A study was conducted on twinning in relatives of consecutive triplet sets in the Åland Islands in the years 1740-1939. The incidence of twinning in sibships of triplets was extremely high, 80/1000 (56/1000 before and 143/1000 after the triplet maternity). In Finland as a whole, 1905-1954, the twinning rate was among mothers of triplets 38/1000, ie, about 2.6 times the rate in general population, and was higher after (48/1000) than before the triplet maternity (34/1000). In the sibships of fathers of triplets there was a low rate of twinning (below 10/1000) both of same-sexed (SS) and of opposite-sexed (OS) triplets. Among sibships of mothers of OS triplets the twinning rate was 18/1000 and among mothers' sibships of SS triplets 26/1000. The series of triplet families from both Åland and Finland as a whole indicate a considerably higher frequency of twinning on the maternal than on the paternal side. The sibships of OS triplets in Finland have higher twinning rates than sibships of SS triplets (50/1000 vs 27/1000). In sibships of triplets, not only the DZ but also the MZ twinning rates were approximately twice as high as those in the general population. The triplet rates in Finland were increasing strongly with maternal age and were in the last century among mothers of 30-39 years of age considerably higher than among mothers from this century. This, in combination with higher mean parity, may explain the high rates of multiple maternities in sibships of triplets in the past. The rate of triplet maternities seems to be more sensitive to sociodemographic changes than the rate of twin maternities. Mothers of triplets in Finland had a high frequency (more than 40%) of prenuptially conceived firstborn children. This, and a short protogenesic interval indicate that triplet-prone mothers are more fecundable, ie, they conceive with greater ease and/or may have a better physical condition than other women for completing a gestation with multiple embryos.

A preliminary series of quantitative genetic models was applied to a subset of longitudinal height data, spanning birth to maturity, gathered from twin families in the Louisville Twin Study. Descriptive Cholesky factor parameterization was found to give more satisfactory results than did a system of constraints based on a model of developmental transmission of a time-constant and time-specific factors. The results from application of two autosomal sex-limitation models are contrasted with those from a model specifying both autosomal and sex-chromosomal patterns of inheritance. The latter model was more conducive to parameter reduction. Although these models do not constitute conclusive tests of autosomal sex-limitation versus sex-linkage, the more parsimonious model is consistent with previous research suggesting a stature locus on the long arm of the Y chromosome. Heritability of height is estimated at about 90% or greater from 6 years of age on. Substantial and fairly constant longitudinal genetic correlations are found from 3 years of age on. Shared environmental effects unrelated to parental height were seen for birth length, corrected for gestational age, to height at 3 years of age, but these are not satisfactorily differentiated from possible twin effects in the present sample. The genetic consequences of assortative mating are emphasized since failure to take assortment into account can lead to overestimation of shared environmental effects and under-estimation of genetic effects. The results indicate that about 20% of within-gender variability for mature height can be attributed to the genetic consequences of assortment, even though the phenotypic marital correlation of 0.22 is quite modest. The importance of testing the assumption of multivariate normality underlying the application of the method of maximum-likelihood is also highlighted.

In previous papers a stochastic model of the ageing process has been proposed. Some genetic parameters (redundance, repair) have been used to explain the observed differential predisposition to the process and family heredity. Because the process is basically due to effective random mutations, any individual of the population would be predisposed differently to ageing according to the structure of his/her genome. In the present paper, the previous model is generalized to take into account an additional genetic parameter, namely, the stability against random mutations, defined as the probability that a random mutation in a codon would produce no mutation in the corresponding protein. Estimation problems connected with the model are approached on the basis of twin data in maximum likelihood estimation as well as in bayesian framework. Some comparisons between the two methods are reported.

The simultaneous analysis of means and covariance structures is applied to longitudinal twin data. Body weight was measured on six occasions in a sample of young female MZ and DZ twins. When average body weight at the first measurement occasion, as well as the increments in weight at later occasions, are specified in the genetic part of the model that also adequately explains the covariance structure, a good fit is obtained. In this application the increase in body weight at each occasion is weighted by the square root of the genetic variance innovation terms that represent the new genetic variance entering into the process.

The population-based twin registries of Denmark, Finland, Norway and Sweden represent an extraordinary resource for scientific research. Although each register has its own history and composition, they share certain common qualities. All the Nordic countries have a long tradition of population registration, a high standard of living and health-related registers of high quality. The large size of the registers means that they are uniquely placed for representative studies of rare occurrences. Examples of studies that these registers make possible are illustrated with data from the Finnish Twin Cohort, which in its first phase consisted of over 17,000 like-sexed twin pairs born before 1958. It has been recently expanded to include multiple births between 1958 and 1986 (nearly 23,000 sets) and their first-degree relatives. During this period the DZ/MZ ratio decreased, while an increase in MZ twinning rates was observed between 1974 and 1986.

Multiple birth rates in entire Japan were analyzed using vital statistics for 1951 to 1988. The triplet rate was nearly constant from 1951 to 1974, where the rate per million births was 58, then increased with the year up to 1982 (104), and decreased up to 1984, and suddenly increased thereafter (109 in 1987). The average rate of quadruplets per million births from 1951 to 1968 was 0.93, then increased with the year up to 1975 (7.5), and decreased until 1984 and suddenly increased thereafter (10.6 in 1987). The rate of quintuplets was 0.77 per million births during the period from 1975 to 1987. The higher multiple birth rate since 1975 was attributed to the higher proportion of mothers treated with ovulation-inducing hormones in Japan. Since 1985, higher multiple birth rates might be partially attributed to in vitro fertilization. The stillbirth rates for male triplets gradually decreased from 1960 to 1978 and thereafter remained constant at a little higher level except in 1988, whereas the rates for females gradually decreased with the year. The overall stillbirth rates decreased to 1/4 for triplets and to 1/5 for quadruplets during the 37-year period from 1951. The overall stillbirth rate of quintuplets was 0.60 (51/85) during the period 1975-1987.

The effect of migration on pairwise concordance for disease was assessed in 11,154 twin pairs of the Finnish Twin Cohort Study by comparing the pairs living in the same province to the pairs which members were living in different provinces of Finland. The cumulative incidence of psychosis and hypertension for the years 1972-1985 were analyzed. The cumulative concordance of psychosis for those MZ twin pairs living in the same province were higher than for those MZ pairs living further apart. Similar findings were found among DZ pairs for psychosis. The cumulative concordance of hypertension was only slightly higher among those MZ and DZ pairs living in the same province compared with pairs living in different provinces. These results indicate an overestimation of concordance of psychosis caused by selective migration. This bias in twin studies is likely to influence heritability estimates in a sample of limited geographical area.

In an attempt to improve our understanding of the factors that affect human twinning, we further developed the models given by Hellin (1895) and Peller (1946). The connection between these models and our own model (“Fellman's law”) were studied. These attempts have resulted in a more general model, which was then applied to data from Åland Islands (1750-1939), Nîmes (1790-1875), Stuttgart (about 1790-1900) and Utah (1850-1900). The product of the mean sibship size and the total twinning rate can be considered as a crude estimate of the expected number of sets of twins in a sibship. The same can be said about the twinning parameter in our model. These estimates are in good agreement. If we consider twinning data only, we obtain the geometric distribution, and log (Nk), where Nk is the number of mothers with k twin maternities, is a linear function of the number of recurrences. Graphically, this property can easily be checked. For sibships containing three or more sets of twins, all four populations show higher values than expected, particularly the populations from Stuttgart and Utah, which data also show poor agreement according to a χ2-test. A more exact model would demand more detailed demographic information, such as distribution of sibship sizes, age-specific twinning rates and temporal variations in twinning.

The osberved number of mothers in Åland with several recurrences of multiple maternities shows a considerable excess over the expected number as predicted by Peller's rule. The parameters in our model can be estimated by the maximum likelihood method and the obtained model fits the data better then Peller's model.

Segregation and twin disease concordance analyses have assumed a theoretical underlying liability following a multivariate normal distribution. For reasons of computation, of incorporation of measured explanatory variables, and of testing of fit and assumptions, newer analytical methods are being developed. The regressive logistic model (RLM) relies on expressing the pedigree likelihood as a product of conditional probabilities, one for each individual. In addition to logistic regression modelling of measured epidemiological variables on disease prevalence, there is modelling of vertical transmission, of transmission of unmeasured genotypes and of sibship environment. This paper discusses methods for the analysis of binary traits in twins and in pedigrees. Some extensions to the RLM for pedigrees which include twins are proposed. These enable exploration of twin concordance in the context of the twins' common parenthood, the sibship similarities within the family, and the twins' similarity in age, sex, genes and environment.

Present attempts to identify genes contributing to coronary heart disease (CHD) risk focus on “candidate genes”. With respect to CHD this could be any gene whose protein product is directly or indirectly involved in atherogenesis, thrombogenesis or thrombolysis/fibrinolysis. Genes that exhibit associations with absolute risk factor levels may be referred to as “level genes” to distinguish them from “variability genes”, which are genes involved in establishing the framework within which environmental influences may cause risk factor variation. In a series of persons recruited from the Norwegian Twin Panel, confirmatory evidence for level gene effect with respect to apolipoprotein B (apoB) concentration was found with an XbaI polymorphism in DNA at the apoB locus corresponding to residue 2,488 in the mature protein. Evidence for variability gene effect with respect to apoB as well as body mass index emerged with DNA variants in the 3′ part of the apoB gene. Level gene effect with respect to apolipoprotein A-I (apoA-I) and high density lipoprotein (HDL) cholesterol as well as apparent variability gene effect with respect to total and LDL cholesterol were detected with a DNA polymorphism at the cholesteryl ester transfer protein (CETP) locus. The first example of interaction between normal genes in determining risk factor level was uncovered in analysis of the apolipoprotein E (apoE) polymorphism and a restriction fragment length polymorphism (RFLP) at the low density lipoprotein receptor (LDLR) locus. An LDLR gene identified by presence of a PvuII restriction site eliminated completely the well known effect of the apoE4 allele on cholesterol level. Finally, in families where high Lp(a) lipoprotein level (a well established risk factor for CHD) segregated as a Mendelian trait, very close linkage with an RFLP at the plasminogen locus was established and DNA variation at the LPA locus reflecting varying numbers of a structure homologous to the “kringle IV” region of plasminogen was uncovered.

We consider three alternative parametric models to describe genetic and environmental influences on smoking initiation and smoking persistence. Under the single liability dimension model, the same genetic and environmental influences which determine smoking initiation also influence smoking persistence. Under the independent liability dimensions model, independent initiation and persistence dimensions determine onset of smoking, and persistence in those who become smokers. The combined model also postulates separate initiation and persistence dimensions, but allows for the possibility that some smokers are so low on liability to smoke on the initiation dimension that they become ex-smokers for this reason. Reanalyses of London twin data published by Eaves and Eysenck support the single liability dimension model. We discuss the difficulty of reconciling this finding with the hypothesis that nicotine dependence is a major determinant of smoking persistence, but caution that sample sizes in the London twin study were small.

Respiratory sinus arrhythmia (RSA) has been shown to be a sensitive index of vagal cardiac control. We studied the genetic and nongenetic influences on individual differences in RSA in a sample of 160 adolescent twins. RSA was measured during rest and across two different tasks. Results show that heritability is task dependent. The amount of genetic variance is the same, however, during rest and task conditions. Because nonshared environmental variance decreases during tasks, heritability is larger for RSA measured under more stressful conditions than for RSA as measured during rest. Multivariate models assessed the continuity of the genetic and environmental influences and show genetic influences to be the same across different conditions, while environmental influences are different. More specifically, a one-factor model is found for genetic influences and a second-order autoregressive model for the environmental factors.

The dizygotic twinning rate in Italy, adjusted for maternal age and parity, declined from 8.9/1000 in 1949-1953 to 6.0/1000 in 1978-81, but may have stabilized or recovered slightly in the early 1980s. The long decline was greatest in the oldest mothers, and the recent increase, if any, has been principally in first births and in women under 30. The monozygotic rate rose from 3.8/1000 in 1969-73 to 4.3/1000 in 1982-85, having been relatively stable prior to that period. No good explanation for these trends is apparent.

Despite increasing scientific evidence for a causal role of tobacco smoking in lung cancer and coronary heart disease, an alternative hypothesis was put forward several decades ago. The constitutional hypothesis has stated that there are genetic or other common factors, which predispose both to smoking and disease. A critical test of this hypothesis was considered the pattern of occurrence of disease in monozygotic (MZ) twin pairs in which one is a smoker and the other never has been a smoker. Initial twin studies found only small differences in the mortality of smoking and nonsmoking twins of discordant pairs. In the Finnish Twin Cohort, a population-based panel of adult like-sexed twin pairs, a questionnaire study carried out in 1975 permitted identification of 2488 twin pairs discordant for cigarette smoking. Analyses of total mortality and mortality due to coronary heart disease and lung cancer indicate that the smoking members of discordant MZ pairs are at higher risk than their nonsmoking cotwins; increased mortality of smoking cotwins was also found on 21-year follow-up of smoking-discordant pairs in Sweden. Incidence data and noninvasive studies of atherosclerosis in the Finnish sample provide confirmatory evidence for the causal role of smoking in the etiology of coronary heart disease and lung cancer.

The Minnesota Study of Twins Reared Apart has conducted comprehensive medical and psychological assessments of monozygotic (N = 49) and dizygotic (N = 25) twin pairs, separated early in life (average age of separation = 0.3 and 1.1 years, respectively) and reared apart during the formative years (average age of reunion = 30.3 and 37.2 years, respectively). The twins are administered two special mental ability batteries. The Hawaii Battery (H-B), supplemented by several Educational Testing Service tests, is administered toward the beginning of the assessment week. The Comprehensive Ability Battery (CAB) is administered toward the end of the assessment week. All data are age- and sex-corrected. The average MZA and DZA intraclass correlations for the 15 H-B subtests were 0.45 and 0.34, respectively, and the average MZA and DZA intraclass correlations for the 13 subtests of the CAB were 0.48 and 0.35, respectively. Biometric model-fitting of these data indicate an average heritability of about 0.50. Data for groups of subtests in the Verbal, Spatial, Perceptual Speed and Accuracy and Memory domains were compared to a meta-analysis of the special mental ability findings in the ordinary twin literature. The Spatial domain appears to yield the highest and the Memory domain the lowest heritabilities.

A case-control study was conducted to evaluate the association between maternal age and parity and dizygotic (DZ) and monozygotic (MZ) twinning, respectively. Mothers of all twins born alive in Denmark in 1984 or 1985 were included into the study as cases, and a random sample of mothers of singletons were controls. Data on maternal age at delivery and number of previous births were collected from the Danish Medical Birth Register. By means of logistic regression analysis, a significant and positive association was found betwen maternal age and DZ twinning, and a significant and negative association between parity and MZ twinning. No association was found for parity and DZ twinning, nor for maternal age and MZ twinning. The study suggests that human MZ twinning has predictors too, as has DZ twinning. The finding has implications for zygosity classification in future twin research.

Many studies of embryogenesis and fate of twin pregnancies are invalidated because zygosity is not determined definitively, or is assumed on the basis of inadequate criteria. This paper briefly reviews methods of zygosity determination. It reports published results and a new series of twins in which zygosity was determined by DNA fingerprinting. Implications for methods of prenatal diagnosis of zygosity are discussed in the context of the occasional need for intervention in twin transfusion syndrome or in twins discordant for major malformations. Definitive zygosity and placental anatomy (number of chorions and amnions) is discussed as the firm substrate for studies of normal and abnormal twin development.