Twins

Developmental Mechanisms

Twins are classified as either dizygotic or monozygotic. Dizygotic twins (also called fraternal twins) arise from two separately fertilized eggs, or zygotes. In humans, usually only one egg is released at a time from a woman's ovaries. When two are released, both may become fertilized by separate sperm and implant in the uterus. Dizygotic twins develop separate placentas and amniotic sacs. They may be of the same or different sexes. In the absence of reproductive technology interventions, dizygotic twinning occurs in approximately three of every thousand human births, a rate that increases with maternal age, varies with ethnic group, and is probably influenced by genes that control pituitary function. Various types of assisted reproductive technologies routinely create dizygotic twins, triplets, and higher numbers of offspring.

Monozygotic twins (also called identical twins) arise from a single fertilized egg. At some point after the zygote begins to divide, the cell mass splits into two, creating two embryos from one. Monozygotic twinning occurs in approximately 0.25 percent of human births. Monozygotic twins are always of the same sex. If the cell mass splits before about day five after fertilization, the two embryos will develop with separate placentas and separate amniotic sacs. This occurs in about two-thirds of human monozygotic twins. Between day five and about day nine, splitting leads to two amniotic sacs but one placenta. This occurs in about one-third of Monozygotic twins. Twins that split after day nine will share the amniotic sac. Splitting that late also increases the likelihood that the twins will not separate completely and will develop into conjoined (Siamese) twins.

Monozygotic versus Dizygotic Twins

Because monozygotic (MZ) twins develop from a single fertilized egg, they begin life with exactly the same set of genes. In this respect, they are clones—organisms whose genes are identical. As discussed below, however, they may accumulate genetic and other differences during development.

In contrast, dizygotic (DZ) twins are no more genetically close than any pair of siblings. While it is commonly said that siblings share half their genes, this is incorrect for two reasons. First, the random nature of meiosis and fertilization means that two siblings could end up with many, or few, genes from a particular parent in common. Second, there are many human genes for which there is only one common form, or allele. Therefore, any two people will share many alleles, regardless of their relationship. Only those genes with more than one allele form the basis of human genetic variation. These are the real focus of the question about gene-sharing in siblings. Of these variable genes, siblings (including dizygotic twins) on average share half.

Because dizygotic twins are the same age, they may share more of their environment than would two siblings of different ages. For instance, because they are likely to be engaged in similar activities, dizygotic twins are more likely to have similar environmental exposures (including behaviors, diet, hobbies, exposure to infectious agents, and exposure to chemicals)—whether at home, at school, or in the community—than two siblings of different ages and different activity patterns. It is this similarity of environment but difference of genes that makes them a useful contrast to monozygotic twins, whose environments and genes are largely identical.

Similarities and Differences between Monozygotic Twins

The fertilized egg cell that gives rise to MZ twins begins life with a single set of genes, and so we might predict that every cell that arises from it would be exactly identical. However, small differences between daughter cells may accumulate throughout embryonic development and later in life. The earliest difference may be in the mitochondria each inherits. Mitochondria are the cell's power plants and contain a small amount of DNA. Some of the hundreds of mitochondria in a cell may contain mutations. If the cells that create the two twins carry different mitochondrial genes, even identical twins will be genetically different. Mutations can also accumulate during embryonic development, or after birth, either in the mitochondrial genes or the genes in the nucleus. Such mutations may have a significant effect: Some types of cancer are due to mutations accumulated during one's lifetime, often through exposure to environmental chemicals or radiation.

For the vast majority of genes, though, MZ twins are exactly identical. Nonetheless, twins do experience slightly different environments, even when reared together, and any early differences between them may be accentuated by families members, or by one another, leading to development of very different personalities.

Amazing Twin Similarities

Some of the most tantalizing clues to the genetic basis of human personality and behavior come from studies of MZ twins reared apart since birth. Such twins have the same genes but, presumably, different environments. A major study of more than 100 such twin pairs showed some remarkable coincidences. A pair of twins meeting for the first time at age thirty-nine each arrived wearing seven rings, two bracelets on one wrist, and a watch and one bracelet on the other wrist. Another twin pair discovered they each had dogs named Toy, had married and divorced women named Linda, remarried women named Betty, and named their sons James Allan and James Alan.

While these coincidences are amazing, it is important to remember that many more twin pairs in this study did not have such parallel lives or habits. Such stories are curious and provocative but cannot by themselves tell us about the relative contributions of genetics and the environment in shaping personality, behavior, health, or other aspects of the self.

Twin Studies and Concordance

Insight into such questions can be gleaned by several types of studies that compare twins. Comparison of MZ twins reared apart is one type of study but is hampered by the extreme rarity of such twin pairs. Another type of study, comparing MZ twins to DZ twins, is more commonly done, because there are many hundreds of thousands of such twin pairs worldwide. Data on twins have been collected by numerous research groups who have created large and growing databases (registries) that can be mined for information.

Determining a characteristic called concordance plays a crucial role in most such studies. A twin pair is said to be concordant for a trait if both members show it. If neither twin shows the trait, the pair is also concordant, but for the absence of the trait. For instance, twins are concordant for Alzheimer's disease if both develop it. They are discordant if one does have the disease but the other does not.

If a trait is strongly influenced by genes, more MZ twin pairs should be concordant than DZ twin pairs, because MZ twins share more genes. Comparing concordance rates between the two groups, and applying some mathematical analysis, allows researchers to estimate the genetic contribution to a trait, as shown in Table 1.

Twin Registries

Twin studies can have several starting points. Some investigators begin simply by trying to identify twins who will volunteer to be part of a particular research study. Often twins are sought by advertising for twins with the particular disease of interest. This approach has the advantage of simplicity, as twins identify themselves to the research team.

However, twins who volunteer may differ in some important way from those who do not volunteer, and this could affect the conclusions drawn from the study. For example, MZ twins are more likely to volunteer, in general, than DZ twins are. This tendency to volunteer for twin studies among MZ twins is probably because being a twin is a more central part of the identity of MZ pairs than DZ pairs. Also, twins concordant for a particular disease are more likely to volunteer than those without the disease are. If both influences are at work in the same study, more concordant MZ twins than DZ twins may be identified, not because there is an actual difference in concordance between MZ and DZ twins (and thus a genetic effect at work), but because more concordant MZ twins volunteered for the study. If this pattern of volunteerism is mistaken to represent the true pattern of the disease in all twins, an inappropriate conclusion that the disease has genetic causes could result.

Other twin registries attempt to identify all twins within a particular population. One approach is the statewide or national twin registry. All twin births in the region are reported to a central registrar. This results in a more complete picture of all twin pairs in these populations. Examples include the statewide Virginia and Minnesota twin registries in the United States and many national twin registries, including those in the United Kingdom, Australia, the Scandinavian countries, Germany, Belgium, the Netherlands, Italy, and Sri Lanka.

Twin registries have also been assembled from among special populations. Examples in the United States are registries assembled from military records (the World War II Veteran Twins Registry and the Vietnam Era Twin Registry) and from Medicare files (the U.S. Registry of Elderly African-American Twins). In these registries, likely adult twins were identified by searching records to identify individuals with identical dates of birth, birthplaces, and surnames. These individuals were then contacted to verify whether they actually constituted a twin pair. Registries may also be established by identifying twin births within a health maintenance organization (such as the Kaiser Permanente Twin Cohort, in California).

Each registry varies in the amount of contact with registrants. In all, individual contact is strictly monitored to preserve the privacy of each twin. Every research proposal must be approved by a panel to assure the scientific value of the project, the justification for doing the study in twins, and to ensure that the privacy and safety of individual twins will be protected.

Twin registries can be useful starting points for investigating many questions about the genetic and environmental determinants of a trait. Records linkage studies involve no personal contact with the twins. Instead, information in the twin registry is "linked" electronically to information in another database, such as a national health insurance database or a cancer registry. In this way, twins with a particular health problem can be identified, and concordance estimates can be calculated. Similarly, information collected for each twin at registration can later be used to investigate certain kinds of questions without ever contacting the individual twins. On the other end of the spectrum, twins may be asked to volunteer for physical examinations, blood tests, radiological studies, or interviews. Depending on the questions asked, such studies may be useful for comparing concordance, or for identifying risk factors or modifying factors for a trait.

Twin Studies to Investigate the Cause of Parkinson's Disease

An example of the use of investigations in twins to understand more about a disease is provided by recent work in Parkinson's disease. Parkinson's disease (PD) is a progressive neurodegenerative disease causing slowness, tremor, and problems with walking and balance. PD is rare before age fifty but becomes more common thereafter, with increasing age. The cause of PD has long been debated. Both genetic and environmental causes have been suggested, but neither has been definitively shown. Researchers turned to studies in twins to determine the relative contribution of genes and environment to the disease.

The first studies identified twin pairs by recruiting through physicians and PD patient organizations. Studies in the United States, the United Kingdom, and Germany identified 103 pairs, of which only thirteen were concordant for PD. In Finland, forty-two twins with PD were identified by records linkage, but among these was only one concordant pair—a DZ pair. No study had convincingly demonstrated greater monozygotic than dizygotic concordance for the disease, and in all studies the preponderance of twin pairs were discordant for disease. These findings supported an environmental cause of PD. Nonetheless, the advent of molecular genetics prompted great interest in investigations of genetic causes of disease and prompted the resurgence of the hypothesis that all PD had a genetic cause. To address this, a study in a large, unselected cohort—the National Academy of Sciences/National Resource Council (NAS/NRC) World War II Veteran Twins Registry—was undertaken.

In the mid-1950s, the Medical Follow-up Agency of the Institute of Medicine of the NAS/NRC established a registry of approximately 32,000 Caucasian male twins, all of whom were born between 1917 and 1927 and were veterans of the U.S. Armed Services. In all, 161 twin pairs were identified, twenty-one of which were concordant for PD, as shown in Table 2. In those few pairs with early-onset PD, concordance was greater in MZ pairs. In those with more typical PD, beginning after age fifty, there was no difference in MZ and DZ concordance.

These findings suggest a strong genetic determinant for early-onset disease but predominantly environmental causes in more typical late-onset disease. One caveat is the narrow age range of the twins, who were sixty-seven to seventy-seven years old when studied. Since PD is a late-life disorder, PD in some twins may have been missed with an examination at only one time point. To overcome this, a second evaluation is in progress.

Risk-Factor Investigations in Twins

Studies of twin pairs discordant for disease can be useful for identifying risk factors for disease. Since both genetic and environmental factors are extensively shared by twins, particularly by MZ twins, case-control studies can be particularly powerful. In such a study, each twin is interviewed with regard to specific environmental factors—such as occupation, lifestyle factors, illnesses or injuries, and diet—prior to the onset of the disease in the affected twin. The presence of these factors in the twin with the disease is compared to the twin without disease. An association of an environmental factor with the disease suggests this factor may be causally related. Factors more common among the unaffected twins suggest that the factor may protect against the development of the disease.

Environmental influences on PD have been investigated by studying discordant twin pairs. PD has repeatedly been found to be more common in people who do not smoke cigarettes. Some have proposed that some people are genetically predisposed to both Parkinson's disease and smoking, while others suggest cigarette smoking somehow prevents the degeneration that leads to PD. In two studies of discordant twin pairs, cigarette smoking was more common in the twin without Parkinson's disease, especially in the MZ pairs. Because monozygotic twins are genetically identical, this pattern tips the scales in favor of a direct biological action of cigarette smoke.

As medicine focuses more on early intervention or prevention, it becomes important to identify those persons at risk for a particular condition. This can be a problem if there is no diagnostic test. In discordant twin pairs, the unaffected twin is more likely to be "at risk" for a particular condition, whether due to shared genes or environment, than would be true for two nontwins. Therefore, studying the unaffected "at risk" twin may help to clarify what features are useful for predicting those who later will develop a particular disease. For example, in the PD twin study, the unaffected twins are being studied prospectively with brain imaging tests that may show early evidence of injury to the brain area damaged in PD. If abnormalities on this test are found to precede the development of PD, this could provide a useful method of early detection. When treatments to slow or stop the onset of PD are available, individuals with imaging abnormalities may receive intervention before symptoms develop.

Results from Twin Studies of Other Disorders and Conditions

The twin study method has been used to try to determine the extent of genetic or environmental influence on a wide variety of traits and conditions. Among these are sense of humor, which appears to be largely environmentally determined, as MZ and DZ pairs have similar concordance. Examples of other diseases in which MZ concordance exceeds DZ concordance, suggesting a significant genetic component, include addictive behaviors such as cigarette smoking and alcohol drinking, mental illnesses such as schizophrenia, as well as stroke and certain types of high blood pressure. Twin studies of many other disorders are ongoing.

Conclusion

Twin studies provide a unique approach to investigating the determinants of a disease or condition. A single twin study cannot absolutely determine the importance of genetic or environmental factors. However, the twin study method, in combination with other approaches, can be a powerful tool for unraveling the causes of disease.

SEE ALSO BEHAVIOR; FERTILIZATION; GENE AND ENVIRONMENT; GENE DISCOVERY; INHERITANCE PATTERNS.

Caroline M. Tanner

and Richard Robinson

Bibliography

Bouchard, T. J., et al. "The Sources of Human Psychological Differences: The Minnesota Study of Twins Reared Apart." Science 250 (1990): 223-228.

Segal, Nancy L. Entwined Lives: Twins and What They Tell Us about Human Behavior. New York: Plume, 2000.

Wright, Lawrence. Twins: And What They Tell Us about Who We Are. New York: John Wiley & Sons, 1997.

Internet Resource

Minnesota Twin Family Study. University of Minnesota. <http://www.psych.umn.edu/psylabs/mtfs/default.htm>.

WHY STUDY TWINS?

• To estimate the relative contributions of genes and environment to the cause of disease by comparing MZ to DZ concordance;
• To investigate environmental determinants of etiology in discordant twin pair studies;
• To investigate environmental influences on disease course in concordant twin pair studies;
• To characterize "presymptomatic" or "at risk" states by studying the unaffected twins in discordant pairs.