Malnutrition induces gut atrophy and increases hepatic fat infiltration: studies in a pig model of childhood malnutrition

Introduction

Malnutrition in developing countries affects 165 million children under 5 years of age [1]. These children exhibit stunted growth, and although the prevalence of stunting is decreasing, the actual number of stunted children in some parts of Africa, for example, is increasing due to population increase [2]. The most complicated form of malnutrition is severe acute malnutrition (SAM), which affects 20 million children under the age of 5 years and is estimated to result in approximately 1 million deaths every year [3].

SAM occurs in one of two forms, marasmus or kwashiorkor. Children with kwashiorkor exhibit muscle wasting, hair and skin changes, apathy and generalized edema, whereas marasmic children are characterized mainly by severe muscle wasting. The poor understanding of the fundamental biology of severe acute malnutrition in children may stem from the facts that most studies are cross-sectional and that prospective studies are generally not conducted for ethical reasons. In addition, as this particular patient group is extremely vulnerable, rapid advancement of understanding is complicated. Although the classical work by Keys et al. [4] remains some of the best information available on temporal changes in adult volunteers subjected to prolonged malnutrition, it may not adequately reflect temporal pathological changes in children.

While in malnourished adults, there is evidence of changes in body weight, hematology and clinical biochemistry, very little is known of changes in organ function and morphology. Therefore, to advance our understanding, it is essential to combine growth markers of malnutrition with estimations of organ function. Understanding organ function and morphology following prolonged malnutrition is a prerequisite to identifying optimal refeeding strategies that do not further compromise the clinical situation. This may be especially important to gut and liver function, as they are the keys to propdigestion and metabolism. Several animal models, including rats [5-8], dogs [9], monkeys [10-12] and pigs [13-16], have been used to study malnutrition. However, to the best of our knowledge, none have shown how temporal changes during malnutrition are associated with intestinal and hepatic function and morphology.

We sought to describe the course of malnutrition from a normal physiologic state to a state of SAM, including growth measurements and functional and structural changes in the intestine and liver. Changes in the function and structure of the small intestine and gut microbiota in particular have recently been suggested as causal factors in the development of kwashiorkor [17]. A shift in the gut microbiota induced by malnutrition may favor endotoxin serotypes more likely to cause edema and liver damage [18,19].

Materials and methods

Results

Discussion

We established a pig model of child malnutrition with clinical signs similar to children with SAM. Using a classification system of growth similar to that used for children, this study showed severe stunting in malnourished pigs, with a CRL of 70.8% relative to the CRL in the REFERENCE pigs. As the malnutrition in this particular study affected both body weight accretion and linear growth, the degree of wasting (i.e., body weight relative to CRL) indicated that two-thirds of the MAIZE pigs were normal to mildly wasted, while one-third was moderately to severely wasted, as severe wasting was defined as having a weight-for-height below 70% of that of the REFERENCE pigs (Table 2). A weight-for-height between 70% and 80% was classified as moderate wasting, and a weight-for-height above 80% was considered normal [24]. Although two-thirds of the MAIZE pigs still had a weight-for-height above 80% after 7 weeks, data from a separate study indicated that the weight-for-height continues to decrease if maize feeding is continued up to 10 weeks (Hother et al. unpublished results). In that study, all MAIZE pigs became moderately or severely wasted, as the weight-for-height for some was lower than 60%. However, in the present study, the piglets displayed characteristics of severe undernourishment and stunting with various degrees of wasting, classifying the model as moderately acute malnutrition with severe stunting.

In the malnourished group, there were clinical signs of anemia, and the hair coating was typical of animals that are not thriving, with long coarse hairs. The clinical signs of anemia were confirmed by decreased levels of hemoglobin, hematocrit and mean cell volume. Anemia is also commonly observed in children suffering from SAM [25]. The cause of microcytic anemia, as observed in the MAIZE pigs, was most likely iron deficiency, which is consistent with the low iron content of the diet (Table 1). However, anemia in kwashiorkor is usually normocytic and may reflect an anemia due to a lack of not only iron but also B vitamins [26-28]. Collectively, the clinical and hematological indices from the MAIZE pigs were more like children with marasmus than children with kwashiorkor, although some elements of kwashiorkor were present, such as hypoalbuminemia, wasting, stunting, anemia and fatty liver. We did not observe any signs of edema in the MAIZE pigs, which would have been a key feature of kwashiorkor.

Having validated the degree of malnutrition in this pig model via growth, biochemical and hematologic measurements, we were then interested in studying the association with functional and structural pathological changes in the intestine and liver. There was clear mucosal atrophy in both the proximal and distal small intestine, with blunt villi, as is observed in malnourished infants with tropical enteropathy. It was somewhat surprising that this clear differencein mucosal morphology was not paralleled by functional endpoints such as specific digestive enzyme activity. In general, no differences were observed in brush border enzyme activity between the MAIZE and REFERENCE pigs, except for decreased lactase and ApA activity in the distal part of the small intestine in the MAIZE group. However, the trend for the MAIZE pigs was to have lower levels of peptidase activity in both the proximal and distal parts of the small intestine. The notion that malnourished pigs may have reduced protein digestive capacity was indirectly supported by in vivo measurements of lower protein digestibility of the maize (68.2%) relative to table values (80-88%) (data not shown). Reduced enzyme activity and protein digestive capacity is a well-known phenomenon in anorexic post-weaning piglets and is thought to play an important role in post-weaning diarrhea, as undigested nutrients pass through the small intestine and enter the colon, where they become substrate for fermentation by resident microorganisms. A large amount of fermentative activity in the colon may cause accumulation of small organic acids, including lactic acid, which may exacerbate diarrhea symptoms. To put this observation in perspective, feeding of malnourished children with mucosal atrophy and lower digestive capacity may be a contributing factor to malabsorption and subsequent diarrhea development [29]. However, this may be more pronounced when gut atrophy is accompanied by intestinal infection and inflammation, as is observed in anorexic weanling pigs. For standardization reasons, we chose pigs that were free of specific pathogens (SPF) and observed them in a high-hygiene environment to study the isolated effect of malnutrition without any bias introduced by infection. However, malnourished children often have infectious comorbidity, which may exacerbate malnutrition symptoms [30,31]. In pigs, it has been shown that reducing infectious pressure by treating anorexic weanlings with mucosal atrophy with antibiotics improves digestive enzyme activity [32], which suggests that being malnourished per se only marginally decreases specific digestive enzyme activity.

The intestine represents an organ system that has a high need for an adequate supply of dietary nutrients provided as enteral nutrition. This is evident when examining the first-pass metabolism in the intestinal mucosa of several amino acids, which can be up to 80% of enterally supplied amino acids [33-35]. Together with first-pass metabolism, several other factors are important in maintaining optimal gut health, including a stable microbial flora, optimal barrier function and adequate motility. Low food intake, which was observed in the MAIZE group, may result in inadequate enteral stimuli, leading to decreased gastrointestinal motility and stasis of luminal content. Decreased motility has been associated with small intestinal bacterial overgrowth in a number of diseases [36-38] and may also be an important factor in malnutrition. It is of particular interest that small intestinal bacterial overgrowth can increase the production of bacterial lipopolysaccharides and endotoxins, which are associated with pathological changes, including the development of fatty liver [19].

Microscopic evaluation of the hepatic tissue revealed lesions consistent with fatty liver. Hepatocytes had a vacuolated appearance but with no clear distinction or presence of vacuolated cells around the central veins or in the periportal area. Livers from children with kwashiorkor are also infiltrated with fat that has a periportal distribution [39], and in other experiments with animals with SAM, the fat distribution of the liver has been shown to be periportal [40]. The periportal area is where processes related to fatty acid oxidation and oxidative energy metabolism occur [41,42]. The liver triglyceride contents were 2.8% in the MAIZE pigs versus 2.0% in the REFERENCE pigs. Although the difference is significant, it would not qualify as clinical hepatic steatosis, which would require a triglyceride content of at least 5% [43]. The liver enzymes ALAT and ASAT, although not specific to the liver, presented a varied picture. ALAT was increased in the MAIZE pigs, but the concentration was above the normal levels for both the MAIZE and the REFERENCE pigs. ASAT levels were increased in the REFERENCE pigs at some time points, but the levels were within the normal range throughout the intervention for both groups. Bilirubin levels, which are indicative of cholestasis, were also increased in the MAIZE pigs but were still within the normal range. Albumin, which is synthesized by the liver, was markedly decreased in the MAIZE pigs during the intervention; however, this may have been due to the restricted dietary protein intake, which affected de novo synthesis of albumin in the liver. After 7 weeks on a maize diet, the plasma concentration of albumin was 2.01 g / 100 ml, while others have reported concentrations as low as 0.32 g / 100 ml [44], 0.7 g / 100 ml [16] and 1.14 g / 100 ml [45] for periods of protein restriction ranging from 8 to 28 weeks. Collectively, the histopathologic lesions and the increased triglyceride contents, along with the differences in concentrations of hepatic enzymes, bilirubin and albumin, all indicate that the livers of the MAIZE pigs were affected by malnutrition in a manner similar to that observed in malnourished children.

Animal models of SAM and more specifically kwashiorkor have been conducted with various degrees of success and various clinical outcomes [14-16,45]. In studies investigating low-protein diets as a way to induce kwashiorkor, lack of growth or even cessation of growth has commonly been observed in pigs fed low-protein diets [44,46]. In our study, the MAIZE pigs fed a diet containing 9% protein (Table 1) increased their body weight by only 9.8% during the 7-week study period. In another study, feeding a diet containing 3% protein for 8 weeks resulted in an increase in weight of 50%, from 7.5 kg to 11.3 kg, whereas the weight of controls increased by a factor of 3.5. Body composition was not determined in that study, but for the low-protein group, protein synthesis was reduced in the skin, jejunal mucosa and liver [14]. In another study, a gradual decrease in the dietary protein content from 25% to 2.2% over a period of 24 weeks showed that it was not until week 16 that growth ceased and the pigs started to lose weight [45]. Pigs fed diets with either 4.5% or 6.5% protein also ceased to grow, and after a period of 12-16 weeks, signs of kwashiorkor became evident [15]. However, observed growth rates in these studies are difficult to compare because of differences in protein source and quality, the genetic potential of the animals for growth and the methods of handling and housing. One major reason for the very low growth rate observed in the current study, compared to growth rates observed in other studies, is that we did not fortify the low-protein maize diet with vitamins or minerals, so the diet was inadequate in not only protein but also a wide range of other type II nutrients, resulting in growth failure [47]. Feeding with only maize flour was conducted to mimic real-life situations in which young children are often fed a cereal-based diet with limited variety. The aim was not to evaluate how single nutrients in the maize diet would give rise to specific symptoms but rather to test the effect of a uniform diet over a period of time. Although maize flour is commonly ingested among infants, other staples such as sorghum or cassava could also have been chosen.

In conclusion, pigs fed pure maize flour for 7 weeks developed moderate acute malnutrition resembling marasmus based on assessment criteria similar to what is used for malnourished infants. Having validated the degree of malnutrition, we demonstrate that the digestive system, including the small intestine and liver, are primarily affected in their structure, whereas functional endpoints were affected to a lesser extent at this stage of malnutrition.

Disclosure of conflict of interest

None.