Breast-feeding and infant mortality in Norway 1860-1930

Using medical records from maternity clinics in the two Norwegian cities Oslo and Bergen, the effect of lactation on infant mortality during the period 1860-1930 is examined, comparing those who were and were not breast-fed in a total of about 6900 live born infants. The mortality of children not breast-fed was nearly three times that of those who were breast-fed. In a Cox regression analysis the infant's year of birth and the mother's marital status were found to influence mortality in addition to lactation. Children born to unmarried mothers experienced a mortality about twice that of those born to married mothers, both during on-going lactation and in the absence of lactation, up to about 1915. Children born in Oslo had a slightly higher mortality than those born in Bergen. The duration of lactation was found to have a continuing protective effect on infant survival after weaning--the longer the duration, the lower the mortality after cessation of lactation.     

Birth spacing,breastfeeding, and early child mortality in a traditional Indian society: A hazards model analysis

Abstract

There are few studies of the interrelationships among breastfeeding, child spacing, and child mortality in traditional societies that incorporate extensive controls for social and demographic characteristics of the mother and child. In this paper, we investigate the impact of breastfeeding and the length of the preceding birth interval on early child mortality (defined as a death in the first two years of life) using data from a traditional society of India. Multivariate hazards models are used to analyze the data. Most prior analyses related the impact of breastfeeding duration to the duration of child survivability by taking breastfeeding as a fixed covariate. The present study has a methodological focus in the sense that breastfeeding information from retrospective survey data is treated as a time‐dependent covariate both as a status variate as well as a duration—with empirical findings compared across the two specifications. The effects of postpartum amenorrhoea and various other demographic and socioeconomic characteristics of mother and child are also studied. The results suggest that breastfeeding duration has a strong impact in reducing the relative risk of early child mortality; but it does not explain the effect of the length of the preceding birth interval on early child mortality.     

Breast-feeding, birth interval and child mortality in Bangladesh

The 1975-76 Bangladesh Fertility Survey data show little evidence that breast-feeding is the intermediate factor through which birth intervals influence child survival in Bangladesh. Preceding birth interval, subsequent pregnancy and breast-feeding duration each have an independent influence on early mortality risk. Within a specific interval the risk of dying decreases with increase in duration of breast-feeding, and also with an increase in the time between the index birth and the next pregnancy. The death of the immediately preceding child in infancy has a significant negative effect on the survival chance of the index child at ages 1-5 months. However, death of the preceding child appears to have a significantly positive effect on the survival chance of the index child between ages 9 months and 5 years; this may be related to competition between siblings.     

Parent characteristics and sex differential infant mortality: the case in Tennessee

Females survive males at most ages in virtually all modern industrialized societies. However, stratification of a sample by socioeconomic indicators shows that white infant male and female postneonatal mortality rates converge as the mothers' or parents' education rises and as the mothers age. These results are congruent with the parental selection hypothesis and with other findings that suggest that families at the pinnacle of the local social structure invest heavily in males and that parents nearing the end of their reproductive career try hard to effect the survival of all offspring.     

Female reproductive development: A hazards model analysis

Abstract

In this paper we applied a recently developed statistical technique, hazard modelling, to the study of menarche and adolescent subfecundity. Using data collected in Bangladesh in 1976 and 1977, which included anthropometric and socioeconomic status measures, we found that age and weight were significant predictors of menarche; no other measure of growth significantly added explanatory power. Weight was at least as good as other more complex measures, including the fat index proposed by Frisch. Low SES was associated with later age at menarche and appears to act through slower growth. Whenever weight was included in an analysis, the SES effects were not significant. Finally, although low SES tends to delay menarche, it seems to have little effect on the interval of adolescent subfecundity. Indeed, late maturers exhibit some degree of “catching up” in that the length of the subfecundity interval is negatively related to age at menarche.

The study also has a methodological focus. We demonstrate that use of hazard models permits more detailed and penetrating findings to be extracted from the kinds of retrospective data traditionally collected as well as from longitudinal data of the types only recently available. In particular, multivariate analysis of age at menarche and the subfecundity interval, and their relations to anthropometric time‐varying covariates as well as socioeconomic indicators, is shown to be feasible and informative.     

Analyzing infant mortality with geoadditive categorical regression models: a case study for Nigeria

In this paper, we analyze infant mortality in Nigeria based on the data set from the 1999 Nigeria Demographic and Health Survey (NDHS). We investigate spatial patterns at a highly disaggregated level of Nigerian states and consider non-linear effects of mother's age at birth. Time to the occurrence of a child's death can intuitively be considered to be categorical in nature and the determinants of a child's death may differ in different age groups. Thus, it may be desirable to investigate separately the death of a child in the first month and in the remaining 11 months of the first year of life. To avoid selection bias, the data set used for this case study is based on information on children who were born 12 months preceding the survey. Inference is Bayesian and is based on Markov chain Monte Carlo (MCMC) techniques. We find that spatial variation and the determinants of death indeed differ considerably for the two age groups considered.     

Covariates of age at first birth in Guyana: a hazards model analysis

Analysis of data from the Guyana Fertility Survey on the trends and covariates of age at 1st birth among various birth cohorts of women ever in union indicates that an early entry into union is associated with young age at 1st birth and higher number of children born. Multivariate analysis showed that women with higher education, urban residence, and entry into 1st birth compared to others, and that young women are delaying their 1st birth for longer durations than older women. Work status of women before 1st birth and the starting age of union seem to be the 2 major contributory factors for age at 1st birth. Noticeably, the role of education has changed and is now more significant among younger cohorts than among older ones for 1st birth timing.     

A hazards model analysis of breastfeeding variables and maternal age on return to menses postpartum in rural Indonesian women

There is mounting research evidence that the duration of lactational amenorrhea is dependent on the infant's suckling input. Multivariate techniques, including the proportional hazards model, offer an effective methodological approach for sorting through the variables that contribute to a process as complex as breastfeeding. This approach was utilized on a sample of 382 mothers who participated in the Ngaglik Study, a longitudinal investigation of maternal health and nutrition, infant development, child spacing, and fertility trends in Central Java, Indonesia. 3 primary breastfeeding variables--average number of nursing episodes during the day, average number of nursing episodes during the night, and average minutes per episode--were obtained from monthly interviews with study subjects, 260 of whom experienced return to menses while breastfeeding. The mean and median durations of amenorrhea were 17.3 and 16.4 months, respectively. The reported total number of suckling bouts per 24 hours averaged 8.85, with an average of 8.23 minutes per nursing episode. Amenorrhea duration ranged from 19.2 months in mothers who nursed 6 or more times during the day-time hours to 12.2 months in mothers who nursed an average of 6 minutes or less per episode. The variable of minutes per nursing bout has the greatest effect on return to menses, while the average number of day-time feeds has the least; the number of night-time feeds is intermediate. When age was introduced into the model, the effects of the nursing variables on return of menses remained constant relative to 1 another but the increment in the risk of menstruating increased with younger age. In summary, this analysis indicates that low intensity breastfeeding with 3 or fewer episodes reported at night, 6 or fewer episodes reported for the day, 6 or less minutes reported per nursing episode, and younger age all increase the risk of early postpartum resumption of menses.     

Variation and classification of human age patterns of mortality: analysis using competing hazards models

The structure of variation in human mortality patterns is explored using a five-parameter competing hazards model and standard multivariate taxonomic procedures. The data consist of 281 national life tables representing a wide range of environmental and cultural regions of the world. A general or average age pattern of mortality was generated for the entire sample using multiple regression procedures. A K mean cluster analysis conducted on the residuals of the regression analysis identified seven distinct models of mortality that differ in characteristic ways from the general pattern. Four of the seven clusters have age patterns of mortality similar to the north, east, south, and west regions of the Coale and Demeny model life tables. The remaining three clusters represent regions of the world and age patterns of mortality that are not represented in the Coale and Demeny model life tables.     

Multivariate analysis of infant and child mortality in Java and Bali

The data used in this analysis come from the 1976 Indonesian Fertility Survey, part of the World Fertility Survey. The data are examined 3 times, fitting them to models which include different combinations of independent variables. The dependent variables are: 1) the proportion of children born between 5 and 15 years before the survey who died before their 1st birthday, for infant mortality; and 2) among those alive on their 1st birthday, the proportions who died before reaching their 5th birthday, for child mortality. The figures indicate that the chance of dying for children who were 1st born, born shortly after a previous child, whose previous sibling had died, who lived in rural areas, or had parents who were young and with little education, was greater than for children without these characteristics. In all 3 models used, the greatest net effects are attributed to the survival of a preceding sibling or the length of the preceding interval. Birth order does not have a significant gross effect on infant mortality, but the net effects are significant because of the control on maternal age. Education of both parents has significant effects, but these are overshadowed in magnitude by the demographic variables. Maternal education has a greater influence in determining differences in child mortality than was found for infant mortality. Father's education also has a significant independent effect, but mainly for 1st births. It is uncertain whether these variables are measuring the effect of schooling as such, or other characteristics such as economic status or various social roles adopted by people with different levels of education. The variables distinguishing urban from rural status shows significant gross effects which are greatly reduced when controls for other variables are introduced in the model which includes all births. That is to say, the difference in the survival chances of a child in the city is more a function of the education of its parents, and the associated demographic variables than city residence as such. Access to medical services is quite probably the main element in these differences. The findings are weakened to some extent by the lack of satisfactory data on household economic status which might have provided a better base for indirectly discerning the effects of nutrition and sanitation on mortality at young ages.     

Teenage fertility, socioeconomic status and infant mortality

This paper examines the relationship between infant mortality and a complex measure of socioeconomic status for evidence of diminution. In data on counties in the US with a minimum of 20 infant deaths over the 5-year period 1971-75, no evidence of a declining relationship between socioeconomic status and infant mortality was found. Both level of community affluence and racial composition of the population exerted direct effects on levels of infant deaths. In addition, both socioeconomic status and racial composition exhibited indirect effects which operated through teenage childbearing. When total infant mortality was subdivided, teenage fertility serves as a mediating variable in the link between socioeconomic status and neonatal mortality, but not for the postneonatal components. Given the nearly equivalent total effect of socioeconomic status on infant mortality, it is concluded that the classic division into neonatal (supposedly a function of biological and genetic agents) and postneonatal (traditionally attributed to social and environmental agents), may be too crude to allow the contemporary effects of the socioenvironmental milieu to be evaluated effectively.     

Breast-feeding, diarrhoea and sanitation as components of infant and child health: a study of large scale survey data from Ghana and Nigeria

Using Demographic and Health Survey datasets from Ghana and Nigeria, this study examined whether the protective effects of breast-feeding are greatest where the poorest sanitation conditions prevail. It was found that mixed-fed infants aged between 0 and 11 months tend to have a higher risk of diarrhoea than fully breast-fed children, while the risk of diarrhoea among weaned infants is twice that of mixed-fed infants. The probit regression models employed in the analysis were used to predict the probability of diarrhoea associated with each breast-feeding pattern for both 'poor' and 'good' sanitation areas. It was found that the risk of diarrhoea among mixed-fed infants in the poor sanitation areas tends to be high while the same risk among fully breast-fed infants tends to be minimal. In essence, the health risks of mixed feeding are real, particularly for infants aged less than 7 months, and are even worse for those weaned before 6 months of age.     

Ranking bioprocess variables using global sensitivity analysis: a case study in centrifugation

The selection of appropriate operating conditions for bioprocessing is complex due to the large number of interacting stages and variables. Bioprocesses also operate under tight regulation and therefore tools to assist bioprocess design are of significant utility. Conventional approaches for the analysis of variable sensitivities are inadequate. We propose the use of global sensitivity analysis to determine the level of importance of each variable and their interactions. Once key variables have been determined, the designer may focus on the most significant subset. Two case studies are used to demonstrate the applicability of the approach. Each is based on centrifugation and determines the impact of flow-rate, feed viscosity, density difference and particle size, while performance is assessed by supernatant clarification. Significant differences in sensitivities were found between the two studies due to the different feed material properties. Variable sensitivities were found to be system-specific and provide insight for potential operating strategies.     

Influence of birth weight on differences in infant mortality by social class and legitimacy.

OBJECTIVE--To investigate the influence of birth weight on the pronounced social class differences in infant mortality in Britain. DESIGN--Analysis of routine data on births and infant deaths. SETTING--England and Wales. SUBJECTS--All live births and infant deaths, 1983-5. MAIN OUTCOME MEASURE--Mortality in infants by social class, birth weight, and legitimacy according to birth and death certificates. RESULTS--Neonatal and postneonatal mortality (deaths/1000 births) increased with social class. Neonatal and postneonatal mortality was 4.2/1000 and 2.3/1000 respectively for social class I and 6.8/1000 and 5.6/1000 respectively for social class V. Mortality was lower among births registered within marriage (postneonatal 3.5/1000; neonatal 5.2/1000) than among those jointly registered outside marriage (5.1/1000; 6.4/1000); mortality was highest in those solely registered outside marriage (7.2/1000; 7.0/1000). For neonatal mortality the effect of social class varied with birth weight. Social class had little effect on neonatal mortality in low birthweight babies and increasing effect in heavier babies. For postneonatal mortality the effect of social class was similar for all birth weights and was almost as steep as for all birth weights combined. CONCLUSION--Birth weight mediates little of the effect of social class on postneonatal mortality.     

Deforestation and infant mortality: Evidence from Indonesia

Factors expected to change concurrently with forest loss—such as economic activity and air pollution—shape human health in different ways, making it difficult to ex ante predict the net impact of deforestation. This paper investigates the infant mortality effects of prenatal exposure to high biomass forest loss in Indonesia, a country with rich forest reserves increasingly being subjected to high levels of deforestation. Indonesia officially bans clearing in areas with high biomass natural forests (referred to henceforth as ‘protected forests’), yet these forests face illegal logging. The analysis uses a fixed effects approach, essentially tracking how mortality responds to protected forest cover changes over time within districts. Results suggest that protected forest loss favors survival among all infants. However, there is variation in the protected forest loss-infant mortality relationship by pregnancy order or gravidity—while children born from women’s higher order pregnancies are less likely to die when exposed to deforestation, children born from first pregnancies experience an increase in their risk of death. Potential mechanisms such as overall air pollution, economic activity and perinatal health care do not appear to explain the gravidity-specific effects of deforestation in protected areas. However, the observed pattern of results suggests that effects are being channeled through malaria—the disease, which is likely to increase with forest loss, tends to disproportionately infect women during their first pregnancy, thus causing greater harm to the children born from these pregnancies.