White Newcastle upon Tyne schoolchildren born in 1962 were examined at approximately half-yearly intervals from 9 to 17 years of age. During the last year of compulsory education losses to follow-up became selective with respect to socio-economic background, and the results reported here are restricted to 564 boys and 669 girls seen regularly between the ages of 10 and 15 years at least. Newcastle adolescents were slightly shorter and lighter than the London children on whom Tanner's British Standards were based. The longitudinal pattern of skinfold changes in boys was different from that observed in girls. Differences in height attained between occupational groups were well established by the age of five years, and the same was true of differences between family-size groups for children with fathers in manual occupations. It is inferred that differences in height between occupational and family-size groups resulted more from differences in leg than in trunk length. At age 15, the correlations of height attained with maternal height were 0.42 and 0.43, and with birthweight 0.25 and 0.28, for boys and girls respectively. Adjustment, by regression, of heights attained for birthweight, age at peak height velocity and maternal height diminished differences between occupational and family-size groups. After such adjustment, heights significantly lower than average were found only among subjects with fathers in unskilled manual occupations and with large families. Among subjects with fathers in manual occupations, subscapular and triceps skinfolds and arm circumference decreased significantly with increasing family size.

669 boys and 753 girls born in 1962 and living at home were observed at approximately half-yearly intervals from 9 to 17 years of age. Mean ages of reaching various developmental stages were calculated for voice change and facial hair in boys, menarche and breast development in girls, and for axillary hair in both sexes. Peak height velocity and age at which it occurred were also calculated for both sexes. Means were generally in good agreement with those found about 15 years previously in the Harpenden Growth Study. Most pubertal changes in the present study showed gradients by social class; in boys, development was later in those with fathers in unskilled manual occupations, whereas in girls the greatest differences, where found, were between those with fathers in non-manual occupations and the remainder.

Data from Frisch and Revelle (1971) and Frisch et al. (1973) are used to show that some of their biological conclusions relating body composition to age at menarche follow directly from arithmetical operations on the two observed items of information: weight and height. This emphasizes the dangers of drawing biological conclusions from derived, as opposed to observed, data. The hypothesis that an age-unrelated critical body weight triggers off menarche is examined in relation to preliminary data from a longitudinal study of adolescent girls. It is unacceptable because: (a) mean weights at menarche showed an upward trend with increasing age; (b) for given body weights the proportion of girls starting to menstruate increased with age; (c) at all ages the variation of body weight at menarche was as large as that among non-menstruating girls: (d) only 41 per cent of girls started to menstruate at weights 48 +/- 5 kg; and (e) eleven of the 20 girls whose weight exceeded 48 kg at the age 10-5-11 years did not start to menstruate although 1-1-5 years had elapsed since they attained the "critical" weight. Although no direct body composition data are presented the behaviour of skinfold suggests that body composition is unlikely to be the primary trigger of menarche.

Age at menarche was recorded for 699 white Newcastle upon Tyne girls taking part in a longitudinal study of growth and development during puberty. Girls from large (5+) families started menstruation about five months later than girls from small families. This effect was observed only among girls with fathers in manual occupations. The occurrence of menarche had a peak in January and a possible secondary peak about August. Height and to some extent weight increased with the age at menarche, while arm circumference and skinfolds tended to decrease. The correlation of age at menarche with weight at given height was -0.13 and that of age at menarche with height at given weight was 0.38. The correlation of age at menarche with skinfolds and arm circumference for given weight were of the order of -0.3 and for given height of the order of -0.2. The time lapse between the age at peak height velocity (PHV) and the age at menarche decreased with increasing subscapular skinfold at PHV.

No abstract