Original Article

Journal of Bone and Mineral Metabolism

, Volume 29, Issue 1, pp 62-70

First online:

Urinary pentosidine and plasma homocysteine levels at baseline predict future fractures in osteoporosis patients under bisphosphonate treatment

  • Masataka ShirakiAffiliated withResearch Institute and Practice for Involutional Diseases
  • , Tatsuhiko KurodaAffiliated withPublic Health Research Foundation
  • , Yumiko ShirakiAffiliated withResearch Institute and Practice for Involutional Diseases
  • , Shiro TanakaAffiliated withDivision of Clinical Trial, Design and Management, Translational Research Center, Kyoto University
  • , Tsuyoshi HiguchiAffiliated withDepartment of Obstetrics and Gynecology, Hirosaki University School of Medicine
  • , Mitsuru SaitoAffiliated withDepartment of Orthopedic Surgery, Jikei University School of Medicine Email author 

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access


To clarify what kind of risk factors predict incident fractures in patients treated with bisphosphonates, the authors investigated the relationship between baseline characteristics and incident vertebral fracture in Japanese osteoporosis patients undergoing bisphosphonate treatment. This was a multi-center follow-up study conducted at three centers, in which a total of 251 Japanese patients with osteoporosis (mean age 70.5 years) from the three centers were followed for 3.2 ± 2.0 years. Baseline data, including pre-existing fractures, bone mineral density in the lumbar spine (LBMD), bone metabolic markers, urinary pentosidine, and plasma homocysteine, were evaluated. Changes in LBMD, bone turnover markers, and incident fractures after the treatment were followed. Sixty-one patients developed incident vertebral fractures; this group of patients was older and had lower LBMD, a higher prevalent vertebral fracture number, and higher homocysteine and pentosidine levels than patients who did not develop incident vertebral fractures. Changes in LBMD, urinary N-terminal telopeptides of type I collagen (NTX), and bone-derived alkaline phosphatase showed no significant association with the occurrence of vertebral fractures. Cox’s proportional hazard model demonstrated that age, prevalent fracture, pentosidine, and homocysteine were independent predictors of the incident vertebral fracture rate under bisphosphonate treatment. Higher baseline levels of pentosidine and homocysteine in osteoporosis patients are potential risk factors for incident vertebral fractures when these patients are treated with bisphosphonates. Further clarification is needed to explain why such patients have higher fracture susceptibility.


Pentosidine Homocysteine Bisphosphonates Osteoporosis Fracture