FIFTEEN YEAR TRENDS IN BODY MASS INDEX IN A LARGE POPULATION OF AMERICAN CHILDREN: 1980-1994 338

Abstract

In recent years there has been considerable interest in and speculation about changes in body composition of American children. Because of costs and other difficulties associated with skinfold and densiometric measurement, body mass index (BMI) has been employed frequently in the estimation of fatness in large populations. Most reports of trends, however, have had to rely on estimates derived from extrapolations from two or three data points separated by as many as 10 years. This study was designed to monitor year-to-year age-specific BMI scores in a large (N=500,000), relatively stable population of 6 to 14 year-old boys and girls over a 15 year period. Cross sectional random samples averaging 11,000 subjects were drawn annually from 1980 through 1994. All states were represented and regional distribution percentages closely approximated those of the 1990 U.S. census of school age youngsters. Mean BMI scores were calculated by age and gender for each year and subjected to ANVOA and linear trend analysis. Significant positive linear trends(P≤.01) were observed in all age groups for both boys and girls. Age related means for 1984 and 1986 did not differ significantly (P≤.05) from those reported for corresponding years in the National Children and Youth Fitness Surveys I & II. In general, absolute increases in BMI between 1980 and 1994 were greater among girls than boys. The mean increase for all girls was 0.93 units while that for boys averaged 0.81 units. Despite variability among age groups there was no consistent relationship between age and magnitude of increases. It is concluded that this population experienced a meaningful change in BMI across the 15 year period from 1980 to 1994 indicating a gradual increase in body fatness for all age groups.