Loss of Bone Mineral Density with Depot-Medroxyprogesterone Acetate

Abstract & commentary

Source: Tang OS, et al. Contraception 1999;59:25-29.

Synopsis: The loss of bone in users of depot-medroxyprogesterone acetate is probably not clinically significant.

Tang and colleagues from the university of Hong Kong report a case-control study of bone mineral density in long-term users of depot-medroxyprogesterone acetate. In this study, 67 Chinese women who had used depot-medroxyprogesterone acetate for 5-15 years were compared to 218 control women. The results of bone density were measured by the dual-energy x-ray absorptiometry (DEXA) method. Bone mineral density in the hormone-using group ranged from 1.5% per year in the spine to 3.5% in Ward’s triangle. These patients were older than the usual age groups studied in contraception studies. The mean age was 42.8 years, with a range of 34-46 years. A relationship was demonstrated with duration of use and, in the spine, the older the patient, the greater the bone loss.


There continues to be concern that the long-term use of depot-medroxyprogesterone acetate for contraception is associated with bone loss because of blood levels of estrogen that are relatively lower, compared to a normal menstrual cycle. Bone loss has been documented in several cross-sectional studies, including one in adolescent girls.1,2,3 In contrast, a case-control study in Thailand found no bone loss in users of this method of contraception.4 Indeed, the investigators from Thailand also performed a longitudinal, prospective study, and this too could not detect loss of bone density, although the method measured bone in the forearm.5 The Thailand investigators suggested that previous studies were explained by inadequate control of factors that affect bone, such as smoking and alcohol intake. The investigators from the University of Hong Kong, however, controlled these factors and still reported bone loss. They further point out that in the study from Thailand depot-medroxyprogesterone acetate was used for only three years. Because there is a greater loss with greater duration of use, they argue that this would explain the Thailand results. Nevertheless, there is another small prospective study that demonstrated stable forearm bone density over a six-month period.

Obviously, this disagreement is confusing and probably reflects to an important degree different ethnic groups and the cross-sectional design of the studies. It is important to note that bone density measurements in women who stop using depot-medroxyprogesterone acetate indicated that the loss was regained even after long-term use.7 It is most likely that individuals who use this method of contraception will be exposed to adequate hormonal support of their bones for a period after discontinuing the method. The important question, therefore, is whether this method of contraception has any effect on the risk of fracture in the latter stages of life. At this point, we have no data on the incidence of fracture in past users, and this concern will require ongoing surveillance of past users. In my view, it is unlikely that bone loss occurs sufficiently to raise the risk of osteoporosis-related fractures later in life. At the present time, this should not be a reason to avoid this method of contraception.


    1. Cundy T, et al. BMJ 1991;303:13-16.

    2. Cundy T, et al. Obstet Gynecol 1998;92: 569-573.

    3. Cromer BA, et al. J Pediatr 1996;129:671-676.

    4. Virotamasen P, et al. Asia Oceania J Ob Gynae 1994; 20:269.

    5. Teneepanichskul S, et al. Contraception 1997;56:1-3.

    6. Naessen T, et al. Contraception 1995;52:35-39.

    7. Cundy T, et al. BMJ 1994;308:247-248.

The following statements are true regarding the relationship between depot-medroxyprogesterone acetate and bone density except:

    a. The data on bone density in users of depot-medroxyprogesterone acetate are derived only from retrospective case-control studies.
    b. Most studies have indicated greater bone loss with increasing duration of depot-medroxyprogesterone acetate use.
    c. The amount of bone loss is not great and may not be clinically important in terms of future risk of fractures.
    d. Smoking is not a confounding bias in most of the studies.