Self-reported cold data have been analyzed for 95 pairs of identical twins aged 14–64 who took part in a double-blind trial of vitamin C tablets. One member of each twin pair took a tablet containing 1 g vitamin C and the other took a well-matched placebo each day for 100 days. In the total sample there was no effect of vitamin C in preventing colds. However, subdivision of the data showed a significant preventive effect of the placebo in the 51 pairs living together and an equal and opposite preventive effect of the vitamin C in the 44 pairs living apart. The placebo effect in the pairs living together may be attributed to the large proportion who wrongly perceived which treatment they were taking. The reduction of colds in the vitamin C group of the pairs living apart was about 20%. There were significant correlations between cold symptoms reported and the personality trait of neuroticism. No side effects or substantive changes in serum biochemistry could be attributed to the vitamin C dose.

The expected correlation between polar body twins depends upon the frequency and distribution of chiasmata occurring during oogenesis. We have estimated the expected values of these correlations using assumptions believed to be biologically valid. The expected value of the overall correlation between first polar body twins is estimated to be 0.38 and that between second polar body twins is estimated to be 0.51. There is little difference between the correlation of DZ twins and that expected for second polar body twins.

The correlation for a specific locus depends upon its distance from the centromere for both types of polar body twins. This effect is especially marked for second polar body twins. For second polar body twins, loci at or near the telomere will have a correlation close to that expected for first polar body twins. It is concluded that only those marker genes that lie at or close to the centromere are appropriate for the detection of polar body twinning.

The application of multiway contingency table analysis to the multivariate analysis of concordance ratios in twin studies is developed. The approach is illustrated by data on smoking and alcohol use in Finland and Sweden. This approach can enable the assessment of the effect of other variables on the concordance ratio and thus allow estimates of genetic effects on the trait under study. Hypotheses on relationships between genetic effects and other variables can be tested. After hypothesis testing, model fitting of the best hypothesis can be carried out.

The rates of infant mortality of twin individuals were 4.38% and 7.76% for mothers healthy and nonhealthy after delivery, respectively, and the difference is significant at the 0.01 level. The lowest infant mortality rate was seen in the mothers with paid work during pregnancy (4.56%), followed by the mothers engaged only in housekeeping (4.72%) and by those self-employed during the pregnancy (4.99%). Infant mortality rate for MZ twins decreased with increased monthly expenditure of the household, whereas the rate for DZ twins remained constant with expenditure. Socioeconomic factors still affect the infant mortality of twins, and the infant mortality rate can be improved.

The rate of infant mortality of triplet individuals (deaths under one year of age) was computed using 34 sets of triplets born in the first half of 1974. The rates were 8.82%, 9.68%, and 10.34% for the first-, second-, and the third-born triplets, respectively. For males and females the rates were 8.33% and 10.34%, respectively, and the difference was not significant. The rate decreased with gestational age up to 32-35 weeks. For those with heavier weight at birth (≤2,000 g) the rate was lower (0%) than for those with lighter weight (<2,000 g, 8.16%), but the difference is not significant (P = 0.087). Infant mortality rate of triplets decreased with increased monthly expenditure of the household.

The best time for twin delivery is at 38 or 39 weeks gestation. Systematic Caesarean section or induction is decided upon during these weeks. Thanks to this policy, the good results formerly obtained have been maintained and an important decrease in emergency Caesarean sections has been noted thus reducing the risks inherent in this method. However, a randomized trial is necessary to confirm these results.

Anti-A and anti-B titers were measured in 60 twin pairs and their families. Family data shows high correlations of antibody titer values for first degree relatives but no significant spousal correlations.

In a recent paper in this journal [5], Wilson offers a revision of his previously published method [4] for diagnosing the zygosity of twins from bloodtype findings. Wilson notes that his revised method will yield the same results (given the same estimates of gene frequencies) as will the methods described by Smith and Penrose [2], Sutton et al [3], and Lykken [1], and he advocates his new method, over the others, as being “simplest and most direct.” He also reinterprets various quantities employed by Lykken and proposes a different index of the efficiency of a genetic marker for zygosity diagnosis. In this note, I attempt to clarify the relationship between Wilson's revised method and my own, so that the reader can more accurately assess their relative advantages.