The World Health Organization defines undernutrition as the outcome of a variety of pathological conditions that result from deficiencies in absorption, transportation or utilization of nutrients by cells of the body. This type of nutritional disorder mainly affects infants and children of pre-school age [1]. It is estimated that more than one quarter of the global population of children are currently affected by protein-energy undernutrition (PEU), and that 148 million (26.7%) of the 555 million pre-school children living in developing countries are underweight for their age, while 180 million (32.5%) exhibit low height for their age [2,3].

In Brazil, the prevalence of undernutrition was reduced by half during the period 1996–2007 [4] but, owing to social inequality, the condition remains a public health problem, especially in the northern and north-eastern regions of the country [5,6]. According to a nutritional survey conducted in 2005 in the semiarid region of Brazil, the prevalence of undernutrition (stunting) among children living in the State of Alagoas was greater than in other north-eastern states. High amongst the factors influencing this situation was the level of absolute poverty in Alagoas, which remains the highest of any state in the country [7,8].

Undernutrition in children is often accompanied by vitamin and mineral deficits. Indeed, ion deficiency anaemia (sideropenic anaemia) is the most frequent nutritional deficiency worldwide, closely followed by hypovitaminosis [9]. Moreover, in addition to intestinal parasitism, inadequate food intake is considered an important factor in the aetiology of nutritional insufficiency [10]. Chronic undernutrition in early life is associated with diminished intellectual capacity, poor scholastic performance, reduced physical capacity and alterations in metabolic functions, particularly those relating to endocrine processes. Such adaptations tend to diminish the utilization of nutritional substrates in order to maintain the basic life processes, with a concomitant reduction in resources required for normal growth [11,12].

Undernourished children typically present high levels of growth hormone (GH), but low plasma levels of insulin-like growth factor 1 (IGF-1), and this is the main cause of growth deceleration [13]. Furthermore, undernourished children exhibit high levels of cortisol, which promotes protein catabolism and, consequently, delays linear growth [14]. The respiratory quotient (RQ) values in children with chronic undernutrition are larger in comparison with those of eutrophic children, and this represents a further strategy to save energy by reducing lipid oxidation, thus explaining the association between PEU and the accumulation of fat in the liver [11].

The first indication that severe undernutrition could cause hepatic lesions and alterations in the lipid profile was reported by Snapper [15] in 1965 following studies on adults who had experienced interuterine undernutrition during World War II. The relationship between pre-natal exposure to undernutrition and alterations in the lipid profile of adults has since been confirmed in a number of investigations carried out in developed countries [16–18]. However, studies concerning the possible association between hepatic alterations and chronic undernutrition in childhood are scarce and dated [19–21]. In the present study, the biochemical profiles of chronically undernourished children submitted to a semi-confined regime of nutritional rehabilitation in the Centre for Rehabilitation and Nutritional Education of the State of Alagoas (CREN/AL) have been determined in order to test the hypothesis that chronic undernutrition causes changes in lipid profile in pre-school children.

Details of the study were submitted to and approved by the Committee of Ethics in Research of the Universidade Federal de Alagoas (protocol no. 009580/2007-26). Written informed consent was obtained from the parents or legal guardians of participating children prior to the commencement of the study.

One child of the original 80 selected moved out of the area and, hence, failed to complete the study. Of the remaining 79 children, 59 (75%) were classified as moderately undernourished (HAZ values in the range −2.01 to −2.94) and 20 (25%) were severely undernourished (HAZ between −3.0 and −4.41). The diet supplied by CREN/AL, which had been adapted to the requirements for rehabilitation, complied with the DRI standards [24] and provided an average total caloric value of 1,250 kcal and 40.2 g of protein to both the moderately and severely undernourished groups [25] (Table 1).

The socioeconomic conditions of the families were homogenous (Table 2) in that their incomes were low, their living conditions were generally precarious (i.e., overcrowded brick dwellings with no floor covering, and reliance on an untreated water supply and septic tank drainage), while most mothers had little schooling and undertook no paid employment. The likelihood of children being severely undernourished increased significantly in families with numerous (7–12) children (Table 2).

Most children presented intestinal parasitism but no statistical differences could be detected between the two groups in this respect (Table 3). The overall prevalence of iron-deficient anaemia was 44.3%, although the proportion of anaemic children was significantly greater within the severely undernourished group than for those presenting moderate undernutrition. Hypovitaminosis A was not detected in children of either group. There were no differences between the groups regarding the levels of salivary cortisol, with the majority of children presenting normal concentrations of this hormone. With regard to IGF-1, however, the levels decreased significantly as the degree of undernutrition increased (Table 3).

Lipid profile analysis demonstrated that 98.9% of the children presented dyslipidaemia. Additionally, levels of HDL were statistically correlated with the degree of undernutrition, such that the likelihood of occurrence of low levels of HDL in children with moderate undernutrition was smaller than for those with severe undernutrition. According to the results of the multivariate analysis (Table 4), the variables most strongly associated with severe undernutrition were untreated water supply, anaemia, low HDL and low IGF-1.

Similar to previous reports [12,36], the present study demonstrated that poor socioeconomic conditions, but most especially reliance on an untreated water supply, strongly influence the development of PEU. Indeed, the chances of children becoming severely undernourished increases by some 600-fold in families that do not treat their drinking water. It is noteworthy, however, that no cases of hypovitaminosis A were detected in the studied population, a finding that contrasts with the results of a previous investigation involving a low-income population [37]. This difference may be explained by the fact that the children in the present study were registered in the National Program of Supplementation of Vitamin A, sponsored by the Brazilian Ministry of Health [38], as confirmed by notes on the vaccination cards of the participants. In contrast, however, the prevalence of iron-deficient anaemia among the studied population was relatively high even though the diet supplied by CREN/AL was perfectly adequate with regard to this mineral. It is assumed that the occurrence of anaemia in infancy derives from the combination of an exceptionally high requirement for iron, imposed by growth on a mineral-poor diet, and the high frequency of infections and parasitic diseases [39,40]. In the present case, the high prevalence of parasitism (81%) together with chronic undernutrition may have contributed to the high prevalence of anaemia detected, reflecting a situation that has been previously observed in a poor population in Maceió [10].

Undernutrition causes great metabolic stress to the body, which typically manifests as an increase in cortisol and a reduction in the resting metabolic rate (RMR) in order to conserve energy [11,12]. In contrast to previous studies [11,41], however, the levels of cortisol observed in the studied population were found to be normal, most likely because the children had already received dietetic treatment for 90 days. It is evident that the supply of a diet that is balanced, particularly with respect to protein, is crucial for the recovery of normal levels of cortisol. Regarding the levels of IGF-1, the present study confirms the severity of undernutrition since a reduction in level of this important growth factor gives rise to serious structural deficits in severely undernourished children. Similar findings have been reported previously for children living in impoverished areas of the State of Sao Paulo, Brazil [42,43].

Together with endocrine alterations, PEU produced modifications in the lipid profiles of children that were similar to those reported in adults who had been exposed to intrauterine undernutrition [44]. It has been demonstrated previously that prenatal undernutrition directly influences the development of altered lipid profiles that are more atherogenic in the long term. Such changes are probably caused by the diminution or inactivation of active liver receptors, or by overproduction of very-low density lipoprotein (VLDL) and LDL, or by defects in the expression of lipoprotein lipase [17,18].

Previous studies involving adolescents with stature deficit have revealed that reductions in RMR are compensated by increased RQ and decreased lipid oxidation [45,46]. It is possible, therefore, that the children who participated in the present study may be accumulating abdominal fat and developing an atherogenic lipid profile by virtue of the alterations described herein. Recent reports [47,48] have confirmed the association between deficient GH response and greater visceral adiposity, dyslipidaemia, insulin resistance and increased cardiovascular risk in overweight adolescents. More specifically, one study revealed a positive association between the levels of plasma IGF-1 and those of HDL-cholesterol in non-diabetic adults [49].

In the present study, dyslipidaemia was characterised by the predominance of low HDL levels (86.1% of the children) together with hypertriglyceridemia. Indeed, both conditions are correlated, although low HDL is caused by an acceleration in catabolism and not by a decrease in the synthesis of these particles [50]. In order to explain the relationship between hypertriglyceridemia and low HDL, two main hypotheses have been put forward, namely: (i) the reduced activity of lipoprotein lipase may hamper the maturation of HDL particles, and (ii) the increased activity of the protein that promotes the transfer of cholesteryl esters from HDL to triglyceride-rich lipoproteins may reduce the levels of HDL particles [51]. Either of these hypotheses could explain the high prevalence (49.4%) of low HDL concomitant with hypertriglyceridemia in the studied population. It is assumed, therefore, that the high frequency of dyslipidaemia observed in the present study was a consequence of the adaptation to chronic undernutrition, although there are currently no studies to support this supposition. Future monitoring of the treated children is, therefore, of vital importance in order to establish if the adequate diet provided by CREN/AL allows recovery of nutritional status and reversal of the atherogenic lipid profile.

The chronic undernutrition experienced by the study population resulted in endocrine adaptations that gave rise to modifications in lipid profile.