© 2022 MJH Life Sciences and Contemporary OB/GYN. All rights reserved.
Hormonal implants are generally thought to be safe for bone, although whether a young adolescent would reach their genetically determined PBM while using a contraceptive implant is not known.
Case. SL is a 14.5-year-old girl with heavy periods, dysmenorrhea, and acne seeking menstrual management. Her medical history includes 2 traumatic lifetime upper-extremity fractures and lactose intolerance. Menarche occurred at 12.5 years of age and periods occur every 27 to 32 days, and she routinely misses a few days of school each month because of her bothersome periods. SL takes no medications, including herbal agents or supplements. SL’s mother also had heavy, painful periods as a teen that improved as she got older. She was also recently diagnosed with osteopenia. SL has a boyfriend of 5 months and she confidentially shares that she would like contraception “just in case.” Her mother is open to SL starting a hormonal agent to improve her periods. They ask you what her options are. SL’s story is not atypical. In taking a patient-centered, reproductive justice approach to contraceptive counseling, all options should be offered to SL without provider bias. Of course, there are many prescribing considerations for any adolescent. Clinicians must assess for medical contraindications to contraceptive use while also counseling patients about agent efficacy, potential side effects, confidentiality, access, and ease of and duration of use. Patients and guardians often ask about the potential long-term effects of contraceptive use on fertility and cancer risk, yet rarely ask about the potential skeletal impacts of any given contraceptive. However, clinicians should consider and discuss what is known about the bone health implications of hormonal contraceptive use in adolescence1-4 given that (1) 40% to 60% of one’s adult bone mass is accrued in adolescence; (2) a 10% increase in bone mass accrual during adolescence can potentially reduce adult fracture risk by 50%; and (3) Medicare cost estimates in the United States related to adult osteoporosis and fracture care exceed $50 billion dollars annually, including indirect costs such as productivity loss and caregiving.5
Collagen and noncollagenous proteins form an organic matrix upon which bone mineral, composed primarily of calcium and phosphate, is deposited to form bone. Osteoblasts, osteoclasts, and osteocytes are the cells responsible for forming and resorbing bone, as bones grow in length and diameter, increase in bone mineral density (BMD), change shape, and respond to the daily insults of childhood and adolescence, an overall process known as bone remodeling.6 Bone mass (BM) has the potential to accrue throughout childhood and into young adulthood to a genetically determine peak bone mass (PBM), at which time the skeleton has reached its adult form and a stable state of bone mineral content. While 60% to 80% of one’s potential PBM is genetically predetermined, there are extrinsic factors that can negatively impact attaining optimal PBM, including chronic illness, medication exposure, substance use, physical inactivity, nutritional deficits, and abnormal hormonal milieu.7,8
Maximal rates of BM accrual occur during the dynamic phase of puberty associated with peak height velocity driven by multiple hormones including, but not limited to, sex steroids, growth hormone (GH), and insulin-like growth factor-1 (IGF-1). Estrogen is particularly influential with estrogen receptors found on all skeletal cell types. Estrogen promotes the survival of osteoblasts, the bone building cells, and has a pro-apoptotic effect on osteoclasts, the cells responsible for bone resorption. Net BM accrual depends upon the balance between formation and resorption. During adolescence, bone formation should exceed bone resorption. It is suggested that by late adolescence, over 90% of one’s PBM has been attained.7 Notably, estrogen also influences the pubertal growth spurt via the GH/IGF-1 axis as well as growth plate closure and the cessation of linear growth.9 Given the importance of estrogen in bone physiology, any conditions or treatments altering physiologic estrogen production can negatively impact bone health.
The best clinical indicator of bone health is fracture history. Individuals with a history of fracture resulting from low-impact mechanisms of injury should raise concern for skeletal fragility and warrant a bone health evaluation. While adolescent fractures are not uncommon, with risk estimates ranging from 6% to 40% for girls up to aged 19 years, few pediatric studies evaluate fracture incidence as a primary study end point.10,11 Dual-energy x-ray absorptiometry (DXA) is the clinical and research tool most commonly used to assess BMD, acting as a surrogate marker of bone health. An adolescent’s BMD, as measured by DXA, is compared to a normative database of age- and sex-matched peers to generate a Z-score. A Z-score of 0 suggests average BMD for age, while a Z-score of –2.0 or less means that the adolescent’s BMD is 2 or more standard deviations lower than the average BMD of age- and sex-matched peers in the comparative database. BMD accounts for the majority of bone strength with the primary cause of adult osteoporotic fracture being reduced BMD from age-related bone loss and/or a failure to achieve optimal PBM by young adulthood.12
Unlike measures of height and weight, obtained frequently in childhood and adolescence, BMD measures by DXA are not routinely tracked over time in otherwise healthy individuals. Therefore, for most healthy children and adolescents it is not possible to know their BMD trajectory toward their PBM. Importantly, per the International Society for Clinical Densitometry, “In patients with primary bone disease, or at risk for secondary bone disease, a DXA should be performed when the patient may benefit from interventions to decrease their elevated risk of a clinically significant fracture, and the DXA results will influence that management,” meaning a DXA is warranted only if its results will inform clinical decision-making.13
The risks and benefits associated with the medical care we provide should be weighed and discussed with patients and, when appropriate, with their caregivers. From a pure bone health perspective, allowing the body to do what the body naturally does best is physiologically ideal. This includes supporting a normally functioning hypothalamic-pituitary-ovarian (HPO) axis. Many hormonal contraceptive agents prevent pregnancy and/or alter menstrual patterns by altering this HPO axis. Therefore, such hormonal contraceptive agents have the potential to negatively impact BM accrual in adolescence. Scientifically determining such impact is challenging on multiple levels given numerous available and changing contraceptive formulations, variable contraceptive adherence over time, frequently needed confidentiality in adolescent care, confounding extrinsic impacts on bone, and the inability to randomize and use placebo controls in adolescent contraceptive research. Therefore, available evidence must be interpreted in light of these limitations. Contraceptive options discussed will focus on what is known regarding their bone health impacts, which in no way is meant to diminish the significant roles these agents play in adolescent and adult reproductive health care. There is nothing more important than providing an individual with the tools they desire along their reproductive health journey, including menstrual management and contraception. It can also be argued that preventing the psychosocial impacts and known BM loss associated with 1 or more pregnancies and breastfeeding in the adolescent years may far outweigh any negative bone impacts of contraceptive agents themselves.
Depot medroxyprogesterone acetate (DMPA) is the only hormonal contraceptive agent with a black box warning from the US Food and Drug Administrative regarding its adverse effects on bone.14 This warning was informed by several studies demonstrating BM loss during DMPA use.15-19 The BM loss associated with DMPA is attributable primarily to its effects on the HPO axis resulting in a marked hypo-estrogenic state resulting in increased osteoclast activity at the level of bone. Interestingly, DMPA is also thought to bind to bone glucocorticoid receptors, thereby decreasing osteoblast proliferation.20 The overall effect favors bone resorption over formation, which is counter to normal adolescent bone physiology. Of note, studies show that BM loss is most significant in the first year of use and is at least partially reversible following DMPA cessation.21,22 In older adolescents and adult women who have achieved or nearly achieved PBM, this is reassuring. However, in younger adolescents tasked with gaining considerable BM over several years, the implications of BM loss on future PBM potential are likely more significant. It is unlikely that younger adolescents treated with DMPA for prolonged periods of time attain their genetically predetermined PBM following cessation of DMPA. In regard to fracture risk, in a case-control study of adult DMPA users, there was an association between fracture and prolonged current and prior DMPA use; however, similar studies have not be conducted in adolescents who have very different bone physiology.23 Despite such concerns, an individual’s DMPA use should not be time-limited if it is the best option for them and there is no formal recommendation to assess bone density in DMPA users.24 Lastly, in a convenience sample of older adolescents and young adults with cerebral palsy, there was no difference in BMD Z-scores by DXA in those with and without DMPA exposure. The authors concluded that this population with known BMD concerns should not be denied access to DMPA if it is the best option for them per patient-centered decision-making—a relevant conclusion for all patients.25
There have been few studies regarding the bone health impacts of long-acting, reversible contraceptive (LARC) methods in adolescents. Cromer et al prospectively assessed vertebral BMD changes in adolescent controls (n =17) and users of the NORPLANT implant system (n = 7) at 1 year and again at 2 years (n = 4 controls, n = 3 Norplant users).15 Both groups were of similar age (15.2 years vs 15.5 years for Norplant users and controls, respectively). Reassuringly, BM accrual was similar over time in both groups. The authors propose that estrogen levels are not suppressed as much with this contraceptive implant as they are with DMPA. Such an estrogen threshold hypothesis has been supported by numerous studies and reviewed by Hadji et al in a paper regarding the bone health effects of progestins.26 Additional studies evaluating the 68-mg etonogestrel implant Implanon and the 75-mg levonorgestrel implant Jadelle was primarily conducted in adults. Evidence suggested a reduction in BMD at the midshaft ulna with BMD stability at the distal radius over 18 months of implant use.27 When comparing BMD change in adult women with Implanon vs a nonhormonal intrauterine device (IUD), there were gains in BM in both groups at the spine. Mean estradiol levels were notably higher in the Implanon cohort at all measured time points over 2 years. This is reassuring when considering how important estradiol is for BM accrual in adolescence. It is important to note that routine clinical practice recommendations do not include obtaining DXA scans or estradiol levels in implant users.
There are no studies of the BMD impacts of hormonal IUDs in adolescents. Hormonal IUDs do not significantly suppress endogenous estrogen production. The amenorrhea that is frequently achieved with hormonal IUDs is likely secondary to direct endometrial effects rather than to significant HPO axis suppression.= In a cross-sectional study of adults using the 52-mg levonorgestrel IUD Mirena, there was no difference in forearm BMD compared with adults using a nonhormonal IUD.27 In this same cohort followed prospectively, forearm BMD was stable over time.24 Importantly, in case-control analysis, there is no association between hormonal IUD use and fracture, and some evidence supports a reduced fracture risk in hormonal IUD users.29,30
The bone health impact of progestin-only pills (POPs) in adolescents is similarly understudied. Once again, referring to the estrogen threshold hypothesis, POPs likely do not suppress the HPO axis enough to lower estrogen levels below a threshold that would be harmful to bone.26 That said, whether a young adolescent would accrue as much BM on POP therapy compared with no therapy at all is unknown. Importantly, there are no known associations between fracture incidence and POP use.29
Combined hormonal contraceptives (CHCs) similarly exert their contraceptive and menstrual management impacts by altering the HPO axis, resulting in suppression of endogenous ovarian estradiol production, although not to the extent that DMPA does. Additionally, combined oral contraceptive (COC) use negatively impacts the liver’s production of IGF-1 and is associated with elevated biological markers of bone resorption.31,32 The demonstrated impact of CHCs on adolescent bone health is challenged by multiple available drug formulations and a lack of randomized, placebo-controlled trials. Additionally, studies are flawed by the fact that control groups are often younger than subjects on COCs, and early adolescent bone physiology is drastically different from that of later adolescence, making meaningful conclusions about BM change between cases and controls difficult. All totaled, studies and systematic reviews suggest that treating adolescents with COCs containing ≥ 30 μg ethinyl estradiol (EE) may support bone health better than low-dose (≤ 2-μg EE) pills.15,19,30,32-36 Additionally, the greatest impairment of COC use on BM accrual is thought to be within the first few years following menarche when BM accrual is most significant.32,37 Of note, in a large retrospective study of women from the United Kingdom with COC use, fracture incidence was lower in those who had used COCs compared with those who had not, and longer duration of COC use was associated with lower fracture incidence.38
Very few studies have evaluated the bone health impacts of contraceptive patches and vaginal rings. Massaro et al assessed spinal BMD in young adult women using a contraceptive patch (norelgestromin 150 µg and EE 20 µg) vs a vaginal ring (etonogestrel 120 µg with EE 15 µg) and found no significant change in BMD over 1 year in either group.39 A pilot study with 5 adolescents using the contraceptive patch and 5 similarly aged controls found no change in BM over 1 year in the patch users compared with a nearly 4% increase in total-body bone mineral content in controls.40 IGF-1 levels were notably the same in the patch users compared with the controls.
Regardless of contraceptive used, all clinicians should be counseling adolescents to optimize their bone health. This includes engaging in weight-bearing activity as is possible; avoiding substance use; and having a healthy, adequate, and well-balanced diet. Very few adolescents achieve the recommended daily allowance (RDA) for calcium and vitamin D.41 A healthy individual will maintain blood calcium levels in a normal range at the expense of their bone density if they are not ingesting adequate calcium or are unable to adequately absorb the calcium they do ingest as a result of vitamin D deficiency. The RDA for calcium is 1300 mg daily for 9 to 18-year-olds, which is slightly more than four 8-ounce glasses of milk each day. Since calcium is not easily absorbed in large quantities (up to 500 mg is absorbed at a time), intake must also be spread throughout the day, whether through dietary intake or supplements. Reviewing the Center for Young Women’s Health resource on calcium intake with patients and families can be eye-opening. Similarly, it is very hard to meet the RDA for vitamin D through diet alone, which is why vitamin D deficiency is common.42 Although there are no recommendations to check vitamin D levels in otherwise healthy individuals, meeting the RDA for vitamin D is necessary to support overall adolescent bone health, and some individuals may need more than the RDA to achieve vitamin D sufficiency.43
After reviewing her options, SL is most interested in pursuing a hormonal IUD for menstrual management and for contraception when she needs it. Her mother was concerned about the potential bone health impacts of the other contraceptive agents given her own diagnosis of osteopenia. SL likes that she wouldn’t have to remember to take a medication. Given her lactose intolerance, she avoids dairy, and after reviewing the youngwomenshealth.org website, she and her mother plan to incorporate more calcium and vitamin D into her diet and/or through supplements to support healthy bones.
Optimizing child and adolescent bone health can reduce morbidity and mortality associated with adult osteoporosis, so considering and discussing bone health with adolescents is essential. Unfortunately, it is challenging to fully understanding the impacts of all available hormonal contraceptives on adolescent bone health given inherent limitations in research to date, yet it is our job to use available evidence to inform our patient-centered approach to reproductive health care.
DMPA is associated with BM loss during use. Prolonged use in the first few years following menarche could have the greatest negative impact on bone. It is reassuring that there is substantial BM recovery following DMPA cessation; however, prolonged use in younger adolescents may impair PBM. Prolonged DMPA use may also be associated with adult fracture.
POPs and hormonal IUDs do not substantially suppress the HPO axis and therefore are thought to have no ill effect on BM, although studies in adolescents, especially young adolescents, are not available.
Hormonal implants are generally thought to be safe for bone, although whether a young adolescent would reach their genetically determined PBM while using a contraceptive implant is not known.
Individuals using COCs have demonstrated BM gain during adolescence, although such accrual may be less than what they would have seen without COC treatment. When used, a COC containing ≥ 30 µg EE is likely preferred, if tolerated, from a purely bone health perspective. There is too little evidence regarding the bone impact of vaginal rings and contraceptive patches in adolescents to inform a recommendation.
Engaging in weight-bearing activities as is feasible and ingesting adequate nutrition including meeting the RDA for calcium and vitamin D intakes are important for all adolescents.
The contraceptive agent that is best for a given adolescent is the 1 that they prefer once they have been informed about their use, efficacy, side effects, and risks, and once issues of access, confidentiality, and medical eligibility have been considered. Such information can be overwhelming for patients and caregivers. Clinicians are tasked with the important job of conveying this information in an understandable way without personal bias to support adolescents along their reproductive health care journey.