Nowadays, most twin research studies tend to be focused solely in the developed countries, despite some research on the subject from developing countries that is starting to emerge. Considering the value of investigation of genetic and environmental factors in complex traits, as well as the distribution and heritability of various phenotypes across different populations, the establishment of more national twin registries is quite important (Ohm & Derom, Reference Ohm and Derom2006; Sung et al., Reference Sung, Cho, Cho, Duffy, Kim, Kim and Park2002, Reference Sung, Cho, Lee, Ha, Choi, Choi, Kim and Song2006). As an iterative solution for research, the criteria for selection of twins and method of determination can be essential for twin registries using population-based approaches. The aim of the present article is to describe the methods we used to conduct this study, and to provide baseline results on the nature of the samples identified.
Methods
Setting
The Mongolian Twin Register is based in Songinokhairkhan, one of the nine districts of Ulaanbaatar, the capital city of Mongolia. In 2011, the population of Ulaanbaatar numbered 1.3 million people, which accounts for 50% of the total population of Mongolia. Each district is divided into khoroo, which is the smallest administrative unit in terms of population size (approximately 8,000 people in each). There are a total of 32 khoroos and 240,295 people in Songinokhairkhan, which accounts for 25% of total population of Ulaanbaatar.
Participants
Ascertainment of twins
We randomly selected nine khoroos from the Songinokhairkhan district. The municipal register in each khoroo of the capital contains information such as data on births, names, and deaths of all residents. Thus, these civil registration records were used to identify the source of participants for the twin register. The national registration data lacked the 8-digit individual identification number given to all Mongolians at birth, consisting of the date of birth (first 6 digits, year-month-day). Participants are sorted by last name, birth date, and home address. Persons born on the same day and same place, and given the same surname at birth, as well as living in the same home, are almost certainly twins.
Pilot Study
Randomly selected participants for a pilot study were identified and letters sent to those participants explaining the study in more detail. The participants were offered the right to refuse to participate. They were informed that interviews could take place in their own home or at the research center. Twins were excluded if the individuals said they were not twins; one or both of the pair had died or gone abroad; or there were no twins at the given address. Fieldworkers completed the interview (consisting of four pages, and lasting 20 minutes). A group of eight individuals (mostly students from a medical college) were extensively trained as field workers.
Zygosity Determination
We assessed zygosity using a questionnaire. Six questions to estimate zygosity were asked. Pairs where both twins reported a ‘water drop-like resemblance’, and schoolteachers or friends ‘always or nearly always’ had problems distinguishing between them were categorized as monozygotic (MZ). The pairs who did not report having a ‘water drop-like resemblance’, and were ‘seldom’ or ‘never’ confused by teachers or friends were categorized as dizygotic (DZ). All remaining pairs, including those pairs whose zygosity categories conflicted, were assessed as uncertain zygosity.
Anthropometric Measure and Heritability Estimation
Birth weight, height, and weight were self-reported and body mass index was calculated as the ratio of weight (kg) to height (in meters) squared. The twin birth rates were computed by dividing the number of twins by the total number of births and multiplying the outcome by 1000.
Results
Total births and twin live births in Mongolia from 2001 to 2011 are shown in Table 1. A total of 600,211 births were recorded during this 11-year period. Of these, 11,448 were twin births. The 2001 twin birth rate was 7.8 per 1,000. This rose slightly in 2002, when the twin birth rate was 9.1 per 1,000. The twinning rate rose sharply between 2006 and 2010 (from 9.2 to 10.6) and was fairly stable in 2011.
TABLE 1 Total and Twin Live Births From 2001 to 2011: Mongolia
Total and twin live births from 2001 to 2011 were identified by using the annual report (The report of Health Indicators 2001–2011) of Ministry of Health, Mongolia.
The annual report is based on the birth records that include information on twins vs. singletons.
Those who registered comprise 822 individuals or 411 twin pairs (same sex: male — 178; female — 157; different sex — 76), 2 sets of triplets (same sex: female — 2). The oldest twins and triplets are 81 and 16 years old, respectively; 79.8% of twins are less than 10 years old (Table 2). The mean age was 6.5 ± 9.3 years and 8.1 ± 12.8 years for men and women, respectively, and 52.4% were male. Interviews were performed on 77 twin pairs or 154 individuals. Those twins who were interviewed comprised 39 male pairs (18 pairs classified as MZ, 20 as DZ, and 1 as XZ) and 38 female pairs (22 pairs classified as MZ, 13 as DZ, and 3 as XZ; see Table 2). Mean observed anthropometric values are showed in Table 3. Heritability of height, weight, and birth weight as well as BMI was 0.98, 0.84, 0.45, and 0.31, respectively.
TABLE 2 Age, Sex, and Zygosity Distribution of the Registered and Pilot Study Participants, Number of Pairs (for Twins)
Notes
MZ = monozygotic twin; DZ = dizygotic twin; XZ = zygosity undetermined twin.
*Zygosity estimation was based on questionnaire.
†Zygosity of XZ was assigned either in case questionnaire-based zygosity survey showed discrepancy between the cotwins (i.e., MZ vs. DZ) or both co-twins fell in the category of ambiguous zygosity.
TABLE 3 Heritability and Mean Observed Anthropometric Variables
SD = standard deviation; BMI = body mass index.
Discussion
We report the results of the pilot study for establishing twin registration in Mongolia. The rate of twin births rose sharply from 9.2 to 10.6 per 1,000 births in 2006 and 2010, respectively, and was quite stable in 2011. It is well documented that the twinning rate is much lower in Mongoloid people (Hur & Song, Reference Hur and Song2009; Rushton, Reference Rushton1990) to that in other races (the highest twinning rates found across Central Africa are 18–30 twin sets per 1,000 live births, and twinning rates in the United States were reported as 31.4 per 1,000 in 2009; Fell & Joseph, 2012). In detail, the twin birth rate of countries such as China and Vietnam is 9 twins per 1,000 births on average. There are also very low twin birth rates observed for Japan (7.8 per 1,000 in 2007; Kurosawa et al., Reference Kurosawa, Masuno and Kuroki2011) and Singapore (7 per 1,000 in 2001; Chia et al., Reference Chia, Lee, Cheung, Cheung, Seielstad, Wilcox and Liu2004). Compared with these twin birth rates, the twinning rate in Mongolia is rising and slightly higher.
In recent years, there have been various efforts made to improve civil registration and vital statistics systems in Mongolia. However, despite these efforts, there are still significant levels of under-registration, which can be traced back to the lack of awareness of its importance. Moreover, civil migration statistics are still inadequate or deficient in Mongolia. For this reason, we have attempted to identify twins from municipal record lists. The participants were sorted by last name, birth date, and home address. Persons born on the same day and at the same place, and given the same surname at birth as well as living in the same home were identified as twins. As for the issue of over-sampling of children younger than 10 years in our study, it should be noted that underage children living with their parents or legal guardians are considered as one household. This is how we have been able to easily identify underage twin participants. On the other hand, under-representation of adults in the twin register is because older twins are more likely to have moved away from one another. The method we have used in this study was successful for young twins, but not so for identifying older ones. With this in mind, one of the future aims of the twin register will be to make additional recruitment efforts within the general population, in order to increase the relative number of adolescent and adult twin participants to be comparable to the general population of Mongolia. To our knowledge, this is the very first study to have analyzed whether there is a potential as well as sufficient data for building a twin register in Mongolia.
Conclusion
We aimed to recruit twins within a defined population, and invited a random sample to participate in the study. It has proved possible to build a twin register for research purposes in Mongolia. The potential exists to create an important new resource for twin studies in Mongolia.
Acknowledgments
This study was supported by the National Research Foundation of Korea (NRF; No. 2011-220-E00006, and NRF 2012K2A1A2032536), and the Ministry of Population Development and Social Welfare of Mongolia.
