The main determinants of infant mortality in Nepal

Abstract

Infant mortality has reached a low stable rate in developed countries while it is still high and on a slow decline in developing countries. There are many factors that contribute to the incidence of a high or low level of infant mortality. Although credit for contributing to the lowering of infant mortality has been given to health programs by public health personnel and to the improvement in socio-economic status by social scientists, in a traditional and agricultural country such as Nepal, both these factors are found to influence infant mortality. Data on infant mortality obtained by the 1991 Demographic Health Survey of Nepal are analyzed in this study. A logistic regression model is used for analyzing the data. Several hypotheses are tested to explain the incidence of infant mortality in Nepal. The various reasons for the persistence of high infant mortality and the difficulties in lowering it are discussed. The findings suggest that among all the variables analyzed in the study, parity, place of residence, immunization, and ethnicity influence infant mortality the most.

Introduction

The infant mortality rate,¹ known to be a good and sensitive indicator of the development of a nation, has been a focus of study in Nepal recently. This sensitive indicator is influenced directly and indirectly by a number of factors. In difficult situations, adults and the elderly may be able to survive better than infants, whose immune systems may be less able to cope with the environment. For this reason, infants are affected the most by the availability of health facilities, life style of the family, affordability of good food, sanitation, etc. In developing countries, infant mortality accounts for a relatively higher proportion of all deaths, whereas in the developed countries, it represents an increasingly small segment of total mortality (Klinger, 1985). Furthermore, lower infant mortality means that fewer children need to be born to achieve a certain number of survivals (Weeks, 1994). Thus, infant mortality has a significant role to play in the demographic transition of a nation.

There has been a debate for many years among demographers and public health scientists about the contribution of socio-economic development and the medical facilities provided by public health programs in the reduction of mortality (Arriaga & Davis, 1969; Preston, 1975; Caldwell, 1979). The relatively rapid mortality reduction in Latin America, Africa, and Asia in the mid-20th century was attributed to the use of antibiotics and malaria control (Stolnitz, 1955). In Sri Lanka, 23% of the total national post-war decline in mortality was accredited to malaria control (Gray, 1974). Frederiksen (1968), however, had a different view of the dramatic decline in Sri Lanka's mortality rate in 1946. He provided evidence that economic development, by increasing per capita food consumption, was an important factor. On the other hand, the control of infant mortality through widespread education programs contributing to an increase in literacy among women and health consciousness is believed to be the major factor in the reduction of mortality in the state of Kerala, India (Nayar, 1985).

Caldwell (1986) has a similar view, giving credit to improvements in education and health as the determinants of low mortality in developing countries in the 20th century. But in a country such as Nepal where social norms are complex, socio-economic status is low and medical facilities are often out of reach for most of the population, the socio-economic factor or the public health program alone may not explain the slow pace in reducing infant mortality. In the case of Nepal, these factors are indeed interrelated to one another, and both seem to have an impact on infant mortality. The significant drop in the infant mortality rate from about 250 per 1000 live-births in the mid-1950s to about 130 per 1000 live-births in the mid-1970s is, perhaps, the outcome of the malaria control program of the 1950s and the small-pox eradication campaign of the 1960s (Table 2). Nevertheless, judging from the 1960s and 1970s age distribution in Nepal, a dramatic decline in the death rate was observed in the generation before the country opened the door to international contact and before health programs were introduced (Taylor & Taylor, 1976). Among several other factors, a massive internal migration has been taking place in Nepal in recent decades, mostly from the mountains to the foothills and flatlands (Gurung, 1998). Such movements within the country, especially from rural to urban areas, could also have made a difference in infant survival in Nepal.

Population scientists classify mortality broadly as two types: endogenous and exogenous. Endogenous mortality is presumed to arise from genetic causes such as degenerative diseases (cancer, heart disease, diabetes, etc.) and from causes related to early infancy such as birth injuries, congenital disorders, premature births, and postnatal asphyxia. Exogenous mortality, on the other hand, is presumed to arise from environmental or external causes such as infections and accidents. The former type of mortality has a biological character and dominates the deaths in the elderly population and in infancy, particularly in early infancy. The latter class of mortality is viewed as relatively preventable and treatable (Shryock et al., 1976).

Considering the infant deaths only, when medical care is less available, post-neonatal deaths exceed neonatal deaths (deaths occurring in the first 28 days of birth). On the other hand, in societies in which medical care has achieved low death rates, it is the post-neonatal deaths that have been reduced the most. This results in a majority of all infant deaths being neonatal, caused by physiological and organic weaknesses (Bogue, 1969), of which the most severe cases succumb to death in the early neonatal stage.

The infant mortality rate in Nepal (for the census year 1991) was very high at 97.5 per 1000 live-births (Central Bureau of Statistics, 1995). However, the Population Reference Bureau, Inc., USA (1995) estimates Nepal's IMR as 102 per 1000 live-births, which is third highest in the South Asian region with 138 for Bhutan, 108 for Bangladesh, 91 for Pakistan, 74 for India, 52 for the Maldives, and 19.4 for Sri Lanka. According to a recent estimate among the 75 districts of the country, an infant mortality rate as low as 32 per 1000 live-births is estimated for Mustang district (in the Central Mountain Region), whereas the rate for Mugu district (in the Western Mountain Region) is the highest at 201 per 1000 live-births. Kathmandu district, the capital of Nepal, on the other hand, has an infant mortality rate of 34 per 1000 live-births (RECPHEC, 1997).

According to the 1991 census, Nepalese infants comprised 3.1% of the total population; that is, 565,413 out of 18,491,097 people. Current estimates from various surveys show that as many as 49% of all infant deaths occur in the neonatal period, and a little less than half of these occur in the early neonatal (within 7 days of birth) stage (National Population Commission/UNICEF, 1992). Quite contrary to Bogue's thesis noted above, the high percentage of neonatal deaths in Nepal is not due to a reduction of post-neonatal deaths with more health-care availability, but to causes such as congenital abnormalities, premature births, birth injuries, tetanus infections, and low birth weight (National Population Commission/UNICEF, 1992). In the post-neonatal stage, the most prominent causes of death are acute respiratory infection² and diarrhoea.

Previous studies on infant mortality suggest that among the three geographical regions of Nepal, the mountain area had the highest infant mortality, followed by the Terai (flat areas), with the hills having the least infant deaths (Gubhaju, Choe, Retherford, & Thapa, 1987; Gubhaju, 1991). An earlier study on the mortality differential in rural Nepal also found that in 1975 and 1978 the hill area had a higher child survival rate than the Terai area (Tuladhar & Stoeckel, 1983). Infant mortality was found to be lower for literate mothers than that for illiterate ones (Shrestha, Gubhaju, & Roncoli, 1987; Gubhaju et al., 1987). The risk of dying of first-order births was attributed to the higher proportion of very young mothers (Gubhaju, 1991). Some contradictory results were found regarding the birth orders. Infant mortality was found to decline with higher birth order in one study (Gubhaju et al., 1987), while another study showed a higher risk of dying during infancy for high-order births (Gubhaju, 1991). However, regardless of the birth order and the birth interval, the death of the previous sibling before the birth of a baby raised risk of dying for that baby (Gubhaju et al., 1987; Gubhaju, 1991).

Contrary to the popular belief and findings from other developing countries, indications of sex discrimination were not found in Nepal (Gubhaju (1984), Gubhaju (1991)); in fact, male babies were found to have rather a higher risk of dying (Gubhaju et al., 1987). Estimates of infant mortality rates by sex computed by different individuals and surveys over the past five decades also support this finding (Table 2). Among other results, higher production of cereal per capita depressed infant deaths, while a larger population to health worker ratio and higher female labor force participation increased infant mortality (Shrestha et al., 1987). Table 1 shows the socio-demographic figures for Nepal. Table 2, Table 3 present a short time series of the infant mortality rate in Nepal.

It is clear from these data that the infant mortality rate shows a declining trend. A similar trend is noticed for infant mortality by gender as well as by rural and urban residence. However, the trend is not consistent.

The main purpose of this study is to analyze the determinants and consequences of infant mortality in Nepal with the objectives of analyzing the effects of socio-economic variables such as ethnicity, religion, education of the mother, education of the father, occupation of the mother and occupation of the father, various demographic and health-care-related factors on infant mortality, the rural/urban differential, and the risk of lifestyle factors on infant mortality.

Studies such as this one are important to assess the current situation of the country in terms of the health and survival of infants and how the deaths of infants who are the future manpower of the country can be prevented. This study is also relevant because only a few investigations has been done on this topic. The major work in this area is that of Gubhaju (1991) who used 1976 Nepal Fertility Survey data. However, prenatal visits and immunization variables were not included in the 1976 survey. The current study is based on new data and incorporates some additional variables such as religion, ethnicity, and health-care utilization.

This study will provide valuable insights into unexplored aspects of socio-economic, demographic, and health-related factors on infant mortality in Nepal. This country was isolated from the outside world until 1951. The development of its infrastructure, including educational institutions, hospitals, primary health care, roads, and so on took place only after the early 1950s. This study contributes to a better understanding of how mortality change relates to such factors.