 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ARTICLE |
a Department of Growth and Reproduction GR, University Hospital of Copenhagen, Rigshospitalet, Denmark
b National Centre for Register-based Research, Taasingegade 1, University of Aarhus, Aarhus, Denmark
c Department of Environmental Medicine, University of Southern Denmark, Odense, Denmark
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
METHODS. Patients who were registered with the diagnosis of precocious puberty during the period 1993–2001 were identified through the national patient registry. The background population of children born from 1983 to 2001 were identified through the unique Danish Civil Registration System and subsequently categorized as being Danish (N = 1062333), adopted (N = 10997), immigrating with their family (N = 72181), or being descendants of immigrants (N = 128152). The incidence rate ratio of precocious puberty was estimated by log-linear Poisson regression. All rate ratios were adjusted for age and its interaction with gender and calendar year. P values were based on likelihood ratio tests, and 95% confidence intervals were calculated by Wald's test.
RESULTS. In the study period, 655 children developed precocious puberty during 5627763 person-years at risk. Adopted children were followed during 39978 person-years at risk, during which 45 girls and 6 boys developed precocious puberty. The risk of developing precocious puberty was significantly increased 10 to 20 times in adopted girls compared with girls with Danish background. The risk of developing precocious puberty depended on the country of origin. In children immigrating with their family, the risk of developing precocious puberty was only marginally increased. Older age at adoption significantly increased the risk of precocious puberty in adoptees independent of region of origin. The incidence rate ratio was significantly higher in children adopted after the age of 2. In children immigrating with their family, we found no effect of age at migration.
DISCUSSION. In this large, nationwide, register-based study including 655 cases of precocious puberty, we found that intercountry boys and girls were 10 to 20 times more likely to develop precocious puberty compared with the Danish reference group. Older age at adoption significantly increased the risk of precocious puberty. Uncertainty of the exact age is a well-known problem in adopted children, and systematic underestimation of age might bias the result. However, using the worst-case scenario that all children who according to the Danish Civil Registration System were adopted after 2 years of age were in fact 1 year older, we still observed a highly increased risk of precocious puberty associated with adoption and especially with adoption after 2 years of age. Surprisingly, the risk of precocious puberty was not increased in the large group of children adopted from Korea. One case of precocious puberty was identified among Korean children, whereas >20 cases of precocious puberty would have been expected if the risk for a Korean child was at the same level as observed among adopted children from India and South America. In the study population, 99% of Korean children were adopted before 2 years of age, which may contribute to explaining our finding. In Korea, children appointed for adoption are often living in foster care settings from birth to adoption, whereas most other countries are reported to take care of the children in orphanages before adoption. It can only be speculated whether a relation between preadoption living conditions and later risk of precocious puberty exists. Genetic factors play a key role in the timing of puberty, and large variations in age at menarche are observed worldwide. Age at menarche is reported to be in the same age range in South Korea as in well-off populations in other parts of the world, indicating that the different risk of precocious puberty observed between Korean and other adoptees probably cannot be explained by genetic factors alone. The finding that the risk of precocious puberty was significantly increased among adoptees in contrast to what was seen in children immigrating with their families contradicts a direct effect of migration. An increasing number of studies have shown long-term effects of certain prenatal and postnatal growth patterns, including advancement in pubertal maturation after poor intrauterine growth and catch-up growth during childhood. Different growth patterns and dietary habits between adoptees and children immigrating with their families might contribute to explain our findings. It has been hypothesized that stressful psychosocial factors in infancy and childhood may lead to earlier pubertal maturation. In general, adoptees have experienced several traumatic life events, and it may be speculated that these events alter the susceptibility for developing precocious puberty.
CONCLUSIONS. Foreign-adopted children originating from regions other than Korea had a 15- to 20-fold increased risk of precocious puberty compared with Danish-born children, whereas adoptees originating from Korea had no increased risk of precocious puberty. In addition, children immigrating with their families had no increased risk of precocious puberty. The effect of country of origin might be explained by genetic factors or by different environmental exposures and living conditions in the different countries. Older age at adoption increased the risk for premature onset of puberty, which may suggest that environmental factors influence the risk of precocious pubertal development in adopted children.
Key Words: epidemiology • puberty • precocious puberty • adopted children • Denmark
Abbreviations: PP—precocious puberty • ICD—International Classification of Diseases • CRS—Civil Registration System • IRR—incidence rate ratio • CI—confidence interval
Thousands of children are migrating from developing countries to Western countries through international adoptions, and in Denmark 
1% of recent birth cohorts are foreign-adopted children. Their health, development, and social adjustment have become an increasingly important topic. Although the majority of adopted children thrive,1,2 large-scale register-based studies from Sweden have documented an increased risk of severe mental health problems and social maladjustment in adolescence and young adulthood among adoptees.3
A somewhat overlooked aspect, which could contribute to this undesirable development, is the increased occurrence of early sexual development in adopted children that has been reported from series of patients with precocious puberty (PP) and in cohort studies from Sweden and the Netherlands.4–8 The risk of developing PP was estimated to be 20- to 80-fold increased in 2 earlier studies,7,9 but the epidemiology of PP in adopted children has not been investigated in larger population-based studies. A previous Danish study indicated that the prevalence of registered cases of PP has increased markedly over 40 years among Danish children.10 Theoretically, the high number of children who migrate to Denmark through intercountry adoption could contribute to this finding.
The etiology behind premature sexual maturation in adopted children is yet to be understood, but it has been proposed that genetic/racial differences, prenatal and postnatal growth patterns, and environmental factors such as improved nutrition and exposure to endocrine disrupters may be involved. New observations have indicated that nonadopted children, immigrating with their families, also seem to have an increased risk of precocious sexual development after migration, suggesting environmental factors to be important contributors.7,11
These trends prompted us to use Danish national registers to examine the risk of developing PP in children who migrate to Denmark after international adoptions or immigration with their family and in children who are descendants of immigrants.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Background Population
We used data from the Danish Civil Registration System (CRS), in which all Danish citizens are registered with a unique identifier (CPR [personal identification] number) to identify all children who were born in Denmark between January 1, 1983, and December 31, 2001, and the identity of the mother was known (1296368 persons). The register also stores information on vital status, emigration, disappearance, and CPR numbers of parents. The links to parents are based on the legal relationship, such that for each person registered, the parental links registered represent the legal parents; the registry is updated whenever changes occur. Danish legislation prohibits the CRS from keeping information on previous legal relationships, so for adopted children, for example, the CRS contains information on the legal parents only.14 Information on place of birth, the mother's place of birth, and the mother's place of residence at the time of the child's birth was obtained from the register.
Categories of Study Subjects
We used information on the child's country of birth, the mother's country of birth, and the mother's country of residence at the time of the child's birth to identify demographic characteristics of the study population. An "adopted child" was defined as a child born abroad although the mother (ie, legal mother) was born in Denmark and lived in Denmark at the time of child's birth (n = 10997). A "child immigrating with his or her family" was defined as a child born outside Denmark of a mother who was also born abroad and lived abroad at the time of the child's birth (n = 72181). "Descendants of immigrants" were defined as children born in Denmark by mothers born abroad who were living in Denmark at the time of the child's birth (n = 128152). For adoptees and children immigrating with their family, the country of origin was defined using the child's place of birth, and for descendants of immigrants, the origin was defined by using maternal place of birth. "Children with Danish background" were defined as children born in Denmark of a mother born in Denmark who lived in Denmark at the time of the child's birth (n = 1062333). "Children born abroad by Danish women" were defined as children born abroad of mothers born in Denmark who were living abroad at the time of the child's birth (n = 12590). Finally, "children who were unclassifiable" were defined as children who were not classified according to the other categories (n = 10115) either because of missing information (n = 4775) or because the mother resided in Denmark at the time of the child's birth but the child was born abroad or visa versa (n = 5330).
Age at arrival to Denmark was defined as the age at which an adopted child or a child immigrating with his or her family was registered in the Danish CRS with a permanent address in Denmark for the first time.
Based on a previous study on the association between immigration and schizophrenia risk15 and the prior wish to estimate the risk of adoption separately for children adopted from Korea and India and those adopted from East and Western Europe, we defined the following regions of origin: Korea, India, Asia, South America, Eastern Europe, Africa, Middle East, Western Europe, Scandinavia (excluding Denmark), North America, Australia, Greenland, Denmark, or unknown.
Statistical Analyses
The incidence rate ratio (IRR) of PP was estimated by log-linear Poisson regression.16 All rate ratios were adjusted for age and its interaction with gender and calendar year. Age and calendar year were treated as time-dependent variables, whereas all other variables were treated as variables independent of time. P values were based on likelihood ratio tests, and 95% confidence intervals (CIs) were calculated by Wald's test.17 This study was approved by the Danish Data Protection Agency.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Effect of Migration and Region of Birth
Table 1 shows the distribution of cases of PP and person-years at risk according to region of origin (place of birth for adopted children, place of birth for children immigrating with their families, and maternal place of birth for descendants of immigrants). Overall, children with Danish background were followed during 4873824 person-years at risk, during which time 471 children developed PP; adopted children were followed during 39978 person-years at risk, during which time 51 children (45 girls and 6 boys) developed PP; children immigrating with their families were followed during 97355 person-years at risk, during which time 23 children developed PP; and descendants of immigrants were followed during 551694 person-years at risk, during which time 94 children developed PP. The majority of adopted children were born in Korea, India, other Asian countries, or South America, whereas children who immigrated with their family and descendants of immigrants mainly originated from Asia, the Middle East, and Europe.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This study confirms a long held suspicion raised by a number of smaller studies in adopted children.4–8 Our study was larger (ie, includes more cases and is population based) and analyzes the effect of age at adoption and region of birth while controlling for potential confounders such as calendar year, age, and gender. Previous studies have mainly reported an increased frequency of sexual precocity in girls.4,5,8,18,19 We found that the risk was increased in boys as well as in girls. The total number of cases in boys was only 6, which is well in accordance with previous studies showing a low incidence of PP in boys.10,20 Because of this low number of cases, we were not able to analyze the effect of country of origin and age at adoption separately by gender. The risk estimate for adoptees was very close to what was found in a preliminary study from Denmark.9 However, our risk estimate was considerably lower than that reported from Belgium, where an 80-fold increased risk was reported.18 Unfortunately, the Belgian registry of adoptions was considered incomplete, which made direct comparison with these data difficult.
In this study we found that older age at adoption significantly increased the risk of PP. Uncertainty of the exact age is a well-known problem in adopted children (ref 4; Danish National Board of Adoption, verbal communication, 2005), and estimation of age is often difficult. Using the worst-case scenario that all children who were adopted after 2 years of age were 1 year older, we still observed a highly increased risk of PP associated with adoption and especially with adoption after 2 years of age. Therefore, it is highly unlikely that our results were severely biased by underestimation of the age of the adopted children.
Studies have shown that growth retardation is correlated to the age at adoption.21,22 The observed effect of age at adoption could putatively be explained if catch-up growth that occurs before 2 years of age (ie, in infancy) is less harmful, because it is a physiologic response to unfavorable prenatal conditions, whereas catch-up growth during childhood, as seen in older adoptees, is an unphysiological phenomenon that is more likely to trigger endocrine responses (eg, increased levels of insulin-like growth factor I, which in turn could stimulate the gonads to produce sex hormones).23,24
The finding that adoptees from Korea had no increased risk of PP was surprising. One case of PP was identified among the large group of adopted Korean children, whereas >20 cases of PP would have been expected if the risk for Korean children was at the same level as that observed among other adopted children. This disparity was not the result of methodologic problems, because children adopted from Korea were not followed mainly at very young ages, where the risk of developing PP is very low10: 52% of the follow-up time of girls adopted from Korea (5038 person-years) was after the age of 6 years. For comparison, only 24% of the total follow-up time of girls adopted from the remaining part of Asia (3306 person-years) was after the age of 6 years. In addition, all results were adjusted for age and its interaction with gender. Therefore, it is highly unlikely that the very low risk associated with adoption from Korea is caused by methodologic errors.
Korea differs in several ways from most other countries that provide children for adoption. Children born by unmarried women are hardly accepted in the Korean society for cultural reasons and are often placed in foster care settings just after birth and until adoption. Most other countries are reported to give children for adoption because of poverty and place the children in state-run or private orphanages before adoption. From South Africa, Colombia, and the Philippines, it is also reported that a proportion of children designated for adoption are living in foster care settings (ref 25; Danish National Board of Adoption, verbal communication, 2005). We were not able to track information about individual preadoption living conditions through the register. It can only be speculated whether a relation between preadoption placement and later risk of PP exist. In addition, children from Korea are adopted at young ages, as confirmed in our study, in which 99% of adoptees from Korea were adopted before 1 year of age, which may contribute to the explanation that children adopted from Korea had no increased risk of PP.
Genetic factors play a key role in the timing of puberty, and large variations in age at menarche are observed worldwide. Age at menarche is reported to be in the same age range in South Korea as in well-off populations in other parts of the world (South Korea: 12.6 years; other Asian countries: 12.1–12.6 years; South America: 12.3–12.7 years), which indicates that the different risk of PP observed between Korean and other adoptees probably cannot be explained by genetic factors alone.11,26–28
We found no increased risk of PP in children immigrating with their family. Age at menarche in Asia and South America (where adoptees were born) are not reported to be very different from the Middle East countries and Eastern Europe (where immigrants were born).11 However, variation of age at menarche between populations should be interpreted with care, because the estimates are highly dependant on the socioeconomic status of the studied population, ongoing secular trends as well as the applied statistical method.29
The peculiar finding that the risk of PP was significantly increased among adoptees in contrast to what was seen in children immigrating with their families may support the hypothesis that genetic factors cannot solely explain the differences in the risk of PP between adoptees and children immigrating with their families. Differences in dietary habits and growth patterns between the 2 groups may contribute to the explanation. Growth is frequently retarded in adopted children at the time of arrival in the foster country, and rapid catch-up growth is very common.5,8,30 In contrast to this, it is likely that growth remains more stable in those children who immigrate with their family, keeping their original dietary habits after migration. An increasing number of studies have shown long-term effects of poor prenatal and postnatal growth, followed by catch-up growth, including advancement in pubertal maturation as well as other metabolic and endocrine disorders later in life.31–34 Different dietary habits and growth patterns in adoptees and children immigrating with their families, therefore, might contribute to an explanation of our findings.
The risk of PP was significantly increased in descendants of immigrants in contrast to what was found in children immigrating with their family, although the 2 groups originated from the same regions. Although the literature is sparse, studies from Germany and England indicate that age at menarche is not markedly accelerated in second-generation immigrants, suggesting that genetic rather than environmental factors determine timing of puberty at the population level.35,36 This could support the notion that descendants of immigrants enter puberty earlier than Danish children because of genetic variations linked to ethnicity, and perhaps the same occurs in children who immigrate with their family. Language barriers, other diseases, social problems, etc might impede referral to more specialized health care services, thereby biasing the risk of PP toward null in the latter group.37,38
Psychological mechanisms have also been proposed as triggers of pubertal maturation. Recent studies support a hypothesis that certain stressful psychosocial factors in infancy and childhood may lead to earlier pubertal maturation.39,40 From this theory, it can be speculated that adoptees, who in general have experienced many stressful life events, are more vulnerable to developing PP compared with children who immigrate with their family. Likewise, the finding that children adopted before 2 years of age were less likely to develop PP could indicate that the duration and sum of stressful events might influence pubertal timing.
Our register-based data have strengths and limitations. The study was based on nationwide unbiased selection of all patients with PP seen at a hospital ward. However, only cases of PP registered in the national patient registry were included. The results can possibly be biased by selective patterns of use of health service and referrals between the different groups.37,38 Theoretically, selective "over-referral" of adopted children could have biased the result. We validated the diagnosis of PP by going through the files of 100 randomly selected cases from the present study population in a former study.10 Records from 8 adopted children were among the 100 case files selected for validation, and in all cases sexual precocity was found, indicating that all referrals of adoptees were well founded.
Classification of the study population into 4 mutually exclusive groups may have some pitfalls; some cases were unclassifiable, and some may have been misclassified. However, we have no reason to believe that misclassifications were systematic in favor of cases or noncases or in favor of certain regions of origin. If nondifferential misclassification has occurred, estimates would be biased toward null.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Older age at adoption increased the risk for premature onset of puberty, which strongly suggests that environmental factors influence the risk of precocious pubertal development in adopted children.
|
ACKNOWLEDGMENTS |
|---|
We thank I. Bjerrum-Bach and J. Nordhoeck for inspiration to conduct this study.
|
FOOTNOTES |
|---|
Address correspondence to Grete Teilmann, MD, Department of Growth and Reproduction GR, University Hospital of Copenhagen, Rigshospitalet, DK-2100 Copenhagen 0, Denmark. E-mail: gteilmann{at}rh.dk
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J.-C. Carel and J. Leger Precocious Puberty N. Engl. J. Med., May 29, 2008; 358(22): 2366 - 2377. [Full Text] [PDF]
|
|
|
|
 |
|
 D. Gunnarsson, P. Leffler, E. Ekwurtzel, G. Martinsson, K. Liu, and G. Selstam Mono-(2-ethylhexyl) phthalate stimulates basal steroidogenesis by a cAMP-independent mechanism in mouse gonadal cells of both sexes Reproduction, May 1, 2008; 135(5): 693 - 703. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C M Schooling, C Q Jiang, M Heys, W S Zhang, X Q Lao, P Adab, B J Cowling, G N Thomas, K K Cheng, T H Lam, et al. Is leg length a biomarker of childhood conditions in older Chinese women? The Guangzhou Biobank Cohort Study J. Epidemiol. Community Health, February 1, 2008; 62(2): 160 - 166. [Abstract] [Full Text] [PDF]
|
|
Read all P3Rs
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
