Objective
To investigate bone mineral density (BMD) and biochemical markers of bone turnover in cerebral palsy patients according to the severity and type.

Method
BMD and biochemical markers of bone turnover were examined in 30 normal children and 57 children with cerebral palsy. They were 10 to 15 years old and divided into 5 groups: Group I, 30 normal children; Group II, 11 with moderate spastic cerebral palsy; Group III, 10 with moderate non-spastic cerebral palsy; Group IV, 24 with bed-ridden spastic cerebral palsy; Group V, 13 with bed-ridden non-spastic cerebral palsy. The bed-ridden cerebral palsy subjects were further divided into two groups: one with treatment of anticonvulsants more than 5 years; the other with no experience of anticonvulsants treatment. BMD and its T-score on the dominant forearm were measured in all subject, and the level of serum osteocalcin and urine deoxypyridinoline were measured in cerebral palsy patients in early morning.

Results
The bed-ridden cerebral palsy children were shorter, weighed less, and also showed significantly lower value of BMD and T-score on the distal radio-ulnar and the distal end of radial bones compared to those of the normal and the moderate cerebral palsy. These parameters were not significantly different between spastic and non-spastic types of same severity of cerebral palsy. There's no difference in the level of serum osteocalcin and urine deoxypyridinoline between each group of cerebral palsy. In cerebral palsy groups, the level of serum osteocalcin remained in the normal range of the same age group of the normal, however, the urine deoxypyridinoline levels were significantly higher than those of the same age groups of the normal. No difference in either BMD or biochemical markers of bone turnover was observed in bed-ridden cerebral palsy groups regardless of anticonvulsants treatment.

Conclusion
A couple of factors accounting for lower BMD in bed-ridden cerebral palsy are as follows: 1) the increase in activity of bone resorption rather than formation, 2) the diminish of muscle use and the decrease of mechanical stresses on the bone. In addition, these results also suggest no effect of anticonvulsants on lower BMD.