|
|
||||||||
Department of Family Relations and Applied Nutrition, University of Guelph, Guelph
Department of Sociology and Anthropology, University of Windsor, Windsor, CANADA
Address reprint requests to: Susan Evers, PhD, Department of Family Relations and Applied Nutrition, University of Guelph, Guelph, ON N1G 2W1, CANADA. E-mail: severs{at}uoguelph.ca
| ABSTRACT |
|---|
|
|
|---|
Methods: annual interviews were completed with parents (primarily the mother) living in economically disadvantaged communities in Ontario who are participating in the Better Beginnings, Better Futures project. Weight and height were measured annually for the children (n=760) beginning in JK. Risk of overweight was defined as body mass index (BMI)
85th to < 95th percentile; overweight was BMI
95th percentile. Parents height and weight were self-reported; BMI
25 was considered overweight.
Results: the risk of overweight among children ranged from 14.1% to 17.5%; the prevalence of overweight increased from 9.9% to 15.2%; 68.2% (15/22) of the children who were overweight in JK were >95th percentile in grade 3. BMI
85th to < 95th percentile or
95th percentile in JK were strongly predictive of overweight in grade 3. Almost 50% of the mothers were overweight.
Conclusions: A high prevalence of overweight was found in young children; and, for a large proportion, their early weight status persisted. Strategies promoting healthy eating and physical activity for both children and parents are essential.
Key words: overweight, obesity, children, body mass index, secular trends
| INTRODUCTION |
|---|
|
|
|---|
The prevalence of obesity is inversely related to socioeconomic status among adult women [12,13]. Among men, however, findings of an association are inconsistent, with reports that increasing income is associated with higher rates of overweight and obesity [13], and other indications that income and obesity are not related [12]. There is little evidence that the prevalence of overweight in children is linked to household income [3,14,15].
We have collected anthropometric measurements from children annually since 1994 as part of the Better Beginnings, Better Futures project. Better Beginnings is a prevention initiative in low income communities in Ontario; details of the design are given elsewhere [16]. In brief, each community receives funding to develop programs focusing on ages 0 to 4 years or 4 to 8 years. In cross sectional analyses of data collected prior to program implementation, we reported a high prevalence of overweight among both parents [17] and children [18,19].
In 1994 we began the longitudinal phase of the study adding three demographically similar comparison communities to control for secular changes that could have an influence on the results. These communities do not receive any funding from the Better Beginnings project; however, the data collection procedures are identical. In communities focusing on ages 4 to 8, we recruited families with children in junior kindergarten (JK), and in the succeeding 3 years we recruited families moving into the neighborhood with a child in the same grade.
With five years of anthropometric measures available, the goals of the present analyses were to determine the annual prevalence of overweight, to see if weight status in JK persisted to grade 3, and, to identify factors associated with overweight in children.
| METHODS |
|---|
|
|
|---|
The Better Beginnings, Better Futures project has been approved by the tri-university (Queen's University, University of Guelph, Wilfrid Laurier University) Human Subjects Committee.
Anthropometric Measures
The protocols for measuring height and weight were developed using established guidelines [20]. Each measurement was taken twice: height to the nearest 0.1 cm using a modified tape measure (Microtoise, CMS Weighing Equipment, London, UK); weight to the nearest 0.5 lb with a strain-gauge digital scale (WonderscaleTM, Health-o-meter, Inc., Bridgeview, Illinois). If the two measures differed by 0.5 cm, or 0.5 lb respectively, a third measurement was required. The average of the two closest measures was recorded. Following the training of interviewers, pilot studies (n=10–15) were conducted in each community. Quality control procedures also included periodic reviews of the measurement techniques with each interviewer throughout the data collection period.
We calculated the body mass index (kg/m2) (BMI) [21]. Overweight was defined as a BMI
95th percentile compared to age-and sex-specific NCHS reference data [22]. Children with a BMI
85th to < 95th percentile were considered at risk of overweight [22]. Overweight in parents was defined as BMI
25 [23].
Analysis
Data were analyzed using the SPSS statistical analysis software (version 11.0, SPSS Inc., Chicago, Illinois). Analyses were considered statistically significant at the .05 level. We used backward stepwise (conditional) logistic regression to assess the association of maternal BMI, maternal education (<high school, completed high school, post-secondary), maternal birthplace (in Canada vs other), maternal age, poverty status (
or > Low Income Cut-off based on city of residence, household size, and annual income [24]), and sex of the child with being overweight at JK. We used the same variables plus child's BMI at data gathering points from JK through Grade 2 to identify factors predictive of overweight in grade 3 (age 9 years).
| RESULTS |
|---|
|
|
|---|
85th percentile between the Better Beginnings and comparison communities; therefore, data were merged for these analyses. The annual prevalence of risk of overweight (BMI
85th to < 95th percentile) ranged from 14.1 to 17.5% over the five years; the proportion who were overweight (
95th percentile) increased from 9.9% in JK to 15.2% in grade 3 (Table 2).
|
|
|
In JK anthropometric data had been collected for 43 fathers and 210 mothers, therefore we used the weight status (BMI < 25 or
25) of mothers only in the logistic regression. Maternal overweight was the only variable predictive of child overweight in JK (OR = 1.13; 95% CI 1.03, 1.23). Only two factors were predictive of BMI
95th percentile in grade 3: being at risk of overweight in JK, and being overweight in JK. Children whose BMI was
85th to < 95th percentile in JK had almost 6 times the risk of being overweight four years later compared to those < 85th percentile (OR = 5.82; 95% CI 2.03, 16.67). The greatest risk, however, was associated with BMI
95th percentile in JK (OR = 108.67; 95% CI 25.1, 467.21). Sex of the child, maternal BMI, marital status (single parent vs married/living with a partner), maternal age, household income (< or
LICO), year of immigration, and maternal education level were not associated with BMI
95th percentile in grade 3.
Based on the data collected as part of the grade 3 parent interview the prevalence of overweight was 66.7% (38 of 57) for fathers and 48.5% (211 of 435) for mothers.
| DISCUSSION |
|---|
Unlike Dwyer et al [1] we did not find a sex difference in the prevalence of overweight but this may not appear until children are older. Carriere [8] noted that 23% of Canadian males aged 12 to 19 years were overweight or obese compared to 13% of females. We are now collecting anthropometric data every three years; future analyses may confirm not only an increase in the overall prevalence of overweight but higher rates among adolescent males. Over 31 % of the children were
85th percentile by grade 3, suggesting an increase in risk of overweight in adolescence.
The proportion of adults whose BMI exceeded 25 is considerably greater than in other reports of obesity in Canada. In the National Population Health Survey, BMI was
25 for 57% of males and 35% of females [26]; the prevalence among parents in the present study was 10% and 13% higher, respectively. Obesity is an established risk factor for both non-insulin dependent diabetes and cardiovascular disease in adults. The long term consequences of overweight in childhood are also alarming. In the Bogalusa Heart Study, 77% of participants whose BMI was above the 95th percentile in childhood had a BMI of 30 or greater as adults [27]. Recently, Sinha et al [28] reported that 25% of obese children had impaired glucose tolerance, and there is evidence of unfavorable lipid profiles and hypertension [29] among children who are overweight. Such findings argue for population-based initiatives to decrease the prevalence of overweight and increase levels of physical activity for both adults and children.
Implementing such initiatives in economically disadvantaged communities presents a considerable challenge. Compared to middle- and upper-income families, barriers to increasing physical activity include a lack of transportation, the cost of equipment and fees for certain sports, a lack of child care for siblings, and a shortage of well-equipped playgrounds. Overweight in children is associated with more hours spent in sedentary activities, such as watching television [2,30]. This kind of activity may be preferred by parents who feel that their neighborhood is unsafe.
Schools can increase the physical activity levels of children through daily physical education. A recent review of school-based interventions to promote physical activity noted that most were more effective at improving knowledge rather than increasing activity levels [31]. Such interventions need to be multi-faceted, incorporating environmental changes and focusing on improving behaviours. Unfortunately, this approach alone is unlikely to have any impact on the weight status or activity levels of parents. The number of men who participated in the annual interviews is small but the prevalence of overweight among women alone is sufficient to warrant an intensive intervention. In the last two years of follow-up, over 40% of the women had a BMI of 25 or greater. Furthermore, parental height and weight were self-reported; thus our estimate of the prevalence of overweight among women is likely an underestimate because of the tendency for females to underreport their weight [32].
In earlier reports from Better Beginnings there was no difference in the energy and macronutrient intake of overweight children compared to those of normal weight either at age 4 [18] or age 8 years [19]. We do not advocate restricting the energy intake of children; they are still growing and already have difficulty meeting requirements for several nutrients such as calcium and folate [18,19]. Instead, the emphasis should be on the development of healthy eating habits. In a survey of the health behaviours of children in grades 4 to 8, 38% reported that they ate potato chips or a similar snack food every day, and 36% consumed soft drinks daily [33]. Both foods are typically available in vending machines in schools. The Ontario government recently established a policy on nutritional standards for foods and beverages sold through vending machines in elementary schools [34] which replaces high fat/low nutrient density foods with low fat/nutrient-rich foods. This is consistent with the recommendation of the US Committee on Prevention of Obesity in Children and Youth to develop nutrition standards for foods and beverages sold in schools [35]. Work is currently underway to develop nutritional guidelines for student nourishment programs (largely breakfast and snack programs) across Ontario. Unlike the US, Canada does not have federally mandated child nutrition programs. Instead, such programs are developed locally; these provincial guidelines will help to ensure that the foods being served are of high nutritional quality. Both initiatives are consistent with the recent US recommendations to develop nutrition standards for food and beverages sold in the schools [35].
The focus of many school-based intervention programs has been to improve both eating behaviours and levels of physical activity. Although some programs have been successful in achieving these goals, there have been few experimental studies [36]. The recent success of the implementation of a coordinated program which included components of the Centers for Disease Control and Prevention (CDC) guidelines for school programs [37] is encouraging [38]. The weight status, dietary intake and physical activity of students in schools with a nutrition program providing healthy menu alternatives were compared to students in schools following the CDC recommendations. Although there were no differences in physical activity outcomes, there were positive changes in weight status and diet. Students in the CDC program schools had half the prevalence of overweight and increased consumption of fruits and vegetables. Thus it is possible to effect significant changes within the school environment that reduce the risk of overweight.
Many parents would also benefit from improved eating behaviours, a reduction in energy intake, and an increase in energy expenditure. A positive change in the diets of parents could result in better eating habits among children. Dietz and Gortmaker [39] advocate a family-based approach based on modifiable influences on diet and activity. A recent review of trials with a family involvement suggests that parental involvement is associated with weight loss in children [40]. While theoretically sound, this presents a challenge to low income families. Interventions promoting increased consumption of fresh fruits and vegetables and reduced intakes of foods high in fat and low in fiber will not work unless economically disadvantaged households have access to appropriate foods. Increasing physical activity may well present an even greater challenge among largely sedentary and overweight adults.
Anthropometric measures are being collected in grade 6 and again in grade 9 for this cohort. We will have an opportunity to see if weight status at JK persists to these later ages. School-based interventions need to be evaluated for both the short-term and long-term impact on weight status, eating and activity behaviours. Evaluations of such programs have assessed short-term goals and there is an urgent need to determine whether improvements in behaviours can be sustained. If so, this would argue for implementing changes to the school environment on a permanent basis.
| ACKNOWLEDGMENTS |
|---|
|
|
|---|
Received September 20, 2004. Accepted April 14, 2006.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |