Study design: Cross-sectional study to evaluate bone mineral density (BMD) and fracture history after spinal cord injury (SCI). Objectives: To determine frequency of osteoporosis and fractures after SCI, correlate extent of bone loss with frequency of fractures after SCI, and determine fracture risk in SCI patients. Setting: The Hines Veterans A airs Hospital in Hines, Illinois, USA. Methods: Femoral neck BMD was measured in 41 individuals with a history of traumatic or ischemic SCI using dual-energy X-ray absorptiometry (DEXA Lunar Whole Body Densitometer Model). Results: Twenty-®ve patients (61%) met the World Health Organization (WHO) criteria for osteoporosis, eight (19.5%) were osteopenic, and eight (19.5%) were normal. Fracture after SCI had occurred in 14 patients (34%). There were signi®cant di erences between the femoral neck BMD and SCI duration in patients with a fracture history compared to those without. For patients in the same age group, each 0.1 gm/cm 2 and each unit of standard deviation (SD) (t-value) decrement of BMD at the femoral neck increased the risk of fracture 2.2 and 2.8 times, respectively. Considered simultaneously with age, duration of SCI, and level of SCI, BMD was the only signi®cant predictor of the number of fractures. Conclusion: Osteoporosis and an increased frequency of fractures occur after SCI. Measurement of femoral neck BMD can be used to quantify fracture risk in SCI patients.

Seven clinically healthy, nondiabetic (ND) and four Type II diabetic (D) men were assessed for circadian rhythms in oxidative "stress markers." Blood samples were collected at 3h intervals for approximately 27 h beginning at 19:00h. Urine samples were collected every 3 h beginning with the 16:00h-19:00h sample. The dark (sleep) phase of the light-dark cycle extended from 22:30h to 06:30h, with brief awakening for sampling at 01:00h and 04:00h. Subjects were offered general hospital meals at 16:30h, 07:30h, and 13:30h (2400 cal in total/24h). Serum samples were analyzed for uric acid (UA) and nitrite (NO) concentrations, and urine samples were assayed for 8-hydroxydeoxyguanosine (8-OHdG), malondialdehyde (MDA), and 8-isoprostane (ISP). Data were analyzed statistically both by the population multiple-components method and by the analysis of variance (ANOVA). The 24h mean level of UA and NO was greater in D than in ND subjects (424 vs. 338 micromol/L and 39.2 vs. 12.7 microM, respectively). A significant circadian rhythm in UA (p = 0.001) and NO (p = 0.048) was evident in ND but not in D (p = 0.214 and 0.065). A circadian rhythm (p = 0.004, amplitude = 8.6 pmol/kgbw/3h urine vol.) was also evident in urine 8-OHdG of ND but not of D. The 24h mean levels of ND and D were comparable (76.8 vs. 65.7 pmol/kgbw/3h urine vol.). No circadian rhythm by population multiple-components was evident in MDA and ISP levels of ND subjects, or in 8-OHdG, MDA, and ISP in D. However, a significant time-effect was demonstrated by ANOVA in all variables and groups. The 24h mean of MDA and ISP in D was significantly greater than in ND (214 vs. 119 nmol/3h urine vol. and 622 vs. 465 ng/3h urine vol.). The peak concentrations of the three oxidative "stress markers" in urine, like those of serum NO, occurred early in the evening in both groups of men. This observation suggests a correlation between increased oxidative damage and increased rate of anabolic-catabolic events as evidenced by similarities in the timing of peak NO production and in parameters relevant to metabolic functions.

No abstract