METHODS Cross sectional and prospective observational study in Japanese postmenopausal women. OBJECTIVE The aim of the study was 2-fold. The first was to investigate what kind of comorbidities could be found in conjunction with back pain in Japanese postmenopausal women. The second was to investigate whether significant relationship between baseline back pain and future fracture exists or not. BACKGROUND Back pain has been reported to be associated with vertebral degeneration or vertebral fracture. However, there has been no available data that indicates the relationship between back pain and future fracture risks. METHODS The subjects who visited their practitioner were examined for their prevalent back pain or pains in other site. Bone mineral density, body height, body weight, and serum parameter were measured at baseline, and comorbidities were investigated by interview. Fragility fractures were also assessed at baseline and then followed up with 1- to 2-year intervals. The correlation between back pain and baseline characteristics was investigated by logistic regression analysis. The hazard ratio of back pain to future vertebral fracture was estimated by multivariate Cox regression analysis. RESULTS A total of 899 postmenopausal ambulatory women (62.5 +/- 10.3 years old) were enrolled and 81 subjects were dropped out from the study within 1 year. The remaining 818 postmenopausal women (62.1 +/- 10.3 years) were followed-up for 5.7 +/- 3.5 years. Compared to the group without pain, the group with back pain had significantly higher age, lower bone mineral densities at lumbar spine and hip, and higher number of prevalent vertebral fractures. The back pain was significantly associated with rheumatic arthritis (odds ratio [OR]: 2.01, P < 0.05), prevalent vertebral fracture (OR: 4.60, P < 0.001) and osteoporosis (OR: 2.14, P < 0.001). A total of 189 future fractures were observed, of which the most frequent was vertebral fractures (78.3%). The fact that baseline back pain was a significant risk factor for time-dependent vertebral fractures (hazard ratio: 1.62, 95% confidence interval: 1.16-2.27, P = 0.005) was demonstrated by the Cox hazards model after adjusting for traditional risk factors, such as age, bone mineral density, and prevalence of vertebral fractures. CONCLUSIONS The data obtained in this study indicated that the back pain is significantly associated with osteoporosis and rheumatoid arthritis and that it can be useful predictor for future vertebral fracture risk in Japanese postmenopausal women in clinical settings.