Protein-Energy Malnutrition

J D L Hansen. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas. Volume 1. Cambridge, UK: Cambridge University Press, 2000.

Protein-energy malnutrition (PEM) is the current term for a group of nutritional diseases related to dietary protein and energy (calorie) intake. These diseases are most frequently seen in infants and young children in developing countries but may be a feature of famine or the result of illness for people of all ages throughout the world. Research during the twentieth century has considerably clarified the causes and manifestations of what are now known as dietary-related effects on the growing or mature individual. PEM includes conditions known in the medical world as kwashiorkor, marasmus, and growth retardation in children. Related to PEM are pellagra, starvation, and protein malnutrition. Infection, debilitating disease, and surgical procedures are frequently complicated by PEM. It is, therefore, a factor of importance in determining morbidity and mortality and has to be taken into account by health-care personnel at all levels.

Historical Concepts of Protein and Energy

Early reports of what may have been PEM lack the clinical, pathological, and biochemical details that make identification certain. The history of PEM is thus confined to the nineteenth and twentieth centuries, and it is only in the last 50 years that clarification of the various forms that PEM can manifest has emerged.

In his book Protein and Energy, Kenneth J. Carpenter has provided a detailed survey of nutritional science as it was known in the period from 1614 to 1893 (Carpenter 1994: 1-99). Of particular interest in relation to later discoveries is that the first “balance studies” were carried out by Italian scientist S. Santo-rio in 1614. He weighed his food and drink as well as his excreta (urine and feces) and measured changes in his own weight. There was an unexplained daily disappearance of 5 pounds of material that he attributed to a breakdown of body tissue that was then secreted through the skin as insensible perspiration; the losses were made good by the nourishment ingested. This was only a more quantitative restatement of Galen’s view in the second century that “[o]ur bodies are dissipated by the transpiration that takes place through the pores in our skins that are invisible to us; therefore we need food in a quantity proportionate to the quantity transpired” (Carpenter 1994: 1).

Anton Lavoisier’s work in the late eighteenth century (1770 to 1790) made him the “Father of Nutritional Science.” His first contribution was the recognition of the distinction between compounds that could change their character and simple substances or elements (e.g., carbon, hydrogen, nitrogen, oxygen, and others). His second contribution was an understanding that combustion and respiration involved similar processes of oxidation that could explain the phenomenon of “animal heat.”

Among those who followed Lavoisier was Jean-Baptiste Boussingault, who published the first table of the nitrogen content of foods in 1836. The protein radical was discovered just two years later by Gerrit Mulder and was considered to be the essential ingredient for both body building and physical activity. By the end of the nineteenth century, however, it was realized that protein was not the main or obligatory source of energy and that it is the oxidation of carbohydrates and fatty acids on which we rely for continued physical work.

Definition and Nomenclature

Description of the various syndromes that we now include in PEM began around 1850, and in the subsequent hundred years, much confusion in terminology arose. However, E. Kerpel-Fronius, a well-known Hungarian pediatric investigator who bridged both the pre- and post-1950 eras, clarified the old and new nomenclature, which has made it possible to identify references in early case reports (Kerpel-Fronius 1983: 30-4). He showed that the contradictions in terminology were rooted in regional differences, age of weaning, local foods, and prevalence of infections, and he classified the various types of malnutrition we now call PEM in the following way:

  1. Hypoalbuminemic forms (low serum proteins)
  2. Edematous
  3. Kwashiorkor
  4. Mehlnährschaden
  5. Hypoalbuminemic forms without edema
  6. Dry forms without hypoalbuminemia
  7. Underweight (dystrophic infants)
  8. Stationary stage
  9. Repairing stage with retardation in height (stunted infants)
  10. Marasmus (atrophy)
  11. Moderately severe
  12. Severe forms
  13. Severest form in young infants (athrepsia or, decomposition in the classic texts of pediatrics)

Reference to this classification assists considerably in identifying in older literature the various forms of what is now known as PEM. A current definition of PEM is the Wellcome classification shown in Table IV.D.7.1, which refers to children (Wellcome Trust Working Party 1970), whereas a broad current understanding of the nutritional basis of PEM is as follows:

  1. Protein quantity and/or quality (amino acid pattern of the protein) intake that is below the minimal requirements for growth and health, with or without an energy intake that is less than energy expenditure on muscle activity, heat production, growth, and other energy requirements.
  2. Excessive loss of protein and energy in diarrhea and acute and chronic diseases. (Waterlow 1992: 152-8)

PEM can, of course, be complicated by mineral deficiencies (e.g., of sodium, potassium, calcium, phosphorus, or iron), by deficiencies of trace elements (e.g., zinc or chromium), and by vitamin deficiencies (e.g., of vitamins A, D, C, or K). Much of this understanding has come about through intensive worldwide research during the last 50 years, which is discussed in the remainder of this chapter. We begin with marasmus.

Marasmus 1850 to 1950

“Marasmus” means “wasting away of the body” (from the Greek marasmos) and is a term applied mainly to infants and children (the term “wasting” is employed for similarly afflicted adults) (Figure IV.D.7.1). It occurs when the diet is grossly deficient in energy. Such a diet also necessarily fails to meet protein requirements. Marasmus may become manifest in wholly breast-fed infants when the milk is quantitatively insufficient, but more frequently it occurs after early weaning to diluted or low-energy bottle feeds or cereal paps. In the age group from 1 to 5 years, marasmus occurs when food of any kind is in short supply, as in conditions of war, civil unrest, famine, extreme poverty, or just lack of care for the child. Often, it is produced by starvation that occurs during treatment of diarrhea or other infections, but it can also stem from severe weight loss brought on by chronic pyogenic disease, tuberculosis, syphilis, AIDS, and tropical infestations.

The presenting symptoms are failure to thrive, irritable crying, or apathy. Diarrhea is frequent, and the child has a shrunken or wizened appearance, even though it is ravenously hungry. The degree of under-weight for age is extreme, and the muscles are weak and atrophic. Currently, marasmus is diagnosed when the weight is 60 percent or less of the 50th percentile of the National Center for Health Statistics (NCHS) standards. However, based on prewar literature, Kerpel-Fronius wrote that if the degree of wasting reaches 35 to 40 percent of the average body weight, recovery is impossible (Kerpel-Fronius 1947).

An interesting description of the clinical signs of marasmus was given in a speech by Charles Dickens at a fund-raising dinner for the Hospital for Sick Children on February 9, 1858. After discussing the tens of thousands of children who were dying because of poverty and sickness, he described a tour of the old town of Edinburgh, where lived some of the city’s worst-lodged inhabitants. In one of the most wretched dwellings, “there lay, in an old egg box, which the mother had begged from a shop, a little feeble, wasted, wan, sick child. With his little wasted face and his little hot, worn hands folded over on his breast, and his little bright, attentive eyes looking steadily at us” (Dickens 1956: 607).


Exact figures of the prevalence of marasmus are difficult to obtain because of the confusion of nomenclature and diagnosis. But marasmus must have been a leading cause of morbidity and mortality among infants and preschool children during the latter half of the nineteenth century and the early part of the twentieth century—and in Europe and North America as well as the rest of the world. A large percentage of the victims were illegitimate foundlings, who represented from 15 to 45 percent of all newborn infants in most European capitals. In the poor hygienic conditions of the “foundlings’ homes,” death claimed between 30 and 90 percent during the first year of life.

Similar rates of mortality occurred in North American institutions. H. D. Chapin (1915), examining 11 foundling institutions in New York, discovered a death rate of 422 per 1,000 for children under 2 years of age—compared with a community-wide figure of only 87 per 1,000. In 1921, Oscar M. Schloss, the author of the annual report of the Infants’ Hospital in Boston, called attention to “the frequent relationships of both malnutrition and acute infections to infant deaths” and stressed their importance as fields of research (Smith 1983: 138). Marasmus and atrophy do not appear as diagnoses in lists of patient admissions from the period 1883 to 1913, but figures for debility, dysentery, and diarrhea are high. In 1882, more than one-third of admissions were for “debility.”

With the two world wars, marasmus was especially prevalent in besieged cities of Europe, such as Warsaw, Leningrad, and Budapest. But by 1950, and since then, marasmus resulting from poverty and sickness has become very rare in the so-called developed world. It remains, however, a nutritional problem in the developing countries of Africa, the Middle East, South and Central America, and Southeast Asia (Bellamy 1996: 98).


From 1850 to 1950, numerous medical research reports on marasmus appeared, and many of these have been outlined, at least in part, in recent publications (Kerpel-Fronius 1983; Hansen 1991). Early in the twentieth century, German and French pediatricians stressed that the fate of undernourished infants depended on whether they escaped infection: It was thought that a rapid decline in the weight curve, ending in severe marasmus, could seldom be caused just by semistarvation (Kerpel-Fronius 1983: 64-8).

In a 1905 analysis of the bodies of marasmus victims and normal infants, a striking difference was discovered in fat content. Fat content in marasmic infants was only 1.7 percent of body weight, compared with 13.1 percent in normal babies. In 1920, body water in marasmic infants was found to be increased from 70 to 80 percent, and it was noted that the brain, kidneys, and skeleton were relatively preserved in marasmus.

Ten years earlier dextrose saline had been used for collapsed and dehydrated cases—including those of infantile atrophy, as recorded in an article about the activities of the Boston Floating Hospital (Beaven 1957). Losses of potassium, sodium, and chloride in diarrhea had been demonstrated by K. Schmidt in 1850, L. F. Meyer in 1910, L. Tobler in 1911, and I. Jun-dell in 1913 (Darrow 1946). In 1915, it was confirmed that increased water, fat, chloride, sodium, and potassium losses occurred in loose stools of infants, including those with marasmus (Holt, Courtney, and Fales 1915). The authors of this study pointed out that in therapy, potassium and magnesium were needed in addition to water and sodium, but it was not until 1946 (when it was proved that potassium could safely be added to intravenous fluids) that this advice was followed. This addition cut mortality from 32 to 6 percent (Darrow 1946; Govan and Darrow 1946).

Kwashiorkor to 1954

Marasmus was the first syndrome of what we now know as PEM to become the focus of wide interest, concern, and research, especially between 1850 and 1950. But from 1935 onward, kwashiorkor became another intensively studied and important nutritional disease. The term “kwashiorkor” was introduced by Cicely Williams, who wrote: “The name kwashiorkor indicates the disease the deposed baby gets when the next one is born, and is the local name in the Gold Coast for a nutritional disease of children, associated with a maize diet” (1935: 1151-2). Williams explained (1973) that the word comes from the Ga language of Accra, Ghana, and J. C. Waterlow (1991) has identified at least 31 distinct vernacular names in tropical Africa. Other regions also have words that doubtless frequently mean kwashiorkor. Among the common names in English, for example, are nutritional dystrophy, infantile pellagra, nutritional edema, and wet marasmus. In Spanish-speaking countries, kwashiorkor was referred to as a multiple deficiency state—distrofia pluricarencial.

Before naming the disease, Williams had published a paper giving its clinical description (Williams 1933). Afflicted infants were of preschool age (1 to 4), and their diet generally involved breast feeding with supplementary feeds of maize paps low in protein content. On examination, there was edema, wasting diarrhea, sores of mucous membranes, desquamation of the skin on the legs and forearms, and a fatty liver. The disease was uniformly fatal unless treated. This description provided by Williams fits perfectly with our current clinical definition of kwashiorkor, except that since 1933, retardation of growth has been more emphasized and there is now detailed knowledge of the changes in function of various systems and organs (Hansen and Pettifor 1991).

In a 1952 World Health Organization (WHO) report, the name kwashiorkor was applied to the syndrome in Africa, and the relationship of the disease to a low-protein diet was firmly established (Brock and Autret 1952). This study and report had been initiated by the Joint FAO (Food and Agriculture Organization)/WHO Expert Committee on Nutrition at its first session in Geneva in October 1949. The committee had found that one of the most widespread nutritional disorders in tropical and subtropical areas was an ill-defined syndrome known by various names such as kwashiorkor, malignant malnutrition, polydeficiency disease, and so forth. It was resolved that WHO conduct an inquiry into the various features of kwashiorkor, and, subsequently, the FAO was asked to cooperate. J. F. Brock (WHO) and M. Autret (FAO) traveled extensively throughout Africa over a period of two months in 1950, after which they concluded “that kwashiorkor is the most serious and widespread nutritional disorder known to medical and nutritional science” (Brock and Autret 1952: 72).

In subsequent years, FAO and WHO sponsored other studies in Central America, Brazil, and southern India, and research units in various parts of the world began to concentrate their efforts on determining the etiology, pathogenesis, treatment, and prevention of kwashiorkor.

The similarities and differences between marasmus and kwashiorkor soon became evident and gave rise to intensive debate. D. B. Jelliffe proposed the term “protein calorie (energy) malnutrition” (PEM) to cover the spectrum of syndromes that range from marasmus to kwashiorkor (Jelliffe 1959). This concept was a major contribution in understanding the variations of this group of nutritional diseases.

A history of kwashiorkor appeared in 1954 (Trowell, Davies, and Dean 1954). It discussed an early description of the disease, written by J. P. Correa in 1908 in the Yucatan. Waterlow, in his recent article on the history of kwashiorkor, found an even earlier description from Mexico by F. Hinojosa in 1865 (Waterlow 1991: 236). H. C.Trowell and his colleagues (1954) made it apparent that up to 1954, kwashiorkor had a worldwide distribution. They supported their contention by listing approximately 250 publications from Africa, Asia, Europe, North and Central America, and South America that contained details on established or probable cases of the disease (Trowell 1954). Most of these dated from between 1920 and 1950, and some early reports from German workers on Mehlnährschaden are of especial interest (Czerny and Keller 1925-8). This term, which is best translated as “damaged by cereal flours,” was used to indicate infant malnutrition resulting from imbalanced (excess starch) feeding habits, and the researchers described a clinical picture similar to that of kwashiorkor. In Germany at the time, it had become popular to use cereal gruels instead of milk when a child had gastrointestinal difficulties. The gruels were usually made without milk, and the return of loose stools when milk was once again added to the diet was too often regarded as an indication of sensitivity to milk. The many similarities between Mehlnährschaden and kwashiorkor in terms of etiology, pathology, and treatment suggest that these German researchers—in a developed country—may well have been describing kwashiorkor in the early part of the century.

In their book on kwashiorkor, Trowell and colleagues (1954) present a fascinating historical description of the puzzling features of the disease that had confounded investigators and triggered controversies. Among these were the similarities of kwashiorkor to pellagra, the fact that many cases did not have skin rashes, that fatty livers were found in nearly all cases, the role intestinal parasites might play in the disease, and the various forms of treatment. However, what emerged clearly was that lack of protein in the diet was an important cause—a lack caused by dependency on foods that supply adequate carbohydrate but little protein, such as cassava and cereal foods like rice, corn, millet, and sorghum. A distinction between kwashiorkor and infantile pellagra was based on a difference in the distribution of dermatosis (in kwashiorkor the diaper area; in pellagra on exposed areas such as the face, hands, and feet) and a failure of kwashiorkor victims to respond to nicotinic acid unless a high-protein diet was simultaneously given (Trowell et al. 1954: 118-19).

The treatment recommended for kwashiorkor in 1954 was cow’s milk, given in a concentrated form with little lactose and less fat. However, the exact nature of the factor(s) responsible for bringing about improvement in the children’s condition had not been determined. Serum biochemistry showed that treatment increased serum albumin, cholesterol, non-specific esterase, and cholinesterase. But the completeness of recovery and the ultimate prognosis could not at that time be assessed. Moreover, the authors felt that many children suffered from a mild form of kwashiorkor that could not be accurately defined and needed much further investigation (Trowell et al. 1954).

Earlier, the distinction between marasmus and kwashiorkor in tropical countries had been made in an important monograph, in which children with fatty liver disease (kwashiorkor) were distinguished from those who were undernourished but had no evidence of liver damage (Waterlow 1948). The clinical manifestations of the second group were retarded growth and loss of body fat and resembled the condition known in Europe as infantile atrophy or marasmus. Cases of marasmus were mostly below 60 percent of expected weight for age—the current criterion for a marasmus diagnosis.

Kwashiorkor Since 1954

Following the authoritative reports of Brock and Autret (1952) and Trowell and colleagues (1954), there was intense research into PEM on a worldwide basis. This research was funded by international and national agencies now aware of the high infant and child morbidity and mortality occurring in developing or underdeveloped areas. In addition, academic institutions and individual researchers alike were stimulated to look into such questions. The results have been summarized in recent authoritative publications that have brought our knowledge of PEM up to date (Waterlow 1992; Carpenter 1994).

Dietary Treatment of Kwashiorkor

By 1954, it was clear that milk—as a source of protein—induced recovery, although mortality was still high in seriously ill cases. Brock (personal communication) posed the questions: What was it in milk that brought about recovery? Was it the protein in milk, and if so, what factors or amino acids of that protein initiated it? Were the other constituents of milk, such as the fat, carbohydrate, minerals, trace elements, vitamins, or as yet unknown factors, important for recovery?

A series of clinical trials and balance studies, conducted at Cape Town, South Africa (a nontropical area), from 1953 to 1956, concentrated on what exactly brought about the cure for kwashiorkor and established that a vitamin-free synthetic diet of 11 mostly essential amino acids, glucose, and a mineral mixture could cure the skin rashes, regenerate serum albumin concentration, improve appetite, and eliminate the edema suffered by kwashiorkor victims (Hansen, Howe, and Brock 1956). It was further shown that potassium deficiency as a result of diarrhea and poor intake of potassium-containing foods was an important cause of edema. In fact, edema could resolve without change in serum albumin concentration if potassium depletion was corrected (Hansen 1956).

These studies ended the mystery of what milk contained that initiated recovery from the disease by establishing that protein (amino acid) deficiency was an essential feature of kwashiorkor and that there was no unknown factor involved. Nonetheless, energy deficit was subsequently emphasized by many authors, but often children with kwashiorkor had enjoyed adequate energy intake and, as with all nutritional deficiency disorders, concurrent vitamin, mineral, and trace-element deficiencies can cause added complications. Milk contains all of these elements except for iron and was thus shown to be an ideal food with which to treat children suffering from kwashiorkor.

Because milk is not universally available, however, there was much interest during the 1950s in the question of whether plant proteins could provide a satisfactory substitute. Nitrogen balance studies during recovery from kwashiorkor (Hansen et al. 1960) revealed that nitrogen (protein) retention was very strong with a milk diet. On a maize (corn) diet, nitrogen retention was much less, but it was greatly improved by the addition of the amino acids missing in maize (lysine and tryptophan), or a legume (pea flour), fish flour, or milk (Hansen 1961). The Institute for Nutrition in Central America and Panama (INCAP) successfully developed a mixture of corn, sorghum, cottonseed flour, and yeast that had good results. It was commercially produced as “Incaparina” in Guatemala and other Central American countries, but it became too expensive for the people who needed it most (Carpenter 1994: 173-5). The extensive study and work with “Incaparina” did, however, prove that commercial vegetable mixtures could be used as a weaning food to promote growth and prevent kwashiorkor.

Between 1955 and 1975 there were numerous other efforts to find substitutes for milk (Carpenter 1994: 161-79). A Protein Advisory Group (PAG) was established by WHO in 1955 and subsequently supported by FAO and the United Nations International Children’s Emergency Fund (UNICEF) to stimulate worldwide research into high-protein foods that might close the so-called protein gap between developed and underdeveloped countries. However, in 1974, a challenging article by D. S. McLaren called into question the importance of protein in the prevention of PEM and stressed that energy depletion was at least as important as lack of protein, if not more so. He argued that marasmus was a more widespread disease than kwashiorkor and that too much emphasis had been placed on—and too much money invested in—the production of protein-rich food mixtures,” … whilst children were lost in the unchecked scourge of malnutrition” (McLaren 1974: 95). Although much controversy followed this article (Carpenter 1994: 180-203), the emphasis on the production of high-protein foods waned, and interest became focused on improving food quantity rather than quality (Waterlow and Payne 1975).

Unfortunately, the debate that continues to the present day on the relative importance of protein and energy has often lost sight of the earlier concepts of kwashiorkor and marasmus. Marasmus implied wasting from overall energy lack or starvation, whereas kwashiorkor was characterized by a low-protein diet that frequently had an adequate energy component. In between the two extremes is marasmic kwashiorkor, which has features of both. Milder cases that manifest only growth retardation (Table IV.D.7.1) can result from a lack of either protein or energy or a combination of the two. A current explanation of the dietary background of PEM is that variations in energy intake, total protein intake, and “quality” of protein (the amino acid pattern) are responsible for the individual clinical forms of PEM.

A protein intake of less than the minimal requirements will result in low serum proteins (hypoalbuminemia) and failure of growth, even in the presence of adequate energy intake (Hansen 1990). Unfortunately, in many parts of the world, most of the dietary protein comes from a single source, often a cereal. Cereals have the disadvantage of being low in total protein content and lacking in essential amino acids, such as lysine (in the case of wheat) or both lysine and tryptophan (in the case of maize). Populations subsisting solely on these foods are thus at risk of energy and protein depletion, and children in particular are at risk of PEM in one form or another. Inevitably, vitamin, mineral, and trace-element deficiencies can complicate PEM in varying degrees, as does infection.

In a review of much new work on protein and energy requirements, Waterlow (1992: 229-59) has concluded that contrary to much that has been published in recent years, some weaning diets in developing countries contain marginal amounts of protein, even when consumed in quantities that satisfy children’s energy needs. Such a marginal diet may satisfy the protein needs of many—perhaps most—children, but not all. Any group of children, as of adults, appears to have a range of protein (and energy) requirements. On marginal intakes, children at the upper end of the range will be at risk. This does not conflict with a controversial finding (Gopalan 1968) that there was no difference, quantitative or qualitative, between the diets of children who developed kwashiorkor or marasmus.

The Liver and Kwashiorkor

A well-described characteristic of kwashiorkor is a fatty infiltration of the liver (Williams 1933). As already mentioned, Waterlow even described kwashiorkor as “fatty liver disease” and distinguished it from cases of undernourishment (marasmus) that showed no evidence of fatty infiltration (Waterlow 1948). In 1945, the Gillman brothers in South Africa published a paper on the successful treatment of fatty liver with a powdered stomach extract, “Ventriculan” (Gillman and Gillman 1945). They referred to their patients as cases of infantile pellagra and, using the liver biopsy technique, observed that such infants had greater or lesser amounts of fat in the liver (without infection) and that the fat accumulation resolved with successful treatment. The Gillmans felt that “Ventriculan” supplied an essential substance, but other investigators could not confirm their findings and speculated that it was the protein in the diet that was producing the cure.

At that time, there was a high prevalence of cirrhosis of the liver in Africa, and it was thought that suffering PEM early in life might be an underlying cause. This suggestion was refuted, however, by a five-year follow-up study of kwashiorkor cases in a non-tropical environment, which demonstrated that there was complete recovery of the liver with no residual cirrhosis (Suckling and Campbell 1957). In 1969, it was found that there was a connection between fatty liver and serum lipoprotein concentrations (Truswell et al. 1969), and it was hypothesized that fat accumulates in the liver because of the failure of fat transport out of the liver—a failure resulting from the impaired synthesis of apolipoprotein B, which, in turn, is a consequence of protein deficiency. This remains the current view (Waterlow 1992: 61-5).

PEM and Infection

A classic work on the synergistic association of malnutrition and infection was published in 1968 (Scrimshaw, Taylor, and Gordon 1968). In the case of PEM, the condition predisposes to other diseases, but diseases can also bring on the condition. For example, children with PEM are particularly susceptible to respiratory and gastrointestinal infections, whereas measles frequently precipitates severe PEM. An intriguing question is whether PEM interferes specifically with the protective immune responses or whether the generally poor environmental conditions associated with PEM (which implies frequent exposure to infection) means the child with PEM has less metabolic reserve to resist infection. For example, although exposure to measles will infect a well-nourished child and a child with PEM equally, the well-nourished child will survive with some weight loss, whereas the child with PEM, already under-weight, becomes severely ill and frequently dies.

The various and complex ways in which immunity to infection can be impaired by PEM has been recently reviewed (Waterlow 1992: 290-324). Although in some communities the relationship between PEM and infection is linear, in others there is a much weaker association. But what is clear is that a child with severe PEM is seriously at risk of infection in any community. An early observer of reduction of cell-mediated immunity in PEM (Smythe et al. 1971) also noted reduction in the weight of the thymus gland as well as a reduced size of the spleen, lymph nodes, tonsils, appendix, and Peyer’s patches.

Much interest has also been shown recently in the role of vitamin A deficiency in the susceptibility of PEM victims to infection, especially respiratory disease and diarrhea (Sommer, Katz, and Tarwotjo 1984). Earlier, it was demonstrated that some patients with kwashiorkor had dangerously low levels of plasma vitamin A (Konno et al. 1968), and treatment of measles with large doses of vitamin A has given good results (Hussey and Klein 1990), which, in conjunction with widespread immunization, means that measles is no longer the threat to the life of PEM victims that it was in 1969 (Morley 1969).

Diarrhea has always been a clinical characteristic of kwashiorkor and marasmus, both as a precipitating factor in a marginally malnourished child and as a continuing recovery-retarding drain of electrolytes, energy, and protein. The organisms and viruses responsible have been well defined (Waterlow 1992: 297), but in the case of PEM victims, frequently no pathogens are isolated. Balance studies on patients recovering from kwashiorkor have revealed a remarkably high daily fecal loss (500 to 1,000 grams [g] per day compared with a normal figure of 100-150 g/day), part of which was found to have been caused by lactose intolerance as a result of secondary lactase deficiency in the duodenum (Bowie, Brinkman, and Hansen 1965; Bowie, Barbezat, and Hansen 1967).

At the time, this discovery was thought to be a breakthrough in the cause of the diarrhea in PEM; further experience, however, revealed that lactose intolerance is not universal in PEM, although it can explain the severe diarrhea that frequently occurs when PEM cases are treated with milk. Diarrhea also occurs in PEM because the gastrointestinal tract atrophies and becomes paper-thin and almost transparent. The mucosa of the intestine has a reduced absorptive surface, and electron microscopy reveals considerable disorganization of the intracellular architecture (Shiner, Redmond, and Hansen 1973). Marked improvement occurs within a few days of treatment.

Looking back on the last 50 years of research on diarrhea in PEM, it is apparent that infection, intestinal atrophy, lactose intolerance, and immunological deficiencies all play their part. Recently, the advent of AIDS has particularly affected the immunological defenses in infected children, resulting in diarrhea, severe wasting, and marasmus or kwashiorkor. In a summing up of all the recent evidence, it can be said that there is a causal relationship between a state of malnutrition (PEM) and diarrhea morbidity and mortality (Waterlow 1992: 313, 339). The same may be said for respiratory disease (pneumonia) and measles, but not for malaria, which has little or no relation to the state of nutrition (Waterlow 1992: 333). Confounding factors in morbidity and mortality are vitamin A deficiency, breast feeding, sanitary facilities, and the mother’s education, caring capacity, and availability.

PEM and Body Composition

The profound physical changes in marasmus (wasting) and kwashiorkor (edema) stimulated research into body composition when new techniques became available after 1950. In Waterlow’s (1992) extensive review of the subject, he shows the inconsistencies between different studies and points out that there is still no agreed-upon understanding of the mechanisms of fluid-retention edema. There is a considerable loss of muscle mass and of fat, particularly in marasmus, and as a result, there is an increase of total body water as a percentage of body weight both in kwashiorkor with edema and in marasmus without edema.

Based on evidence available so far, the difference between the two could be that children with edema have more extracellular fluid, which is probably related to the extent of potassium depletion (Mann, Bowie, and Hansen 1972). Kwashiorkor children with edema have lower total body potassium than marasmus cases without edema. It is of interest that the increase of total body water—and of extracellular water as a percentage of body weight—represents a reversion to an earlier stage of development. This means that the weanling child with PEM has the composition and size of a younger child (Hansen, Brinkman, and Bowie 1965). Total body protein is severely depleted in PEM victims, and compared with normal children of the same height, there is a greater deficit of total protein than of body weight. Cellular protein is greatly depleted, but collagen (structural protein) is little affected (Picou, Halliday, and Garrow 1966). The brain is relatively well preserved when compared with other organs in PEM. However, computed tomography has recently shown there is some reversible shrinkage of brain mass (Househam and De Villiers 1987).

PEM in General

Growth Retardation and PEM

A constant feature of kwashiorkor has been growth retardation, occurring in children even before the disease is recognized. Weight, height, and bone development are all affected. In the second half of the twentieth century, anthropometric indices—weight, height, weight-for-height, arm circumference, and skinfold thickness—have been greatly refined and used extensively in the assessment of health and disease. In the Wellcome classification of PEM (Table IV.D.7.1), weight is used as a basis of defining differences between the various syndromes, which has proved most valuable in comparing PEM in different communities and countries.

F. Gomez, R. Ramos-Galvan, S. Frenk, and their colleagues in Mexico were the first to divide deficits in weight-for-age into three categories of severity, based on Harvard growth standards (Gomez et al. 1956). This classification had the drawback that it combined in one number the figures for height-for-age and weight-for-height (Waterlow 1992: 189) and was not widely adopted. The Harvard growth charts for height and weight of children in the United States proved valuable as a standard against which to compare the growth of children with PEM. They were used in the initial Wellcome classification of PEM (Table IV.D.7.1), although later they were superseded by charts from the U.S. National Center for Health Statistics (NCHS), which—though similar—were more thoroughly worked out in terms of statistics. These charts were accepted and published by WHO as international standards (WHO 1983).

There have, however, been controversies concerning the use of these standards, ranging from questions about possible ethnic and environmental influences on growth to debates over the desirability of using national growth charts. Yet environmental and income differences have enormous effects in local surveys (Wittmann et al. 1967), which makes the use of local standards impractical, especially as there is often a secular trend toward improvement in disadvantaged groups. Currently, for the community and individual assessment of children with PEM, cutoff points on the international charts are employed; for example, children who are below the third percentile, or 2 standard deviations below the mean figures for height, weight, or weight-for-height indices, are suspect for PEM (Waterlow 1992: 212-28).

In clinic and field assessments, the “Road to Health” weight charts have revolutionized preschool health assessment even among the most unsophisticated populations. These charts are issued to mothers at the birth of their children and updated at each visit to a primary care center, physician, or hospital. D. Morley (1973) had demonstrated the value of continuously monitoring weight gain and of making the mother responsible for keeping the record. This is an interesting example of the practical procedures that grew out of the early observations of the growth of African children with kwashiorkor (Trowell et al. 1954: 70-3).

The retardation of growth caused by kwashiorkor immediately raised questions about its reversibility. Provided there were improvements in nutrition and environment, could a preschool child regain his or her genetic potential for growth? An early prospective study, started in 1959, monitored children admitted to hospitals with kwashiorkor for a subsequent period of 15 years (Bowie et al. 1980). This study showed that growth retardation resulting from severe PEM is reversible if environment and food intake are reasonably adequate during the prepubertal years. In similar studies, the increment in height was very much the same as that achieved by children in the United States, regardless of the degree of stunting at 5 years (Waterlow 1992: 195-211). As Waterlow has discussed, however, where catch-up has not occurred, stunting caused by long-continued protein and energy lack can lead to functional consequences, for example, a reduction in absolute capacity for physical work. For a given workload, people who are small, even though fit, are at a disadvantage. Thus, research on body growth in PEM has gone on to stimulate investigation into many interesting aspects of human development and function.

PEM and Intellectual Development

In 1963, a pioneering follow-up study of marasmic infants in South Africa focused on the possible effects of PEM on brain size and intellectual development (Stoch and Smythe 1963). This sparked ongoing worldwide research and intense political interest. Decreased brain weight in a state of malnutrition was reported from East Africa in 1965 (Brown 1965), and, as mentioned earlier, computed tomography has demonstrated that there is a reversible shrinkage of brain mass in kwashiorkor (Househam and De Villiers 1987). The more difficult assessment of the effect of PEM on intellectual development is confused by the interaction of nutrition per se and other environmental factors such as poor social conditions, nurturing, education, and environmental stimulation. In addition to protein and energy, other nutrients—iron, potassium, trace elements, or vitamins, to name a few—may be deficient, and this, too, may affect mental development (Grantham-McGregor 1992).

There is, however, some evidence that PEM does not necessarily cause permanent damage to the intellect. Follow-up studies have failed to demonstrate differences in intellectual development between exkwashiorkor patients, their nonaffected siblings, and other children from the same environment (Evans, Moodie, and Hansen 1971). Planned stimulation in hospital has produced recovery of cognitive development (to a normal level) in kwashiorkor children (Cravioto and Arrieta 1979: 899), and a recent study in Jamaica showed that a nutritional supplement provided significant benefits to stunted children between 9 and 24 months of age, as did stimulation alone—but the two together had the best result (Grantham-McGregor 1992).

A WHO symposium in 1974 concluded that in spite of the widely held opinion that PEM in early life permanently jeopardizes mental development, the evidence to support this contention was scanty. Twenty-three years later, this still appears to be the case, but it does seem probable that there is an interaction between malnutrition and other environmental factors, especially social stimulation, and that a child’s intellectual status is the result of this interaction. However, good nutrition in the first two years of life enables an underprivileged child to make better use of what stimulus there is in the environment. This kind of beginning has a long-lasting effect on intelligence even if nutrition after this period is less than optimal (Evans et al. 1980).

The Social Background of PEM

Trowell and colleagues (1954: 49-51) pointed out that PEM has always been associated with poor communities and occurs among the most depressed social classes. These researchers also mentioned social customs at the time of weaning, and of course Williams interpreted the word kwashiorkor to mean the “disease of the displaced child”—displaced by the next baby (Williams 1935). In addition to these basic factors, overpopulation and the movement of rural people to urban areas were considered important. A. Moodie (1982), a social worker who devoted her working life to the study of the background of PEM, as seen in a nontropical area of Africa, has reviewed the literature, including some of her own studies. She noted the following as constituting the essential background of PEM:

  1. Economic inadequacy or poverty—rural or urban.
  2. Lack of sophistication and knowledge loosely termed ignorance and the cultural factors underlying this state.
  3. Problems of overpopulation and too large families. (Here the mother suffers and it is through the mother that child nutrition is mediated. In all programs for prevention the well-being of the mother should be a priority.)
  4. Social disorganization, especially illegitimacy and alcohol abuse.
  5. High incidence of infection and diarrhea.
  6. Early weaning and decline of breast feeding.

In an urban study funded by WHO, low income was found to be critical (Wittmann et al. 1967). In another investigation, sociological and cultural factors such as the working mother appeared to be more important (Shuenyane et al. 1977). There have been many studies throughout the world showing what Waterlow has termed the multiplicity of causes. He grouped them under three headings: lack of food, infection, and psychosocial deprivation (Waterlow 1992: 9-11). It has been pointed out that economists and other planners are now recognizing that nutritional indicators (for example, growth retardation or the clinical features of PEM) provide a more sensitive, objective, and easily collected measure of socioeconomic development than conventional indicators such as per capita income (Church, Burgess, and Burgess 1991: 569-70).

Prevention of PEM

With the high prevalence of PEM in some areas of developing countries (marasmus afflicts 5 to 10 percent and stunting 30 to 60 percent of the under-5 population), prevention has received much attention from international and government agencies. These organizations have been aided by economists and social scientists as well as by nutritionists and health scientists, and the literature has grown enormously (Waterlow 1992: 361-92).

The basic strategy for achieving “health for all” by the year 2000 is Primary Health Care (PHC). This plan includes growth monitoring, health education, maternal and child health care, family planning, immunization against major infantile diseases, and appropriate treatment of common diseases, like oral rehydration treatment for diarrhea. There is no doubt that these measures, implemented by PHC teams, have had an overall effect on preventing PEM, which has shown a decline in prevalence in many countries (Bellamy 1996: 54). The details of nutrition inputs in primary care are well described in a recent publication (Church et al. 1991). The problem of PEM needs a holistic approach. Health teams and public-health authorities can reduce mortality and morbidity with active programs of clinical care, rehabilitation, and food supplementation, but they cannot affect the prevalence of PEM resulting from underlying socioeconomic and cultural realities. Physicians must have the active, integrated, and effective cooperation and assistance of economists, agriculturalists, and governments to eradicate PEM.

Treatment of PEM

In a remarkable but little-quoted paper, published in an obscure journal, skimmed lactic acid milk was shown to reduce the 40 to 60 percent mortality of severe kwashiorkor to 20 percent (Altmann 1948). The author stressed the importance of small feeds initially until appetite returned. He also stressed the danger of severe dehydration that causes 5 out of 6 of the deaths. A colleague at the same hospital later found that the use of intravenous fluids in dehydrated cases cut mortality by half (Kahn 1959: 161-5). The same principles of feeding and rehydration are followed today (Waterlow 1992: 164-86), with perhaps more emphasis on oral rehydration. Refinements include supplements of potassium, magnesium, zinc, vitamin A, folic acid, and iron. Blood transfusion is used only for very severely anemic children. Because of the frequency of infections, antibiotics are routinely given and, in tropical areas, malaria and other infestations have to be dealt with, using appropriate therapy and prophylaxis. Mortality of severe PEM cases should now be less than 10 percent. Less severe cases of PEM respond well—with negligible mortality—to diets providing adequate energy and protein.

History of the Cause of PEM

Waterlow (1992) has concluded that there is no reason to abandon the concepts put forward many years ago, namely, that kwashiorkor develops when the diet has a low protein-energy (P/E) ratio and that when energy is limiting, the end result is marasmus. To this it should be added that individual children vary in their requirements of nutrients. If the P/E ratio is marginal, protein or energy may be a limiting factor for some children but not for others. Research has shown that total protein intake and the quality of the protein (the amino acid pattern) are also important. In areas where PEM occurs, vitamin, mineral, and trace-element deficiencies can complicate the basic syndrome to varying degrees, as does infection.

Recently, it has been hypothesized that all the serious features of kwashiorkor—edema, fatty liver, infection, and mortality—can be explained by an excess of free radicals (Golden and Ramdath 1987). This theory has not yet been firmly established, and in any case, children still have to be short of protein or energy before they become susceptible to free-radical excess. Another theory is that kwashiorkor results from aflatoxin poisoning (Hendrickse 1984). Aflatoxins come from fungi growing on improperly dried nuts. However, aflatoxins cannot be blamed for kwashiorkor and marasmus that occur where there is no aflatoxin contamination of the diet.

PEM in Perspective

In the nineteenth and twentieth centuries, PEM, as a particularly important nutritional deficiency in infants and children, has become recognized and better understood. This has come about through extensive worldwide observation and research, and recent emphasis has been on the effect of PEM on health, growth, and intellectual development.

As we enter the twenty-first century, we have the knowledge, if not always the means, to limit the prevalence of PEM in individuals and communities at risk. A looming danger is the “demographic trap,” the situation that arises when population growth exceeds growth in food production or availability (King and Elliot 1994: 528). Signs of the trap are already present in several countries, particularly in Africa (Bonneux 1994). If this situation is not addressed by fertility control, PEM is likely to remain with us for the foreseeable future.