Last Updated: 2007-04-25
Introduction and Definition
The study of identical twins is very useful in allowing for the identification of similarities caused solely by environmental factors as opposed to genetic influences. The study of different sex twins can provide comparisons with different sex siblings (the same genetic relationships as different sex twins but different environments).
In the Manitoba database, twins are defined as two children born on the same day to the same mother. Two types of twins exist:
(fraternal). Monozygotic twins come from the same egg and are 100% genetically related. Dizygotic twins come from different eggs and are therefore no more genetically similar than any other non-twin sibling. Fraternal twins, non-twin siblings, and other first-degree relatives have a genetic correlation of approximately half that of identical twins
(Duncan et al., 2001).
Birth records used at the Manitoba Centre for Health Policy (MCHP) do not distinguish between monozygotic and dizygotic twins. Based on earlier descriptive studies (such as Conley et al., 2003), the twin data used at MCHP are comprised of approximately 25 percent monozygotic pairs and 75 percent dizygotic pairs.
Twin studies are used in most analyses allocating individual differences into genetic and environmental components as they are extremely helpful in eliminating unobserved heterogeneity across families
(Duncan et al., 2001; Oreopoulos et al., 2006).
Specifically, Oreopoulos et al. (2006) compare whole population, sibling, and twin findings in order to overcome concerns that twins are a select sample of the population and that inference from this sample is not, therefore, generalizable to the broader population.
Bjorklund, Jantti, and Solon (2005)
suggest the use of twins to determine the roles that nature and nurture have on the development of a child. This approach is based on the premise that "the greater correlations typically observed for monozygotic twins occurs mainly because [they] ... have identical genes, whereas the genes of dizygotic twins ... are correlated only in the same way as the genes of non-twin siblings (Bjorklund et al., 2005, p. 146).
Manitoba research, like that of others using somewhat different measures, finds twin correlations to be substantially higher than those of non-twin siblings. For example, the overall magnitude of Manitoba same-sex twin correlations on the Language Arts achievement index (0.738) proved very similar to those reported in reviews of IQ correlation studies
(Devlin et al., 1997; Duncan et al., 2001).
Variables that can be used to match twins include:
the Personal Health Identification Number (PHIN) of the mother;
the postal code of the patient's residence;
the hospital ID number from which the twin was discharged; and
the month of the hospital admission.
Since one family may have more than one set of twins, the existing family number cannot be used to uniquely identify a set of twins. In order to specifically identify twin groups, "twinID" numbers were created and assigned to twin pairs.
See information on
Twins - SAS code
(internal access only).
The SAS code was used to identify twins by calculating births to the same mother on the same day. Therefore, twins born on two different calendar days (e.g., born on either side od midnight) would not be captured by this coding.
Twins with different gestation periods
Of the total 677 twinID numbers in the birth cohort from 1978 to 1985 (excluding 1983) that continuously reside in the province, 17 twinIDs had different gestation periods. Since twins are born at roughly the same time, gestation periods for a set of twins should not vary. In the statistical analyses (using a family fixed effects model), coefficient estimates were found because of the variations in gestation periods. A coefficient estimate of zero should have been found for gestation period of the twins.
To eliminate the coefficient estimates, each twin pair would have to be treated as a single ID with only one gestation period. This meant that 17 twinIDs required alterations. It was decided that when making the alterations to the twinIDs, the longer of the two given gestation periods would be used. For example, if a twin set had gestation periods 38 weeks and 36 weeks, the gestation period for the twinID became 38 weeks. As well, if one of the gestation periods for a set of twins was missing, the gestation period for the twinID became the one that was given.
- Bjorklund A, Jantti M, Solon G. "Influences of nature and nurture on earnings variation: A report on a study of various sibling types in Sweden."
In: Bowles S, et. al. (eds).
Unequal Chances: Family Background and Economic Success.
Princeton University Press; 2005.
- Conley D, Strully KW, Bennett NG.
The Starting Gate: Birth Weight and Life Chances.
University of California Press;
- Devlin B, Daniels M, Roeder K.
The heritability of IQ.
1997;388:468-471. [Abstract] (View)
- Duncan GJ, Boisjoly J, Harris KM.
Sibling, peer, neighbor, and schoolmate correlations as indicators of the importance of context for adolescent development.
2001;38(3):437-447. [Abstract] (View)
- Lawlor DA, Mishra GD.
Family Matters: Designing, Analysing, and Understanding Family-Based Studies in Life Course Epidemiology.
Oxford University Press;
- Oreopoulos P, Stabile M, Walld R, Roos LL.
Short, medium, and long term consequences of poor infant health: An analysis using siblings and twins.
J Hum Resour
2008;43(1):88-138. [Abstract] (View)
- Roos LL, Walld R.
Neighbourhood, family, and health care.
Can J Public Health
2007;98(Suppl 1):S54-S61. [Abstract] (View)