Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
Depot medroxyprogesterone acetate (DMPA [Depo-Provera®, Upjohn, Kalamazoo, MI]) is an effective, convenient, and economical contraceptive option that has been in use worldwide for at least three decades. Its approval by the U.S. Food and Drug Administration as a contraceptive in 1992 may have led to increasing use of this injectable progestin-based method in the U.S. In a 1995 survey, 1 approximately 2% of women using contraception reported injection (ie, DMPA) as their current method. Among those age 15–24 years, this method was the third most common contraceptive, behind oral contraceptives and condoms. Since an initial report by Cundy et al 2 in 1991, there has been concern over the possible adverse effects of DMPA on bone mass attainment and retention. Some subsequent studies of this association report lower bone mineral density (BMD) among DMPA users, 3–13 whereas others do not. 14–20 There have been few longitudinal studies among current users 3,4,11,13–15 and, to our knowledge, only one evaluation of the effects of DMPA discontinuation. 3
We report the results of a 4-year prospective study of the effects of DMPA use and discontinuation on bone mineral density in a population-based cohort of 457 reproductive-age women.
This study was conducted at Group Health Cooperative, a non-profit health maintenance organization located in Washington State.
The study cohort was recruited between November 1994 and April 1996. Using Group Health's computerized databases, we identified potential participants from the ages of 18–39 years who had received DMPA injections and who resided in a defined area near the Group Health research clinics. Each month we selected from the ambulatory care database all women who could be considered possible new users (no evidence of previous DMPA injections in the past 12 months), along with a random sample of prevalent users. A comparison group of women not exposed to DMPA was randomly sampled from the Group Health membership database, frequency matching on age and clinic.
All study procedures were reviewed and approved by the Group Health Cooperative and University of Washington Human Subjects Committees. Written informed consent was obtained from participants at the baseline clinic visit. Study methods (exclusions, baseline measurements, and questionnaire items) have been detailed previously. 10 Briefly, participants were recruited and screened by telephone, excluding women who were currently pregnant or who reported conditions known to alter BMD. Oral contraceptive use was not an exclusion for comparison-group women because the effects of this hormonal exposure on bone density are unclear. 21,22 Instead, we enrolled a larger comparison group to permit us to assess the possible effects of oral contraceptive use.
At the baseline visit, participants’ height and weight were measured, and the baseline questionnaire was reviewed. The questionnaire included items on dietary intake, 23 family history, lifestyle factors, and demographic characteristics. Bone mineral density at the posterior-anterior lumbar spine (L1–L4), proximal femur (total hip), and whole body was measured using dual-energy x-ray absorptiometry (Hologic QDR 2000, Hologic, Inc., Bedford, MA). 10 In our assessment of reproducibility, the mean absolute percentage difference of 1.8% for 15 duplicate hip measurements was consistent with that reported in the literature. 24
During the 36-month follow-up period, bone density measures and questionnaires were completed every 6 months. Food frequency questionnaires and laboratory specimens for indices of bone turnover 25 were collected every 12 months. Because the calibration system for one densitometer began to malfunction between June and August of 1996, scans from this 2-month period (N = 86) were not used in the analyses. Repair of the densitometer caused a shift in in vivo measurements of 0.74%, −1.39%, and −2.24% for the spine, hip, and whole body, respectively. Measurements taken after the calibration repair were adjusted by these percentages.
The primary study outcome was change in bone mineral density (expressed in gm/cm2) for the spine and total hip during the 36-month follow-up. We examined mean BMD for DMPA users, discontinuers, and the unexposed comparison women. The data were subsequently evaluated in age-specific strata (18–21, 22–29, and 30–39 years) to see whether DMPA exerted differential age-related effects on bone.
We examined the effects of DMPA use and discontinuation on BMD change in a repeated measures multiple regression model that allowed individuals to change their exposure status over the course of the study. A marginal generalized estimating equations approach was used to evaluate the change in BMD for each 6-month interval, out to 36 months, using all available bone density measurements. 26 Consequently, women who missed a visit re-entered the analyses at the next visit they completed. DMPA exposure was entered as a time-dependent covariate, classifying participants as users, discontinuers, or non-users at the start of each 6-month interval. Covariates that were found to be associated with either bone density or with DMPA use were evaluated for inclusion in the model. An independent working correlation structure was assumed because time-dependent covariates (as well as fixed baseline covariates) were included.
Of the eligible DMPA exposed and unexposed women (317 and 473, respectively), 58% of each group enrolled. The study cohort consisted of 457 women, 183 receiving DMPA injections and 274 not. We excluded 16 women in the DMPA unexposed group who had some prior use of DMPA, and we excluded one DMPA exposed participant who was found to have metabolic bone disease. Thus, our final analysis group consisted of 182 DMPA-exposed and 258 DMPA- unexposed women.
All women enrolled as DMPA-exposed at baseline were receiving the standard contraceptive dose (150 mg every 3 months). Duration of use at baseline ranged from 1 to 133 months, with a median of 11.3 months. Forty-three women (24%) were new users of this method, that is, they were enrolled after their first injection and before their second.
At baseline, DMPA-exposed women were more likely than unexposed women to be young, to be nonwhite, to smoke, to have been pregnant, and to be physically active (Table 1). They were less likely to have a history of fracture in a female relative and to have education beyond high school. Baseline bone mineral densities for the spine and total hip were lower among DMPA-exposed women. Among the 43 new users, mean BMD for the spine and total hip was only slightly lower than for DMPA-unexposed women (data not shown).
The percentages of women completing clinic visits at 12, 24, and 36-months of follow-up were 87%, 76%, and 67%, respectively. The principal reasons for not completing the study were attempting pregnancy or becoming pregnant (13% and 8% for DMPA-exposed and -unexposed women, respectively); these women became ineligible because our study protocol precluded exposing them to radiation. Moving out of the area was the second most common reason for loss of participants. Women who did not complete the 30- or 36-month follow-up visits were more likely to be below the age of 25, to be nonwhite, and to have lower baseline bone density at the spine (data not shown).
Of the 182 women who began the study as DMPA users, 110 (60%) new and prevalent users discontinued their injections during the follow-up period, with 44% discontinuing within the first 6 months after enrollment. Women who discontinued DMPA were followed for a mean of 15 months after discontinuation (range 6–30 months). These women were more likely than women who continued DMPA to be white, current smokers, to have a female relative with a fracture, to have lower educational attainment, and to weigh more (data not shown). The three most frequently noted reasons for discontinuation were weight gain (19% of discontinuers), nonspecific side effects (18%), and no longer being sexually active (11%).
The mean BMD for the 89 women using oral contraceptives did not differ from the 169 other DMPA-unexposed women, so we combined them into a single comparison group.
Effects of Continuous DMPA Use on Bone Density
Mean BMD for women who continued to receive DMPA injections declined at both the spine and the hip throughout the study follow-up period, with the greatest loss occurring during the first 12 months of enrollment (Figure 1). Among DMPA-unexposed women, mean BMD increased slightly. Expressed as annualized mean percentage change, DMPA users experienced a decrease of −0.87% at the spine and −1.12% at the hip, compared with +0.40% and −0.05% at the spine and hip, respectively, among non-users.
In age-specific analyses, mean BMD fell among DMPA users in all three age groups (18–21, 22–29, and 30–39 years) and increased slightly among non-users (Figure 2).
After adjustment for baseline variables (ethnicity, gravidity, and amenorrhea) and time-dependent variables (age, weight, smoking, alcohol and calcium intake, and fracture in a female relative), the continuous DMPA users experienced bone loss for each 6-month interval (Table 2). The rate of BMD change also differed notably from that of non-users at both anatomic sites. Adjusted mean BMD change at the spine was −0.0053 gm/cm2 in DMPA users compared with +0.0023 gm/cm2 in non-users. Adjusted mean BMD change at the hip was −0.0060 compared with −0.0002 gm/cm2, respectively.
We also evaluated the effects of cumulative duration of use. The adjusted mean change in BMD for each 6-month interval decreased with increasing cumulative use of DMPA (P-value for grouped linear trend = 0.002 for spine and <0.0005 for hip) (Table 3). Thus, women with ≤12 months’ exposure to DMPA (the majority of whom were new users) lost bone at a faster rate than longer-term users. In women using DMPA, BMD at the hip and spine continued to decline through 4 years of cumulative DMPA exposure. Mean bone density change for each 6-month interval was greatest at the hip.
Effects of DMPA Discontinuation on Bone Density
During the time of their exposure to DMPA, women who eventually discontinued injections had patterns of bone loss similar to those in women who continued receiving injections throughout follow-up. After discontinuation, substantial BMD gains were seen. At the end of the available follow-up time, mean BMD for discontinuers was not markedly different from DMPA non-users for either anatomic site (Figure 1). Mean annualized percentage change in BMD was +1.41% per year at the spine and +1.03% per year at the hip among DMPA discontinuers. This compared with +0.40% and −0.05% per year, respectively, among non-users.
Women who discontinued DMPA had gains in mean bone density for each 6-month interval, and these gains differed from the non-user comparison group (Table 2). At the spine, adjusted mean BMD change was +0.0067 gm/cm2 among discontinuers compared with +0.0023 gm/cm2 among non-users. At the hip, adjusted mean BMD change was +0.0035 compared with −0.0002 gm/cm2, respectively. Overall, mean bone gain for the spine among DMPA discontinuers was similar to bone loss among continuous users for each 6-month interval. But, for the hip, the amount of bone loss for each 6 months among continuous users was greater than bone gain among discontinuers, leaving some question as to whether bone density is entirely restored at this important anatomic site.
In the multiple regression model that examined the effects of discontinuation by duration of DMPA use and time since discontinuation, DMPA discontinuers experienced steady gains in BMD for as long as we were able to measure, regardless of duration of DMPA use at the time of discontinuation (Table 3).
Results for whole-body BMD were in the same direction as the changes reported for the spine and hip, but were smaller in magnitude (data not shown).
This 4-year longitudinal study addresses, within the same cohort, the related questions of how use and discontinuation of DMPA injectable contraception affect bone density.
Does DMPA Use Decrease Bone Density?
Bone loss among continuous users of DMPA occurred throughout the follow-up interval, with adjusted longitudinal changes in mean BMD that were much greater than BMD changes among non-users (Figure 1 and Table 2). We saw much lower BMD at baseline for 18–21-year-old DMPA users, despite a lower median duration of use. DMPA users 30–39 years of age, although similar to unexposed women at the outset, experienced notable declines in BMD during follow-up (Figure 2).
After 4 years of DMPA use, the bone density at the hip decreased approximately 0.06 gm/cm2. This is 6.0% of the mean value, or 0.5 standard deviations of peak bone mass, which is similar to the 4.8% of bone lost during lactation 27 or the 5.3% of estrogen-related bone loss during menopause. 28 In postmenopausal women this degree of bone loss would increase the risk of hip fracture by about 50%. 29,30
In showing an overall decrease in bone density with continued DMPA exposure, our longitudinal results accord with those from cross-sectional 2,3,5–10,12 and other prospective studies. 3,4,11,13 However, not all studies report this association. 14,15,17–20 This may be due to wide variation in the age of subjects, duration of DMPA use, or anatomic sites that were evaluated. The prospective investigations that include comparison groups have found greater decreases in bone density among DMPA users relative to non-users. 3,4,11,13,14 However, previous prospective studies have been considerably smaller (≤155 women), with 2 years or less of follow-up.
Are the Effects of DMPA on Bone Reversible?
Many changes in the body's hormonal milieu exert transient effects on bone. Thus, it is of great interest to know whether bone density is restored after cessation of DMPA use. On this important question, the data from our large group of DMPA discontinuers are particularly germane. By approximately 30 months postdiscontinuation, BMD values were similar to those of unexposed women (Figure 1). However, our adjusted 6-month changes in BMD suggest that BMD for the hip is recovered more slowly than for the spine (Table 2). Our age-specific data show that, because of their large deficit at baseline, 18–21-year-old DMPA discontinuers still had lower BMD values than non-users after 30 months of follow-up. Further investigation of the effects of DMPA effects in adolescents is warranted.
In the only other prospective study of the effects of DMPA discontinuation, Cundy et al.3 reported a 3.4% annualized increase in mean spinal BMD among 14 long-term DMPA users (median use 10 years) after discontinuation, with no changes noted in 18 control women or in 22 women who continued DMPA. They noted a smaller 0.8% increase at the femoral neck. The annualized rate of change at the spine is somewhat higher than the +1.41% we noted, but their results at the hip are quite similar to ours (+0.8% compared with +1.03%).
Strengths and Limitations
In evaluating these results, the current study offers a number of strengths. The prospective cohort design, in which both incident and prevalent users of this method were enrolled, allowed for assessment of both short- and long-term changes in bone density. The population-based sampling frame permitted selection of DMPA-exposed and unexposed comparison group women from the same defined population. We controlled for the confounding effects of other factors and considered the age- and duration-specific effects of DMPA on bone. Finally, the large study group and lengthy follow-up allowed for detailed assessment of DMPA discontinuation as well as DMPA exposure.
These data also have limitations. We experienced attrition from the cohort primarily through the requisite study termination of women who became pregnant or began seeking pregnancy. However, we adjusted for covariates associated with partial follow-up in the analyses, and all available data for these women are included in the results.
Our study design and recruitment approach did not allow a pre-DMPA exposure assessment of bone density. We did see, however, that the group of participants who were enrolled after only one injection (with a median of 78 days between injection and baseline BMD measurement) had bone densities that were very similar to those of non-users.
We observed that DMPA discontinuers continued to show BMD increases through to the maximum follow-up attainable (30 months postdiscontinuation;Figure 2). Still longer follow-up in more women would be needed to clarify the complete time course of BMD changes after DMPA discontinuation. In women with other causes of hypoestrogenemia, BMD values increase with normalization of estrogen levels, but the extent of recovery depends on the clinical situation. For example, after prolonged lactation BMD continues to show recovery up to 18 months. 31 Young women with anorexia who are treated with estrogens show minimal restoration of bone mass. 32 Similarly, after 8 years, amenorrheic athletes show only partial recovery of BMD despite normalization of estrogen levels. 33
Osteoporosis is a growing public health problem. Reducing exposure to factors that may inhibit bone mass attainment and maintenance is a key prevention strategy. Our evaluation of the effects of DMPA on bone in healthy premenopausal women supports the conclusion that this exposure acts to decrease bone density, with effects that vary by age, duration of use, and anatomic site. However, the observation that bone density increases substantially for many women after DMPA discontinuation provides a measure of reassurance to those who select this contraceptive method.
We thank Donna Edgerton, Cathy Hutchison, Kim Caudill, Jane Grafton, Kay Hooks, Ryan Finholm, Cheryl Falls, and Carol McDonald for their many contributions to this study, and we also thank our study participants.
1. Abma JC, Chandra A, Mosher WD, Peterson LS, Piccinino LJ. Fertility, family planning, and women's health: new data from the National Survey of Family Growth. Vital Health Stat 1997; 1995: 1–114.
2. Cundy T, Evans M, Roberts H, Wattie D, Ames R, Reid IR. Bone density
in women receiving depot medroxyprogesterone acetate
for contraception. BMJ 1991; 303: 13–16.
3. Cundy T, Reid I. Depot medroxyprogesterone and bone density
. BMJ 1994; 308: 1567–1568.
4. Cromer BA, Blair JM, Mahan JD, Zibners L, Naumovski Z. A prospective comparison of bone density
in adolescent girls receiving depot medroxyprogesterone acetate
(Depo-Provera), levonorgestrel (Norplant), or oral contraceptives. J Pediatr 1996; 129: 671–676.
5. Cundy T, Cornish J, Roberts H, Elder H, Reid IR. Spinal bone density
in women using depot medroxyprogesterone contraception. Obstet Gynecol 1998; 92: 569–573.
6. Paiva LC, Pinto-Neto AM, Faundes A. Bone density
among long-term users of medroxyprogesterone acetate as a contraceptive. Contraception 1998; 58: 351–355.
7. Gbolade B, Ellis S, Murby B, Randall S, Kirkman R. Bone density
in long term users of depot medroxyprogesterone acetate
. Br J Obstet Gynaecol 1998; 105: 790–794.
8. Tang OS, Tang G, Yip P, Li B, Fan S. Long-term depot-medroxyprogesterone acetate and bone mineral density. Contraception 1999; 59: 25–29.
9. Bahamondes L, Perrotti M, Castro S, Faundes D, Petta C, Bedone A. Forearm bone density
in users of Depo-Provera as a contraceptive method. Fertil Steril 1999; 71: 849–852.
10. Scholes D, Lacroix AZ, Ott SM, Ichikawa LE, Barlow WE. Bone mineral density in women using depot medroxyprogesterone acetate
for contraception. Obstet Gynecol 1999; 93: 233–238.
11. Berenson AB, Rickert VI, Grady JJ. A prospective study of the effects of oral and injectable contraception on bone mineral density. Obstet Gynecol 2000; 95: S6.
12. Petitti DB, Piaggio G, Mehta S, Cravioto MC, Meirik O. Steroid hormone contraception and bone mineral density: a cross-sectional study in an international population. The WHO Study of Hormonal Contraception and Bone Health. Obstet Gynecol 2000; 95: 736–744.
13. Berenson AB, Radecki CM, Grady JJ, Rickert VI, Thomas A. A prospective, controlled study of the effects of hormonal contraception on bone mineral density. Obstet Gynecol 2001; 98: 576–582.
14. Naessen T, Olsson SE, Gudmundson J. Differential effects on bone density
of progestogen-only methods for contraception in premenopausal women. Contraception 1995; 52: 35–39.
15. Tang OS, Tang G, Yip PS, Li B. Further evaluation on long-term depot-medroxyprogesterone acetate use and bone mineral density: a longitudinal cohort study. Contraception 2000; 62: 161–164.
16. Virutamasen P, Wangsuphachart S, Reinprayoon D, Kriengsinyot R, Leepipatpaiboon S, Gua C. Trabecular bone in long-term depot-medroxyprogesterone acetate users. Asia Oceania J Obstet Gynaecol 1994; 20: 269–274.
17. Taneepanichskul S, Intaraprasert S, Theppisai U, Chaturachinda K. Bone mineral density during long-term treatment with Norplant implants and depot medroxyprogesterone acetate
: a cross-sectional study of Thai women. Contraception 1997; 56: 153–155.
18. Taneepanichskul S, Intaraprasert S, Theppisai U, Chaturachinda K. Bone mineral density in long-term depot medroxyprogesterone acetate
acceptors. Contraception 1997; 56: 1–3.
19. Orr-Walker BJ, Evans MC, Ames RW, Clearwater JM, Cundy T, Reid IR. The effect of past use of the injectable contraceptive depot medroxyprogesterone acetate
on bone mineral density in normal post-menopausal women. Clin Endocrinol (Oxf) 1998; 49: 615–618.
20. Merki-Feld GS, Neff M, Keller PJ. A prospective study on the effects of depot medroxyprogesterone acetate
on trabecular and cortical bone after attainment of peak bone mass. BJOG 2000; 107: 863–869.
21. Polatti F, Perotti F, Filippa N, Gallina D, Nappi RE. Bone mass and long-term monophasic oral contraceptive treatment in young women. Contraception 1995; 51: 221–224.
22. Recker RR, Davies KM, Hinders SM, Heaney RP, Stegman MR, Kimmel DB. Bone gain in young adult women. JAMA 1992; 268: 2403–2408.
23. Kristal AR, Feng Z, Coates RJ, Oberman A, George V. Associations of race/ethnicity, education, and dietary intervention with the validity and reliability of a food frequency questionnaire: the Women's Health Trial Feasibility Study in Minority Populations. Am J Epidemiol 1997; 146: 856–869.
24. Melton LJ 3rd, Khosla S, Atkinson EJ, O'Fallon WM, Riggs BL. Relationship of bone turnover to bone density
and fractures. J Bone Miner Res 1997; 12: 1083–1091.
25. Ott S, Scholes D, LaCroix A, Ichikawa L, Yoshida C, Barlow W. Effects of contraceptive use on bone biochemical markers in young women. J Clin Endocrinol Metab 2001; 86: 179–185.
26. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a generalized estimating equation approach. Biometrics 1988; 44: 1049–1060.
27. Sowers M, Corton G, Shapiro B, et al
. Changes in bone density
with lactation. JAMA 1993; 269: 3130–3135.
28. Recker R, Lappe J, Davies K, Heaney R. Characterization of perimenopausal bone loss: a prospective study. J Bone Miner Res 2000; 15: 1965–1973.
29. Taylor BC, Schrieiner PJ, Stone KL, Cummings ST, Nevitt MC, Ensrud KE. Long-term prediction of incident hip fracture risk in older white women. J Bone Miner Res 2001; 16: S139.
30. Cummings SR, Black DM, Nevitt MC, et al
. Bone density
at various sites for prediction of hip fractures. The Study of Osteoporotic Fractures Research Group. Lancet 1993; 341: 72–75.
31. Sowers M, Randolph J, Shapiro B, Jannausch M. A prospective study of bone density
and pregnancy after an extended period of lactation with bone loss. Obstet Gynecol 1995; 85: 285–289.
32. Grinspoon S, Thomas E, Pitts S, et al
. Prevalence and predictive factors for regional osteopenia in women with anorexia nervosa. Ann Intern Med 2000; 133: 790–794.
33. Keen AD, Drinkwater BL. Irreversible bone loss in former amenorrheic athletes. Osteoporosis Int 1997; 7: 311–315.