-up at examination centers. It cannot be excluded that characteristics of such a selected sub-population changed over time, especially in light of the fact that information on obesity was released to the media. Overweight children (or their parents) might have felt stigmatized and more reluctant to have their weight checked. Another hypothesis is that leveling off could represent a ‘natural’ plateau, given the distribution of childhood overweight risk factors in the population, including genetic, environmental and socioeconomic characteristics. This also includes early risk factors such as growth and diet during the first years of life. Moreover, overweight is generally more prevalent in lower socioeconomic groups; improvement in socioeconomic conditions in France might have occurred in 2000 and would have resulted in the reported stabilization of childhood overweight. This hypothesis is particularly difficult to validate. In France, no significant decrease in poverty rates or increases in inequalities appeared to occur during that same period (L Goutard and J Pujol, unpublished data). Moreover, in our sample, economically disadvantaged children were purposely over-represented across the years of the survey.
Trends in overweight prevalences in the subgroups
According to gender, overweight prevalence trends matched those of the overall sample, with an increase in prevalence up until 1998–1999, and stabilization later on. The increase during the 1996–2001 period, as shown by regression analysis, is not interpretable as stabilization occurred before 2001. In agreement with our results, studies in the United States7 and Sweden10 showed stable trends in both genders. In Switzerland and in another Swedish study, the prevalence of overweight decreased in recent years among 10-year-old girls, whereas no change was observed in boys.5, 9 More data are needed to confirm or refute the absence of a difference in prevalence trends between boys and girls to better understand overweight dynamics and develop adequate prevention programs.
By age category, overweight prevalence trends matched those of the overall sample. The increased prevalence in 6–10 year olds during the 1996–2001 period, as shown by regression analysis, is not interpretable as stabilization occurred before 2001. Regression analysis between 2001 and 2006 showed a significant difference between age groups, with negative trends in younger children. National surveys showed global stability in overweight and obesity frequency in various age classes from 2000.22 Heightened public awareness about the issues of obesity over the past few years13 might have contributed to the prevention of development of overweight or obesity in young children, whereas older children are less likely to be receptive to nutritional information. Moreover, losing excess weight often requires individual care plans. These differences could explain why the stabilization observed was more clear-cut in younger children.
In non-disadvantaged children, the increasing prevalence of overweight was stabilized in 1998, as in the whole sample. On the other hand, in the economically disadvantaged group, increasing prevalence stabilized only more recently, starting in 2001. This stabilization, confirmed by the comparison of linear regression slopes between 1996–2001 and 2001–2006, coincided with the launch of the PNNS. Overweight trends were shown to be associated with discrepancies in socioeconomic status in the 2000s in France.22 In Sweden, widening differences between the lower and upper socioeconomic groups were observed between 1999 and 2003 in boys, whereas in girls the social gradient was similar in both years.10 In France, some PNNS measures were implemented in school settings, including the improvement in school meals, cessation of morning snacks and removal of vending machine.13 These actions were applicable to all children, but it is possible that disadvantaged children in particular benefited from these, whereas non-disadvantaged children were in a family environment where awareness of the importance of childhood overweight was present prior to creation of the PNNS. In addition, a booklet presenting information and recipes on how to ‘eat healthy’ on a small budget were made available to disadvantaged families. Finally, the awareness that disadvantaged individuals are difficult to reach through health campaigns has led the PNNS to pay special attention to this group for each action developed. For example, national food-based guides directed at the general population contained a target specifically for individuals with financial problems. It is important to note that although both groups stabilized, disadvantaged children continued to have higher rates of overweight than non-disadvantaged ones. Higher rates of overweight in children from poor families or with a lower socioeconomic status are commonly found.9;23, 24
Strengths and limitations
One strength of our study lies in continuous data collection over a long time period (11 years), which enabled accurate estimation of the annual prevalence and evaluation of trends. In addition, this sample was large and presented different gender, age and economic status characteristics that led us to evaluate differences in trends between specific groups. The health insurance that proposes such health examinations concerns about 85% of the French population. Another strength of this study is the use of measured rather than reported anthropometry and the use of the same clinical standard procedures during the entire period at all centers involved in collecting data. Our study, however, presents some limitations. First, our sample was limited to the central/western part of France and relied mainly on subjects who presented at the examination centers. This population is therefore not randomly selected and cannot be considered strictly representative of the overall 6- to 15-year-old French population. However, this bias can be considered relatively limited, as we were interested in studying trends in overweight prevalence during a given time period rather than in obtaining absolute estimates of overweight prevalence. Second, sample characteristics might have changed over the period of study. To increase comparability of our sample for each year of the study, overweight prevalences were standardized on age, gender and economic status. Finally, as data from the health examination centers were anonymous, we could not take into account the fact that the same child may have had a free check-up at different time points. In young people, however, the period between two examinations is relatively long, thereby limiting this bias.
The results of our study show stabilization of the prevalence of overweight in French central/western 6- to 15-year-old children starting during 1998–1999, following an increase during 1996–1998. The stabilization that was observed in all the subgroups studied is encouraging, and its persistence for 8 years supports the existence of a true plateau rather than a random fluctuation. The stabilization coincides with increasing information on childhood overweight in France. The PNNS, set up in 2001, might have helped to maintain prevalence rates, though other hypotheses have to be taken into account in this stabilization. It must be kept in mind that many efforts remain necessary to prevent overweight, as its prevalence remains high in children in general, as well as in those of economically disadvantaged families in particular.
World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 2000; 894: i-253.
Lobstein T, Frelut ML . Prevalence of overweight among children in Europe. Obes Rev 2003; 4: 195–200.
Wang Y, Lobstein T . Worldwide trends in childhood overweight and obesity. Int J Pediatr Obes 2006; 1: 11–25.
International Obesity Task Force the European Association for the Study of Obesity. Obesity in Europe: the case for action, 2002. Available at http://www.iotf.org/media/euobesity.pdf (accessed 20 January 2009).
Aeberli I, Amman RS, Knabenhans M, Zimmermann MB . The national prevalence of overweight in school-age children in Switzerland has decreased between 2002 and 2007 (abstract). Int J Obes 2008; 32: 214.
Lioret S, Touvier M, Dubuisson C, Dufour A, Calamassi-Tran G, Lafay L et al. Trends in child overweight rates and energy intake in France from 1999 to 2007: relationships with socioeconomic status. Obesity 2009 doi:10.1038/oby.2008.619.
Ogden CL, Carroll MD, Flegal KM . High body mass index for age among US children and adolescents, 2003–2006. JAMA 2008; 299: 2401–2405.
Salanave B, Péneau S, Rolland-Cachera MF, Hercberg S, Castetbon K . Stabilization of overweight prevalence in French children between 2000 and 2007. Int J Pediatr Obes (in press).
Sjoberg A, Lissner L, Albertsson-Wikland K, Marild S . Recent anthropometric trends among Swedish school children: evidence for decreasing prevalence of overweight in girls. Acta Paediatr 2008; 97: 118–123.
Sundblom E, Petzold M, Rasmussen F, Callmer E, Lissner L . Childhood overweight and obesity prevalences levelling off in Stockholm but socioeconomic differences persist. Int J Obesity 2008; 32: 1525–1530.
Ebbeling CB, Ludwig DS . Tracking pediatric obesity: an index of uncertainty? JAMA 2008; 299: 2442–2443.
Lobstein T, Baur LA . Policies to prevent childhood obesity in the European Union. Eur J Public Health 2005; 15: 576–579.
Hercberg S, Chat-Yung S, Chauliac M . The French National Nutrition and Health Program: 2001–2006–2010. Int J Public Health 2008; 53: 68–77.
Nutbeam D . Evaluating health promotion––progress, problems and solutions. Health Prom Int 1998; 13: 27–44.
Lobstein TJ, James WP, Cole TJ . Increasing levels of excess weight among children in England. Int J Obes Relat Metab Disord 2003; 27: 1136–1138.
Schutz Y, Woringer V . Obesity in Switzerland: a critical assessment of prevalence in children and adults. Int J Obes Relat Metab Disord 2002; 26 (Suppl 2): s3–s11.
Gordon CC, Chumlea WC, Roche AF . Stature, recumbent length, and weight. In: Lohman TG, Roche AF, Martorell R (eds). Anthropometric Standardization Reference Manual. Human Kinetics Books: Champaign, IL, 1988. pp 3–8.
Cole TJ, Bellizzi MC, Flegal KM, Dietz WH . Establishing a standard definition for child overweight and obesity worldwide: international survey. Br Med J 2000; 320: 1240–1243.
INSEE. Recensement général de la population française. Imprimerie Nationale Paris. Available at 1999, http://www.recensement.insee.fr/RP99/rp99/page_accueil.paccueil (accessed 20 January 2009).
Jougla E . Tests statistiques relatifs aux indicateurs de mortalité en population. Rev Epidemiol Sante Publique 1997; 45: 78–84.
Rolland-Cachera MF, Spyckerelle Y, Deschamps JP . Evolution of pediatric obesity in France. Int J Obes 1992; 16 (Suppl 1): s5.
Nutrition et activité physique. Surpoids et obésité chez l'enfant. In: Salines E, Danet S (eds). L′état de santé de la population en France, rapport 2007. Direction de la Recherche des Etudes de l'Evaluation et des Statistiques, Ministère de la Santé: Paris, 2008. pp 100–101.
Phipps SA, Burton PS, Osberg LS, Lethbridge LN . Poverty and the extent of child obesity in Canada, Norway and the United States. Obes Rev 2006; 7: 5–12.
Shrewsbury V, Wardle J . Socioeconomic status and adiposity in childhood: a systematic review of cross-sectional studies 1990–2005. Obesity 2008; 16: 275–284.
S Péneau and S Hercberg designed the study. S Péneau drafted the paper. S Péneau, B Salanave, L Maillard-Teyssier, MF Rolland-Cachera, A-C Vergnaud, C Méjean, S Czernichow, K Castetbon, S Vol, J Tichet and S Hercberg critically revised the paper for important intellectual content. S Vol and J Tichet were in charge of data collection. S Péneau, B Salanave and L Maillard-Teyssier analyzed the data. All authors have read and approved the final version.
Conflict of interest
The authors declare no conflict of interest.
About this article
Cite this article
Péneau, S., Salanave, B., Maillard-Teyssier, L. et al. Prevalence of overweight in 6- to 15-year-old children in central/western France from 1996 to 2006: trends toward stabilization. Int J Obes 33, 401–407 (2009). https://doi.org/10.1038/ijo.2009.31
- public health
Trends in the prevalence of overweight, obesity and underweight in French children, aged 4–12 years, from 2013 to 2017
Public Health Nutrition (2020)
Journal de Pédiatrie et de Puériculture (2018)
DIFFERENCES IN THE PREVALENCE OF OVERWEIGHT AND OBESITY IN 5- to 14-YEAR-OLD CHILDREN IN KRAKÓW, POLAND, USING THREE NATIONAL BMI CUT-OFFS
Journal of Biosocial Science (2018)
Frontiers in Public Health (2018)