Commentary: The catch-up dilemma— relevance of Leitch's ‘low–high’ pig to child growth in developing countries

Isabella Leitch's paper1 is exciting reading in many ways. Our main difficulty in writing this commentary was choosing a single topic from her several original ideas that have major relevance to the nutritional problems of developing countries today. We opted to discuss the possible hazards of catch-up growth. Leitch's analogy of the growth of pigs and humans is thought-provoking. Based on research on pigs who were starved early in life and then fed appropriately, she observed that ‘skeleton and muscle will not grow as they would have done if they had had the opportunity at the right time, and the extra food will be used mainly to lay on fat’. She then uses this analogy of the ‘low–high’ pig to propose that humans who suffered malnutrition in early life would be better off by remaining thin than by putting on weight: ‘growth potential is not something that one can speed up and slow down ... and get the same result in the end’. She concludes by arguing that this condition ‘can probably be prevented only by the continuation of a spartan regime throughout life, which seems a bit hard’. Leitch's ‘low–high human’ has short legs, since leg length is more heavily compromised by malnutrition than trunk (or crown-rump) length. Her discussion of the aesthetic appeal of long legs and of its associated health benefits is both amusing and scientifically challenging. Recent data suggest that low–high humans are indeed becoming increasingly prevalent, as obesity rapidly escalates in developing countries2,3 affecting individuals who also faced childhood malnutrition. Although small size at birth is not usually associated with greater adiposity in adults,4 children who are stunted in early childhood may have a greater risk of later obesity in countries undergoing the nutrition transition. Studies from Brazil, China, Russia and South Africa suggest that stunted children have a 2–8 times greater risk of becoming overweight.5 In Guatemala, stunted children are particularly prone to central fatness as adults.6 Leitch's low–high human might well be referred to as the ‘stunted–obese’ human. At this stage, it is useful to distinguish two types of catch-up growth. Early catch-up refers to children who are born small but catch up in weight and height in infancy or early childhood. Late catch-up affects children who may or may not be born small, but who became stunted in infancy or childhood and catch up later in weight but not in height, becoming ‘stunted– obese’. In less developed countries, about one in every four newborns is growth-retarded.7 Growth monitoring programmes provide nutritional advice to mothers whose children are underweight or not gaining weight as expected.8,9 For children aged 6 months or older, this implies providing additional energy-dense and micronutrient-rich foods to promote catch-up growth.10 Since measuring child length in first-level health facilities is problematic, weight-for-age is used to assess risk and progress. Growth promotion programmes do not differentiate early from late catch-up and provide feeding advice to all malnourished children under 5 years. They have been strongly endorsed by the international paediatric and nutrition communities, and by agencies such as WHO and UNICEF during the last 30 years. Leitch is concerned primarily with the long-term effects of late catch-up. Her hypothesis is consistent with the rapid increase in chronic disease that is already taking place in developing countries,11–13 and with the well-known detrimental effects of trunk and upper body fatness.14 However, recent literature arising from the Barker hypothesis15 suggests that early catch-up may also be detrimental. This is cause for much greater concern. Leitch's paper was visionary by anticipating, by three to four decades, what later became known as the Barker hypothesis.15 Studies summarized below further advance this hypothesis by suggesting that catch-up growth in childhood may affect adult health. Some of this evidence is summarized below. It should be noted that some studies do not separate catch-up growth in small newborns from above average growth in all children, nor test for interaction between birth size and later growth. A 1995 report from England16 suggested that fast growth in the first year of life was associated with lower coronary morbidity. However, at least four other studies reported opposite trends. In a Finnish cohort, boys who were thin at birth but whose weight caught up had increased coronary heart disease mortality;17 for girls, catching up in height rather than weight was associated with higher coronary morbidity and mortality.18 Another Finnish study showed that subjects who were in a higher quartile of body mass index at age 7 than their birthweight quartile appeared to have an increased risk (odds ratio = 2.3; 95% CI : 0.9–5.4) of metabolic syndrome—high blood pressure, dyslipidaemia and insulin resistance—in adulthood.19 In England, infants who gained more than 0.67 z-score of weight or length for age between birth and 2 years were fatter and had more central fat distribution at age 5 years.20 Other studies report on growth from birth to adulthood. In Wales, the highest risks of coronary heart disease were for subjects whose birthweight was in the lowest tertile and their adult body mass index in the highest tertile.21 In Sweden, men who were light at birth (<3250 g) but whose adult height was above the median presented higher blood pressure.22 Another relevant English study is the 60-year follow-up of the Boyd-Orr study, cited in Leitch's paper. Low-income British children who received food supplementation in the 1930s had a 13% (non-significant) increase in adult mortality compared to non-supplemented children. The study lacked statistical power to detect small but important differences in mortality.23 Several other publications addressed the long-term consequences of low birthweight on chronic diseases. However, as Lucas et al.24 pointed out, in many of these the association only become apparent when adult body size was adjusted for. This suggests that relative growth or change in size—and not small size at birth per se—was the factor associated with chronic disease, since: or Results of these studies suggest that early catch-up in both weight and height may affect adult health. Although the evidence supporting the fetal origins hypothesis is compelling, important methodological concerns have been raised24,25 and additional research is needed to provide solid ground for policy making, particularly regarding the effects of catch-up. We will now review some of the evidence that catch-up growth has short-term advantages. Although the literature on catch-up growth itself is limited, there is strong evidence on the association between malnutrition, morbidity and mortality. Weight-for-age deficits markedly increase the risk of death in children under 5 years. A meta-analysis26 showed that, compared to children with weight-for-age of 80% or more of the NCHS growth reference, the relative risks of dying were equal to 11 for children below 60% of the reference, 3 for children between 60 and 69%, and 2 for those between 70 and 79%. Reviews on nutrition and morbidity also confirm that malnourished children are at a much higher risk of developing serious infections.27–29 These findings have been interpreted as indicating that catch-up growth is associated with beneficial health outcomes, as it would shift children who are born small to a lower risk category. However, stronger evidence would arise by comparing children who (1) were born small and remained small, (2) were born small and caught up, (3) were born large and failed to grow, and (4) were born large and remained so. We are not aware of any studies reporting on morbidity or mortality outcomes for these four groups of children. Eighteen years ago, we started a population-based birth cohort study in Southern Brazil,30,31 and Leitch's paper led us to re-analyse these data. Children born in 1982 were followed up in 1984 at an average age of 20 months, and again in 1986. Being small for gestational age (SGA) was defined as a birthweight below the 10th centile of weight for gestational age,32 and hospital admissions during 1985 were obtained by maternal recall in 1986. Table 1 shows admission rates according to SGA status and to weight gain from birth until 1984. Hospitalizations were significantly less frequent for SGA children with above-average growth than for those with below average growth; the former even appeared to have lower rates than non-SGA children with slow growth. These preliminary results should be treated with caution, as there is a possibility of additional confounding as well as of Berkson bias, as doctors may be more likely to admit malnourished children. Summarizing, there is strong evidence that heavier infants and children present lower morbidity and mortality in developing countries, but data on the beneficial health effects of catch-up growth are limited. Ideally, one would like to have information on the short- and long-term consequences of catch-up growth from the same cohorts. When visiting a 25% sample of our cohort at the age of 15 years, we calculated the prevalences of overweight—above the 85th percentile of the sex and age-specific body mass index9—for the four groups of children in Table 1. These prevalences were 9.3% for SGA children with slow growth, 16.7% for those with fast growth, 16.0% for non-SGA children with slow growth and 27.2% for those with fast growth. This finding highlights the fact that early catch-up may be associated with both short-term benefits and with long-term risks. In 1950, Leitch raised the challenging hypothesis that individuals who failed to grow optimally in early life might be better off by remaining small than by catching up. Fifty years later, her early programming hypothesis is one of the most challenging areas of research in public health. Few nutritionists would disagree with Leitch regarding the harmful effects of late weight catch-up in stunted children, producing ‘stunted–obese’ adults with central body fat distribution. But there will be less consensus on whether or not early catch-up may be detrimental. Although Leitch did not go so far as proposing that fetal undernutrition would be best followed by relatively slow growth in infancy and childhood, some recent evidence suggests that early catch-up may also result in higher adult morbidity and mortality. On the other hand, the international nutrition and public health communities strongly support catch-up growth for small newborns, and there is ample, albeit mostly indirect, evidence that it is beneficial in the short term. Assuming that both the positive effects of catch-up on child health and its negative impact on adult health are confirmed, there are additional issues to be considered. Firstly, the trade-off between serious childhood infections and late-onset chronic diseases will depend on the epidemiological setting—in poor countries where infectious disease mortality is high, catch-up would have a positive net effect whereas in developed countries the risks of chronic diseases would probably outweigh the benefits of catch-up on child health. Secondly, economists have brought into public health the idea of ‘discounting’: most individuals prefer benefits in the short rather than in the long term.33 Therefore, a positive impact on child survival would have stronger appeal than a negative impact on adult mortality of similar magnitude. This further highlights how difficult it will be to weigh the pros and cons of catch-up growth. Thirdly, it remains to be seen whether it will ever be possible to implement selective dietary restriction to children who are small at birth. Leitch herself acknowledged that maintance of ‘a spartan regime throughout life’ would be ‘a bit hard’. Surely, promoting adequate fetal growth seems far more appropriate than recommending strict diets to children who were born small. Given current knowledge on child health in developing countries, it seems reasonable to continue to promote growth for small infants and young children. A major priority, however, is to re-analyse existing datasets from both developed and developing countries to assess the full impact of catch-up growth on health both on the short and long term. Let us follow what may be interpreted as Leicht's plea for epidemiological research: ‘Lay aside the microscope and look at things with the naked eye’. Frequency of hospital admissions in 1985 according to birthweight for gestational age and to weight gain from birth (1982) to the average age of 20 months (1984) Frequency of hospital admissions in 1985 according to birthweight for gestational age and to weight gain from birth (1982) to the average age of 20 months (1984) We would like to thank Bernardo Horta and Paulo Orlando Monteiro for their inputs to this commentary.