Levels, Patterns and Determinants of LBW in Developing Countries
Morbidity and Mortality Consequences of LBW in Neonates and Infants
Growth in LBW Children
Long-term Consequences of LBW: The Foetal Origins of Disease Hypothesis
At least 17 million infants are born every year with LBW, representing about 16% of all newborns in developing countries. Nearly 80% of all affected newborns with LBW at term are born in Asia (mainly south-central Asia, with Bangladesh having the highest incident rate in the world5); about 15% and 11% are born LBW at term in middle and western Africa respectively, and approximately 7% in the Latin American and Caribbean region.3 The geographical incidence of LBW at term in selected Asian and African countries (Figures 2 and 3 respectively) confirm that many developing countries exceed the internationally recommended cut-off levels which should trigger public health action. Incident rates of >15% for LBW and >20% for IUGR indicate that LBW at term is a major public health problem. Population-wide interventions aimed at preventing LBW at term are therefore urgently required.6
Figure 2. Incidence of LBW at term in selected Asian countries
Source: de Onis et al. (1998) Eur J Cl Nutr 52(S1):S5.
Figure 3. Incidence of LBW at term in selected African countries
Source: de Onis et al. (1998) Eur J CI Nutr 52(S1):S5.Historically, because valid assessment of gestational age is often not available in developing countries, evidence of LBW has often been used as a proxy to quantify the magnitude of IUGR. The incident rates for LBW at term conservatively estimate IUGR because when all infants below the 10th percentile of the birthweight-for-gestational-age reference are considered, approximately 24% or 30 million newborns in developing countries would be affected each year. Major constraints to deriving this estimate include both the quantitative and qualitative limitations of the available birthweight data.6 Most of the data available from different parts of the world are from clinic or hospital deliveries, whereas, in some regions of Africa and south-east Asia most infants are born at home and are not measured. There is a need to determine whether data from hospital-born infants in developing countries are representative of the large population born at home.6
Prematurity and IUGR are the two main causes of LBW. The majority of LBW in developing countries is due to IUGR, while most LBW in industrialized countries is due to preterm birth.7 In many cases, the causes of prematurity are unknown; they may include high maternal blood pressure, acute infections, hard physical work, multiple births, stress, anxiety, and other psychological factors.8,9 Causes of IUGR are complex and multiple, but center on the foetus, the placenta, the mother, and combinations of all three. For instance, growth will be retarded in utero if the placenta is abnormally small or blocked causing insufficient nutrients to reach the foetus. The maternal environment is the most important determinant of birthweight, and factors that prevent normal circulation across the placenta cause poor nutrient and oxygen supply to the foetus, restricting growth. These factors may include maternal undernutrition, malaria (where it is endemic), anaemia, and acute and chronic infections (such as sexually transmitted diseases and urinary tract infections).9 Also associated with IUGR are primiparity; multiple gestation; foetal, genetic or chromosomal anomalies; as well as maternal disorders such as renal diseases and hypertension.10 Cigarette smoking and pre-eclampsia cause the highest relative risks for IUGR in industrialized countries, while alcohol and drug use may also restrict foetal growth.8, 11
Major determinants for LBW in developing countries, however, are poor maternal nutritional status at conception, low gestational weight gain due to inadequate dietary intake, and short maternal stature due to the mothers own childhood undernutrition and/or infection (Box 2).12 Because maternal undernutrition is a major determinant of LBW in developing countries, high rates of LBW should be interpreted not merely as an indicator of undernutrition, morbidity and mortality for the newborn, but as an urgent public health warning that women of childbearing age are undernourished as well. Countries with higher percentages of LBW infants generally have a higher percentage of women with low body mass index (BMI) and a higher percentage of underweight children.13,14 To address these issues successfully, the underlying and basic causes of LBW in developing countries such as household food security, maternal and child care, access to and quality of antenatal and other health services, sanitation and hygiene, education, gender discrimination and poverty must be included in any long-term strategies for prevention (Figure 4).
Box 2. Determinants of LBW in Developing Countries
Maternal undernutrition - a major determinant of LBW in developing countries as evidenced by the following nutritional deficiencies:
à Low gestational weight gain
Other etiologic determinants include:
à Young maternal age (adolescent)
Figure 4. Causes of malnutrition
Source: UNICEF (1997) The Care Initiative: Assessment, Analysis and Action to Improve Care for Nutrition, UNICEF: NY.
LBW is generally associated with increased morbidity and mortality, impaired immune function, and poor cognitive development for neonates (newborns 1-28 days of age) and infants. Infants born LBW are at risk to develop acute diarrhoea or to be hospitalized for diarrhoeal episodes at a rate almost two to four times greater than their normal birthweight counterparts.15-18 Infants who are LBW risk contracting pneumonia or acute lower respiratory infections (ALRI) at a rate almost twice that of infants with normal birthweight; and more than three times greater if their weight is less than 2000 g.17-20 LBW is also implicated as a contributor to impaired immune function which may be sustained throughout childhood.21-23
The risk of neonatal death for infants who are LBW weighing 2000-2499 g at birth is estimated to be four times higher than for infants weighing 2500-2999 g, and ten times higher than for infants weighing 3000-3499 g.24 In Brazil, 67% of all infants dying during their first week of life are LBW infants; in Indonesia the rate is 40%; and in the Sudan the rate is 35%. Infant mortality (less than one year of age) due to LBW was slightly lower: 47% in Brazil and 19% in Indonesia.25-27 LBW infants during the post-neonatal period (>28 days of age) also have high mortality rates -and in some cases their risk may be greater than those for LBW infants during the neonatal period.5,28 LBW accounted for 69% of the ALRI deaths in India, and it is estimated that in Bangladesh, almost half of the infant deaths from pneumonia or ALRI and diarrhoea could be prevented if LBW were eliminated.5,29
Do LBW infants grow normally? What are the consequences of LBW on body size, composition, strength and cognitive development? Attaining full growth potential is especially important for women and girls in order to break the intergenerational cycle of LBW and have fewer delivery complications. Maternal height is not only a reflection of genetic make-up, but also reflects her dietary history. From societal, community and individual standpoints, adolescents and adults born with LBW generally have less strength and lower lean body mass resulting in decreased work capacity and lost productivity, which may cost nations billions of dollars.1,30
When growth restriction in utero occurs early in pregnancy, infants exhibit symmetrical (or proportional) growth with length, weight, head and abdominal circumference all below the 10th percentile reference for a given gestational age (stunting). When growth restriction in utero occurs late in pregnancy, the infant exhibits asymmetrical (or disproportionate) growth with a normal length and head circumference, but low weight due mainly to a lower proportion of visceral and fat tissue (wasting).7,31 Neonatal mortality rates are reported to be higher among asymmetrical IUGR infants, but if they survive, they have a better prognoses for long-term growth and development than that for symmetrical IUGR infants. IUGR infants catch-up partially in growth relative to their appropriate birthweight counterparts during their first one or two years of life. Thereafter, IUGR children maintain their place in the distribution and neither catch-up nor fall further behind. They remain about 5 cm shorter and 5 kg lighter as adults. Premature infants (who are usually asymmetric LBW), who survive their first year, have a much better prognosis in terms of future growth than IUGR infants. Despite their earlier disadvantage, preterm children gradually catch-up with their appropriate birthweight, term counterparts. Premature infants and IUGR infants should be studied as separate groups because they show different patterns of growth, morbidity and mortality. From a programmatic viewpoint these differences have enormous implications for intervention strategies and limitations of the approach of nutritional recovery of IUGR infants in early childhood.10,25,30,32-36
Neurological dysfunction is often associated with attention deficit disorders, hyperactivity, clumsiness, and poor school performance. Neurologic dysfunction, when present, seems to affect IUGR boys more than girls, and children of lower socioeconomic circumstances. If IUGR infants are symmetrical and head growth is affected, there seems to be more of an impact on neurological function and it is not clear whether interventions directed toward these infants will improve their outcome. For asymmetric IUGR infants, preventing asphyxia should reduce the prevalence of major and minor handicaps, especially cerebral palsy and mental impairment frequently seen in these infants.37,38 IUGR is a much larger public health problem in developing countries than in industrialized countries and the outcomes are more likely to be aggravated by obstetric complications and perinatal problems, and later by poor health and nutrition as well as psycho-social deprivation.37
In developing countries children are exposed to poor nutrition, high levels of infections, and other conditions of poverty, thus, their long term development is dependent to a large extent on the quality of their environment. It is difficult to isolate the effects of IUGR from these factors in relation to cognitive development. Cognitive deficits appear to change over time. For instance, when IUGR infants were examined, no differences were found during the first year of life, but differences emerged during two and three years of age; and then differences disappeared at four to five years. Deficits have been found in children with very low birthweights, the smallest size, or with early IUGR (growth restriction prior to 26 weeks gestation). Since LBW occurs more often in deprived environments, it can serve as a marker for the associated poor outcomes throughout life. A length deficit at an early age (stunting) would be the best predictor of motor and mental development deficits.39,40
The foetal origins of disease hypothesis states that foetal undernutrition at critical periods of development in utero and during infancy leads to permanent changes in body structure and metabolism. These changes result in increased adult susceptibility to coronary heart disease (CHD) and non-insulin dependent diabetes mellitus (NIDDM). There is also growing evidence that those adults born with LBW suffer an increased risk of high blood pressure, obstructive lung disease, high blood cholesterol and renal damage. Thus, a poorly growing foetus is an undernourished foetus prone to reduced growth, altered body proportions, and a number of metabolic and cardiovascular changes. It has been suggested that these changes are adaptations for foetal survival in an inadequate nutritional environment, and that these changes persist post-natally, contributing to adult chronic disease when nutrients are plentiful.4
The foetal origins of disease hypothesis, also known as the Barker hypothesis, was generated by David JP Barker and colleagues of the MRC Environmental Epidemiology Unit of the University of Southampton. Barkers group was puzzled that CHD was the most common cause of death among certain men who otherwise had low risk characteristics, i.e., they were slim, non-smokers, and had low blood cholesterol. This suggested that the etiology of CHD needed further exploration. The group speculated that foetal undernutrition during the first trimester may result in a proportionately small (symmetrical or stunted) infant prone to haemorrhagic stroke. Foetal undernutrition during the latter stages of pregnancy may result in a disproportionate (asymmetrical or thin) infant prone to CHD and an increased risk of insulin resistance, or a short infant prone to CHD and thrombotic stroke.41
The foetal origins hypothesis originated in the 1980s when Dr. Barker replicated a study from Norway which demonstrated a strong correlation between infant mortality rates (IMR) at the beginning of the century with current death rates from CHD. The author of the Norwegian study suggested that because infant mortality is a sensitive indicator of the quality of the immediate post-natal environment, perhaps growing up in poverty causes some sort of deficit which results in a lifelong vulnerability to aspects of an affluent adult lifestyle such as a high fat diet.42 Dr. Barker found a similar correlation between IMR and death from CHD in England and Wales but suggested, however, that since CHD was more closely correlated with neonatal mortality than with post-neonatal mortality, CHD may find some of its roots in IUGR as reflected by LBW.43 The historical evolution of the Southampton groups research included a move from geographical associations, to associations related to individuals, then to biological risk factors. Birth records from Hertfordshire, UK, were first used to study mortality in relation to birthweight. The study showed that the highest death rates were in men and women who had the lowest birthweights, and death rates fell as birthweight increased. This pattern was specific for CHD and chronic obstructive lung disease.44 A similar pattern was found for biological risk factors (hypertension and impaired glucose tolerance (IGT) and diabetes) for CHD in men: the highest rates were in men who had been small infants. More than 20% of men whose birthweights were lower than 2500 g had abnormal glucose tolerance, compared with under 10% of those weighing more than 4000 g at birth.45 These study results have now been replicated by several groups in many different countries including the USA, Sweden, Finland, India and China.41,46-53
The Barker research group also proposed a foetal programming hypothesis in which there is a brain-sparing reflex that, in an undernourished foetus, diverts or conserves the blood flow to the head, while simultaneously reducing the blood flow to the liver, pancreas and kidneys. This results in a reduced secretion of growth hormones, insulin and other endocrine changes which leads to CHD and NIDDM in adulthood (Figure 5). The Barker theory remains hypothetical since no causal relationships have yet been established, only associations. Two other explanations for the association between LBW and adult disease include the confounding explanation (Figure 6) and a genetic explanation (Figure 7). The confounding explanation suggests that LBW is a marker for poor socioeconomic status: poor people have smaller infants who are more likely to smoke, be exposed to stress, grow up with inadequate nutrition and become obese - all factors which cause CHD.54 The genetic explanation, on the other hand, suggests that if an individual has a gene for insulin resistance, this would lead to LBW, and the same genetic pre-disposition would lead to an increased risk of adult diabetes and CHD.55
Figure 5. LBW and adult disease Foetal Programming Hypothesis
Source: Fall (1999) Personal communication.
Figure 6. LBW and adult disease confounding explanation
Source: Whincup (1997) Diab 40:319.
Figure 7. LBW and adult disease genetic explanation
Source: Hattersley (1999) Lancet 353:1789.Lack of information on possible confounding lifestyle and environmental factors, limitation of the initial Barker studies to two populations in the UK, the retrospective nature of the observations, and differences in study methodologies all underscore the need to establish a core research protocol to investigate a longitudinal relationship between LBW due to poor foetal growth and disease in later life.56 The foetal origins theory appears to be of greatest relevance to developing countries where mean birthweights remain low and rates of LBW are high. Many of these countries are experiencing a nutrition transition which includes changes in dietary intake, physical activity and body composition. The nutrition transition refers to a shift to diets high in total fat, sugar, and refined grains; it includes a more sedentary lifestyle; and increased use of tobacco products. Simultaneously an epidemiological transition is occurring in these countries. This is evident by a shift away from the high prevalence of infectious disease and undernutrition as causes of mortality to a high prevalence of chronic and degenerative diseases - conditions made worse by the nutrition transition. This raises urgent concerns regarding prevention of the already burdensome and growing epidemic of CHD in these countries (Figure 8) because LBW, especially in association with increased body fat, either as an adult or as a child, leads to insulin resistance, and an increased risk of CHD. Regardless of the controversy over the foetal origins theory, the fact remains that the effects of malnourishment at different stages of gestation are poorly understood.41 The foetal origins theory leaves the scientific community with unanswered questions, although waiting for these and other answers should not delay the programme implementation of those interventions that have already been shown to be, or are likely to be, efficacious against low birthweight.57 The role of adequate pre-pregnancy weight has been established as a determinant of LBW in developing countries, so improvement in nutrition of young girls and women is very probably one important step toward the prevention of LBW and its accompanying disease burden.
Figure 8. Coronary heart disease probability of death ages 15-60 years
Source: Murray & Lopez (1994) Bull WHO 72:447.
* estimated numbers of deaths in thousands (1990 - sexes combined)