Cohort studies in humans have suggested that the peak bone mass attained at skeletal maturity may be programmed in utero. To investigate which aspects of bone development might be influenced in utero, we utilised a rat model of maternal protein insufficiency, which has previously been used to demonstrate the fetal origin of adult hypertension. In rodents, a growth plate remains present throughout life, even after longitudinal growth ceases. Generally, the height of the growth plate is related to the rate of bone growth. Fast growing bones have maximal height growth plates, and as bone growth slows down the height decreases until it remains stationary. The aim of this study was to compare the morphology of long bones in aged rats that had been subjected to protein insufficiency in utero with that of controls. Rat dams were fed either an 18% casein control diet or a 9% casein low protein diet from conception until the end of pregnancy. The offspring were fed a normal diet until death (~72 weeks), when bone density was measured by dual energy X-ray absorptiometry (DEXA) and the tibiae and femurs were processed for histology. The offspring of rats from the low protein group had a significantly lower bone mass, as assessed by DEXA. The major differences in bone structure were found in the growth plates, which were very irregular without the usual zones of resting, proliferating and hypertrophic chondrocytes. A number of unusual cellular events were noted to have taken place subsequent to cessation of growth, including: a) elimination of all chondrocytes in a number of regions, resulting in vast acellular areas; b) formation of chondroid bone and/or transdifferentiation of chondrocytes to bone-forming cells in other regions; c) partial resorption of those latter regions while the acellular regions were not resorbed; d) ‘horizontal’ apposition of bone against a smooth metaphyseal edge of the growth plate. To compare the growth plates from the low and high protein groups semi-quantitatively, the degrees of the above features were scored. In addition, the heights of the growth plates were were assessed by two independent measurements. In the low protein group, the height of the growth plate were found to be significantly greater (p<
0.001). Additionally, the growth plates from this group of animals were observed to be more irregular with regards to all the features outlined above. These findings are consistent with the hypothesis that growth trajectory and bone mass are programmed in early life. The increased height of the growth plate in animals undernourished in utero may reflect the cessation of growth at an earlier age. The increased irregularity of the growth plate in this group of animals may infer an earlier onset of age-related changes within the growth cartilage.
In a prospective trial we performed MRI of the spine and hind brain in 31 patients with scoliosis of onset between the ages of four and 12 years. In eight patients (26%) there was a significant neuroanatomical abnormality; there were six cases of Chiari-1 malformation associated with a syrinx, one isolated Chiari-1 malformation and one astrocytoma of the cervical spine. Four of these patients had left-sided curves. There were no clinical features which could reliably identify those patients with abnormalities on MRI. In particular, the unilateral absence of abdominal reflexes was found to be non-specific (1 of 8 of patients with neuroanatomical abnormalities (12.5%) In view of the established risks of surgical correction of scoliosis in the presence of undecompressed syringomyelia and the possible improvement that may follow decompression of the foramen magnum, we feel that MRI of all patients with scoliosis of juvenile onset should be obligatory.
