Medical complications are frequently seen in patients with anorexia nervosa (AN) and bulimia nervosa (BN).Loss of bone mineral density (BMD), though not often the initial focus of eating disorder treatment, can persist long after apparent recovery from AN. It is a well-established consequence of poor nutritional intake, and is particularly concerning for the developing bodies of adolescent patients. Adolescence is a critical time for bone development, because 40% to 60% of peak bone mass accrues during these years.Unfortunately, the peak age of onset for AN also occurs during adolescence. The time of onset, type and duration of the eating disorder, as well as the degree of nutritional deficit, will determine whether peak adult bone mass can be achieved.
Since low bone density is most often associated with AN, BN is not considered a risk factor.If bone accrual is interrupted by the malnutrition associated with AN, life-long adverse sequelae to bone health can occur. These include long-term increased risk of fractures, osteoporosis, and limited longitudinal growth, impairing eventual height. Although osteoporosis is considered an adult disease, there is increasing evidence that factors in childhood contribute to its development and morbidity in adulthood.3
Osteoporosis is defined as a “reduction in bone mass and disruption of bone architecture, resulting in reduced bone strength and increase of bone fractures.”Osteopenia is also defined as reduced bone mass, but this is less severe than in osteoporosis.  Among adults, the type of bone density is determined by bone densitometry, defined by a T-score of -1 to -2.5 for osteopenia, and -2.5 or lower for osteoporosis.However, these definitions are not applicable to adolescent and pediatric populations. According to a systematic review of bone health in AN, there is a paucity of available data regarding bone health and prevalence of poor bone health in adolescents.2
Pathophysiology of Low Bone Marrow Density
Bone is dynamic, with ongoing formation of new bone as well as continuing resorption of existing bone. In AN, malnutrition leads to an uncoupling of bone formation and bone resorption. Thus, high rates of bone resorption with decreased bone formation lead to rapid bone loss, potentially resulting in osteopenia and osteoporosis.This is secondary to disruption of hormonal levels and regulatory processes of the hypothalamic-pituitary-adrenal (HPA) axis and hypothalamic-pituitary-gonadal (HPG) axis. Severe malnutrition causes an adaptive state of hypercortisolemia. High cortisol levels directly inhibit the HPG axis, impair calcium absorption and renal processing of calcium, inhibit osteoprotegerin (OPG) secretion, which normally inhibits bone reabsorbing osteoclasts, and increases the receptor activator of nuclear factor kappa B ligand  (RANKL), which also increases osteoclast activity.Furthermore, the states of low energy and hypercortisolemia both have an inhibitory effect on the HPG axis, leading to a hypogonadal state, which is characterized by low estrogen levels and amenorrhea.
Estrogen normally inhibits bone resorption by blocking secretion of inflammatory cytokines and RANKL. In males, low testosterone levels similarly have an adverse effect on bone density.2  Pubertal changes and growth during adolescence are directly dependent on rising levels of sex hormones, which in turn stimulate increased secretion of growth hormone (GH) and insulin-like growth factor (IGF-1). GH and IGF-1 are essential for bone accrual. In patients with AN, studies show a state of GH resistance.This in turn leads to further decreased bone formation. IGF-1 levels are thus reduced in AN, which inhibits osteoblast bone-forming activity.
Diagnostic Imaging in Adolescents
The preferred method for assessing the state of bone health in AN is with dual-energy x-ray absorptiometry (DXA) scans. This provides an image of two-dimensional areal bone mineral density (aBMD) and bone mineral content (BMC). Any individual with a 9- to 12-month history of AN should have a DXA scan, with follow-up every 2 years while their eating disorder is active.Appropriate interpretation of these values in adolescents relies on the use of Z-scores, as opposed to T-scores, which represent standard adult reference points for older patients.  DXA scans are seen as the preferred screening method because of their relatively low cost, low degree of radiation exposure, and widespread availability.
Adolescent Data Review
Several studies have demonstrated increased fracture risk in adults with AN, but few have focused on fracture risk in adolescents. Faje et.al (2014) were the first to examine the risk of childhood and adolescent fractures. The authors compared the risk among 310 patients with AN compared to normal-weight controls. The average age of the subjects was 16 years, with a mean duration of AN of almost 2 years.  Faje et al. found increased fracture prevalence and lower aBMD in those with AN compared to controls.Though a single aBMD may not be enough to extrapolate the fracture risk for a patient with AN, the overall increased fracture risk in patients with AN seems to be related to bone strength and accrual disruptions in adolescents.  Shepherd et al. (2018) found that bones are 10% smaller in adolescents with AN compared to healthy children. Their retrospective study of 111 patients under the age of 20 was the first to examine bone accrual in relation to linear growth.7
One study analyzed limitations of aBMD as a focal measurement of bone health. Singhal et al. (2018) conducted the first study to evaluate both bone microarchitecture and strength with high-resolution peripheral quantitative computed tomography (pQCT). The authors found that microarchitectural changes in the bone may precede measurable deficits in aBMD in females with AN.Their results further illustrate that aBMD does not fully capture all the detrimental effects of malnutrition affecting the bone, especially in younger patients. They also found  that adolescents and young adults with AN had thinner and more porous cortices, lower trabecular bone volume fraction with lower trabecular number and spacing, and higher levels of marrow adipose tissue. Higher levels of marrow adipose tissue have previously been associated with greater fracture risks.
Treatment Recommendations
Weight restoration remains the primary goal of treatment for AN.Present data supports that negative energy states increase bone resorption independent of other hormonal influences.Therefore, normalization of nutritional intake and weight gain are the upmost priorities for treatment of an adolescent with AN, and are necessary to stave off further bone mineral loss. Beyond weight restoration, other treatments to improve BMD in adolescents are somewhat controversial.
There are some other potential treatments that may be considered. Physical activity is generally viewed as being protective of bone health in older, normal-weight individuals. Current evidence, however, has not been able to substantiate such a protective effect in the context of the low-energy state and amenorrhea that characterize AN.2  Caution is often recommended,  with risk/benefit consideration, before allowing aerobic physical activity in patients with AN. Exercise can interfere with the primary goal of weight restoration, and may be deleterious to bone microarchitecture at low body weights.10  DiVasta et al. (2016) examined whether low-magnitude mechanical stimulation (LMMS) contributes to normalization of bone turnover. Noting that bed rest is often required for critically ill patients with AN, the authors wondered if the suppression of bone turnover associated with bed rest could be attenuated by LMMS. In their randomized, double-blind, placebo-controlled trial, 100 females aged 13-21, admitted to the hospital with medical complications of AN, were assigned to one of two groups: (1) an LMMS platform or (2) a placebo platform. The LMMS platform delivered small vibrations, and the placebo platform made an identical noise but did not vibrate.
DiVasta and colleagues found that participants assigned to the LMMS platform had stabilization in markers of bone formation without adverse effects on weight restoration. In contrast, those who used the placebo platform exhibited significant decreases in bone formation markers.10 These findings indicate that LMMS may be a safe, noninvasive, non-pharmacologic method to maintain bone health.
Pharmacologic treatment is often used in adult populations for low BMD, but limited data exist for adolescent patients with eating disorders. Hormonal replacement is incorrectly recommended as a means of preventing bone resorption associated with the hypogonadal condition seen in eating disorders. Often, bone loss in AN is inaccurately compared to the osteoporosis of postmenopausal women. A recent study found that the bone loss in AN evolves as a result of different pathophysiological mechanisms. In this study, a comparison of bone microarchitectures using high-resolution pQCT of young AN patients, postmenopausal subjects, and controls found that the reduction of trabeculae is rapid in AN and comparable to that in postmenopausal women, but the cortical and subcortical  bones were less compromised in AN.11 Additionally, great caution is imperative when discussing hormonal replacement in adolescents since increased hormone levels (estrogen and testosterone) can lead to premature closure of growth plates. Overall, a systematic review by Robinson et al. (2017) found discrepancies in outcomes when exogenous hormone replacement was used. The review noted one study that found physiologic replacement doses of transdermal 17-bestradiol increased spinal and hip BMD in adolescent women with AN, although “complete ‘catch-up’ to a comparable BMD in health controls did not occur.12 Physiologic estrogen replacement achieved by transdermal administration does improve BMD in adolescents with AN, though weight gain appears to improve BMD more.9 A number of studies have suggested that the use of estrogen/progesterone combination oral contraceptives has been ineffective and could also be detrimental to bone health, due to suppression of endogenous gonadal secretions as well as suppression of systemic IGF-1 secretion.In contrast, the “physiologic” doses achieved through transdermal application may be effective because this does not suppress IGF-1 and it avoids hepatic first-pass metabolism. According to the American College of Obstetricians and Gynecologists, the combination of transdermal estradiol and cyclic oral progesterone for sexually active adolescents, to treat loss of BMD, is not effective for preventing pregnancy. 13
Bisphosphonates are also used in adult women with AN  to treat their osteoporosis but there are few studies of bisphosphonate use in adolescents with AN. Golden et al. (2005) conducted a double-blind, randomized trial comparing alendronate with placebo in 32 adolescent females. They found that in the treatment group, there was a positive independent effect on BMD, increasing it at the femoral neck and lumbar spine. However, the authors concluded that weight restoration remained the most important determinant of BMD and recommended that until additional studies could demonstrate efficacy, the use of bisphosphonates should be limited to controlled trials.14 The adverse side effects associated with bisphosphonates limit their use in adolescent females.12Bisphosphonates have a long half-life and can be released slowly from bone over a period of years.There is also a teratogenic risk, which is important to consider. 12 Yet, it is interesting to note that bisphosphonates have been used for years in adolescent patients with osteogenesis imperfecta without reports of significant adverse effects.1  There are, as yet, no studies of teriparatide or denosumab in adolescent patients with AN.
Conclusion
There are some notable limitations in reviewing currently available data for adolescents. First, there are more available data regarding medical complications in adults than in adolescents. Of the studies that did focus on adolescent populations, most were retrospective in nature and only evaluated female patients with AN. There are very little data on adolescent males with AN. Additionally, there are other limitations when conceptualizing and defining BMD abnormalities in children and adolescents. One interesting point is that the International Society of Clinical Densitometry (ISCD) recommends that the term “osteopenia” be limited to adult patients with mild deficits in bone mass. In fact, a DEXA Z-score of £-2 is labeled as “below the expected range for age” for adolescents (Misra et al., 2015, p. 3).  The Pediatric Position Development Conference (PDC), a subgroup of the ISCD, limits diagnoses of osteoporosis in children and adolescents by requiring a BMD Z-score of £-2 and at least one vertebral compression fracture or the presence of a significant fracture history (fractures of two or more long bones by the age of 10 years or three or more long bone fractures at any age up to 19 years). The DXA scan itself may not be as useful in adolescents and, as studies mentioned earlier have indicated, it will be important to consider microarchitecture in future studies that evaluate bone outcomes in AN.
Use of HR-pQCT may provide more accurate assessment of bone health, and may be more accurate in predicting prognosis as well as for reliably tracking outcomes throughout treatment. 8, 15All studies have commented on the need for more longitudinal data and for judicious deliberation before initiating medicinal treatments.  But, in the meantime, some assessment of the state of bone health in patients with a history of AN is a prudent practice and is the first step to initiate further treatment considerations, given the high prevalence of low BMD in adolescents with AN.
References
  1. Mehler PS, Andersen AE. Eating Disorders: A Guide to Medical Care and Complications(3rd ed.). Baltimore, USA: Johns Hopkins University Press, 2017.
  2. Misra M, Golden N H, Katzman DK. State of the art systematic review of bone disease in anorexia nervosa. International Journal of Eating Disorders.2015; 49(3), 276–292.
  3. Steelman J, Zeitler P. Osteoporosis in pediatrics. Pediatrics in Review.2001;22(2), 56-65. doi:10.1542/pir.22-2-56
  4. Masi L. Epidemiology of osteoporosis. Clinical Cases in Mineral and Bone Metabolism.2008; 5(1): 11-13
  5. Karaguzel, G, Holick, M. (2010). Diagnosis and treatment of osteopenia. Reviews in Endocrine and Metabolic Disorder. 2010;11(4), 237-251.
  6. Faje AT, Fazeli P K, Miller KK, et al. (2014). Fracture risk and areal bone mineral density in adolescent females with anorexia nervosa. International Journal of Eating Disorders.2014; 47(5), 458–466.
  7. Shepherd S, Kyriakou A, Shaikh MG, et al. Longitudinal changes in bone parameters in young girls with anorexia nervosa. 2018; 116, 22–27.
  8. Singhal V, Tulsiani S, Campoverde K J, et al. Impaired bone strength estimates at the distal tibia and its determinants in adolescents with anorexia nervosa. 2018;106, 61–68.
  9. Swenne I, Stridsberg M. Bone metabolism in adolescent girls with eating disorders and weight loss: independent effects of weight change, insulin-like growth factor-1 and oestradiol. Eating and Weight Disorders – Studies on Anorexia, Bulimia and Obesity.2014; 20(1), 33–41.
  10. DiVasta AD, Feldman HA, Rubin CT, et al. The ability of low-magnitude mechanical signals to normalize bone turnover in adolescents hospitalized for anorexia nervosa. Osteoporosis International.2016; 28(4), 1255–1263.
  11. Milos G, Häuselmann H, Krieg M, et al. Are patterns of bone loss in anorexic and postmenopausal women similar? Preliminary results using high resolution peripheral computed tomography. 2014; 58, 146–150.
  12. Robinson L, Aldridg, V, Clark E M, et al. (2017). Pharmacological treatment options for low bone mineral density and secondary osteoporosis in anorexia nervosa: a systematic review of the literature. Journal of Psychosomatic Research.2017; 98, 87–97.
  13. ACOG Committee Opinion No. 702: Female Athlete Triad. (2017). Obstetrics & Gynecology.2017; 109 (6), 511-1516.
  14. Golden NH, Iglesias EA, Jacobson MS. Alendronate for the treatment of osteopenia in anorexia nervosa: a randomized, double-blind, placebo-controlled trial. Journal of Clinical and Endocrinologic Metabolism.2005; 90(6):3179–3185.
  15. Roggen I, Vanbesien J, Gies I, et al. Forearm bone mineralization in recently diagnosed female adolescents with a premenarchal onset of anorexia nervosa. International Journal of Eating Disorders, 2016; 49(8), 809–812.