Contents - Previous - Next

Discussion of papers by Price et al. and Nilsson et al.

Lodeweyckx began the discussion by summarizing briefly how she, as a pediatrician, is considering possible causes, when she sees a growth-retarded child. Skeletal growth, according to her approach, can be affected by five categories of interacting factors: (1) Nutrient supply and absorption (malnutrition and some gastro-intestinal diseases have to be considered here); (2) metabolism (in a very broad sense including renal and cardiac function, iron deficiency, etc.); (3) neuro-endocrine control (the hormones playing an important role include GH, thyroxine and T3, sex steroids, insulin and growth factors, which can, in turn, all be affected by psycho-social deprivation, blindness, etc.; (4) chromosomal aberrations (e.g. Turner and Klinefelter syndromes); and (5) diseases of cartilage and bone (leading generally to disproportionate growth retardation).

GH is synthesized continuously in the pituitary gland but secreted intermittently, about every three hours. This pulsatility in GH secretion is essential for growth. GH binds to its receptors which exist in almost all tissues. Part of it is then cleaved off and can be measured in the circulation as GH-BP. Another part is internalized in the cell where it produces its effects (which are often referred to as post-receptor phenomena), like the induction of IGF-I and IGF-II. One of the main methodological problems at the moment is that, whereas one can determine the level of hormones in circulation, one does not yet have satisfactory methods to determine their concentration in tissues. Pre and post-receptor phenomena need not always be synchronous, i.e. GH-BP and IGF-BP do not always vary in the same sense. Internalized GH can be externalized again. GH secretion and action is influenced by T4 and sex steroids.

The role of parathyroid hormone (PTH) was discussed. Traditionally, it has been seen primarily as promoting bone resorption, but one now knows that its effects are far more complex. Its receptor messenger expression is very high in the growth plate, but we do not yet know what functional consequences this might have. Pathologic conditions with elevated PTH result in increased bone turnover, but not increased stature. The bone that is being produced under such conditions is abnormal and fractures easily. Children with hypoparathyroidism show hypocalcaemia and abnormalities in bone formation, but they are not stunted. PTH thus seems to play an important role as far as the quality of the bone and its calcification are concerned, but not in longitudinal growth. The PTH level in human breast milk is about 20 times as high as in maternal blood. PTH, like GH, acts differently depending on whether it is provided continuously or intermittently. Given intermittently, 5-6 times a day as breast-fed infants receive it in their mother's milk, it has an anabolic effect.

Another issue that was discussed is the timing and synchronization of chondrocyte and osteoblast activities during the rehabilitation of malnourished children and how they could be monitored. Osteocalcin levels, which are low in malnourished children, increase during rehabilitation, but do not seem to correlate directly with growth rates (see also discussion after Robins' paper). Alkaline phosphatase levels are more indicative of osteoid calcification and bone turnover than of processes in the growth plate. We therefore unfortunately do not seem to have any good markers of bone or cartilage formation yet.

Effects of vitamin D deficiency are age-dependent, because they affect primarily the bones which, at that particular age, are growing fastest (in very young children the skull, later the wrists and ribs). When vitamin D deficiency is corrected, one can see a clear line of mineralization move through the growth-plate 4 to 6 weeks after the beginning of therapy.

Mechanical forces are important in bone formation. If one neurectomizes the limb of an animal before it is developed, the shape of the long bones will be recognizable but quite different.

In early phases of PEM and in kwashiorkor, GH is usually increased and IGF-I is low, whereas GH-BP levels are normal. Experiments in malnourished rats showed that GH receptors in the liver were decreased. GH is internalized with the receptor and exerts its effects within the cell, but part of it is re-exposed to the extracellular region, and the hormonal levels that can be observed are mainly the result of post-receptor phenomena. In children observed in Chile, who failed to grow from birth and were called 'nutritional dwarfs' and in marasmic children, who have been malnourished over a longer period of time, GH is usually decreased. The age of the children, the relative supply of energy and proteins in the food and the duration of PEM thus all seem to have an effect on the hormonal picture that results. In well-nourished children, the administration of amino acids stimulates GH secretion; this is used as a provocation test. In the early phases of PEM, GH levels are usually already so high that amino acid administration has no additional effect.

As mentioned before, GH is secreted intermittently, with a peak about every three hours. Growth velocity is related to the amount of GH secreted during peak periods (peak amplitude), not peak frequency. In a small number of malnourished Pakistani children, Karlberg observed reduced peak amplitudes, but the correlation between amplitudes and growth rates was not very impressive (0.3-0.5).

An article recently published by Lampl, Veldhuis & Johnson (1992) suggests that linear growth occurs in small spurts, with days to months without growth between them. Even though this notion seems compatible with observed fluctuations in biochemical markers of bone resorption and deposition, most discussants doubt that length measurements at such short intervals can be sufficiently standardized and made with sufficient accuracy to distinguish growth from measurement error.

There is more convincing evidence that growth varies seasonally in some children; about 25% of Swedish children show marked seasonal variations in growth, whereas about 25% do not show any seasonal variations. The extent of seasonal variations in growth, how ever, has no influence on the amount of growth per year, nor is there any seasonal change in the frequency or amplitude of GH secretion.

A recent hypothesis postulates that a small-molecular-weight BP (BP3) competes with large-molecular-weight BPs bound to IGF-I. Relative amounts of small- and large-molecular-weight BPs show diurnal cycles and seem to respond to fasting and feeding as well as the macronutrient composition of the diet. In the fed state, the level of small-molecular-weight BPs is lower and has little effect, whereas in the fasting state their level is higher and they interfere with IGF-I and its biological role.


Lampl M, Veldhuis JD &c Johnson ML (1992): Saltation and stasis: a model of human growth. Science 258, 801-803.

Contents - Previous - Next