Abstract

Background

The Infant Mortality Rate assesses the deaths occurring in children under one year of age and serves as an important health indicator in mapping the living conditions of a given society. The aim of the present study is to evaluate the epidemiological trends of infant mortality in Brazil from 2000 to 2020, based on the most prevalent chapters of the International Classification of Diseases.

Methods

This is an ecological and retrospective study conducted using secondary data collection. The country of origin for the study is Brazil. The dependent variable of the study is infant mortality, specifically referring to children under one year of age. The independent variables include sex, color, and birth weight. The period analyzed spans from 2000 to 2020. The statistical processing was conducted using Joinpoint and GeoDa.

Results

In the analyzed period, the most prevalent causes of infant mortality, classified according to the ICD, were: conditions originating in the perinatal period, congenital malformations, infectious and parasitic diseases, respiratory diseases, and external causes of morbidity and mortality. It is noteworthy that the majority of the causes of death in children under one year of age exemplified in this study exhibited a declining trend, except for congenital malformations, deformities and chromosomal anomalies. In addition, external causes of mortality varied greatly throughout the historical series and, when analyzing more specifically the categories in this chapter, it was found that transport accidents are the main cause of death in children under one year of age. When observing the spatial distribution of infant mortality by RIAUs, it is noticeable that the Boa Vista region presented the highest infant mortality rate in the analyzed historical series, exceeding 3,795 deaths per 100,000 infants.

Conclusions

Brazil has made progress over the years in public policies aimed at reducing infant mortality in the country, particularly through the consolidation and intensification of Primary Health Care coverage, which provides health promotion, protection, and recovery by identifying the main health risks and needs of each locality. However, addressing the health determinants that influence infant mortality requires actions that go beyond the health sector.

Background

The reduction of infant mortality is part of the Sustainable Development Goals (SDGs) proposed by the United Nations (UN), which is a pact aimed at increasing life expectancy, achieving universal coverage, and ensuring access to quality healthcare. This initiative seeks to eliminate preventable deaths of newborns and children under five years of age, reducing neonatal mortality to 12 per 1,000 live births and under-five mortality to 15 per 1,000 live births before 2030 [1].

The Infant Mortality Rate (IMR) assesses the deaths occurring in children under one year of age and serves as an important health indicator in mapping the living conditions of a given society [2]. Therefore, it is treated as a priority by most countries, including Brazil, which has gained prominence in the international arena over the past few decades for successfully reducing the IMR significantly. In the year 2000, the rate was approximately 29.0 per thousand live births, improving to 17.22 in 2010 and 13.8 in 2015 [3].

However, despite the decrease in this indicator, the situation remains concerning when compared to developed countries, such as Japan, which has infant mortality rates approximately 3 to 6 times lower than those of Brazil [4]. Furthermore, there is significant heterogeneity across regions within Brazil; for instance, Distrito Federal, in the Midwest region, exhibited an infant mortality rate of 8.5 deaths per thousand live births in 2019, while the highest rate in the country was observed in Amap谩, located in the North region, with 22.9 deaths per thousand live births [5].

Territorial inequalities are strongly correlated with indicators of morbidity and infant mortality, as these are related to the quality of life of populations and involve multifactorial relationships. Among the factors influencing these indicators are complications associated with preterm birth, issues during delivery, inadequate vaccination, non-sterilized childbirth, substandard breastfeeding practices, stillbirth, and socioeconomic factors [6].

In this perspective, the study is justified by promoting the analysis of infant mortality in Brazil, contributing to the implementation of health strategies and actions aimed at effectively and efficiently reducing this indicator through the detailed examination of its causes and associated factors. Furthermore, the study focuses on the use of the ICD-10, which is a systematic and widely used tool for classifying diseases and health conditions, providing a precise framework for categorizing the causes of infant mortality, allowing for a detailed assessment of the different types of diseases, complications, and conditions that lead to death.

The aim of the present study is to evaluate the epidemiological trends and spatial analysis of infant mortality in Brazil from 2000 to 2020, based on the most prevalent chapters of the International Classification of Diseases (ICD-10).

Methods

This is a time-trend ecologic study conducted using secondary data collection. The country of origin for the study is Brazil, which is composed of five regions: North, Northeast, Southeast, South, and Central-West, and has a population of approximately 190听million inhabitants, according to the demographic census carried out by Brazilian Institute of Geography and Statistics (IBGE) in 2010. The study utilized data collected on October 17, 2024, from the Mortality Information System (SIM), managed by the Department of Informatics of the Unified Health System (DATASUS), which can be accessed through the following link: .

It is noteworthy that the data is entered into the electronic system through a notification logistics for deaths, carried out by healthcare professionals following the completion of the Death Certificate. This is a legally mandated document issued by a physician, which includes the cause of death according to the ICD, as well as other important information for the composition of the individual鈥檚 epidemiological profile. The data is sent by state and municipal health departments to DATASUS.

However, it was observed that many data on the individuals鈥� characteristics were not recorded by professionals when filling out the Death Certificate, called 鈥渋gnored鈥� in DATASUS, which may compromise the analysis of some independent variables.

The dependent variable of the study is infant mortality, specifically referring to children under one year of age. The independent variables include sex, color, and birth weight. The period analyzed spans from 2000 to 2020. This time frame was selected because it has complete and consistent data for all months, as the system requires sufficient time to be updated.

After data collection, adjustments were made to the database using Microsoft Excel^庐, where it was verified that there was no missing data and the database formatting was adjusted for statistical analysis. At this stage, all deaths in each chapter were also added, enabling the mapping of the five most prevalent chapters in the age group and period studied.

Subsequently, statistical processing was conducted using Joinpoint (version 4.9.0.0, Surveillance Research, National Cancer Institute, USA), an open-source and freely accessible software that employs linear regression techniques to adjust rates based on join points in the time series.

Joinpoint was chosen for being a well-established statistical software in the market and widely used for analyzing studies in the field of epidemiology and related areas, allowing for differentiated insights into changes over time. The methodologies used were the Segmentation Regression Model, which divides the time series into segments and fits a straight line (linear regression) for each segment, as well as Joinpoint Detection, a method that automatically detects the points where the trend of the variable changes, and Change Point Tests to assess the significance of trend changes.

The outputs are generated from the Annual Percent Change (APC) and the Average Annual Percent Change (AAPC), which indicate significant variations in the analyzed rates and present the data through epidemiological curves constructed from the rate (y-axis) and the study period (x-axis). The significance of the tests is demonstrated by the Monte Carlo Permutation model.

The rates were calculated by dividing the infant mortality by the number of live births and multiplying the result by 1,000 (1).

$${\rm{M=(infant\:mortality/live\:births)}} \times {\rm{1}}{\rm{.000}}$$

(1)

For better visualization of the data, when analyzing the trend of the most incident groups of mortality due to external causes, the coefficient (K) used was 100,000 live births (2).

$${{\rm{M}}_{\rm{g}}}{\rm{ = (infant\:mortality/live\:births)}} \times {\rm{100}}{\rm{.000}}$$

(2)

The groups are sets of diagnoses defined based on similarities among themselves, according to ICD-10. In Chapter XX, which addresses external causes of morbidity and mortality, the following diagnostic groups are included: transport accidents (V01-V99), falls (W00-W19), drowning and accidental submersions (W65-W74), exposure to smoke, fire, and flames (X00-X09), poisoning, intoxication by or exposure to harmful substances (X40-X49), self-inflicted injuries (X60-X84), assaults (X85-Y09), and all other external causes (W20-W64, W75-W99, X10-X39, X50-X59, Y10-Y8).

It is necessary to study the trend of external causes, because this public health quest is based on the investigation and control of external causes for two reasons: (1) because they are considered avoidable deaths and (2) because they are costly for the Unified Health System and the Social Security System, but both do not depend only on effective public policy, but also on educational and cultural changes.

The group on self-inflicted injuries was excluded from the analysis due to a lack of data. Similarly, the group of all other external causes was also excluded from the analysis because it did not specify the causes of death, thus avoiding bias in the study as it is a confounding factor.

For the geoprocessing of infant mortality by Brazilian states, the software GeoDa (version 1.22.0.4, Luc Anselin) was utilized, which offers various tools for spatial analysis. The data from this study were analyzed using the Natural Breaks technique, as the histogram presented a right-skewed distribution. The algorithm used in this type of analysis aims to minimize the variance within groups and maximize the differences between groups. Among the advantages, it is possible to highlight the mapping of values that are not evenly distributed across the distribution, resulting in more balanced classes and internal homogeneity within the clusters.

The spatial analysis was carried out based on the Intermediate Regions of Urban Articulation in Brazil (RIAUs), which are geographical boundaries drawn from neighboring municipalities that have sociodemographic similarities between them.

Considering that this is a study based on secondary data and public domain information, and since it does not use personal data, it was not necessary to submit it for review by the Ethics and Research Committee, as stipulated by Resolution No. 510/2016 [7].

Results

According to the results obtained in this study, as evidenced in Fig.听1, the most prevalent chapters of infant mortality during the analyzed period were, respectively: (1) Chapter XVI - certain conditions originating in the perinatal period; (2) Chapter XVII - congenital malformations, deformities, and chromosomal anomalies; (3) Chapter I - certain infectious and parasitic diseases; (4) Chapter X - diseases of the respiratory system; (5) Chapter XX - external causes of morbidity and mortality.

Fig. 1

Infant mortality in Brazil, between 2000 and 2020, according to the most prevalent chapters of the ICD-10. Brazil, 2024

It is noteworthy that the majority of the causes of death in children under one year of age exemplified in this study exhibited a declining trend, particularly Chapter XVI (Fig.听1a), which refers to certain conditions originating in the perinatal period. These conditions remain the leading cause of death in childhood, but there was a reduction of 4.5 deaths per 1,000 live births when comparing the years 2000 and 2020, especially during the period from 2003 to 2011, with an APC of -3.26.

Chapter XVII (Fig.听1b) addresses congenital malformations, deformities, and chromosomal anomalies. This chapter shows a gradual increasing trend, particularly from 2011 onwards, reaching a peak in 2019 with approximately 3 deaths per 1,000 live births. The period that demonstrated the highest growth trend was between 2000 and 2004, with an APC of 2.39.

Infectious and parasitic diseases (Fig.听1c) exhibited a downward trend, with significant reductions in mortality, particularly between 2005 and 2009, showing an APC of -12.47, and between 2009 and 2018, with an APC of -4.53. However, in the last three years of the analysis (2018鈥�2020), an increase was observed, with an APC of 1.79. Conversely, mortality due to respiratory diseases (Fig.听1d) also demonstrated a gradual reduction each year, culminating in a significant decline in 2020, with an APC of -19.30.

In contrast to the homogeneous and linear trends presented by the previous causes of death, Chapter XX, which addresses external causes of mortality (Fig.听1e), exhibited significant annual variations in the mortality rate. Through linear regression, it is evident that between 2010 and 2013 there was a growth trend, with an APC of 2.97. Conversely, the period from 2018 to 2020 shows the most significant reduction in mortality, with an APC of -4.33.

Table听1 presents the statistical analysis of the linear regression conducted in Joinpoint regarding infant mortality according to the chapters of the ICD-10. The analysis showed that the AAPCs of all chapters were statistically significant, with a p-value鈥�<鈥�0.05. Furthermore, several APCs also demonstrated statistical significance in line segments.

Table 1 Joinpoint analysis of infant mortality rate in Brazil according to the chapter of ICD-10, 2000 to 2020. Brazil, 2024

Table听2 presents the relative frequency (%) of the profile of children who died between 2000 and 2020 in Brazil, according to the five most prevalent chapters of infant mortality. It is noteworthy that male children were more frequently represented across all causes of death, with a more significant percentage difference concerning perinatal, respiratory, and external causes. Furthermore, Black and Brown children exhibited higher mortality rates due to perinatal, infectious, and respiratory diseases. In contrast, White children had higher mortality rates associated with chromosomal disorders and external causes.

Table 2 Relative frequency of characteristics of infant mortality in Brazil according to the chapter of ICD-10, 2000 to 2020. Brazil, 2024

Furthermore, nearly 30% of the children who died from perinatal diseases were born with extremely low birth weight, between 1.10 and 2.20 lbs, and more than 42% of the children who died from chromosomal disorders were born below the ideal weight. In the other chapters, the majority of children had an ideal birth weight, above 5.51 lbs. However, birth weight was disregarded in many cases, complicating the analysis of this characteristic.

Regarding specifically deaths from external causes, which rank among the five most prevalent chapters of infant mortality, Fig.听2 presents the fragmentation of these deaths by groups in the population per 100,000 live births. In this perspective, it is possible to observe that transportation accidents are the leading cause of death for children under one year of age (Fig.听2a). The data indicate a trend of increase until 2011, followed by a regression with an AAPC of -6.76.

Fig. 2

Trend analysis of the mortality rate due to external causes by ICD-10 group, between 2000 and 2020, in Brazil. Brazil, 2024

Regarding assaults, the second leading cause of infant mortality from external causes, there was an increase, particularly between 2010 and 2013, with an AAPC of 17.49. However, it has shown a declining trend in recent years, with an AAPC of -6.15 (Fig.听2b). The groups of falls (Fig.听2c), drownings and accidental submersions (Fig.听2d), and exposure to smoke, fire, and flames (Fig.听2e) all demonstrate a trend of reduction in the historical series analyzed, with particular emphasis on falls, which have exponentially decreased the number of deaths.

Regarding the group of poisoning and intoxication due to exposure to harmful substances, there has been a progressive reduction in cases; however, starting in 2018, a trend of increase has been observed. It is noteworthy that this group has a low incidence of mortality compared to other groups of external causes of death.

Table听3 shows the Joinpoint analysis of mortality rate due to external causes by ICD-10 group. It is evident that there is no recurring pattern across all chapters, highlighting that, although the trends may appear similar, the behaviors differ in terms of periods of decline, increase, or stability in mortality curves.

Table 3 Joinpoint analysis of mortality rate due to external causes by ICD-10 group, 2000 to 2020. Brazil, 2024

When observing the spatial distribution of infant mortality by RIAUs, it is noticeable that the Boa Vista region presented the highest infant mortality rate in the analyzed historical series, exceeding 3,795 deaths per 100,000 infants (Fig.听3). Following this, clusters of RIAUs emerge in the South, Southeast, and Midwest regions.

Fig. 3

Spatial analysis of the mortality rate from external causes, between 2000 and 2020, in Intermediate Regions of Urban Articulation (RIAUs). Brazil, 2024

Discussion

Monitoring infant mortality allows for tracking changes over time that may be influenced by health, social, economic, and organizational factors, such as issues related to gestation, childbirth, and the postpartum period, lack of access to healthcare services, inadequate sanitation, low education levels, and income [2]. Another factor identified is early weaning and the introduction of complementary foods before six months of age [8]. In the United States, although a 22% reduction in infant mortality was observed from 2002 to 2021, there was a 3% increase from 2021 to 2022 [9].

Data from the 鈥淣ascer no Brasil鈥� survey, a national inquiry on childbirth assistance conducted in 2011 and 2012 across 266 maternity hospitals, concluded that the majority of deaths occurring during the perinatal period were associated with prematurity, low birth weight, maternal risk factors, congenital malformations, and perinatal asphyxia. Furthermore, it was observed that these causes of death primarily arise from failures in the care provided during prenatal, childbirth, and neonatal periods [10], as observed in Fig.听1a, conditions originating in the perinatal period stand out as the leading cause of infant mortality.

Regarding mortality related to congenital malformations, deformities, and chromosomal anomalies (Fig.听1b), there has been a growing increase in deaths of this nature, establishing it as a significant public health issue globally, and in Brazil, it ranks among the leading groups of causes of infant mortality [11]. However, the detection of congenital anomalies during prenatal care has become more common with advancements in technology and increased accuracy of detection. This may be related to the fact that children identified early may have more severe and easily detectable anomalies through available examinations, particularly when surgical intervention is involved [12].

Even with the reduction of mortality from infectious, parasitic, and respiratory diseases over the analyzed period (Fig.听1c and d), this indicator remains present in society and reveals significant challenges. It is essential to adopt measures to reduce the incidence of deaths in this age group [13]. In African countries, respiratory infections and severe sepsis are common causes of respiratory failure and mortality in children under five years of age [14].

In the global context, a study conducted in the United States in 2017 revealed that the main causes of infant mortality in the country were congenital malformations, low birth weight, sudden infant death syndrome, maternal complications, and unintentional injuries. These findings differ in part from the Brazilian reality, which has yet to address issues associated with socioeconomic and sanitary factors, such as infectious, parasitic, and respiratory diseases [15].

Related to deaths from external causes, there has been an increase, particularly between 2007 and 2016 (Fig.听1e), highlighting the need for social concern and mobilization. These incidents have repercussions on families and the social environment, penalizing children in their growth and development phases. The occurrence is associated with immaturity, curiosity, growth, and development, necessitating increased vigilance [16]. A study that gathered data from 2007 to 2017 also observed the same trend, with external causes being responsible for a growing increase in infant mortality [17].

External cause mortality presents itself as a significant public health issue worldwide, affecting all age groups and genders, with an unequal distribution within the population. There are few studies dedicated to external cause deaths in children and adolescents, despite these being preventable deaths in a vulnerable population, whose right to life is guaranteed by law and is the responsibility of society as a whole [18].

In children under 1 year of age, according to the grouped causes in Chapter XX, external causes of morbidity and mortality, the most frequent underlying cause of death was the group of other respiratory-related risks, which include inhalation of gastric contents, inhalation and ingestion of food, and objects causing obstruction of the respiratory tract. In the age group of 1 to 4 years, the most affected are those involved in transportation accidents [16].

Regarding factors linked to the reduction of mortality in Brazil, it is important to highlight the initiatives within the Unified Health System (SUS) aimed at addressing this issue, such as family planning, prenatal care, vaccination coverage, assistance during childbirth and delivery, guidance on breastfeeding, neonatal consultations in the first week of life, and child health monitoring. However, intersectoral actions are still necessary for public policies aimed at reducing health inequities [19].

The improvement of socioeconomic factors, income levels, and care during prenatal, childbirth, and postnatal periods is essential and has a significant impact on reducing infant mortality, particularly among males. Studies indicate that male infants face greater challenges in survival during the first year of life, which corroborates the findings of this study (Table听2), where males lead in mortality across all studied causes of infant death. The reasons for this difference are related to biological factors [20].

In this regard, data released by the IBGE demonstrated that, in 2015, in Brazil, the probability of male infants not surviving until one year of age was 14.9 per thousand live births, while for female infants, this probability was 12.7. When comparing this with the perspective of a developing country like Brazil, a study conducted in India between 2015 and 2016 found that infant mortality disproportionately affects male children and is directly associated with poverty and illiteracy, as well as low birth weight and maternal age below 18 years [21].

When observing an underdeveloped country like Nigeria, a high infant mortality rate is evident. A study conducted in 2003 showed that 101 out of 1,000 live births do not survive to their first year, with the highest mortality rate among boys. Furthermore, the study highlighted that infant mortality is associated with low maternal education levels, lack of assistance during childbirth, poverty, inadequate sanitation, lack of access to clean water, multiple pregnancies, and advanced maternal age [22].

Regarding birth weight data, this study demonstrated that children with low birth weight have a higher prevalence of infant mortality (Table听2), which is strongly related to prematurity or intrauterine growth restriction [23]. This factor is largely associated with low levels of socioeconomic development, poor quality of maternal and child care, and maternal health complications [24].

Regarding ethnicity, this study identified the relative frequency of mortality, which showed a small percentage of deaths among Indigenous children (Table听2). However, since this population is smaller than others, the proportion becomes significant. In this perspective, a study conducted in Brazil between 2007 and 2017 demonstrated that the highest neonatal mortality during the studied period was among Indigenous individuals (13.97/1,000 live births), followed by Whites (9.42/1,000), Browns (8.41/1,000), Blacks (7.24/1,000), and Yellows (5.34/1,000) [17].

It is also noted that although there is a reduction in the IMR and increasing investment in improvements in demographic, socioeconomic, and health factors in the Brazilian population, significant differences among Brazilian regions persist. Therefore, the implementation of intersectoral policies is necessary to achieve better results and the targets set by the UN by 2030 [25].

Regarding this, it is possible to observe marked disparities in the spatial analysis of the infant mortality rate due to external causes, with the formation of clusters mainly in the South, Southeast, and Central-West regions (Fig.听3). A study found that, in recent decades, regions with higher economic development showed more significant numbers of mortality due to external causes and chronic diseases, which may be associated with the process of urbanization and lifestyle [26].

However, it is further emphasized that the RIAU Boa Vista, located in the state of Roraima in the North region, holds the highest infant mortality rate due to external causes in Brazil, according to the results of this study (Fig.听3). Furthermore, it is known that deaths from preventable causes are strongly associated with socioeconomic factors such as income, environment, food security, basic sanitation, and healthcare coverage [27,28,29].

In this context, the literature highlights that Roraima has experienced significant population growth in recent decades. A rapid demographic expansion has made the state, since 1980, one of the states with the highest population growth rates, resulting in a reconfiguration of the urban space [30]. This scenario aligns with the rapid urbanization process, where the planning of actions is focused on the short term and is still incipient in addressing the social demands of the population [31].

Another important factor, as mentioned, is the coverage of healthcare services, with Primary Health Care being of great relevance in caring for infants. In this regard, Roraima stands out for having high coverage rates of the Family Health Strategy, with only four municipalities not reaching 100% coverage [32]. However, according to the results found, Primary Health Care coverage did not lead to a significant reduction in deaths due to external causes, highlighting the need for investment in improving the quality of care [33].

It should also be considered that Roraima has a large Indigenous population, and studies indicate that Indigenous children have a 60% higher risk of dying before the age of one compared to non-Indigenous children [34].

The limitations of the study were primarily due to the lack of recent data on infant mortality in Brazil, restricting the study to the year 2020, which may compromise the analysis of the current reality and cause delays in health assessment and subsequent decision-making. Furthermore, due to the manual entry of information, both from the Death Certificate and the data sent by municipal and state health departments, there may be human errors in this process. Additionally, the completion of certain fields, such as some independent variables like color and birth weight, is not mandatory on the Death Certificate. Such circumstances may be considered limitations of the study.

Conclusions

The results of this study revealed that the five most prevalent causes of death in children under one year of age are due to perinatal diseases, malformations, parasitic diseases, respiratory diseases, and external causes. These conditions show a reduction over the analyzed time series, with the exception of congenital malformations and external causes, which exhibit a growing trend. Furthermore, it was found that male children are more susceptible to infant mortality.

Brazil has made progress over the years in public policies aimed at reducing infant mortality in the country, particularly through the consolidation and intensification of Primary Health Care coverage, which provides health promotion, protection, and recovery by identifying the main health risks and needs of each locality. However, addressing the health determinants that influence infant mortality requires actions that go beyond the health sector, necessitating a set of intersectoral strategies aimed at promoting health and well-being.

Based on these results, it is important to highlight the need to strengthen comprehensive and preventive care in child health consultations within the Brazilian public health services, with a focus on continued education for parents and family members. Furthermore, the importance of investing in public policies to prevent deaths related to external causes is emphasized, especially those aimed at preventing traffic accidents and violence. Actions such as improving traffic conditions, road infrastructure, enhancing and enforcing traffic laws, child violence prevention programs, with a special focus on domestic violence, and improving urban safety can significantly contribute to reducing child mortality due to external causes in Brazil.