Human Reproduction, Vol. 15, No. 1, 118-122,
January 2000
© 2000 European Society of Human Reproduction and Embryology
Bone mineral density during long-term use of the progestagen contraceptive implant Implanon® compared to a non-hormonal method of contraception
1 Streekziekenhuis Koningin Beatrix, Winterswijk, The Netherlands, 2 Research and Development, NV Organon, Oss, The Netherlands, 3 Chilean Institute of Reproductive Medicine, Santiago, Chile and 4 Department of Obstetrics and Gynaecology, University of Oulu, Oulu, Finland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
An open, prospective, comparative study was done in healthy women, aged between 18 and 40 years, to study the effects of long-term etonogestrel treatment on bone mineral density (BMD). The control group used a non-hormone-medicated intrauterine device (IUD). The BMD was measured using a dual energy X-ray absorptiometry instrument. Measurements included the lumbar spine (L2–L4), the proximal femur (femoral neck, Ward's triangle, trochanter) and distal radius. The period of treatment was 2 years and 44 women in the Implanon® group and 29 in the IUD group provided data. Groups were comparable at baseline with respect to age, weight, body mass index, BMD and 17ß-oestradiol status. Changes from baseline in BMD in the Implanon® group were not essentially different from those in the IUD group. There was no relationship between 17ß-oestradiol concentrations and changes in BMD in this study population. The results of the present study indicate that Implanon® can safely be used in young women who have not yet achieved their peak bone mass.
Key words: bone mineral density/contraceptive implant/etonogestrel/3-ketodesogestrel/17ß-oestradiol
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Implanon® is a single-rod implant (NV Organon, Oss, The Netherlands) made of an ethylene vinyl acetate copolymer (EVA) with a core containing 68 mg of etonogestrel (3-ketodesogestrel). The implant has a length of 40 mm and a diameter of 2 mm and is provided in a sterile disposable applicator for subdermal insertion. Contraceptive action is mainly by inhibition of ovulation and lasts for 3 years (Mäkäräinen et al., 1998
).
Ovarian oestrogen production is suppressed to early follicular phase concentrations, especially during the first 6 months of use. Upon continued use, oestrogen concentrations show an increase but, due to the ovulation inhibition also present in the third year, oestrogen lacks the cyclical peaks (Mäkäräinen et al., 1998). Oestrogen deficiency leads to bone loss in pre-menopausal women or into achieving a lower peak bone mass than their age-related peers, who have a normal oestrogen status. Hypothalamic amenorrhoea in young women is associated with low serum 17ß-oestradiol concentrations and low bone mass. In situations where amenorrhoea is induced by medical treatment such as administration of depot medroxyprogesterone acetate (DMPA), a lowered oestrogen status is also observed. During use of Implanon®, amenorrhoea occurs in ~20% of women (Croxatto et al., 1999). It is therefore important to examine the bone mineral density (BMD) of the women using this implant, together with the oestrogen concentrations, for possible associations. A randomized design would have been most desirable from a scientific point of view. However, the contraceptive choice of the women being offered such diverse methods as an implant or an intrauterine device (IUD) has to be respected and does not allow randomization.
The study presented here is a prospective, comparative study of a progestagen-only hormonal contraceptive implant and a non-hormone medicated IUD, for the duration of 2 years.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In all, 79 women who chose a long-term form of contraception were informed about the study. Women interested were offered the choice between Implanon® and a non-hormone-medicated IUD. Moreover, women already fitted with an IUD were also allowed to be enrolled. Permission to carry out the study was obtained from the health authorities and the Ethics Committees of the three countries/centres. All women gave informed consent in writing. The ratio of Implanon®: IUD use was chosen to be 1.5:1. Women had to fulfil a set of inclusion and exclusion criteria. The most important were: age between 18 and 40 years, good physical and mental health, not suffering from a condition (present and history) affecting bone metabolism, not on medication affecting bone metabolism, and a weight between 80 and 130% of ideal (Metropolitan Height and Weight Tables, 1983
). Women with significant scoliosis, hindering the BMD measurements, were also not enrolled. Present use of >2 units of alcohol per day and smoking >10 cigarettes a day were exclusion criteria. Likewise, women engaging in rigorous exercise, such as the marathon or triathlon, were not included. To avoid the risk of significant baseline differences between treatment groups, we aimed at a balance in age and weight categories per centre. In this way, two important confounders could be taken into account at enrolment. The initial laboratory assessments included measurement of calcitonin, parathyroid hormone (PTH), cholesterol, triglycerides and prolactin. Results had to be within the laboratory reference range, to allow inclusion into the study. During 4 successive weeks, oestradiol was measured twice a week, to get an impression of the women's pre-trial oestrogen status. This schedule was repeated at months 12 and 24, or at the moment when the women decided to stop participation in the study. At month 6, only one measurement in total was done, for logistic reasons. The three study centres comprised one urban (Santiago, Chile), one rural (Winterswijk, The Netherlands), and one suburban area (Oulu, Finland). The women were of Latin American and European ethnicity.
The Lunar DPX software Lunar Corporation, Madison, WI, USA, allowed comparison to different reference populations. For analysis of our data we used the Lunar company's female-USA/Europe dataset. To apply to our data, average BMD values of the age brackets 20–29 and 30–39 years were taken.
BMD measurements (g/cm2) were transformed into standard deviation (SD) scores thus providing z scores (equivalent to T-scores), by means of the following equation:
 |
 |
BMD measurements were done in the anteroposterior position at the following anatomical sites: lumbar spine (L2–L4), proximal femur (femoral neck, Ward's triangle, trochanter) and the distal radius. These measurements were performed at baseline, after 6 months of treatment, after 12 and 24 months. If a woman wished to discontinue treatment or participation in the study, she was asked to have a `final' BMD measurement done if the previous one had been done >6 months ago.
Changes in the z-score of the BMD were compared between treatment groups, using analysis of covariance, with centre, age, weight at baseline, and treatment as covariates.
The primary parameter was the change in BMD z-score at `last measurement' i.e. the last measurement done in each subject, either at month 24 or upon discontinuation.
Scatterplots were used to evaluate the relationship between (change in) oestradiol and (change in) BMD, and the relationship between weight changes and changes in the BMD. The change in body weight was compared between treatment groups with the Cochran–Mantel–Haenszel test, adjusted for centre.
The study had adequate power (
5%, ß 20%) to detect a treatment difference in change in z-score of 0.30 between Implanon® users and IUD users.
The dual energy X-ray absorptiometry (DEXA) instruments (Lunar Corporation, Madison, WI, USA) at the three different centres were calibrated at the start of the study and at the end of the study. This was done in addition to the routine calibrations done by the staff of the respective centres. In order to have a proper comparison over time, the three study centres should not have shown a difference in equipment calibration during that period. Calibration was done by Bona Fide (Madison, WI, USA) staff using the European Spine Phantom, serial number ESP100. Analysis of the European Spine Phantom scans was performed by the Bona Fide company, who reported their conclusion in writing.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Of the 79 women informed about the study, 76 actually received treatment. Of these, 46 received the implant and 30 an IUD, or were already IUD users. The intention-to-treat group consisted of 44 implant and 29 IUD users because at least one post-baseline measurement was needed. Overall, the two groups were well balanced with respect to age, height, weight and body mass index (BMI). Table I
presents relevant baseline characteristics of the two treatment groups. The majority of women had been pregnant in the past and had one or more children. The contraceptive history showed that particularly the IUD group was homogeneous. Upon entry into the study, 29 women (out of 30) were already fitted with an IUD. In 90% of the IUD users, this was also their previous method of contraception, while 24% of the implant users had been oral contraceptive (OC) users in the past and 26% were past IUD users. The washout period in case of OC use was 1 month. The treatment groups differed with respect to smoking. In the Implanon® group, 41.3% were smokers compared with 23.3% in the IUD group. Women were not allowed to smoke more than 10 cigarettes a day. Smokers were evenly distributed over the three centres.
|
Oestradiol concentrations at baseline were comparable between the two groups (Table I
). There was no correlation between oestradiol concentrations and BMD at baseline (Figure 1). Individuals with the lower BMD did not necessarily have a low oestrogen status and vice versa.
|
Figure 2
shows the median and 5 and 95 percentiles of the oestradiol serum concentrations over time, based on the average individual concentrations per timepoint. The single measurement at month 6 was left out. The timepoints usually covered the mean of eight samples, i.e. 4 consecutive weeks of measurement. The pattern reveals slightly higher median oestradiol concentrations in the implant group, when compared with the IUD group. BMD was generally slightly higher than that seen in the reference population used by the Lunar equipment and was highest in women from the rural area of Winterswijk (The Netherlands) (Table II).
|
|
BMD at baseline in g/cm2 and in z-score, followed by changes over time in z-scores are presented in Table III
. BMD measurements at baseline were comparable between treatment groups for the anatomical sites of the body measured. For the distal radius absolute changes in g/cm2 are given, since no reference population data were available for the distal radius, and therefore z-scores could not be provided. The most important single measurement to be considered was `last measurement'. This measurement included all subjects and was therefore in no way influenced by discontinuers. In the implant group ~20% of the subjects discontinued in the 2-year period, with the highest percentage in The Netherlands (37%) and the lowest in Chile (13%). Of the IUD users, only 7% discontinued. At `last measurement' the mean number of days of exposure to Implanon® was 642 (SD 198.9), whereas the IUD group participated for 700.3 (SD 157.5) days.
|
The clinically significant mean decrease of 1 SD (z-score –1) was not nearly reached at any point. In general, mean increases from baseline were seen except for `femoral neck' in the Implanon® group and `Ward's triangle' in the IUD group, where small decreases were observed. Results from the covariance analysis at the last measurement indicated that the increase in BMD adjusted for centre, weight and age at baseline, was in general slightly greater in the Implanon® than in the IUD group. This applied to all sites of the body measured, except for the femoral neck, where the BMD in the Implanon® group showed a decrease. At none of the anatomical sites was the difference between the IUD and the Implanon® group statistically significant. Table IV
shows the results from the analysis of covariance. The estimated differences between the two treatment groups with 95% confidence intervals for the change from baseline in BMD z-score are presented.
|
Changes in z-scores over time at L2–L4 and the femoral neck are shown in Figures 3 and 4
. There was no progression of `bone loss' in those women with the lowest BMD at baseline (data not shown).
|
|
During the study period, there was a small mean increase in weight in the Implanon® group (+1.9 kg), while there was no change in weight in the IUD group. In those who had an increase in BMD, weight increases as well as decreases occurred. The same applied to those who had a decrease in BMD. Moreover, there was no relationship between oestradiol at baseline and change in z-score of the BMD, neither was there a relationship between change in oestradiol and change in BMD z-score (data not shown).
The calibration of the three Lunar DPX DEXA instruments remained consistent. In the opinion of the Bona Fide company, any significant changes seen in the BMD results were not attributable to instrument calibration changes.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The results of this comparative study show that long-term treatment with etonogestrel does not adversely affect BMD. Following an early report (Cundy et al., 1991
), concern arose regarding long-term treatment with progestagen-only agents and medroxyprogesterone acetate (MPA) in particular. Use of MPA is associated to a high degree with amenorrhoea and low oestradiol concentrations in the young women using it, resulting in reduced BMD. Amenorrhoea and oestrogen deprivation are interconnected in the case of hypothalamic amenorrhoea and possibly also in the case of MPA administration. During use of the etonogestrel implant, where amenorrhoea is seen in ~20% of the subjects (Croxatto et al., 1999) this condition is not interconnected, as was shown previously (Mäkäräinen et al., 1998). In this study, amenorrhoea was associated with constant concentrations of oestradiol and the percentage of women affected varied over the reference periods from 5% (reference period 2) to 17% (reference period 6). These oestradiol concentrations could be in the early as well as in the late follicular phase range, but were not fluctuating. Taking into account all women who had amenorrhoea in our study, there was no relationship with the direction of change of BMD.
During use of the levonorgestrel implant Norplant®, reports show that there is no adverse effect on BMD (Naessen et al., 1995; Cromer et al., 1996). In another study where Norplant® implants were compared with long-term use of DMPA, no difference in effects was found regarding distal and ultradistal radius measurements (Taneepanichskul et al., 1997). More recent studies in long-term (i.e. at least 1 year) users of DMPA showed somewhat conflicting results. One study (Gbolade et al., 1998) concluded that in spite of long lasting amenorrhoea no adverse effect on BMD was shown. In contrast, another report (Cundy et al., 1998) reiterated that DMPA is associated with a significant reduction in BMD even after correction for confounding factors such as smoking. Also, in a population-based cross-sectional study in the USA (Scholes et al., 1999), it was concluded that particularly in young women aged 18–21 years BMD might adversely be affected by administration of DMPA.
Trabecular bone is most sensitive to oestrogen deficiency. Our study comprises extensive measuring of anatomical sites with high trabecular bone content such as lumbar spine, trochanter, Ward's triangle and femoral neck, with the latter site of more mixed composition. The site of the body measured with the highest content of cortical bone in our study was distal radius. During use of Implanon® there was a slight (but not consistent) decrease in BMD of the femoral neck. At no time was a mean decrease observed which even approached the clinically significant magnitude of 1 SD (WHO, 1994). Decreases in BMD in the femoral neck and Ward's triangle, as observed in the Implanon® group and IUD group of the present study, may be expected in view of the age of the study population. This was already indicated by the Lunar Company's database and confirmed in a study in healthy Finnish women (Laitinen et al., 1991). Results of the measurements at L2–L4 and of the femoral neck showed that there was no accelerated bone loss in those women who presented with lowered BMD at baseline. The lumbar spine and the femoral neck are of most predictive value for future fracture risk (Cummings et al., 1993; Marshall et al., 1996). The distal radius showed a consistent small mean increase in BMD when compared to the baseline situation. This increase was of similar magnitude as with Norplant® (Naessen et al., 1995). The predominantly `trabecular sites' showed small increases from baseline. During use of Implanon® there is a small mean increase in body weight over time, although the positive results on BMD are not attributable to this increase in body weight. Results of the present study indicate that Implanon® can safely be used in young women who have not yet achieved their peak bone mass.
|
Notes |
|---|
5 To whom correspondence should be addressed at: Clinical Development Department, NV Organon, PO Box 20, 5340 BH Oss, The Netherlands
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Cromer, B.A., McArdle Blair, J., Mahan, J.D. et al. (1996) A prospective comparison of bone density in adolescent girls receiving depot medroxyprogesterone acetate (Depo-Provera), levonorgestrel (Norplant), or oral contraceptives. J. Pediatr., 129, 671–676.[Web of Science][Medline]
Croxatto, H.B., Urbancsek, J., Massai, R. et al. (1999) A multicentre efficacy and safety study of the single contraceptive implant Implanon®. Hum. Reprod., 14, 976–981.
Cummings, S.R., Black, D.M., Nevitt, M.C. et al. (1993) Bone density at various sites for prediction of hip fractures. Lancet, 341, 72–75.[Web of Science][Medline]
Cundy, T., Evans, E., Roberts, H. et al. (1991) Bone density in women receiving depot medroxyprogesterone acetate for contraception. BMJ, 303, 13–16.
Cundy, T., Cornish, J., Roberts, H. et al. (1998) Spinal bone density in women using depot medroxyprogesterone contraception. Obstet. Gynecol., 92, 569–573.[Web of Science][Medline]
Gbolade, B., Ellis, S., Murby, B. et al. (1998) Bone density in long term users of depot medroxyprogesterone acetate. Br. J. Obstet. Gynaecol., 105, 790–794.[Web of Science][Medline]
Laitinen, K., Välimäki, M. and Keto, P. (1991) Bone mineral density measured by dual-energy X-ray absorptiometry in healthy Finnish women. Calcif. Tissue. Int., 48, 224–231.[Web of Science][Medline]
Mäkäräinen, L., van Beek, A., Tuomivaara, L. et al. (1998) Ovarian function during the use of a single contraceptive implant: Implanon® compared with Norplant®. Fertil. Steril., 69, 714–721.[Web of Science][Medline]
Marshall, D., Johnell, O. and Wedel, H. (1996) Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. Br. Med. J., 312, 1254–1259.
Metropolitan Height and Weight Tables (1983) Statist. Bull., Jan–Jun, 3–10.
Naessen, T., Olsson, S-E. and Gudmundson, J. (1995) Differential effects on bone density of progestogen-only methods for contraception in premenopausal women. Contraception, 52, 35–39.[Web of Science][Medline]
Scholes, D., Lacroix, A.Z., Ott, S.M. et al. (1999) Bone mineral density in women using depot medroxyprogesterone acetate for contraception. Obstet. Gynecol., 93, 233–238.[Web of Science][Medline]
Taneepanichskul, S., Intaraprasert, S., Theppisai, U. and Chaturachinda, K. (1997) Bone mineral density during long-term treatment with Norplant® implants and depot medroxyprogesterone acetate. A cross-sectional study of Thai women. Contraception, 56, 153–155.[Web of Science][Medline]
World Health Organization (1994) Assessment of Fracture Risk and its Application to Screening for Post-Menopausal Osteoporosis: Report of a WHO Study Group. WHO, Geneva (WHO technical report series 843).
Submitted on June 23, 1999; accepted on October 4, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. Bahamondes, C. Monteiro-Dantas, X. Espejo-Arce, A. M. dos Santos Fernandes, J. F. Lui-Filho, M. Perrotti, and C. A. Petta A prospective study of the forearm bone density of users of etonorgestrel- and levonorgestrel-releasing contraceptive implants Hum. Reprod., February 1, 2006; 21(2): 466 - 470. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Massai, L. Makarainen, A. Kuukankorpi, C. Klipping, I. Duijkers, and T. Dieben The combined contraceptive vaginal ring and bone mineral density in healthy pre-menopausal women Hum. Reprod., October 1, 2005; 20(10): 2764 - 2768. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 {blacktriangledown}Etonogestrel implant DTB, August 1, 2001; 39(8): 57 - 59. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||