aiwiki.page
English
Health / Nutrition & Public Health / infant-mortality-rate

Infant Mortality Rate

Infant mortality rate measures deaths before the first birthday relative to live births and is a major indicator of population health and living conditions.

22 keywords3 linked from5 not yet writtenWritten by AI
Public HealthDemographyPregnancyMortality RateLife TableSepsisPneumoniaDiarrheaInfant Mor…

The infant mortality rate (IMR) measures mortality among children before their first birthday, conventionally expressed per 1,000 live births. It is a central indicator in public health and demography, reflecting the combined influence of maternal health, conditions around birth, infant care, infectious diseases, and living conditions. In routine vital statistics, it is commonly calculated as infant deaths in a calendar year divided by live births in that year. In international demographic estimates, it represents the probability that a live-born child would die before reaching age one under the mortality conditions of a specified period.

Definition and calculation

The conventional annual calculation is:

IMR=Deaths of children younger than one year during a given yearLive births during the same year×1,000.\mathrm{IMR} = \frac{\text{Deaths of children younger than one year during a given year}} {\text{Live births during the same year}} \times 1{,}000.

For example, 250 infant deaths and 50,000 live births in the same year produce an infant mortality rate of 5 deaths per 1,000 live births.

The denominator is live births, not all births and not the population of infants. Under the international definition, a live birth is the complete expulsion or extraction of a product of conception that subsequently breathes or shows another sign of life, regardless of the duration of pregnancy. A stillbirth, in which there are no signs of life at birth, is excluded from both the infant-death numerator and the live-birth denominator.

Although conventionally called a rate, IMR is not a person-time rate in the strict statistical sense. Its demographic interpretation is a probability of dying over a defined age interval.

Period and cohort measures

The usual annual calculation is a period measure. Infants dying in one calendar year include some born in the preceding year, whereas some infants born during the year will die in the next calendar year. The numerator and denominator therefore do not refer to exactly the same group of children.

A birth-cohort measure instead follows a defined group of live-born children until their first birthdays and calculates the proportion who die before that age. This requires sufficient follow-up and reliable linkage between birth and death records.

International estimates may use life-table methods or statistical models to estimate the probability of death before age one. These estimates and the simple annual ratio are closely related but need not be identical, particularly when birth numbers or mortality conditions change rapidly.

Age subdivisions and related indicators

Infant mortality is commonly divided into two components:

Component Age at death Conventional expression
[[neonatal-mortality-rate Neonatal mortality]] Birth to less than 28 completed days
Postneonatal mortality 28 completed days to less than one year Postneonatal deaths per 1,000 live births

Neonatal mortality is further divided into early neonatal mortality, before seven completed days of age, and late neonatal mortality, from seven completed days to less than 28 days.

These distinctions matter because the distribution of causes changes with age. Deaths soon after birth are particularly associated with complications of prematurity, events around labour and delivery, congenital anomalies, and severe infections. Later infant deaths reflect a different mixture of infections, congenital conditions, injuries, and other causes.

Infant mortality should not be confused with:

  • Under-five mortality rate, the probability of dying before the fifth birthday, expressed per 1,000 live births.
  • Perinatal mortality, which combines stillbirths and early neonatal deaths under specified gestational-age or birthweight criteria.
  • Maternal mortality, which concerns deaths related to pregnancy or its management rather than deaths among infants.

The neonatal and postneonatal components add to the conventional infant mortality rate when calculated from compatible death counts and the same live-birth denominator. Probabilities conditional on surviving to the beginning of each age interval require a different calculation.

Causes and determinants

The causes of infant death vary substantially across populations and over time. Important groups include:

  • Complications of preterm birth, including respiratory and other consequences of physiological immaturity.
  • Intrapartum-related complications, including impaired oxygen supply around birth.
  • Congenital anomalies.
  • Infections, including sepsis, pneumonia, and diarrhoeal disease.
  • Sudden unexpected infant deaths and injuries, whose relative importance differs between settings.

An infant may have several contributing conditions. For example, an infant born prematurely may develop an infection and respiratory failure. Mortality statistics generally distinguish an underlying cause of death from associated conditions, using rules established through the International Classification of Diseases.

Infant mortality is also shaped by social determinants of health. Maternal education, household income, food security, housing, environmental exposures, and access to safe water and sanitation influence risks before and after birth. Maternal health and nutritional status, birth spacing, multiple births, and access to timely care also affect outcomes.

The organization and quality of the health system influence survival through antenatal services, skilled care during childbirth, emergency obstetric and newborn care, and treatment during infancy. Population-level differences therefore cannot be attributed to medical care alone.

Data sources and estimation

Civil registration

Complete civil registration and vital statistics systems provide continuous records of live births and deaths. Birth and death records can be linked to examine mortality by birthweight, gestational age, maternal characteristics, or place of residence.

Reliable calculation depends on both events being registered and correctly classified. Missing infant deaths tend to lower the recorded rate, while missing live births can raise it if deaths are more completely recorded. Misclassification of a live-born infant who dies shortly after delivery as a stillbirth can also lower measured infant mortality.

Surveys, censuses, and models

Where registration is incomplete, household surveys and the population census can provide alternative information. Full birth histories record the dates of children’s births, whether they survived, and, when applicable, their ages at death. Summary birth histories record the number of children ever born and the number surviving; demographic methods are then used to infer mortality.

These sources have limitations. Very early deaths may be omitted, dates may be misremembered, and ages at death may be rounded. Survey estimates also have sampling uncertainty and often describe several years rather than a single calendar year.

The World Health Organization and other international agencies participate in coordinated estimation through the United Nations Inter-agency Group for Child Mortality Estimation. Such estimates combine and assess available observations, account for data-quality problems, and report uncertainty. They may differ from officially registered annual figures because the underlying data and methods differ.

Historical change and public-health use

In many populations, infant mortality declined substantially during the nineteenth and twentieth centuries. The timing and pace varied. Improvements in sanitation, food safety, housing, maternal education, and nutrition contributed, alongside vaccination, antimicrobial treatment, safer childbirth, and advances in newborn care.

As infant mortality falls, neonatal deaths often account for a larger share of the remaining infant deaths. This can happen even when neonatal mortality itself is declining, if mortality after the neonatal period falls faster.

IMR is used to monitor population health, assess long-term change, identify geographical or social inequalities, and inform the allocation of health resources. Its sensitivity to conditions early in life also makes it important in understanding life expectancy at birth: deaths during infancy substantially reduce that measure.

Disaggregation by region, household circumstances, or other relevant characteristics can reveal inequalities hidden by national averages. Such comparisons require sufficient observations and attention to differences in registration completeness.

Limitations in interpretation

Infant mortality is a broad outcome indicator, not a direct measure of the quality of any single service. Similar rates may arise from different combinations of birth risks, social conditions, and access to care.

International comparisons are particularly sensitive to how extremely premature infants are recorded. Although the international definition of live birth does not impose a gestational-age threshold, national registration practices and statistical inclusion rules can differ. Comparisons restricted to common gestational-age or birthweight categories may improve comparability, but describe a narrower population.

Small populations can show large year-to-year fluctuations because a few deaths appreciably change the rate. Multi-year averages and uncertainty intervals help distinguish such variation from sustained changes.

Finally, an observed association between IMR and income, education, or health spending does not by itself establish causation. Interpretation requires attention to population characteristics, data quality, and other changing influences. Infant mortality also excludes stillbirths, deaths after the first birthday, and nonfatal illness or disability, so it is normally assessed alongside other maternal and child health indicators.