Birth statistics for the Johannesburg Metropolitan Region were collected for 757 151 confinements from 1969 to 1989 (467 513 Black, 194 375 White, 67 250 Coloured and 28 013 Indian confinements). From 1969 to 1978 data on the sexes of twins were also collected for 375 203 of the confinements (203 504 Black, 129 631 White, 28 253 Coloured and 13 815 Indian confinements). A twin confinement was defined as two deliveries during one confinement. Twinning rates (TRs), defined as the number of twin confinements per 1000 total confinements, were calculated per year for each population group and from 1969–1978 estimates of the relative proportions of dizygotic (DZ) and monozygotic (MZ) twins were calculated and thus the relative DZTRs and MZTRs. A significant decline in Black and Coloured TRs was observed between 1969 and 1989. A significant decline in Black DZTR but not Black MZTR was observed between 1969 and 1978, the Coloured twin sample was too small to show significant trends over this period. It is probable that the overall decline in Black twinning may be explained by a decline in the DZTR. An analysis of birth statistics for 159 748 confinements (134 504 Black and 25 244 White confinements) collected as part of a prospective study of TRs in the Johannesburg Metropolitan Region from 1988 to 1990, indicated that the Black TR continued to decline at least until the end of 1990. TRs in the Johannesburg Metropolitan Region calculated from City Health Department birth statistics collected from 1988 to 1990 were: 13.8 and 10.77 for the Black and White populations, respectively. TRs for this period calculated from the combined birth statistics of 14 hospitals, nursing homes and maternity clinics across the region were: 12.4 and 10.88 for the Black and White populations, respectively.

The predictive relation between neonatal reactivity and 12-, 18-, and 24-month temperament was assessed for 22 appropriate-for-gestational-age (AGA)/small-for-gestational-age (SGA) twin pairs. Neonatal reactivity included ratings of visual and auditory orienting responses to a bull's-eye, rattle, bell, voice, and face plus voice combined, as well as alertness. Infant temperament ratings were made of emotional tone, activity, attentiveness, and social orientation to staff during ageappropriate activities in the laboratory. There were no group differences in the neonatal or infant ratings. For AGA twins, a pattern of predictive correlations was obtained between neonatal reactivity and temperament at each age. A similar predictive pattern was not observed for the SGA twins. Group differences in rates of maturation related to stresses associated with intrauterine growth retardation influenced the significance of the neonatal variables for later temperament development.

We examined the cumulative prevalences of 22 symptoms thought to reflect immune system function reported in a questionnaire mailed to 7616 Australian twins. The associations between symptoms and demographic variables were expressed in terms of polychoric or polyserial correlations, and a principal components analysis performed. Factors representing underlying propensities respectively to allergic disease, various minor infections, diseases associated with aging such as arthritis, skin disease, and respiratory tract infection were extracted. Possible processes underlying these symptom clusters and the relative strengths and weaknesses of this type of analysis are discussed.