Locally implemented prevention programs may reverse weight trajectories in half of children with overweight/obesity amid low child-staff ratios: results from a quasi-experimental study in France.

BACKGROUND: The aims of the present study were to assess changes in weight status between the first and last year of primary education among children with overweight/obesity in response to locally implemented school-based prevention programs, and to assess the influence of process indicators, expressed as child-staff ratios (CSRs), on these changes. METHODS: To meet the study objectives, a quasi-experimental design was used. Four municipalities that systematically monitored the weight status of schoolchildren and participated in the "Vivons en Forme" program agreed to provide the data available in their school medical service records. The local implementers involved in training sessions were mainly municipal staff in charge of serving midday school meals, which is compulsory in France, and those in charge of designing and facilitating creative, interactive activities at school between and after classes. CSRs were determined by occupation (school catering service/facilitator of extracurricular activities) and training session (healthy eating/physical activity) in each municipality program, and classified as low (1-5 children per adult) or moderate. RESULTS: During the 4 years of primary education, weight status improved in half of the children with overweight/obesity, and worsened in 6.6% of children with overweight/normal weight. In children who remained overweight, the BMI z-score diminished over time. Estimates of the positive 4-year weight changes were related to low CSRs in locally implemented variations of the program. Estimates increased with age and were significantly higher in low-to-moderate CSR multicomponent interventions than moderate CSR single-component intervention (reference). The moderate CSR multicomponent intervention had a similar effect as the reference. The estimated negative weight change decreased with age. CONCLUSIONS: Our findings suggest that training ancillary school staff in experiential-focused interventions for healthy eating and physical activity in locally implemented school-based programs contributed positively to reducing childhood obesity during the four years of primary education without interfering with educational activities. The results also provide preliminary evidence that low CSRs could be pivotal for optimal outcomes, especially in deprived areas.

Parent-offspring correlations in pedometer-assessed physical activity.

BACKGROUND: Physical activity is a major component of a healthy lifestyle in youth and adults. To identify determinants of this complex behavior is an important research objective in the process of designing interventions to promote physical activity at population level. In addition to individual determinants, there is evidence documenting familial influences on physical activity. However, the few studies that have addressed this issue with objective measures did not provide data on parent-offspring physical activity relationships throughout childhood and adolescence. The purpose of this study was to assess familial correlations in pedometer-assessed physical activity. METHODS: We measured ambulatory activity in 286 French nuclear families (283 mothers, 237 fathers, and 631 children aged 8-18 years) by pedometer recordings (Yamax Digiwalker DW 450) over a week. Correlations were computed with their 95% confidence intervals (CI) for spouse pairs, siblings, mother-offspring, and father-offspring. Data were expressed as steps per day and computed both for the full recording period and separately for weekdays and weekends. RESULTS: The correlations were the highest between siblings (r=0.28, 95%CI: 0.17-0.38). Parent-offspring correlations were significant in mothers (r=0.21, 95%CI: 0.12-0.30), especially between mothers and daughters (r=0.24, 95%CI: 0.12-0.36 vs. r=0.18, 95%CI: 0.05-0.31 for sons), but were almost nonexistent in fathers. Correlations were generally higher on weekend days compared to weekdays. Mother-offspring correlations did not decrease with increasing age of children (r=0.17, 95%CI: 0.00-0.34 in 8-11-year-olds, r=0.20, 95%CI: 0.07-0.33 in 12-15-year-olds, and r=0.25, 95%CI: 0.07-0.39 in ≥16-year-olds). Finally, between-spouse correlations were significant only during weekend days (r=0.14, 95%CI: 0.01-0.27). CONCLUSION: Ambulatory activity correlated within families, with a possible mother effect. Mother-offspring correlations remained significant through the transition from childhood to adolescence. Further studies are required to better understand the respective influences of shared activities, parental modeling and support as well as genetic factors on the familial aggregation of physical activity.

Parental body size and early weight and height growth velocities in their offspring.

BACKGROUND: Whereas weight or height at a given age are the results of the cumulative growth experience, growth velocities allows the study of factors affecting growth at given ages. AIM: To study the relationships between parental height and body mass index (BMI) and offspring's height and weight growth during infancy and childhood. STUDY DESIGN: From the FLVSII population-based study, 235 parent-child trios belonging to 162 families examined in 1999. OUTCOME MEASURES: From medical records and previous FLVS examinations, child's height and weight history were reconstructed. Weight and height growth velocities from birth to seven years were estimated from a modelling of individual growth curve and correlated with parent's body size in 1999. RESULTS: Ponderal index and length at birth were significantly associated with maternal but not paternal BMI and height. In the first six months, height growth velocity was significantly associated with maternal stature (at three months: 0.12+/-0.05 and 0.02+/-0.05 cm/month for a 10 cm difference in maternal and paternal height respectively) and weight growth velocity with paternal BMI (at three months: 5.7+/-2.8 and 1.9+/-2.3g/month for a difference of 1 kg/m(2) in paternal and maternal BMI respectively). Between two and five years, height growth velocity was more significantly associated with paternal height whereas weight growth velocity was more closely associated with maternal BMI. CONCLUSIONS: Early childhood growth is characterised by alternate periods associated specifically with maternal or paternal BMI and height. This novel finding should trigger the search for specific genetic, epigenetic or environmentally shared factors from the mothers and fathers.

Relationships of self-reported physical activity domains with accelerometry recordings in French adults.

The objective was to examine the relationships of self-reported physical activity (PA) by domain (leisure, occupational, other) with PA and sedentary time as measured objectively by accelerometry. Subjects were adults with low habitual PA levels from a community in northern France. Among subjects in the lowest tertile of a PA score from a screening questionnaire, 160 (37% males, age: 41.0 +/- 10.8 years, BMI: 25.1 +/- 4.1 kg/m(2), mean +/- SD) completed a detailed instrument (Modifiable Activity Questionnaire), and wore an accelerometer (Actigraph) for seven consecutive days. Relationships between questionnaire domains (occupational, leisure, and "non-occupational non-leisure") and accelerometry measures (total activity and sedentary time) were assessed using Spearman correlation coefficients. In this population, the highest contributor to total reported PA (h/week) was occupational PA. Time spent in non-occupational non-leisure PA ranked second in women and third in men. The most frequent non-occupational non-leisure PA were shopping and household chores. In women, non-occupational non-leisure PA contributed more than occupational or leisure-time PA to total PA energy expenditure (median: 18.0, 9.1, and 4.9 MET-h/week, respectively). Total PA by accelerometry (count/day) was correlated to leisure-time PA in women (r = 0.22, P < 0.05) and to occupational (r = 0.43, P < 0.01) and total reported PA (r = 0.39, P < 0.01) in men (all in MET-h/week). There was an inverse relationship between accelerometry sedentary time (h/day) and non-occupational non-leisure PA (MET-h/week, r = -0.30, P < 0.001). These findings indicate the importance of assessing non-occupational non-leisure PA for a better understanding of how individuals partition their time between active or sedentary occupations.

Downward trends in the prevalence of childhood overweight in the setting of 12-year school- and community-based programmes.

OBJECTIVE: A school-based nutrition information programme was initiated in 1992 in two towns in northern France (Fleurbaix and Laventie, FL) and was followed by a number of community-based interventions. We took the opportunity to measure the outcomes in terms of childhood obesity and overweight over the next 12 years. DESIGN: Repeated, cross-sectional, school-based survey. For the school years beginning in 1992, 2000, 2002, 2003 and 2004, the height and weight of all 5- to 12-year-old children attending school were measured in FL. In 2004, the same assessments were made in two comparison towns with similar socio-economic characteristics but no intervention. SETTING: Fleurbaix and Laventie (intervention towns), Bois-Grenier and Violaines (comparison towns), northern France. SUBJECTS: In 2002, 2003 and 2004, respectively 515, 592 and 633 children were measured in FL (participation rate of 95-98% of all eligible individuals); in the comparison towns, 349 children were measured in the 2004 school year (98% of the towns' school population). RESULTS: After an initial increase, trends in mean BMI and prevalence of overweight started to reverse. Compared with 2002, the age-adjusted OR for overweight in FL was significantly lower in 2003 and 2004 (but for girls only). In the 2004 school year, the overweight prevalence was significantly lower in FL (8.8%) than in the comparison towns (17.8%, P < 0.0001). CONCLUSION: These data suggest that, over a long period of time, interventions targeting a variety of population groups can have synergistic effects on overweight prevalence. This gives hope that it is possible to reverse trends towards increasing overweight by actions at the community level.

Cardiovascular risk factor levels and their relationships with overweight and fat distribution in children: the Fleurbaix Laventie Ville Santé II study.

This study aimed to document for the first time in a general population of French children the prevalence and levels of cardiovascular risk factors and to assess separately in boys and girls whether these risk factors were associated with fat mass distribution independently of subcutaneous overall adiposity. A cross-sectional analysis of baseline data from 452 children (235 boys and 217 girls) aged 8 to 17 years included in a 1999 population-based epidemiologic study (the Fleurbaix Laventie Ville Santé II study) was made. Overweight was defined according to the International Obesity Task Force references and the 90th percentiles of the French body mass index curves. The thresholds of parameters defining cardiovascular and metabolic risks were the 95th percentile of the Task Force Report on High Blood Pressure in Children and Adolescents for blood pressure and those of the American Academy of Pediatrics for lipids. Anthropometric and biological parameters were described by sex and according to overweight status. Partial correlations between cardiovascular risk factors and anthropometric measures of adiposity (body mass index, sum of 4 skinfold thicknesses, waist circumference, waist-to-height ratio) were calculated. Then, these correlations were additionally adjusted for the sum of 4 skinfold thicknesses. High plasma triglycerides, high insulin concentration, and low plasma high-density lipoprotein cholesterol (HDL-C) concentration were associated with all measures of adiposity (|r| > or = 0.20, P < .002). When obese children were excluded, overweight children already had high triglycerides and low HDL-C levels, respectively, 2 and 20 times more frequently than normal-weight children did. Among overweight children, 7.7% had at least 2 risk factors among high blood pressure, high plasma triglycerides or glucose, and low HDL-C concentration vs 0.25% among normal-weight children (P = .002). After adjusting for the sum of skinfolds, an independent association between the risk factors and waist circumference was found in girls. In conclusion, (a) modest excess weight is associated with increased levels of cardiovascular risk factors. (b) In girls, abdominal fat distribution is associated with cardiovascular risk factors, independently of overall adiposity. (c) International definition of abdominal obesity in children is required to standardize studies and to progress in the evaluation of childhood obesity and its consequences.