5.1 Definitions: 

Failure to thrive (FTT), a widely used term in industrialized countries, is not distinct from protein-energy malnutrition encountered in children in developing countries (1). Both terms describe a spectrum of pathologic states resulting from childhood undernutrition. FTT accounts for 1-5% of tertiary hospital admissions for infants and is reported in about 10% of low-income preschool children seen in community-based settings.

FTT is used to describe infants and young children whose growth is substantially less than that of their peers. Although no consensus exists on anthropometric indicators of FTT (2), most commonly used systems for selection of samples of FTT in hospital or clinic studies include children with a weight less than the 5th percentile on reference growth charts (1). The key to the identification of FTT is ascertainment of longitudinal progression of growth with serial measurements over time. A weight decrement of more than two major channels of growth (centiles) from a previously established growth channel or the loss of 10% of an infant's weight is evidence of FTT (1,3). A reduction in growth velocity is a particularly helpful indicator of abnormality in the 5% of otherwise normal children who plot out below the 5th percentile, but follow a growth channel parallel to the normal growth curves.

Acute malnutrition typically results in "wasting," where weight is depressed out of proportion to height. Chronic malnutrition will also affect height and may lead to "stunting," in which weight and height are proportionally decreased. Head circumference is the last parameter to be affected by malnutrition.

Extreme conditions of protein-energy malnutrition are rarely encountered in industrialized countries. Marasmus occurs after severe deprivation, primarily of calories, and is characterized by growth retardation and wasting of muscle and subcutaneous fat. Kwashiorkor results from a protein deficiency exceeding the calorie deficiency and is manifested by edema (secondary to hypoalbuminemia) and muscle wasting.

A useful diagnostic approach recognizes three types of FTT according to the deviance of head circumference, height and weight on growth charts (8) (Table 4).

Head circumference is normal and weight is reduced out of proportion to height (Figure 1). This pattern results from an imbalance between calorie availability and requirements. The majority of patients with FTT fall into this category; adverse psychosocial factors are the most frequent contributors.

Head circumference is normal or enlarged and weight is reduced in proportion to (or slightly more than) the reduction in height (Figure 2). This pattern is mainly represented by children with constitutional growth delay, familial short stature or endocrinopathies. Chronic malnutrition resulting in stunting may also fall into this group (celiac disease). Comparison of the children's height age and bone age with chronologic age (see Section 5.4.2) further identifies three subgroups:

Type IIa -Most frequently bone age and height age are proportionally delayed beyond chronologic age (e.g., constitutional growth delay).

Type IIb -Bone age is in accordance with chronologic age but height age is retarded (e.g., familial short stature).

Type IIc -Bone age is significantly more delayed than chronologic age (e.g., malnutrition).

Head circumference is subnormal and weight and height are proportionally reduced (Figure 3). These children are frequently described as dysmorphic and may present with associated developmental delay or seizures. Patients in this category may have chromosomal abnormalities, intrauterine or perinatal insults (congenital infections, alcohol, drug or anticonvulsant exposure during pregnancy, severe prematurity, birth asphyxia), CNS abnormalities and more rarely a familial phenotype.

The size of an infant at birth is more related to maternal size and intrauterine influences than to genetic factors. Consequently, growth across percentiles (increased or decreased growth velocity) between birth and 2 years of age is to be expected in some children owing to genetic adjustment. Patterns of normal growth in the first two years of life can then present as factitious FTT (2): familial short stature and constitutional growth delay. Term infants born small for gestational age and premature infants can also present diagnostic problems.

Familial short stature is genetically determined, and these children are short throughout life. The final height is determined by mid-parental height, and a readjustment with drop in percentiles may take place in the first two years of age. After this deceleration phase, these children grow normally at constant rates and enter puberty at an appropriate age. Weight in these children is usually proportional to length, and they have no bone age delay. The diagnosis of familial short stature is confirmed on the basis of a normal history and physical examination and if, during follow-up, the child maintains the new growth channel appropriate to his or her genetic potential.

Children with constitutional growth delays are "slow growers" and "late bloomers." They present with marked deceleration of growth in the first three years of life and then follow a lower growth channel into adolescence, when a late pubertal growth spurt occurs and they catch up to their original growth channel (4). The deceleration begins usually between 3 and 6 months, will be greatest in the first two years of life and frequently results in these children falling well below the 5th percentile both for weight and height. These children have a two- to four-year delay in skeletal maturation and will enter puberty late. There is frequently a family history of this type of delayed growth and pubertal development.

Small for gestational age infants are a heterogeneous group who fail to grow in utero (intrauterine growth retardation, or IUGR) as a result of environmental, maternal, placental or fetal factors. Asymmetric IUGR (birth weight disproportionally more depressed than length or head circumference) frequently results from placental insufficiency. These newborns have a good prognosis for catch-up growth if they are provided with enhanced postnatal nutrition. Symmetric IUGR may result from intrauterine infections, chromosomal abnormalities or prenatal exposure to toxins such as alcohol, drugs or anticonvulsants. Infants who are symmetrically growth retarded at birth have a poor prognosis for later growth. Because of the initial small size, the weight gain and growth progression of these patients may give the false impression of FTT; however, if the patient doubles the birth weight by 4 months of age and triples it by 1 year of age, FTT must be excluded.

In premature infants corrected age should be used in growth monitoring or they will be inappropriately labeled as FTT. The age at measurement should be corrected for the number of weeks the child was premature (the difference between 40 weeks and gestational age). Corrected age should be used to 18 months for head circumference, 24 months for weight and 40 months for height (5). Premature infants without serious medical problems may show catch-up growth in the first year of life, whereas more severely affected premature infants may not show catch-up growth but should at least parallel reference curves (1).

Other systems commonly used to categorize malnutrition express the child's weight as a percent of the median weight-for-age (Table 5) or as a percent of the median weight-for-height (Table 6). Percent of the median weight- (or height-) for-age is obtained by dividing the actual measurement by the median value for that age. Median weight-for-height (ideal weight for height) is obtained by extrapolating height horizontally to the 50th percentile and taking the median weight for that age. Comparing the child's height to the median height-for-age gives an idea of the severity of stunting or chronic malnutrition. It must be noticed that these methods applied to the same population may classify children in different grades of malnutrition (7). Therefore, appreciation of the severity of malnutrition should not rely exclusively on these ratios.

On physical examination a clinical appreciation of nutritional status is readily available by inspecting the child's buttocks (flat in malnutrition) and subscapular and limb muscles. Evaluation of the triceps skin fold (index of fat tissue) and mid-arm muscle circumference requires the use of published nomograms and reference tables (8,9). Among other physical signs that may be associated with malnutrition are a distended abdomen (malabsorptive states), edema (hypoalbuminemia) and clubbing (chronic disease). More severe cases may show pallor (anemia), glossitis-stomatitis, pellagroid dermatitis, ecchymosis (vitamin K deficiency) and bone deformities (vitamin D deficiency).

Because poor nutrition and psychosocial factors are the major contributors to FTT, investigations are of very limited diagnostic help. Large series of hospitalized children have documented that only about 1% of laboratory studies performed helped identify an organic etiology to FTT. Furthermore, no test was useful in the absence of a specific indication from the history and physical exam. If the clinical evaluation is normal, children with decreased growth should be followed at regular intervals without extensive investigations; frequently, one of the physiologic variations of growth (Table 4) will be identified. Basic tests may be indicated: urine analysis, CBC and differential, albumin, calcium, phosphorus and alkaline phosphatase. Particularly useful is a stool smear for fat. In cases of moderate and severe malnutrition, nutrient deficiencies should be documented (especially iron, zinc, vitamin D).

Note 1: Recommended daily nutritional intake (RDNI) identifies the adequate intake of essential nutrients judged to meet the needs of all healthy people in Canada. Except for energy, the RDNI is established at two standard deviations above the estimated requirements.