Eating disorders and bone health: A missed opportunity?

These ‘tales’ highlight uncertainties regarding how bone health of patients’ with eating disorders is currently managed within primary and secondary care. We hope to raise awareness of the importance of diagnosing low bone mineral density (BMD) in anorexic patients, and current evidence around treatment options.

Eating disorders are psychiatric conditions associated with a reduction in BMD and fractures. Anorexia nervosa (AN) is condition characterised by low weight due to restriction of energy intake, associated with fear of and/or behaviour that interferes with weight gain. ¹

The lifetime prevalence of AN in women is 2–4%, with AN being twice as common in teenage girls. The typical age of onset of AN is 15 years and, on average, it takes an anorexic patient 5–6 years from diagnosis to recovery. However, up to 30% of patients with AN never fully recover. ²

Loss of BMD is common in those with AN. Ninety percent of community-dwelling adult women with AN have a BMD of ≥1 standard deviation below the young adult mean (T-score <−1), which is traditionally termed osteopenia. Osteoporosis (T score < −2.5) is diagnosed in 40%.^3–5 Consequently, AN is associated with a 3-fold increase in the lifetime risk of fractures. ⁶ Of particular concern are hip fractures, which have a 29% mortality within 12 months, with 53% of hip fracture patients no longer able to live independently. ⁷ Osteoporosis is a leading cause of morbidity and mortality in later years and exerts a significant burden on health and social care costs. The National Institute for Health and Care Excellence (NICE) estimates the annual cost of osteoporosis to the National Health Service to be £4.4 billion. ⁸

The pathophysiology of reduction in BMD in anorexic patients is multifactorial, with hypogonadism, increased cortisol and low insulin-like growth factor-1 (IGF-1) all implicated. ⁹

A woman’s peak BMD is usually acquired by her mid-20s, after which bones undergo continual remodelling until the menopause, when bone resorption begins to outweigh new bone formation, with a consequential loss of bone strength. Given the high prevalence of AN in adolescence, the insult on bone health frequently occurs at a critical stage of skeletal development and, unless managed effectively, can result in morbidity in later life.

The following cases highlight some of the difficulties experienced by anorexic patients and the potential benefits of improving education around the prevention, diagnosis and management of low BMD in these women.

Case study one, is a 17-year-old girl, Katie, who presented to the GP surgery with a 2-year history of amenorrhoea. She reached menarche at 14 years old, and her periods were initially regular. However, after developing AN 2 years ago, Katie became amenorrhoeic. Her BMI was 16 at diagnosis and has been consistently between 16 and 17 since. She was referred to the local eating disorder clinic and has been seen regularly by their team for treatment. Katie came for consultation with her mum with concerns about amenorrhoea and the potential impact of this on her future bone health and fertility. Her case was discussed with the local gynaecology advice and guidance team to ask how they would recommend managing her amenorrhoea, and whether any treatment was advised. The response from the Consultant Gynaecologist was to, ‘put her on a combined oral contraceptive (COC) to induce a monthly bleed’.

There are limited data to guide the effective treatment of low BMD in anorexic patients. Psychological therapies, such as eating-disorder-focused cognitive behavioural therapy (CBT-ED) and specialist supportive clinical management (SSCM) aim to encourage healthy eating and restore body weight. Although weight recovery remains the mainstay of treatment, this often takes time and in the interim other therapeutic interventions may have a role in protecting BMD.

Given that estrogen is known to protect against bone resorption and AN is a hypo-estrogenic state, it would follow that estrogen supplementation would be beneficial in AN. However, prospective studies, with follow-up ranging from 9 to 24 months, have failed to demonstrate benefit of oral estrogen on BMD in AN.^10–13 Estrogen is known to suppress hepatic production of IGF-1, a potent bone-trophic hormone, which is already significantly reduced in AN. It is postulated that the supraphysiological serum estrogen concentration resulting from oral estrogen supplementation may cause further reduction of IGF- 1.^14,15

Lower dose oral and transdermal estrogen have shown more promising results. A recent randomised, placebo-controlled, double-blind study of girls with anorexia nervosa suggests that, when administered at physiological doses, estrogen replacement may protect the bones of those with AN. ¹⁶ During the study, 110 participants were randomised to estrogen replacement at physiological levels or placebo. Those in the estrogen-replacement group received either transdermal 17β-estradiol (100 μg patch twice weekly), if they were >15 years, or escalating doses of ethinyl estradiol (3.75 μg daily, increasing by 3.75 μg every 6 months to 11.35 μg). Oral medroxyprogesterone acetate (2.5 mg daily for 10 days each month) was provided for endometrial protection in those using transdermal estrogen. Follow-up over 18 months with DEXA demonstrated higher BMD in the lumbar spine and hip in the estrogen-replacement group compared to the placebo-control group. Moreover, at 18 months the BMD in those given physiological estrogen-replacement was consistent with normal-weight controls. No comparison was made between transdermal and low dose oral estrogen.

Increase in testosterone levels associated with weight gain in those with AN are predictive of a rise in BMD. ¹⁷ However, data assessing the effect of exogenous testosterone on BMD in AN are few. One trial of 77 patients investigated the effect of transdermal testosterone with and without risedronate on BMD in AN(18). Testosterone was administered at a starting dose of 150 μg/day, with the dose subsequently titrated according to serum testosterone levels. Testosterone did not have a significant effect on BMD over a 12-month period, either when given alone compared with placebo, or in combination with risedronate compared to risedronate alone.

Bisphosphonates have been shown to improve BMD at the spine and hip. ¹⁸ However, special consideration must be made with regards to the use of these medications in women of childbearing age, as bisphosphonates have a long half-life and can adversely affect a foetus.

Current NICE guidance on the treatment of low BMD in people with AN suggest: ¹⁹

1. Consider transdermal 17-ß-esradiol (with cyclic progestogen) for young women (13–17 years) with AN who have long-term low body weight and low bone mineral density with a bone age over 15 years.

2. Consider incremental physiological doses of estrogen for young women (13–17 years) with AN who have delayed puberty, long-term low body weight and low bone mineral density with a bone age over 15 years.

3. Consider bisphosphonates for women (18 years and over) with AN who have long-term low body weight and low bone mineral density.

Case study two, Jane, is a postmenopausal, 56-year-old lady with a long history of anorexia since her teenage years. Jane’s current BMI is 18 and she has a diagnosis of osteoporosis, for which she takes alendronate and Calcichew D3 tablets. She also has depression and takes sertraline. Jane discussed her experience of anorexia and her current menopausal symptoms, which included hot flushes, insomnia, night-sweats and arthralgia. Jane had only recently been offered a dual energy X-ray absorptiometry (DEXA) scan, following a fragility fracture and had not had any previous advice regarding bone health. Following appropriate counselling and blood pressure check, Jane was started on transdermal estrogen with continuous micronised progesterone for her menopausal symptoms and to optimise her bone health. The plan is to provide a follow-up appointment to assess her symptoms and offer her repeat DEXA scan in 3 years.

Current NICE guidance on the treatment of eating disorders suggests considering a scan to assess BMD in the following scenarios:

1. One year of underweight in children and young people, or earlier if they have bone pain or recurrent fractures.

2. Two years of underweight in adults, or earlier if they have bone pain or recurrent fractures.

There is no clear consensus on the optimal frequency of DEXA scans in those with low BMI. However, in Jane’s case, a repeat DEXA scan may be useful in deciding whether she should continue alendronate, given the rare, but serious side effects of long-term anti-resorptive medication. NICE guidance on eating disorders suggests:

1. Repeat BMD scans in those with ongoing persistent low weight, especially when using or deciding whether to use hormonal treatment.

2. Not to repeat BMD scans for people with anorexia nervosa more frequently than once per year, unless they develop bone pain or recurrent fractures.

Given bone loss occurs early in anorexia and can progress rapidly, many physicians are arguing for baseline bone density screening at an earlier stage, when AN or amenorrhoea is present for more than 6 months. This was recently discussed with our local eating disorder clinic, whose current guidance is to offer only the most severe, hospitalised cases a DEXA scan. They explained that they do discuss risks to motivate the patients in terms of weight gain and very rarely prescribe transdermal estrogen and cyclical progesterone.

Case three is a 36-year-old athlete, Anna, who’s BMI has been below 18 since her late teens. Although Anna describes a complex and difficult relationship with food, she has never formally been diagnosed with anorexia. Anna has been amenorrhoeic for many years and has a history of multiple stress fractures. This has stopped her from competing and has significantly impacted her mental health. A DEXA scan revealed osteopenia, with a T-score of −1.5. Following a long discussion, Anna was happy to be prescribed transdermal estrogen 100 μg patches and cyclical micronised progesterone. Anna will be followed up with regular DEXA scans and has also been offered referral to the eating disorder clinic.

Anna has a diagnosis of relative energy deficiency in sport (RED-S). This is a condition that can affect athletes and is characterised by low energy availability, caused by restrictive eating and high intensity training. Similar to AN, low energy availability in RED-S results in amenorrhoea and reduced BMD. The condition is more common in sports that require a low BMI for both performance or from an aesthetic perspective. The diagnosis is often made retrospectively, following multiple stress fractures and poor athletic performance. Given the adverse effect of RED-S on an athlete’s career, it can be hugely detrimental to mental health. Treatment recommendations include weight gain, transdermal estrogen and cyclical micronised progesterone. Treatment with the combined oral contraceptive (COC) is not recommended, as it has not been demonstrated to protect bones and masks menstrual cycle disturbance. ¹² This must be also considered and discussed with the patient when considering the COC as a contraceptive method in athletes.

Case four, Manisha, is a 19-year-old girl who reached menarche aged 13 years and had regular cycles until 18 months ago, when she became amenorrhoeic due to AN. Since this time, her BMI has been consistently between 16 and 17. Manisha was referred to the adult eating disorders clinic, but the referral was rejected because her illness was not severe enough and there was no capacity for her to be seen. Advice was sought from the eating disorders clinic as to how her bones could be protected and weight gain was the suggested course of action. Manisha has no specialist eating disorder support available to her and continues to be seen regularly by her GP. She has had general mental health input for anxiety but has struggled to gain weight. A DEXA scan has been requested.

This case illustrates the lack of support and guidance for GPs on how to approach bone health in those with AN and amenorrhoea. This lack of guidance results in wide variation the management of these patients. We performed a retrospective study of the management of bone health in anorexic patients within primary care. We used EMIS web to search for patients across four local GP practices. The inclusion criteria were women <45 years old, coded as having a diagnosis of AN, with amenorrhoea due to low BMI. The primary outcome of interest was the management plan, specifically whether they had been given HRT or a COC and whether a DEXA scan had been performed.

Fifty-seven patients were included, with average age of diagnosis of AN of 17 years. Eighteen patients (32%) had had a DEXA scan. Of those who had a DEXA, 14 (77%) were demonstrated to have osteopenia or osteoporosis. The majority of these DWXA scans were performed later in life, once BMI had been restored for several years, suggesting that the AN-induced hypothalamic-pituitary-ovarian axis suppression had had a long-lasting effect on BMD.

Sixteen patients (28%) were on a COC; fifteen for contraception, with one patient started on the COC ‘to improve bone health’. No patients were started on transdermal estrogen whilst amenorrhoeic, despite several being managed by the eating disorders service or endocrinology.

It is clear from our study that greater attention needs to be paid to the skeletal health of anorexic patients in both primary and secondary care. Given the prevalence of AN, and the fact that so many of these patients will only ever be managed in primary care, better guidance for GPs on the appropriate management of bone health in AN is required.

Management should be multidisciplinary and individualised, with a combination of psychological strategies to encourage weight gain, estrogen therapy prescribed in line with current guidance and, where indicated, bisphosphonates. DEXA is a useful tool to diagnose low BMD, guide management and assess treatment response.

More good quality research data are needed to further ascertain the benefits of transdermal estrogen and cyclical progesterone on bone health and to guide an appropriate dosing strategy in these vulnerable patients.