Levels of fatness and size attainment.

In a generally low-income group of 4,888 White boys and girls, 740 obese children (above the 85th percentile for the triceps fatfold) stand systematically taller, by 0.64 2 scores, than 840 lean boys and girls (below the 15th percentile for fatness) a difference increasing to 1 S.D. at ages 11-12, suggesting a direct effect of the level of fatness on standing height and a cumulative fatness-related difference in the annual rate of growth. Clinically-obese boys and girls tend both to greater stature and to demonstrated maturational advancement (Wolff, ’55; Quaade, ’55). Within a study population, unselected for the degree of fatness, those children who are fatter (as radiogrammetrically determined) are both larger in body size and advanced in both skeletal and sexual maturation (Garn and Haskell, ’59, ’60). It is not certain, however, whether the fatness-size relationship is part of a developmental continuum, or primarily the contribution of the obese, “Pickwickian” extreme. There is need to ascertain whether lean children are as size-reduced as obese boys and girls are stature-increased, in a sizeable population sample, and in population-context. To resolve this question, we have newly examined (a) triceps fatfold data and (b) stature (standing height) data on 4,888 boys and girls of European derivation, and generally from lower income families. All children were participants in the TenState-Nutrition Survey of 1968-1970, and from the states of California, Kentucky, Massachusetts, Michigan, New York, South Carolina, West Virginia, and Washington. Fatfold data quality has been carefully examined, state by state, and is the subject of a separate publication (Garn, ’73). At each midpoint age, from age 1 to age 12, lean boys and girls (below the 15th percentile for the triceps fatfold) and obese boys and girls (above the 85th percentile for the triceps fatfold) were identified by a special computer program, and then stature-compared. The “cut-off” values for lean and “obese” subjects are given, by age and sex, in the table. Statures were then compared with the median stature values for the group as a whole, and expressed (for each age and sex) as Z scores or standard deviation units relative to the group as a whole. The present data analysis, therefore, consists of 740 lean boys and girls, and an almost equal number of 780 obese boys and girls, and their age-specific statures as compared with the total group of nearly 5,000 children in the total sample from five geographical regions. As shown in the table (table 1) the 780 obese boys and girls, defined in terms of age-and-sex specific fatfolds, tend to be taller than the 740 lean boys and girls, by as much as 6 cm and more. This sizetrend is true without exception for the boys (12 out of 12 comparisons) and nearly so for the girls (11 out of 12 age-comparisons). The differences are highly significant by sign test (23:l as against the chance or 12: 12 distribution). Expressed relative to the stature medians for the total group, the lean boys and girls averaged -0.21 Z scores, i.e., 0.2 S.D. below stature expectancy. In contrast the obese boys and girls (above the 85th percentile for the triceps fatfold) averaged a weighted 0.48 Z scores aboue stature expectancy for age and sex, or nearly 0.5 Standard Deviations. Overall, the rounded Z score differences for stature (comparing the obese with the lean) approximates 0.64 Z for the boys and a closely similar 0.68 Z for the-girls from age 1 through 12. There AM. J. PHYS. ANTHROP., 40: 4 4 7 4 5 0 44 7 448 S. M. GARN, D. C. CLARK AND K. E. GUIRE