This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
This study examined disparities in 1-year mortality rates between infants born to married and unmarried single-parent families, emphasizing the need for targeted health policies.
Methods
Data from 3,298,263 cases, obtained from the South Korea National Statistical Office (2010–2017), were analyzed. T-tests and chi-square tests were used to assess the demographic characteristics of the study group. The number of deaths per 1,000 live births was calculated, and logistic and multivariable logistic regressions were employed to compare infant mortality rates between family types. Additional stratified analyses, based on gestational age and birth weight, further elucidated the relationship between parental marital status and infant mortality.
Results
Infants from unmarried families exhibited a 3.34-fold higher crude odds ratio (OR) for 1-year mortality (95% confidence interval [CI], 2.56–4.36; P<0.001) than that from married families. After adjusting for confounders, the adjusted OR was 1.40 (95% CI, 1.03–1.92). Stratification by gestational age and birth weight revealed crude ORs of 4.62 (95% CI, 3.34– 6.39) in non-preterm infants (≥37 weeks) and 4.76 (95% CI, 3.46–6.56) in non-low-birth-weight infants (≥2.5 kg), highlighting a more pronounced disparity in infants born at or above normal weight and full term. No significant difference (P>0.05) was found in the crude OR for 1-year mortality rates among low-birth-weight (<2.5 kg) or preterm (<37 weeks) infants between the two-family types.
Conclusion
This study highlights the significant disparity in infant mortality rates based on parental marital status, underscoring the need for enhanced social support and tailored policies for unmarried single-parent families.
In recent decades, out-of-wedlock births have increased worldwide due to evolving marital values, shifting social perceptions, increased premarital sex, and various socioeconomic and cultural factors. In the United States, out-of-wedlock births rose from about 5% in 1960 to 35% in 1995 [1], with a similar increase in single-mother families observed in other Western nations, including Britain or Sweden [2]. The number of children born outside marriage has sharply increased, with the Organisation for Economic Co-operation and Development (OECD) indicator averages tripling from 11% in 1980 to nearly 33% in 2007 [3]. The upward trend in sole-parent households across OECD countries is consistent, with projections for 2025–2030 indicating a likely increase from 22%–29% [3]. Out-of-wedlock births in South Korea remain low compared to those in Western countries; regardless, they have gradually increased since the 1990s. In 2019, 6,974 babies—accounting for 2.3% of all births—were born out of wedlock in South Korea [4].
Many existing studies indicate that infants born to unmarried single-parent families have higher mortality rates than those born to married families [5-7,23]. According to previous research focused on 36,608 singleton births from Black mothers in the District of Columbia between 1980 and 1984, infants born to unmarried mothers had a 35% higher neonatal mortality rate (relative risk [RR], 1.35) than that of those born to married mothers [5]. Additionally, a study analyzing 40 million births in the United States, from 1995–2004, found that the 1-year infant mortality rate per 1,000 live births was 2.8 for infants born to married mothers and 5.6 for those born to unmarried mothers. After adjusting for various independent variables, infants born to unmarried mothers exhibited a 35% higher 1-year mortality rate (RR, 1.45; 95% confidence interval [CI], 1.42–1.47) than that of those born to married mothers [6]. In a study involving mothers from Africa, specifically Burkina Faso, Sierra Leone, and Burundi, single mothers were 4.4 times more likely to report neonatal deaths than ever-married women (odds ratio [OR], 4.4; 95% CI, 2.701–7.221). Regarding 1-year infant mortality, single mothers were 3.7 times more likely than ever-married women to experience infant mortality (OR, 3.7; 95% CI, 2.714–5.1436) [7]. Furthermore, a study conducted in India reported that the infant mortality among unmarried women was 125 per 1,000 live births, more than double the rate among infants born to married women [8].
Various studies have analyzed the causes of higher mortality rates among infants born to unmarried mothers compared to those born to married mothers from both physiological and social perspectives. Physiologically, infants born to unmarried mothers are at greater risk of low birth weight, preterm birth, and being small for their gestational age than those born to married mothers [9,10]. Additionally, social determinants—including social policies and capital, social cohesion, material conditions, maternal age, maternal psychosocial factors, maternal receipt of prenatal care, parental health behaviors, individual and neighborhood socioeconomic statuses, living and working conditions, social conditions, environmental conditions, and residential segregation—have been shown to affect infant mortality rates [7,8,11-19].
Infant mortality within the 1st year after birth is an important indicator of maternal and child health and reflects the efficiency and operational status of the national or regional health systems. Infant mortality is closely related to the social and economic environment and correlates with factors such as educational level, poverty rate, and access to maternal healthcare.
Thus far, no study in South Korea has compared mortality rates within 1 year after birth between infants born to unmarried single-parent families and those born to married families. This thesis aims to underscore the importance of health policies and the social attention and support needed for unmarried single-parent families by comparing the 1-year infant mortality rate between married and unmarried groups and by analyzing the factors affecting this disparity.
Methods
Data source and study design
In this study, linkage statistical data on births and deaths of infants aged <5 years, from 2010–2017, provided by the Korea National Statistical Office, was used. In South Korea, all parents are legally obligated to report and register their child within 1 month after birth. When reporting a birth, parents must submit a birth registration form along with a birth certificate, and the registered child is issued a resident registration number. The birth registration form requires the parents’ registration numbers and marital status. A birth certificate is typically issued by the hospital where the baby is born. It contains information such as the place and date of birth, duration of the mother’s pregnancy, birth weight, whether it involved multiple births, and the number of children birthed.
When a child dies in South Korea, parents are also legally required to report the death within 1 month. The death report must include a death certificate, typically issued by the hospital, detailing the deceased child’s resident registration number, birth and death dates, cause, and location of death. The Korea National Statistical Office derived statistical data linking births and deaths of infants under 5 years of age using resident registration numbers, correlating birth registration details with death records. This record formed the data source for this study. The raw statistical data on “Births-Deaths Linkage Database for Children under 5 years old in 2010–2017” includes details such as the year and month of birth, sex, survival status within 5 years, parent’s legal marital status at birth, year of marriage, gestational age, birth weight, place of birth, birth order in cases of multiple pregnancies, total number of births, parents’ residential address at birth, parental ages, occupations, nationalities, education levels, as well as the year, month, day, time, age, place, and cause of death for the deceased child. All data were processed anonymously. The year of marriage and other parental information in the child’s data source were derived from the parents’ registration numbers recorded during the child’s birth registration. Based on the parents’ resident registration numbers, parental information, including their year of marriage, was tracked within approximately 4 months after the child’s birth registration and added to the “Births-Deaths Linkage Database for Children under 5 years old in 2010–2017” database by the Korea National Statistical Office, along with data from the parents’ legal marriage certificate. The presence of the year of marriage in the “Births-Deaths Linkage Database for Children under 5 years old in 2010–2017” data source indicates the legal marital registration of the child’s parents. Consequently, the presence or absence of the parents’ year of marriage in the data source distinguished unmarried and married families.
This study compared the 1-year mortality rates of infants born to unmarried versus married families in South Korea and investigated the effects of various factors, such as maternal age, education level, occupational status, infant birth weight, and gestational age, on these rates. Cases in which the parents’ marital status was unknown, the parents were not native Korean nationals, or the birth-to-death linkage was deemed unreliable were excluded from the dataset.
The initial database of linkage statistics on births and deaths of infants under 5 years (2010–2017) contained 3,500,314 records. Samples with unknown marital status (6,156) or non-native Korean parents (195,733) were excluded. Additionally, 162 samples with uncertain information regarding the link between birth and death were removed (these typically involved orphans found dead without resident registration numbers). The final study population included 3,298,263 individuals: 9,645 children from unmarried families and 3,288,618 from married families. Figure 1 illustrates the derivation of the study population from the overall sample.
Factors related to infant 1-year mortality rates
In this study, the independent variables included the marital status of the parents, maternal age, maternal educational level, maternal employment status, infant gestational age, and infant birth weight. The dependent variable was whether the infant died within 1 year of birth.
According to 2016 data from the South Korea National Statistical Office, 1,539,878 single-parent households were recorded, of which 395,772 (25.7%) were head by unmarried fathers, a smaller proportion compared to households with unmarried mothers. In the study population data, the percentage of missing data for fathers’ ages, education levels, and employment statuses differed significantly from that for the mothers. This discrepancy leads to statistical uncertainty due to the varying rates of missing values between the father and mother groups. Therefore, parental variables related to the father (parental age, educational level, and occupational status) were not used as covariates in the statistical analysis when comparing the 1-year infant mortality rates of unmarried single-parent families versus married families.
The criteria for determining the parental age were as follows: in South Korea, individuals under 19 years are legally minors, making this age a cutoff for parental age. The World Health Organization (WHO) and International Federation of Gynecology and Obstetrics define advanced maternal age as 35 years or older, which is associated with adverse perinatal outcomes, such as low birth weight, preterm birth, neonatal deaths, neonatal intensive care unit (NICU) admissions, and gestational diabetes mellitus [20]. Thus, 35 years of age was used as another cutoff for parental age. Consequently, parental age was categorized into four groups: under 19 years, 19–35 years, 35 years and older, and unknown ages.
The criteria for determining the parental education levels were as follows: The 2020 South Korea Social Trend Study found that 11.8% of mothers of out-of-wedlock children had an education level of middle school graduate or below, 42.8% were high school students or graduates, 38.1% were college students or graduates, and 3.1% had advanced degrees [4]. Therefore, parents’ education levels were categorized into four groups: college graduates or higher, high school students or graduates, middle school graduates or lower, and unknown education levels.
In South Korea, the healthcare system is well-developed, and most births occur within hospitals. Giving birth outside of a hospital can indicate limited access to medical care. Thus, the place of birth was categorized into two groups: hospital and nonhospital.
In terms of parental occupation, the raw statistical data from the “Births-Deaths Linkage Database for Children under 5 years old in 2010–2017,” parents’ occupations were subcategorized into 10 detailed types. Since this study aimed to examine how parents’ economic statuses affect the difference in 1-year mortality rates between infants born to married and unmarried mothers, it is impossible to accurately assess specific economic levels based solely on parents’ occupations. Therefore, in this study, parents’ occupations were subcategorized into three groups—employed, unemployed, and unknown—rather than subdivided into 10 detailed categories, focusing instead on whether parents were economically active.
Regarding the infant’s gestational age, the WHO defines an infant with a birth weight of less than 2.5 kg as low birth weight. Low birth weight was further categorized as: infants weighing less than 1.5 kg being classified as having very low birth weight and those weighing less than 1 kg as having extremely low birth weight. Infants with a birth weight of 4 kg or more are classified as macrosomia. Given that studies have shown different birth outcomes based on birth weight [21-24], this study classified the birth weights of infants into five categories: extremely low birth weight (<1 kg), very low birth weight (<1.5 kg), low birth weight (1.5 to <2.5 kg), normal (2.5 to <4 kg), and macrosomia (≥4 kg).
For gestational age, we classified the infants into five categories based on the standards set by the WHO and the American College of Obstetricians and Gynecologists: extremely preterm (<28 weeks), very preterm (28 to <32 weeks), moderate-to-late preterm (32 to <37 weeks), term (37 to <42 weeks), and postterm (≥42 weeks or more).
Statistical analysis
In this study, to analyze and compare demographic characteristics, a chi-square test was used for categorical variables (including infant sex, place of birth, education level, employment status, and survival status after birth), and the t-test was used for continuous variables (such as parental ages, infant’s gestational age, and infant’s birth weight) (Table 1).
We employed two methods to compare the 1-year mortality rates of infants born to unmarried and married parents: calculating the crude death rate per 1,000 births for both groups, and performing logistic regression to determine the crude OR. In addition to calculating the OR, we used multivariate logistic regression to obtain the adjusted OR (aOR), accounting for variables that could influence mortality rates, such as maternal age, education level, employment status, gestational age, and birth weight. This approach enabled us to evaluate the impact of each adjusted variable on the 1-year infant mortality rate through aORs (aOR1 in model 1, aOR2 in model 2, aOR3 in model 3, and aOR4 in model 4), as shown in Table 2. The crude death rate per 1,000 births was used to compare international organizational metrics, such as OECD Health Statistics and those of the WHO, which is commonly adopted to evaluate maternal and child health conditions worldwide. This statistical method is effective in intuitively presenting the difference in the number of infant deaths within 1 year between married and unmarried families. However, the OR was useful for determining whether the difference in infant mortality between the two groups was statistically significant.
Given that infant mortality rates are significantly affected by factors such as the infant’s gestational age and birth weight [21- 24], we classified infants based on these criteria (<37 weeks group versus ≥37 weeks group; <2.5 kg group versus ≥2.5 kg group) and performed a stratified analysis of 1-year mortality rates (Tables 3, 4). To compare the 1-year mortality rates between infants born to unmarried and married mothers within these strata, we employed the same statistical methods as those used for the overall study population (Table 2).
Supplementarily, we also compared the ORs for each subcategory within the independent variables (parental age, parental education level, parental employment status, infant gestational age, and infant birth weight), as shown in Supplement 1.
Statistical significance was defined as a P-value of less than 0.05, with 95% CIs frequently employed to describe the strength of the associations. All analyses were performed using the Stata MP ver. 18.0 (Stata Corp.).
Results
Demographic characteristics of married group compared to the unmarried group
The demographic and clinical characteristics of the participants are presented in Table 1. Children born to unmarried families and those born to married mothers did not significantly differ by sex (P=0.353).
In contrast, significant differences were observed in the place of birth between the two groups (P<0.001). Among children delivered outside hospitals, 1.19% were born to married families, while 3.16% were born to unmarried families.
The average age of the mothers of children born to married families was 32.07 years, whereas that of mothers of children born to unmarried families was 29.95 years. The average age of fathers in married families was 34.30 years, while that of fathers in unmarried families averaged 33.09 years, showing a significant parental age difference (P<0.001). Cases in which the age was unknown were excluded from the calculations. The percentage of parents under 19 years was 0.12% for mothers in married families, 13.42% for mothers in unmarried families, 0.04% for fathers in married families, and 7.60% for fathers in unmarried families.
Significant differences in the educational levels of mothers and fathers between married and unmarried families were noted (P<0.001). Patients with unknown educational levels were excluded from the analysis. The percentage of parents with educational levels below high school was 24.80% for mothers in married families, 66.85% for mothers in unmarried families, 25.20% for fathers in married families, and 64.79% for fathers in unmarried families.
Regarding parental employment, significant differences were apparent between married and unmarried families (P<0.001). Cases where employment status was unknown were excluded from the calculations. The percentage of unemployed parents showed that 62.92% of mothers in married families, 79.20% of mothers in unmarried families, 0.84% of fathers in married families, and 8.16% of fathers in unmarried families were unemployed.
Notable differences in infant gestational age were observed between married and unmarried families (P<0.001). The percentage of infants born preterm (<37 weeks) was 6.43% in married families and 9.30% in unmarried families.
Additionally, a significant difference in infant birth weight was observed between the two family types (P<0.001). The incidence of low birth weight (<2.5 kg) was 5.47% for infants from married families, compared to 8.03% for those from unmarried families.
Infant mortality within 1 year after birth markedly differed between married and unmarried families (P<0.001). However, no significant differences were observed between the groups for the mortality status within the 2–3-year and 3–4-year age groups, with P-values of 0.727 and 0.274, respectively. The mortality status of the 4–5-year age group significantly differed again (P=0.038). Overall, a significant difference was found when comparing the overall mortality status of infants under 5 years after birth (P<0.001).
1-Year mortality rates of the unmarried group compared to the married group
Overall study population
Table 2 presents a comparison of infant mortality within the 1 year after birth between married and unmarried families in the entire study population. The number of infant deaths within the 1st year per 1,000 live births was 1.7 for children born to married families and 5.7 for children born to unmarried, single-parent families.
Among the overall study population group (n=3,298,263), when calculating the crude OR without adjusting for any variables (maternal age, maternal education, maternal employment, infant’s gestational age, infant’s birth weight), the odds for unmarried families were 3.34 times higher (95% CI, 2.56–4.36; P<0.001) than that for married families.
When adjusting for maternal age in the overall study population, the aOR1 indicated that the odds for unmarried families were 2.23 times higher (95% CI, 1.66–2.99; P<0.001). Following adjusting for maternal education, the aOR2 was 1.80 (95% CI, 1.34–2.41; P<0.001) for unmarried families. Additionally, after adjusting for maternal employment, the aOR3 for unmarried families was 1.79 (95% CI, 1.34–2.40; P<0.001). Finally, after adjusting for infant gestational age and birth weight, the aOR4 for unmarried families was 1.40 (95% CI, 1.03–1.92; P=0.033).
Stratified analysis by infant’s gestational age
For the stratified analysis by gestational age, the overall study population was divided into two groups: the <37 weeks (n=215,453) and ≥37 weeks (n=3,082,810) groups. To compare infant mortality within the 1st year after birth between married and unmarried families, the same statistical methods employed in Section 1 of the overall study population analysis were used (crude death rate per 1,000 births, logistic regression, and multivariate logistic regression). The results are presented in Table 3.
<37 weeks group: When analyzing only the samples within the <37 weeks group (n=215,453), the 1-year infant mortality per 1,000 live births was 13.1 for those born to married families and 20.0 for those born to unmarried, single-parent families, and the crude OR was 1.54 (95% CI, 0.96–2.46; P=0.071). The aOR1 was 0.65 (95% CI, 0.38–1.08; P=0.096), aOR2 was 0.65 (95% CI, 0.38–1.11; P=0.114), aOR3 was 0.70 (95% CI, 0.41–1.20; P=0.197), and aOR4 was 0.82 (95% CI, 0.48–1.41; P=0.470). In the <37 weeks group, there was no significant difference in 1-year mortality between infants born to married and those born to unmarried families (crude aOR1, aOR2, aOR3, and aOR4 >0.05).
≥37 weeks group: When evaluating only the samples within the ≥37 weeks group (n=3,082,810), there were 0.9 infant deaths per 1,000 live births in the 1st year for children born to married families, compared to 4.2 infant deaths per 1,000 live births for those born to unmarried, single-parent families. The crude OR was 4.62 (95% CI, 3.34–6.39; P<0.001), and after sequentially adjusting for variables that influence 1-year infant mortality, the aOR1 was 2.78 (95% CI, 1.92–4.02; P<0.001), aOR2 was 2.15 (95% CI, 1.49–3.09; P<0.001), aOR3 was 2.14 (95% CI, 1.49–3.08; P<0.001), and aOR4 was 2.08 (95% CI, 1.45–2.99; P=0.033). Within the ≥37 weeks group, both the crude ORs and aORs (aOR1–4) for infant mortality at 1 year among infants born to married versus unmarried families showed P-values below 0.05, suggesting a remarkable difference in 1-year mortality rates between infants born to married families and those born to unmarried families.
Stratified analysis by infant’s birth weight
The overall study population was categorized into two groups based on infant birth weight—<2.5 kg (n=180,532) and ≥2.5 kg (n=3,117,731) groups—to conduct stratified analysis based on infant’s birth weight. The same statistical methods used in Section 1 were applied to compare infant mortality within the 1st year after birth between married and unmarried families. The results are summarized in Table 4.
<2.5 kg group: When assessing the samples exclusively within the <2.5 kg group (n=180,532), the number of infant deaths within the 1 year per 1,000 live births was 15.8 for children born to married families and 21.9 for children born to unmarried, single-parent families. The crude OR was 1.40 (95% CI, 0.86–2.27; P=0.171). Following a sequential adjustment for independent variables, the aOR1 was 0.73 (95% CI, 0.43–1.24; P=0.240), aOR2 was 0.71 (95% CI, 0.42–1.20; P=0.202), aOR3 was 0.79 (95% CI, 0.46–1.35; P=0.385), and aOR4 was 0.79 (95% CI, 0.46–1.35; P=0396). Within the low-birth-weight group, both the crude ORs and aORs (aOR1–4) for 1-year mortality rates for infants born to married versus unmarried parents showed P-values exceeding 0.05, signifying no statistically significant difference in 1-year mortality between the two groups.
≥2.5 kg group: When analyzing only the samples within the ≥2.5 kg group (n=3,117,731), there were 0.9 infant deaths per 1,000 live births in the 1st year for children born to married families, compared to 4.3 infant deaths per 1,000 live births for those born to unmarried, single-parent families. The crude OR was 4.76 (95% CI, 3.46–6.56; P<0.001). The aOR1 was 2.83 (95% CI, 1.96–4.10; P<0.001), aOR2 was 2.21 (95% CI, 1.54–3.18; P<0.001), aOR3 was 2.20 (95% CI, 1.53–3.17; P<0.001), and aOR4 was 2.13 (95% CI, 1.48–3.07; P<0.001). In the ≥2.5 kg group, the crude OR and aORs (aOR1–4) for 1-year mortality between infants born to married and unmarried families all had P-values of less than 0.05, indicating a marked difference in 1-year mortality between infants born to married and those born to unmarried families.
Analysis of infant mortality within the 1st year for each independent variable
Additionally, in the overall study population group (n=3,298,263), regardless of the parents’ marital status, we compared infant mortality within the detailed categories of each independent variable (parental ages, parental education level, parental employment status, infant gestational age, and infant birth weight); the results are presented in Supplement 1.
When assessing parental age, for mothers aged <19 years, the OR for infant mortality was 4.59 (95% CI, 3.37–6.25; P<0.001). In mothers aged 35 years or older, the OR was 1.28 (95% CI, 1.20–1.36; P<0.001) compared to those aged between 19–34 years. Similarly, if the father’s age was below 19 years, the OR for infant mortality was 5.19 (95% CI, 3.36–8.00; P<0.001), while for fathers aged ≥35 years, it was 1.17 (95% CI, 1.11–1.23; P<0.001), compared to those between 19 and 34 years.
Regarding parental education level, compared to mothers with a college education, the OR for infant mortality was 1.65 (95% CI, 1.56–1.75; P<0.001) for high school students or graduates and 3.77 (95% CI, 3.23–4.40; P<0.001) for mothers with a middle school education or less. For fathers, the OR was 1.61 (95% CI, 1.52–1.70; P<0.001) for high school students or graduates and 3.20 (95% CI, 2.76–3.70; P<0.001) for those with a middle school education or less, compared to fathers with a college-level education or higher.
Regarding parental employment status, infant mortality was 1.40 times higher (95% CI, 1.32–1.48; P<0.001) among infants whose mothers were unemployed compared to employed mothers. As for fathers, since the OR for infant mortality when unemployed was 1.02 (95% CI, 1.32–1.48; P=0.894) compared to employed fathers, the father’s employment status did not significantly affect for infant mortality. This may be attributable to the higher number of cases with unknown paternal education data than unknown maternal education data in single-parent households, leading to a lack of meaningful results.
Regarding gestational age, compared to the term-birth group (37 to <42 weeks), the extremely-preterm group (<28 weeks) had an OR of 230.95 (95% CI, 215.69–247.29; P<0.001), the very-preterm group (28 to <32 weeks) had an OR of 37.22 (95% CI, 33.96–40.80; P<0.001), and the moderate-to-late-preterm group (32 to <37 weeks) had an OR of 4.87 (95% CI, 4.51–5.26; P<0.001), indicating that the earlier an infant is born, the greater the 1-year mortality risk compared with term-born infants.
For infant’s birth weight, the extremely-low-birth-weight group (<1 kg) had an OR of 261.33 (95% CI, 244.15–279.72; P<0.001) compared to the normal-birth-weight group (2.5 to <4 kg). The very-low-birth-weight group (1 to <1.5 kg) had an OR of 41.93 (95% CI, 37.98–46.29; P<0.001), while the low-birth-weight group (1.5 to <2.5 kg) had an OR of 6.12 (95% CI, 5.68–6.60; P<0.001). This indicated that the lower the infant’s birth weight, the greater the 1-year mortality risk compared with normal birth weight infants.
Discussion
According to the demographic characteristics of the unmarried group compared to the married group (Table 1), all variables (maternal age, maternal education, maternal employment, infant’s gestational age, infant’s birth weight) that could affect 1-year infant mortality and birth outcomes showed significant differences between the two groups (P<0.001).
The proportion of births that occurring outside hospitals was significantly higher in unmarried families, suggesting that single mothers had limited access to medical services.
Additionally, the average age of parents in unmarried families was lower than that of their counterparts in married families. Moreover, the proportions of parents with an educational level of high school or below, the unemployment rate, and the rates of preterm birth and low birth weight among infants were all higher in unmarried families. These findings indicate that all variables influencing infant mortality within 1 year showed significantly worse trends among parents in unmarried families than among those in married families.
As a crucial indicator of maternal and child public health, the rate of death within 1 year after birth was also higher in unmarried families. This finding suggests a disparity in maternal and child health between married and unmarried families, indicating a greater need for government and societal attention and support for parenting and maternal health in unmarried families.
While a significant difference was observed in infant mortality within 1 year of birth between married and unmarried families, no marked difference was noted in the mortality of infants between the two groups during the 1–4-year period after birth (P>0.05). However, a significant difference in infant mortality between the two groups reemerged during the 4–5-year period after birth (P=0.038). This pattern of change, in which infant mortality disparities temporarily disappeared during the first 1–4 years of life and then reappeared after 4 years old, is likely the result of the combined effects of social policy and persistent social discrimination, as discussed in the following sections.
Since 2009, South Korea has implemented a policy providing five free health check-ups for all children up to the age of 6 years. Additionally, beginning in 2014, the free national immunization program was expanded to include a broader range of vaccines, such as the pneumococcal vaccine, provided at no cost to all children nationwide. These complimentary public health initiatives have played a key role in reducing disparities in healthcare access, particularly those rooted in socioeconomic differences between married and unmarried families, by improving early access to essential services such as routine check-ups and vaccinations for unmarried families, especially during a child’s early years. Moreover, during the 2010s, when the statistical data for this study were collected, South Korea’s welfare system for single-parent families primarily focused on supporting unmarried mothers with children under 5 years old, while assistance for older children was limited [25]. This policy orientation, which concentrated support on children under 5 years old, may help explain the reduced disparity in infant mortality rates between married and unmarried families observed during the 1–4 years period after birth. Collectively, these findings underscore the critical role of welfare and social policies in addressing infant mortality inequalities driven by the socioeconomic gap between the two groups.
The reasons for the re-emergence of a significant difference in mortality rates among children born to the two groups after 4 years of age are as follows. Although South Korea’s attitudes toward out-of-wedlock births have gradually improved, surveys indicate that these attitudes remain considerably low. As of 2020, only 25.0% of men and 23.8% of women expressed a positive view of single-parent families, reflecting a societal bias against single mothers [4]. The negative societal perception of single-parent families, along with the inclination to keep distance from them, has been consistently documented in previous studies [25,26]. Unmarried mothers directly experience these negative social perceptions [25-28], with research showing that 89% reported feeling significant discrimination [26]. As their children grow older, unmarried mothers who continue to seek economic assistance or welfare programs often face criticism from civil servants and encounter increasingly negative societal attitudes [25,28]. This disapproval is partly due to the perception that the need for government support becomes less acceptable once children reach a certain age. These negative experiences, particularly in public offices, hospitals, and other environments, foster fear and discouragement, making it more difficult for them to access the necessary assistance or medical care [25-28]. As their children age, the sense of discouragement and fear among unmarried mothers deepened. Furthermore, studies have revealed that unmarried mothers often become socially isolated, losing contact with their original families and biological fathers [25-28]. This isolation compels them to engage in economic activities independently, leaving their children unsupervised. Statistics confirm this trend: while 47.0% of unmarried mothers care for their children at home until the child turns 3 years old, this drops to 13.7% after the child turns 3 years old [25]. Among unemployed single mothers, 37.5% engaged in personal caregiving, but only 16.9% of employed mothers did,showing that more economically active single mothers spend less time directly caring for their children [25]. In summary, unmarried mothers face increasing social and economic isolation as their children grow older, and a lack of social welfare policies for older children exacerbates their struggles [25-27]. Consequently, unmarried mothers are increasingly required to enter the workforce, thereby reducing the time and energy they can dedicate to their children [25]. This situation contributes to a paradoxical increase in the significance of mortality differences among children over the age of 4 years between the two groups, as socioeconomic disadvantages become more pronounced as children grow older. These findings emphasize the importance of eliminating discriminatory attitudes and negative social perceptions from both social and public health perspectives, and underscore the importance of strengthening social policies and welfare systems for older children in single-parent families.
Our study examined disparities in infant mortality rates according to parental marital status by comparing children born to married and unmarried families. Two methods were used for this comparison: the number of deaths within the 1st year per 1,000 live births and the OR (crude OR and aOR1–4).
First, when comparing infant mortality between the two family groups using the number of deaths per 1,000 live births in the overall study population group (n=3,298,263) (Table 2), we found that 1.7 infants per 1,000 live births in married families and 5.7 infants per 1,000 live births in unmarried families died within the 1st year, indicating that infant mortality in unmarried families was 3.35 times higher. When stratifying the analysis by gestational age (Table 3) and birth weight (Table 4), in the <37 weeks group (n=215,453), the number of deaths within the 1st year per 1,000 live births was 1.53 times higher in unmarried families (13.1 in married families versus 20.0 in unmarried families), and in the the ≥37 weeks group (n=3,082,810) it was 4.67 times higher in unmarried families (0.9 versus. 4.2). In the <2.5 kg group (n=180,532), the rate was 1.39 times higher (15.8 versus. 21.9), and in the ≥2.5 kg group (n=3,117,731), it was 4.78 times higher (0.9 versus 4.3) in unmarried families than in married families. Thus, when comparing infant mortality rates using the number of deaths within the 1st year per 1,000 live births, the disparity in infant mortality between married and unmarried families was more pronounced among infants born at or above normal gestational age or birth weight than among those born preterm or with low birth weight.
Second, we compared infant mortality rates between married and unmarried families using ORs. Without adjusting for any independent variables, the crude ORs in the overall study population group (n=3,298,263) (Table 2) was found to be 3.34 times higher in the unmarried group. In the ≥37 weeks gestational age group (n=3,082,810) (Table 3), the crude OR was 4.62 times greater for the unmarried group, and in the ≥2.5 kg birth weight group (n=3,117,731) (Table 4), it was 4.76 times higher for the unmarried group. Across all three groups (overall study population group, ≥37 weeks group, ≥2.5 kg group), the crude ORs indicated that infant mortality within the 1st year was significantly higher for infants born to unmarried families than for those born to married families. Moreover, our findings revealed that when infants were born at or above the normal gestational age (≥37 weeks; crude OR, 4.62) and birth weight (≥2.5 kg; crude OR, 4.76)— thus excluding intrinsic physiological risk factors associated with prematurity or low birth weight that contribute to mortality—the mortality disparity within the 1st year between infants born to married and unmarried families became even more pronounced compared to the overall study population group (crude OR, 3.34). This suggests that the even greater disparity in infant mortality observed among those born at normal gestational age (≥37 weeks; crude OR, 4.62) and with normal birth weight (≥2.5 kg; crude OR, 4.76), compared to the overall study population (crude OR, 3.34), is more likely attributable to social determinants, as infant’s intrinsic physiological risk factors have been minimized in these subgroups (≥37 weeks group, ≥2.5 kg group).
In the overall study population group (n=3,298,263) (Table 2), ≥37 weeks group (n=3,082,810) (Table 3), and ≥2.5 kg group (n=3,117,731) (Table 4), stepwise adjustments for independent variables and potential confounders—adjusted sequentially in the following order: maternal age, education, employment, infant gestational age, and birth weight—resulted in a gradual decrease in the aORs (aOR1→aOR2→ aOR3→aOR4; P<0.05). This finding suggests that all five independent variables significantly influenced (P<0.05) infant mortality in the 1st year. The aOR4, corresponding to model 4, which reflects the adjustments made for all five independent variables, was 1.40 (95% CI, 1.03–1.92) for the overall study population group (Table 2), 2.08 (95% CI, 1.45–2.99) for the ≥37 weeks group (Table 3), and 2.13 (95% CI, 1.48–3.07) for the ≥2.5 kg group (Table 4). All of these aOR4 values remained greater than 1, indicating that even after adjusting for the five independent variables associated with infant mortality in this study, infants from single-parent households continued to have a higher mortality risk within the 1st year than those from two-parent households. This result aligns with those of previous studies conducted in other countries, which also found higher mortality rates among infants from single-parent households [5-8].Based on previous research that has highlighted the influence of various social determinants on infant mortality [7,8,11-19],it is evident that the elevated risk of mortality within the 1st year for infants born to single-parent families—despite adjusting for the five independent variables employed in this study—can be attributed to other social determinants not captured in this study, such as social policies and capital, social cohesion, welfare assistance, living and working conditions, environmental factors, neighborhood socioeconomic status, maternal psychosocial factors, health behaviors, and access to prenatal care, etc. In conclusion, the results indicate that all aOR4 values for the three groups (overall study population group, ≥37 weeks group, ≥2.5 kg group) exceed 1, underscoring the critical need for sustained social attention and support for single-parent families, along with the implementation of appropriate social welfare policies to mitigate disparities in infant mortality among those born to unmarried parents.
In both the preterm (<37 weeks) (Table 3) and low-birth-weight (<2.5 kg) groups (Table 4), we found no significant differences (P>0.05) in 1-year mortality rates between infants born to married and those born to unmarried families, as indicated by the crude OR and aORs (ORs1‒4; P>0.05). This indicates that when infants are born earlier or with lower birth weight, the disparity in 1-year mortality between those born to married and unmarried families becomes less pronounced. Furthermore, this finding suggests that among preterm and low-birth-weight infants, physiological vulnerabilities outweigh the influence of social determinants on 1-year mortality. This finding is further supported by the results in Supplement 1, which show that the difference in infant mortality due to gestational age and birth weight is much larger than the difference caused by socio-determinants (such as parental age, education, and employment status). Additionally, the finding that social factors have less influence on the differences in infant mortality rates between married and unmarried families when infants are born preterm can likely be attributed to South Korea’s welfare policies. Since 2000, the South Korean government has implemented a range of policies to support preterm infants, including financial assistance for medical expenses. These policies include the Medical Expense Support Policy for Preterm Infants (2000), the Program for the Establishment and Operation of Regional Neonatal Intensive Care Centers (2008), local government medical support programs for preterm infants (2009), the Preterm Infant Registration and Management Policy (2010), the Program for the Establishment and Operation of Integrated Care Centers for High-Risk Mothers and Newborns (2014), and the Medical Expense Support Program for High-Risk Pregnant Women (2015) [29]. Given these national efforts, medical expense support and management for families with preterm infants have improved, creating a situation in which physiological factors have a greater impact on the mortality of preterm infants than social determinants. Furthermore, according to a study published in 2006, South Korea had a well-developed neonatal medical care system, with survival rates of 65%–83% for extremely-low-birth-weight infants (<1 kg) and 80%–92% for very-low-birth-weight infants (<1.5 kg) [30]. Taken together, South Korea’s comprehensive welfare policies for preterm infants have enabled even unmarried families to access high-quality neonatal care at reduced costs. Consequently, infant physiological vulnerabilities have become more influential than social determinants in determining mortality outcomes among preterm infants. This result further confirms the important role of government policies and social support in reducing the gap in infant mortality rates between single- and two-parent households.
Additionally, in the overall study population (n=3,298,263), infant mortality was compared across the detailed categories of each independent variable (parental age, parental education level, parental employment status, infant gestational age, and infant birth weight) (Supplement 1). In terms of parental age, infant mortality was highest when the parents were under 19 years of age, had lower education levels, and when the mother was unemployed. The differences in infant mortality due to variations in infant gestational age and birth weight were far greater than those due to social factors (parental age, parental education level, and parental employment status). In particular, when infants were born within 28 weeks of gestation or weighed <1 kg, the OR for mortality was >230 times higher than that of infants born at full term or with a normal birth weight.
Strengths of this study
This study is significant because it represents a retrospective, large-scale, cohort study, which follows a clear chronological order and includes nearly all live births in South Korea from 2010– 2017 (initial raw data: 3,500,314; study population: 3,298,263). It is also the first study in South Korea to compare the 1-year mortality rates of infants born to married and unmarried families.
Limitations of this study
This study excluded 162 cases of “ambiguous matching between birth and death” from the initial raw data. These cases typically involved infants who were deceased without a registered resident identification number, suggesting a higher likelihood that they were born to unmarried families. Therefore, the actual difference in infant mortality between infants born to married and those born to unmarried families may be more significant than that reported in this study.
Additionally, according to Statistics Korea, parental marriage registration information is collected and linked to the child’s data only once: a few months after birth. Since some parents may have registered their marriage after their marriage registration information was incorporated into the child’s data, there is a limitation that the actual number of married and unmarried families in the sample may differ.
Areas for future research
While this study identified a clear difference in infant mortality between married and unmarried families, further research is needed to explore other variables not addressed in our study that contribute to this mortality difference and the extent of their impact. Such research could guide the development of cost-effective and efficient policies for unmarried families by prioritizing variables that significantly impact mortality differences.
Additionally, future studies should analyze and categorize the causes of infant deaths to compare maternal and child health environments between married and unmarried families, providing insights into the role of these environments in infant mortality.
Conclusion
Children born to unmarried families are intrinsically at a higher risk of being born with physiological vulnerabilities, such as a low birth weight and preterm birth, than those born to married families; they also exhibit higher rates of infant mortality. The elevated mortality rates among infants from unmarried families can be attributed not only to physiological factors related to the fetus but also to various social determinants that contribute to these outcomes.
This study examined nearly all births in South Korea from 2010–2017 and found that infants born to unmarried families had a higher mortality rate within the 1st year of life than those born to married families. Even after adjusting for various factors, including maternal age, education, employment, infant gestational age, and birth weight, the 1-year infant mortality rate remained higher along infants from unmarried families, indicating that other social determinants continue to influence these outcomes and require further investigation.
These findings underscore the need for sustained social attention and support for unmarried families, as well as the implementation of relevant social welfare policies. Additionally, there is a pressing need to address and improve societal perceptions of unmarried families to combat discrimination and alleviate the stigma they face, thus making it easier for them to seek social assistance.
Notes
Conflict of interest
No potential conflict of interest relevant to this article was reported.
Funding
None.
Data availability
Contact the corresponding author for data availability.
Author contribution
Conceptualization: SJ, SMP. Data curation: SJ, YC, SMP. Formal analysis: SJ, YC, HK, SMP. Investigation: SJ, SMP. Methodology: SJ, SMP. Software: SJ, YC, HK. Validation: SJ, SMP. Visualization: SJ, YC. Project administration: SJ, SMP. Writing–original draft: SJ. Writing–review & editing: SJ, SMP. Final approval of the manuscript: all authors.
Comparison of infant mortality within the 1st year after birth between married and unmarried families
Variable
Overall study population (n=3,298,263)
P-value
Married family
Unmarried single family
Study population
3,288,618
9,645
No. of infant deaths within the 1st year after birth
5,649
55
Crude death rate (/1,000 births)
1.7
5.7
Crude
1 (Reference)
3.34 (2.56–4.36)
<0.001
Adjusted
Model 1
1 (Reference)
2.23 (1.66–2.99)
<0.001
Model 2
1 (Reference)
1.80 (1.34–2.41)
<0.001
Model 3
1 (Reference)
1.79 (1.34–2.40)
<0.001
Model 4
1 (Reference)
1.40 (1.03–1.92)
0.033
Values are presented as number or odds ratio (95% confidence interval) unless otherwise indicated. P-values are calculated using the chi-square test. Model 1: adjusted for maternal age; Model 2: adjusted for maternal age and education; Model 3: adjusted for maternal age, education, and employment; Model 4: adjusted for maternal age, education, employment, infant’s gestational age, and birth weight.
Table 3.
Comparison of infant mortality within the 1st year after birth between married and unmarried families, stratified by infant’s gestational age
Variable
Gestational age
<37 wk group (n=215,453)
≥37 wk group (n=3,082,810)
Married family
Unmarried single family
P-value
Married family
Unmarried single family
P-value
No. of study population
214,556
897
3,074,062
8,748
No. of infant deaths within the 1st year after birth
2,815
18
2,824
37
Crude death rate (/1,000 births)
13.1
20.0
0.9
4.2
Crude
1 (Reference)
1.54 (0.96–2.46)
0.071
1 (Reference)
4.62 (3.34–6.39)
<0.001
Adjusted
Model 1
1 (Reference)
0.65 (0.38–1.08)
0.096
1 (Reference)
2.78 (1.92–4.02)
<0.001
Model 2
1 (Reference)
0.65 (0.38–1.11)
0.114
1 (Reference)
2.15 (1.49–3.09)
<0.001
Model 3
1 (Reference)
0.70 (0.41–1.20)
0.197
1 (Reference)
2.14 (1.49–3.08)
<0.001
Model 4
1 (Reference)
0.82 (0.48–1.41)
0.470
1 (Reference)
2.08 (1.45–2.99)
0.033
Values are presented as number or odds ratio (95% confidence interval) unless otherwise indicated. P-values are calculated using the chi-square test. Model 1: adjusted for maternal age and infant’s gestational age; Model 2: adjusted for maternal age, education, and infant’s gestational age; Model 3: adjusted for maternal age, education, employment, and infant’s gestational age; Model 4: adjusted for maternal age, education, employment, infant’s gestational age, and birth weight.
Table 4.
Comparison of infant mortality within the 1st year after birth between married and unmarried families, stratified by infant’s birth weight
Variable
Birth weight
<2.5 kg group (n=180,532)
≥2.5 kg group (n=3,117,731)
Married family
Unmarried single family
P-value
Married family
Unmarried single family
P-value
No. of study population
179,757
775
3,108,861
8,870
No. of infant deaths within the 1st year after birth
2,833
17
2,806
38
Crude death rate (/1,000 births)
15.8
21.9
0.9
4.3
Crude
1 (Reference)
1.40 (0.86–2.27)
0.171
1 (Reference)
4.76 (3.46–6.56)
<0.001
Adjusted
Model 1
1 (Reference)
0.73 (0.43–1.24)
0.240
1 (Reference)
2.83 (1.96–4.10)
<0.001
Model 2
1 (Reference)
0.71 (0.42–1.20)
0.202
1 (Reference)
2.21 (1.54–3.18)
<0.001
Model 3
1 (Reference)
0.79 (0.46–1.35)
0.385
1 (Reference)
2.20 (1.53–3.17)
<0.001
Model 4
1 (Reference)
0.79 (0.46–1.36)
0.396
1 (Reference)
2.13 (1.48–3.07)
<0.001
Values are presented as number or odds ratio (95% confidence interval) unless otherwise indicated. P-values are calculated using the chi-square test. Model 1: adjusted for maternal age and infant’s birth weight; Model 2: adjusted for maternal age, education, and infant’s birth weight; Model 3: adjusted for maternal age, education, employment, and infant’s birth weight; Model 4: adjusted for maternal age, education, employment, infant’s gestational age, and birth weight.
References
1. Kennes J, Knowles J. Unmarried births: accounting and equilibrium analysis, 1960-1995. Rev Econ Dyn 2024;52:84-109.
2. Gonzalez L. The determinants of the prevalence of single mothers: a cross-country analysis [Internet]. IZA Institute of Labor Economics 2025;[cited 2025 Mar 11]. Available from: https://doi.org/10.2139/ssrn.769844
3. Organisation for Economic Co-operation and Development. Doing better for families [Internet]. OECD Publishing 2011 [cited 2025 Mar 11]. Available from: https://doi.org/10.1787/9789264098732-en
4. National Statistical Development Institute. Korean Social Trends: Major Changes and Characteristics of Extramarital Birth Behaviors [Internet]. National Statistical Development Institute 2020 [cited 2025 Mar 11]. Available from: https://sri.kostat.go.kr/boardDownload.es?bid=12302&list_no=386936&seq=3
5. Ahmed F. Unmarried mothers as a high-risk group for adverse pregnancy outcomes. J Community Health 1990;15:35-44.
6. Balayla J, Azoulay L, Abenhaim HA. Maternal marital status and the risk of stillbirth and infant death: a population-based cohort study on 40 million births in the United States. Womens Health Issues 2011;21:361-5.
8. Srinivasa DK, Pankajam R. A study of out-of-wedlock pregnancies: outcome and utilization of maternal/child health services. Int J Gynaecol Obstet 1980;18:368-71.
9. Choi Y, Lee G, Song IG, Kim OJ, Park SM. Prevalence of adverse birth outcomes and disparity of unmarried women in South Korea: a systematic review. J Glob Health Rep 2018;2:e2018008.
11. Kim D, Saada A. The social determinants of infant mortality and birth outcomes in Western developed nations: a cross-country systematic review. Int J Environ Res Public Health 2013;10:2296-335.
12. Passarelli-Araujo H. Estimating the effect of socio-economic factors on infant mortality rates in Latin America between 2000 and 2019: a panel data analysis. Public Health 2024;227:232-8.
14. Austin MP, Leader L. Maternal stress and obstetric and infant outcomes: epidemiological findings and neuroendocrine mechanisms. Aust N Z J Obstet Gynaecol 2000;40:331-7.
15. Yun JW, Kim YJ, Son M. Regional deprivation index and socioeconomic inequalities related to infant deaths in Korea. J Korean Med Sci 2016;31:568-78.
16. Guildea ZE, Fone DL, Dunstan FD, Sibert JR, Cartlidge PH. Social deprivation and the causes of stillbirth and infant mortality. Arch Dis Child 2001;84:307-10.
18. Van Malderen C, Amouzou A, Barros AJ, Masquelier B, Van Oyen H, Speybroeck N. Socioeconomic factors contributing to under-five mortality in sub-Saharan Africa: a decomposition analysis. BMC Public Health 2019;19:760.
20. Lean SC, Derricott H, Jones RL, Heazell AE. Advanced maternal age and adverse pregnancy outcomes: a systematic review and meta-analysis. PLoS One 2017;12:e0186287.
21. Mukosha M, Jacobs C, Kaonga P, Musonda P, Vwalika B, Lubeya MK, et al. Determinants and outcomes of low birth weight among newborns at a tertiary hospital in Zambia: a retrospective cohort study. Ann Afr Med 2023;22:271-8.
22. Committee to Study the Prevention of Low Birthweight. The significance of low birthweight. In: Committee to Study the Prevention of Low Birthweight; Division of Health Promotion and Disease Prevention; Institute of Medicine, editors. Preventing low birth-weight [Internet]. National Academies Press (US); 1985 [cited 2025 Mar 11]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK214473/
23. Magnusson A, Laivuori H, Loft A, Oldereid NB, Pinborg A, Petzold M, et al. The association between high birth weight and long-term outcomes-implications for assisted reproductive technologies: a systematic review and meta-analysis. Front Pediatr 2021;9:675775.
24. Cartlidge PH, Stewart JH. Survival of very low birthweight and very preterm infants in a geographically defined population. Acta Paediatr 1997;86:105-10.
26. Kim H. Social exclusion and discrimination against unwed mothers. Gend Cult 2013;6:7-41.
27. Kim H, Kwon JH, Choi HS. A case study on discrimination experienced by unmarried mother. Korean J Fam Soc Work 2012;36:121-55.
28. Sung J, Kim H, Lee M, Park Y. A study of the experiences of unwed mothers in interaction with public service professionals: focusing on the experiences during pregnancy, birth and child caring. J Korea Contents Assoc 2016;16:404-18.
One-year mortality disparities between infants of unmarried and married families in South Korea: a large scale retrospective cohort study
Figure. 1. Study population derivation process.
Graphical abstract
Figure. 1.
Graphical abstract
One-year mortality disparities between infants of unmarried and married families in South Korea: a large scale retrospective cohort study
Characteristic
Married
Unmarried
P-value
Infant sex
0.353
Male
1,689,611
5,001
Female
1,599,007
4,644
Place of birth
<0.001
Hospital
3,249,584
9,340
Others
39,034
305
Maternal age (y)
<0.001
Mean
32.07
29.95
<19
3,836
1,282
19–34
2,567,322
5,334
≥35
717,427
2,939
Unknown
33
90
Maternal education
<0.001
College or higher
2,464,683
3,027
High school
783,993
4,687
Middle school or below
28,848
1,491
Unknown
11,094
440
Maternal employment
<0.001
Employed
1,207,834
1,862
Unemployed
2,049,632
7,088
Unknown
31,152
695
Paternal age (y)
<0.001
Mean
34.30
33.09
<19
1,495
607
19–34
1,931,141
3,711
≥35
1,355,781
3,653
Unknown
201
1,674
Paternal education
<0.001
≥College
2,452,839
2,724
High school
788,292
3,925
≤Middle school
37,912
1,087
Unknown
9,575
1,909
Paternal employment
<0.001
Employed
3,092,516
5,590
Unemployeda)
26,118
497
Unknown
169,984
3,558
Gestational age (wk)
<0.001
<28
7,846
102
28≤ & <32
16,669
87
32≤ & <37
190,041
708
37≤ & <42
3,067,905
8,711
≥42
6,157
37
Birth weight (kg)
<0.001
<1
7,554
79
1≤ & <1.5
12,795
60
1.5≤ & <2.5
159,408
636
2.5≤ & <4
3,000,315
8,564
≥4
108,546
306
Survival status after birth (y)
Death within 1
5,639
55
<0.001
Death in 1< & ≤2
790
4
0.266
Death in 2< & ≤3
485
1
0.727
Death in 3< & ≤4
345
2
0.274
Death in 4< & ≤5
328
3
0.038
Death within 5
7,537
65
<0.001
Survival after 5
3,281,081
9,580
Variable
Overall study population (n=3,298,263)
P-value
Married family
Unmarried single family
Study population
3,288,618
9,645
No. of infant deaths within the 1st year after birth
5,649
55
Crude death rate (/1,000 births)
1.7
5.7
Crude
1 (Reference)
3.34 (2.56–4.36)
<0.001
Adjusted
Model 1
1 (Reference)
2.23 (1.66–2.99)
<0.001
Model 2
1 (Reference)
1.80 (1.34–2.41)
<0.001
Model 3
1 (Reference)
1.79 (1.34–2.40)
<0.001
Model 4
1 (Reference)
1.40 (1.03–1.92)
0.033
Variable
Gestational age
<37 wk group (n=215,453)
≥37 wk group (n=3,082,810)
Married family
Unmarried single family
P-value
Married family
Unmarried single family
P-value
No. of study population
214,556
897
3,074,062
8,748
No. of infant deaths within the 1st year after birth
2,815
18
2,824
37
Crude death rate (/1,000 births)
13.1
20.0
0.9
4.2
Crude
1 (Reference)
1.54 (0.96–2.46)
0.071
1 (Reference)
4.62 (3.34–6.39)
<0.001
Adjusted
Model 1
1 (Reference)
0.65 (0.38–1.08)
0.096
1 (Reference)
2.78 (1.92–4.02)
<0.001
Model 2
1 (Reference)
0.65 (0.38–1.11)
0.114
1 (Reference)
2.15 (1.49–3.09)
<0.001
Model 3
1 (Reference)
0.70 (0.41–1.20)
0.197
1 (Reference)
2.14 (1.49–3.08)
<0.001
Model 4
1 (Reference)
0.82 (0.48–1.41)
0.470
1 (Reference)
2.08 (1.45–2.99)
0.033
Variable
Birth weight
<2.5 kg group (n=180,532)
≥2.5 kg group (n=3,117,731)
Married family
Unmarried single family
P-value
Married family
Unmarried single family
P-value
No. of study population
179,757
775
3,108,861
8,870
No. of infant deaths within the 1st year after birth
2,833
17
2,806
38
Crude death rate (/1,000 births)
15.8
21.9
0.9
4.3
Crude
1 (Reference)
1.40 (0.86–2.27)
0.171
1 (Reference)
4.76 (3.46–6.56)
<0.001
Adjusted
Model 1
1 (Reference)
0.73 (0.43–1.24)
0.240
1 (Reference)
2.83 (1.96–4.10)
<0.001
Model 2
1 (Reference)
0.71 (0.42–1.20)
0.202
1 (Reference)
2.21 (1.54–3.18)
<0.001
Model 3
1 (Reference)
0.79 (0.46–1.35)
0.385
1 (Reference)
2.20 (1.53–3.17)
<0.001
Model 4
1 (Reference)
0.79 (0.46–1.36)
0.396
1 (Reference)
2.13 (1.48–3.07)
<0.001
Table 1. Socio-demographic and birth-death related characteristics of the study population (n=3,298,263)
Unemployed, housewife, or students.
Table 2. Comparison of infant mortality within the 1st year after birth between married and unmarried families
Values are presented as number or odds ratio (95% confidence interval) unless otherwise indicated. P-values are calculated using the chi-square test. Model 1: adjusted for maternal age; Model 2: adjusted for maternal age and education; Model 3: adjusted for maternal age, education, and employment; Model 4: adjusted for maternal age, education, employment, infant’s gestational age, and birth weight.
Table 3. Comparison of infant mortality within the 1st year after birth between married and unmarried families, stratified by infant’s gestational age
Values are presented as number or odds ratio (95% confidence interval) unless otherwise indicated. P-values are calculated using the chi-square test. Model 1: adjusted for maternal age and infant’s gestational age; Model 2: adjusted for maternal age, education, and infant’s gestational age; Model 3: adjusted for maternal age, education, employment, and infant’s gestational age; Model 4: adjusted for maternal age, education, employment, infant’s gestational age, and birth weight.
Table 4. Comparison of infant mortality within the 1st year after birth between married and unmarried families, stratified by infant’s birth weight
Values are presented as number or odds ratio (95% confidence interval) unless otherwise indicated. P-values are calculated using the chi-square test. Model 1: adjusted for maternal age and infant’s birth weight; Model 2: adjusted for maternal age, education, and infant’s birth weight; Model 3: adjusted for maternal age, education, employment, and infant’s birth weight; Model 4: adjusted for maternal age, education, employment, infant’s gestational age, and birth weight.
