TBCS in Brief [EN]

By , 2013-10-10 5:46 下午


We treasure our children’s health, because children are not only the future masters and lifeline of our nation, but also because they, too, have the human rights to health. According to the United Nations Convention on the Rights of the Child, children’s health rights should be treated with the highest standards. Hence, the government should prioritize investment in child health and take actions to reduce children’s health inequalities.

Children’s health in Taiwan has improved dramatically with social and economic prosperity, as well as public health and medical advancement. The infant mortality rate has decreased from 40.4‰ in 1970 to 4.6‰ in 2006; during the same period, the under five mortality rate has also reduced from 99.2‰ in 1970 to 6.6‰ in 2006. Meantime, we have observed that the focus of health policy has gradually shifted away from child health towards emerging topics such as chronic disease prevention and health and social care for the elderly. Consequently, the Department of Health (now Ministry of Health and Welfare ), has merged the Institute of Maternal and Child Health and the Institute of Family Planning into the Health Promotion Administration, and created the Department of Nursing Affairs to be responsible for long-term care.

Yet, as we enter the 21st century, the child health policy in Taiwan has been facing with three new and tough challenges: First, the fertility transition: it is hereby referred as the number of children per woman which has declined from 4 or more per woman to 2 or fewer today. The fertility transition in Taiwan began in the 1950s. By the mid-1980s, the fertility rate fell below the population replacement level . In the late 1990’s, the fertility rate rapidly slid, making Taiwan an lowest-low-fertility-rate country; currently, every woman in Taiwan has only 1.12 children, the lowest in the world. The lowest-low fertility rate implies the rising severity of the trend towards fewer children, and a predictable decrease in labor force. Therefore, we have to advocate for the health of the nation’s future masters.

The second challenge is the social change of Taiwan, particularly those relevant to the children’s growth environment. Between 1980 and 2005, for example, there were radical changes in areas such as poverty, labor force market, marriage, and family composition. To begin, the gap between the rich and the poor has widened, as the Gini coefficient increased from 0.277 to 0.340. In addition, the women’s labor participation rate  in Taiwan has increased from 39.3% in 1980 to 48.1%. Moreover, late marriage and divorce has become increasingly common— the age at the first marriage has been delayed from 27.6 years (male) and 23.8 years (female) to 30.6 years and 27.4 years of age respectively, while the crude divorce rate increased from 0.8 % increased to 2.7 %. It is worth noting that, onwards from the 1990s, transnational marriage (or international marriage, cross-cultural marriage) has become increasingly popular, particularly amongst Taiwanese male. In 2003, transnational marriage reached its peak at one per 3.1 marriages, and it has now been reduced to one per 5.5 marriages. Finally, there are changes in family composition. With divorce and separation prevailing, single-parent families rapidly increased from 5.8% in 1988 to 7.7% in 2004 [11]. Because the growth of children and their health is closely related to the environment they grow in, we must better understand the impact of these social changes, and formulate policies accordingly to promote child health in the new century.

The third and last challenge is the rise of the life course perspective . To an extent, there is another layer of purpose to why we urge you to reconsider and address child health issues. In the past, regardless in academic or practical work, our understandings for the causes of non-communicable diseases (e.g. heart disease, hypertension, diabetes, asthma) have been restricted to our lifestyles or risk factors in adulthood (e.g. high cholesterol, obesity, lack of exercise, smoking). However, in the late 1980s, British epidemiologist DJP Barker proposed “Fetal Programming Hypothesis” (a.k.a. “Barker’s Hypothesis), the aforementioned arguments on “early origins of adult health”  became a major challenge to investigating the cause of non-communicable diseases. Meanwhile, life course epidemiology  came into being. According to the life course epidemiology perspective, health and disease in adulthood are the results of individual, cumulative, or interactive effects from long-term interactions between the organism and social factors during different life phases (e.g. pregnancy, childhood, adolescence, adulthood).

Simply put, childhood health is an important foundation or capital for adult health. Since the pursuit of health is one of the important goals in life, we must place importance on child health. In light of this perspective, the Health Promotion Administration, under the Ministry of Health and Welfare, proposed the “Taiwan Birth Cohort Study” (TBCS) as a warm-up to address the new challenges in child health in the 21st century. In essence, TBCS is a large-scale longitudinal project with three specific aims: (1) record and assess the health trajectory of Taiwanese children in the new century; (2) investigate the influence of social environment on child health; (3) study the relationship between child health and adult health. In the sections below, we lay out the basic concepts and perspectives, planning and design, and the characteristics of the participating children.


Basic concepts and ideas

The topics we wish to explore with TBCS are: the health trajectory of Taiwanese children in the new century, and the determinants of such health trajectory. But what is child health? What are determinants of child health? We aim to use this section to describe some of the basic concepts and ideas that will help the readers understand the planning and design of TBCS. We’d like to begin with child health. According to the World Health Organization’s (WHO) Charter  enacted in 1948, health is “a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity.” Despite criticisms of this definition being “ambiguous” or “too ideal,” it is significant in that it broke through the traditional perspective where health is “the absence of disease.” The WHO definition is also positive and multi-faceted. In 1986, the WHO Ottawa Charter further pointed out that “to reach a state of complete physical, mental and social well-being, an individual or group must be able to identify and to realize aspirations, to satisfy needs, and to change or cope with the environment. Health is, therefore, seen as a resource of everyday life, not the objective of living.” In other words, when we start from the perspective of health promotion, health is not only a status, but also a potential to reach other goals.

TBCS adopted WHO’s definition of health and developed instruments to explore three domains: Growth and Development, Disease and Symptoms, and Health Potential. The Growth and Development domain includes physical, emotional, social, cognitive and lingual changes; the Diseases and Symptoms domain captures abnormalities in the children’s health; the Health Potential domain includes susceptibility, ability to recover, and adaptability.

The instruments were also developed to explore the determinants of child health. According to the socioecological perspective, a series of concentric circles (systems) center the child and expands to the family, the community, and the nation and society. A child is born and developed within this supporting series of concentric systems, where many determinants of health come into play at each system; but because these determinants may be embedded between different systems, each level may interact with and influence each other based on the “mosaic model”.

In general, the determinants of child health can be grouped into three categories: biological, behavioral, and environmental. Biological determinants include gene expression, fetal development and biological constraints, etc.; behavioral determinants include the child’s emotion, beliefs, cognition, attitudes, health behaviors and lifestyles; environmental determinants include a variety of exposure to chemical, physical and biological hazards, as well as social conditions such as social structures, resources, processes, systems, and culture; in particular, although social environment is considered a distal determinant of health, it is the determinant of many proximal determinants of health. Therefore, TBCS is particularly focused on impacts of social environment on children’s health.

The life course perspective  is another foundation of TBCS. As previously mentioned, one of the goals of TBCS is to investigate the relationship between child health and adult health. According to the life course perspective, child health can impact adult health through the critical period model  or risk accumulation model. The critical period model describes that exposure to certain factors at a particular period during childhood may have lifelong or latent impacts on the function or structure of the body systems, tissues or organs. For example, Barker reported the correlations between fetal mal-development in the womb and adult cardiovascular disease, diabetes and hypertension. The risk accumulation model, on the other hand, describes the exposure to risks during different periods over the life course, which results in chain reaction or additive effects that are pathogenic and harmful to the function or structure of the body systems, tissues, and organs.

In Figure 2 we show the conceptual framework of the social environment aspect of TBCS. We assume that socioeconomic status (SES) during pregnancy and childhood, via different paths such as health during childhood and SES in adulthood, can further affect adult health. According to the literature review by Galobardes et al, up to 2003, there were only 29 empirical studies that explored the relationship between childhood SES and adult health. Other than the causes of death and cardiovascular disease, conclusions for other specific diseases are diverged, and new evidence is required.


Planning and Design

The Taiwan Birth Cohort Study officially launched in 2003. Since the early stages of the study, the Principle Investigator has invited Taiwanese scholars and experts from multiple disciplines, including public health, medicine, child developmental psychology, demography, and social work to formulate an multidisciplinary research team. The team is tasked with project timeline planning , sample design , survey instrument development , database establishment , and data analysis and interpretation of results related to the study . In this longitudinal study, we sampled from the 2005 birth cohort (the population) and fielded 4 waves of follow-up surveys (when the sampled children were 6 months, 18 months, 3 years and 5.5 years of age). We will continue to follow up with the samples at least until they turn 21 years old and establish a complete longitudinal database, achieving our research objectives.

To ensure the quality of the large-scale fieldwork, the research team planned pilot studies. In pilot studies we sampled 29 townships and 2,048 individuals from babies born in November and December 2003 with simple random sampling  and conducted interviews at the ages of 6 months, 18 months and 3 years, and 5.5 years. From the piloted results, we were able to review the sampling design, survey instruments, fieldwork feasibility and appropriateness, and use them as the basis of planning and execution of the formal investigation.

Overall, TBCS is mainly divided into three parts: sampling design, survey instrument development and fieldwork. Sampling design and execution are tasked to the Population and Health Research Center  of the Health Promotion Administration. They also assist in obtaining or cascading related information. Survey instrument development is tasked to the research team. The fieldwork is tasked to Asia University’s Operation Center for Community Health Care , entrusted by the Health Promotion Administration. The task assignments and progress of each part is described in the following sections.


Sampling Design

The targeted population for the Taiwan Birth Cohort Study consists of all live births in Taiwan (excluding Kinmen County, Lienchiang County) from January 1, 2005 to December 31, 2005, using the birth document provided monthly by the Bureau of Health (now the Health Promotion Administration) as the sampling frame. In order to have a fully representative sample of the declared live births in 2005 in Taiwan, this study adopted the two-stage stratified random sampling method. Before obtaining the sampling frame, it is impossible to predict the number of birth for each month of the year; hence, the sampling design was created accordingly to the number of live births in the 2002 household registration of all townships in Taiwan. The primary sampling unit of this study was township . To reflect the different degrees of urbanization and birth rate, in the first phase we divided the 369 townships into 12 layers according to the degree of urbanization (districts under special municipalities and provincial cities , county-controlled cities , urban townships , and rural villages ) and total fertility rate  (high, medium, low) for the system to randomly select samples of 85 primary sampling unit (actually containing 89 townships) that are distributed in the northern, central , southern, and eastern regions of Taiwan, as shown in Figure 1-2.

In the phase two, the sample cases (individuals) were directly extracted from the sampled townships. The sample size of each layer is proportional to the population size (probability proportional to size), and evenly distributed between townships in each layer of sample . In other words, the sample size for each township is approximately 300 participants. We assigned the number of births in 2002 for each sample township as the denominator, and the assigned sample size as the numerator, to calculate the probability of withdrawal ; then we sorted the birth notification files based on their birth months and chronological order, and picked out 24,200 individuals based on simple random sampling method, at an average rate of 11.7% withdrawal  rate.


Survey Tool Development

The primary principle of the survey instrument development echoes this experiment’s two fundamental ideas: child health and its determinants. The content of the first to fourth wave of the survey instrument consists of three parts: Infant/Child Health Care Needs Survey (referred to as the Main Survey), Infant and Child Development Scales, and Parents’ Self-Perceived Health Status.

The structure of the Main Survey is broadly divided into five sections: mother, child, family, community, and special scales. In the child section, additional aspects such as health behavior and media exposure were implemented to the third and fourth waves of the survey. In the family section, the aspect of family learning environment was added to the third and fourth waves. In the special scale section, “attachment behavior scale ” and “societal friendliness towards foreign spouse scale ” were added to the third wave; in the fourth wave we added “child temperament scale ” and “child behavior scale (M-CHAT);  in the third and fourth waves we added the “child health potential scale.”

The Infant/Child Development Scale covers four aspects of development: gross motor, fine motor, language communication, and social competency to handle surrounding affairs . In the scale’s design we referenced instruments with established reliability and validity  domestically and internationally; and, after numerous expert discussions regarding the selection of questions and the use of pilot study data, construct validity and internal consistency were established. By sampling another 100 subjects of the same age and having them respond to the infant development scale and take the Bayley Scale of Infant Development –II test, the results demonstrated good criterion validity. As for the Parents’ Self-Perceived Health Status, we used SF-36, a Taiwanese edition, which includes the following eight dimensions: physical physiological function, role limitation due to physiological function, bodily pain, general health, vitality, social functioning, role limitation due to emotions, and mental health.

Moreover, although this study was created for establishing long-term monitoring databases and academic research, it also serves an intervention function for infants and children with special service needs, hence the “infant health care needs assessment and willingness to accept referral services ” scale was attached to the main survey. Before the interviewers completed their visits and survey interviews, they went through the research team’s criteria to check items (e.g. if the child needed but has yet received treatment for congenital deficiency diseases, if child cannot perform any of the five developmental actions or behaviors, if the child has yet received any of the vaccinations) as a preliminary screening for children that may have serious health problems or special healthcare needs. After asking for the respondent’s willingness and permission, children were referred to a medical or welfare unit by the Health Promotion Administration for further assessment and services.



The first to fourth waves of the survey were implemented when the subjects were of ages 6 months, 18 months, 3 years, and 5.5 years old respectively. The fieldwork was completed in the following periods: July 2005 to July 2006, July 2006 to July 2008, January 2008 to January 2009, and June 2010 to July 2011. All four waves of the survey were conducted via in-person interviews, prioritizing the child’s mother as the respondents. If the mother was separated from the child due to a divorce, was deceased, had serious communication issues, was not in town for a period of time, or other reasons that made them unable to interview, the respondent was replaced by the primary caregiver who was most familiar with the child.

In the design of the fieldwork workflow, we aimed for high interview completion rates and effective quality control. The fieldwork workflow consists of three stages: preparation work, fieldwork, and data review and validation. In the preparation stage, in addition to the aforementioned sampling design, development and preparation of survey instruments, we selected more than 80 certified professional interviewers who were trained before the assignment and were assigned to various sampling areas to interview. In addition, the Health Promotion Administration sent letters to the county and city health departments, police departments, regional health offices , household registration offices, township offices and other units in the area where the participants reside, asking to provide our interviewers with assistance as necessary.

In the fieldwork stage, prior to contacting and arranging the first interview visit, the interviewers sent the “Letter to Respondents” signed by the Health Promotion Administration’s Head , explaining the purpose of the survey, the way the samples were selected, the confidentiality of the collected data, and other related query information, etc. to improve respondent’s willingness to participate and cooperate. Upon successful contact with the respondent, the interviewer should have first presented the consent form to the respondents during each four waves of the survey. After the respondents understood their rights to participate and obligations in this study and signed the consent form, the interviewer started the interview.

To reduce withdrawal rate , we established a comprehensive case monitoring and referral system; if the respondent could not be reached by the contact information listed on the birth notification file, the new contact information, found through household registration data at household registration office or through the assistance of a neighbor, a village administrator, or the jurisdiction’s police, was provided to another interviewer at the corresponding area to continue tracking. Furthermore, if during the investigation the interviewer faced interviewing difficulties or had doubts about survey instruments, the supervisors and research team were immediately available to provide advice and recommendations, and simultaneously review completed surveys in order to effectively correct recovered data.

To ensure the quality of the recovered survey data, we also conducted spot checks and logic reviews . The spot check sampling ratio was 20%. The check includes confirming if the interviewers collected in accordance with the protocol for in-person interviews to designated respondents, carefully asked items one by one, page by page, and reviewed the answers based on the corresponding questions. After data entry, using the data’s reasonable range of values and the logical relationship between the item and answer, review criteria were developed, and a computer software was used to execute the logic review. Data from all four waves of survey went through 80 or more review criteria.

Characteristics of the Participating Children

The Taiwan Birth Cohort Study has currently completed the fieldwork for first to fourth waves of the survey, and the result of the overall execution was quite good. The first wave of fieldwork yielded 21,248 cases, an interview completion rate of 87.8%; the second wave yielded 20,172 cases, an interview completion rate of 94.9%; the third wave yielded 19,910 cases, an interview completion rate of 93.7%; the fourth wave yielded 19,721 cases, an interview completion rate of 92.8%. As shown by the completed interview rates for each wave of survey, the retention rate over the first and fourth wave was 77.9% (i.e. 77.9% completed all four surveys). Of the incomplete interview cases in each wave, the primary reason for not having an interview was that the mother or primary caregiver refused an interview. Other reasons include: the whole family migrated and the new address could not be found, respondent could not be reached because other family members refused, or respondent was not in the country for the long term and could not arrange an interview within the fieldwork timeline.

Broadly speaking, the basic characteristics of the children and families did not change much between the first and the fourth wave. For example, in the first wave of the survey sample that completed the interview, there were more males (52.5%) than females (47.5%), and the vast majority was single birth, accounting for 97.4%. As for birth outcomes, 6.9% were low-birth-weight children who were born with less than 2,500 grams at birth, and 8.4% were premature babies who were born in fewer than 37 gestational weeks. Most of the mothers’ age were 25-34 years old (46.1%) and 35 and above (53.0%), with few mothers below the age of 24 (0.9%). In terms of transnational families, the percentage of mothers whose original nationality is not Taiwanese was 13.0% (4.5% from Mainland China/ Hong Kong and Macao region, and 8.8% were of other nationalities). For the fathers’ and mothers’ educational attainment, most of them had a bachelor’s degree or higher, at 45.8% and 45.1% respectively. Parents’ monthly income  of less than $30,000 NT accounted for 11.7% of the participants. Most sampled individuals resided in regions counted as “city or urban townships,” accounting for 45.2%, while “district” and “rural townships” each accounted for approximately 27%.



With the Taiwanese population structure rapidly changing and the increasing international attention to child health, we are facing a major challenge in child health in Taiwan in the 21st century. It is also the time for reflections and actions. TBCS, a large longitudinal study administered by the Ministry of Health and Welfare’s Health Promotion Administration, aims to record and evaluate the changes of child health in modern Taiwan, understand the impacts of social environment on child health, and investigate the relationship between child and adult health. One of the motivations in planning and publishing “The Health Profile of Taiwanese Children in the 21st Century” book series, we used the data collected from the Taiwan Birth Cohort Study to present the overall condition of current Taiwanese children’s health, and to identify practical policy issues in order to formulate response to potential child health challenges in the future.



Comments are closed

Panorama Theme by Themocracy