Data from the final 16 years of a 28-year ongoing survey of menarcheal age are reported. From 1971 onwards, recalled aged at menarche was recorded for all young women entering the University of Warwick. These data show that mean menarcheal age is increasing, a trend which is independent of father's occupation, family size, position of girl in the family, and physique. This continues a similar upward tendency noted in a preceding study in Swansea. It appears that the downward trend to earlier age at menarche of the earlier decades of this century has been replaced by one in the opposite direction.

Data from a study of teenage sexual activity among secondary school girls show the need for a sex education policy as a first step in controlling teenage fertility in Zambia. A large proportion of teenage females enter into close relationships with males at young ages and a high proportion of young females have engaged in sexual intercourse. Most of these sexually active females do not use family planning methods even though a large proportion of them have heard of modern methods. The teenagers receive very little sex education from their parents and a modern institutional sex education programme is needed.

The study compares biological, socioeconomic and behavioural determinants of low birthweight in Cameroon and the United States. Some factors in low birthweight are found to be cross-national, but others are specific to the setting. Positive risk factors of low birthweight in both countries include unmarried motherhood, female sex, multiple births, and preterm births. Outcome of the previous pregnancy is a positive risk factor in the US, but not in Cameroon. Significant negative risk factors include prenatal care visits (in both countries), mother's education (in the US only), births to mothers aged 20–34 and birth orders of 2 or more (in Cameroon only). Separate analyses of all births and the subsamples of singleton births reveal that estimates for the two groups differ only marginally.

Genetic distances between all possible pairs of counties (n = 37) in the state of Lara, Venezuela were calculated using surname frequencies and the Euclidean distance as estimator. In general, Euclidean distances were smaller between counties closer together, and the product moment correlation between geographic and Euclidean distances was 0·49 (p < 0.02). The results suggest that, in Lara, geographic distance has been an important determinant of genetic structure, although topography and roadways also have had an important influence upon this structure.

The link between socioeconomic status and health has long been recognised. This study of deaths among Australian men aged 15–59 years demonstrates that during the 20-year period, 1966–86 the number of premature deaths was dramatically reduced among all socioeconomic groups, primarily as a result of falls in death rates due to heart disease, stroke and trauma. However, the marked differences in death rates according to social class remain, to the extent that if men of all social classes had the same mortality experiences as professional and technical workers the overall death rates for Australian men would be reduced by 60%. Socioeconomic status is the most important indicator of health status among Australians.

Analysis of data on 106,848 marriages in the cities of Bangalore and Mysore, South India, between 1980 and 1989 showed that levels of consanguineous marriage varied between cities through time and by religion. The average coefficient of inbreeding was higher in Bangalore (F = 0·0339) than in Mysore (F = 0·0203), principally reflecting large-scale, post-Independence rural migration into Bangalore. Although there was some evidence of a decline in consanguineous marriages in Mysore, there was no convincing support in either city for earlier projections of a rapid reduction in the popularity of unions between close biological relatives.

The rate of consanguineous mating in north Jordan was 63·7% with an average inbreeding coefficient of 0·0283. First cousin marriages were the most predominant type (58·5%); villagers showed higher levels than city dwellers. The level of female education did not influence consanguinity rate.

Using a sample of 3080 women from Cebu, Philippines, interviewed immediately after childbirth and every 2 months thereafter for 2 years, the determinants of return to coitus are analysed. Eighty percent of the women returned to coitus before the return of menses, and 90% did so before stopping breast-feeding. In hazards models, variables associated with traditional life styles retarded return to coitus. Situational variables (husband resident, crowding, children aged 6 and under), especially husband residence, and biological variables (mother's age, return to menses, and lactation) were significant predictors of return to coitus. The implications of the analysis for the construction of models of birth interval dynamics are discussed.

Sex differences are reported for samples of approximately 1000 16-year-old blacks, Indians and whites in South Africa on ten tests of cognitive ability. Males obtained significantly higher means on non-verbal reasoning, spatial and mechanical aptitude, and females obtained significantly higher means on perceptual speed and memory for meaning (except among the black sample). In general the sex differences in South Africa are consistent with those typically obtained in the United States.

Data from the Cebu Longitudinal Health and Nutrition Survey have been used to examine breast-feeding patterns in the Metropolitan Cebu region of the Philippines. The results show that there is extensive and early use of non-nutritive liquids, resulting in low levels of exclusive breast-feeding. The use of cross-sectional versus longitudinal data, the definition of exclusive breast-feeding, and movements in and out of the exclusively breast-fed category, are considered.