Kiel D P, Felson D T, Hannan M T, Anderson J J, Wilson P W
Division of General Internal Medicine, Brown University Program in Medicine, Providence, RI.
Am J Epidemiol. 1990 Oct;132(4):675-84. doi: 10.1093/oxfordjournals.aje.a115709.
Caffeine increases urinary calcium output and has been implicated as a risk factor for osteoporosis. The authors examined the effect of caffeine on hip fracture risk in 3,170 individuals attending the 12th (1971-1973) Framingham Study examination. Coffee and tea consumption, age, Framingham examination number, weight, smoking, alcohol consumption, and estrogen use were used to evaluate hip fracture risk according to caffeine intake. Hip fractures occurred in 135 subjects during 12 years of follow-up. Fracture risk over each 2-year period increased with increasing caffeine intake (one cup of coffee = one unit of caffeine, one cup of tea = 1/2 unit of caffeine). For intake of 1.5-2.0 units per day, the adjusted relative risk (RR) of fracture was not significantly elevated compared with intake of one or less units per day. Consumption of greater than or equal to 2.5 units per day significantly increased the risk of fracture. Overall, intake of greater than two cups of coffee per day (four cups of tea) increased the risk of fracture. In summary, hip fracture risk was modestly increased with heavy caffeine use, but not for intake equivalent to one cup of coffee per day. Since caffeine use may be associated with other behaviors that are, themselves, risk factors for fracture, the association may be indirect. Further studies should be performed to confirm these findings.
咖啡因会增加尿钙排出量，被认为是骨质疏松症的一个风险因素。作者在参加第12次（1971 - 1973年）弗雷明汉姆研究检查的3170人中，研究了咖啡因对髋部骨折风险的影响。根据咖啡因摄入量，通过咖啡和茶的饮用情况、年龄、弗雷明汉姆检查次数、体重、吸烟、饮酒及雌激素使用情况来评估髋部骨折风险。在12年的随访期间，135名受试者发生了髋部骨折。每2年期间的骨折风险随着咖啡因摄入量的增加而增加（一杯咖啡 = 一个咖啡因单位，一杯茶 = 1/2个咖啡因单位）。每天摄入1.5 - 2.0个单位时，与每天摄入一个或更少单位相比，骨折的校正相对风险（RR）没有显著升高。每天摄入大于或等于2.5个单位会显著增加骨折风险。总体而言，每天饮用超过两杯咖啡（四杯茶）会增加骨折风险。总之，大量使用咖啡因会适度增加髋部骨折风险，但每天饮用一杯咖啡的摄入量则不会。由于咖啡因的使用可能与其他本身就是骨折风险因素的行为有关，这种关联可能是间接的。应进行进一步研究以证实这些发现。
