Endocrinology

Polycystic ovary syndrome: an introductionFunding: The development of this guideline was funded by the Australian Government Department of Health and Ageing, through the Jean Hailes Foundation for Women’s Health on behalf of the PCOS Australian Alliance. Editorial independence: This guideline is editorially independent. The funders were not involved in the development of the guideline and have not influenced the scope or recommendations of this guideline. Polycystic ovary syndrome (PCOS) has recently been shown to affect a striking 12%–21% of Australian reproductive-age women, being more common among those who are overweight or of Indigenous background.1 PCOS can be a frustrating experience for women, a complex syndrome for clinicians and a scientific challenge for researchers, and is a major public health concern. Although reproductive features are prominent, PCOS has potential for major metabolic consequences, including obesity and related type 2 diabetes mellitus (DM2) as well as cardiovascular disease (CVD), all of which are currently national health priority areas.2,3 It also has significant mental health and psychological impact, impairing quality of life (QoL).4,5 Because increased obesity exacerbates incidence, prevalence and severity of PCOS, and weight loss improves reproductive, metabolic and psychological features, lifestyle change should be first-line therapy for PCOS.6 It is estimated that 70% of Australian women with PCOS remain undiagnosed;1 clinical practice is inconsistent;7 psychological issues are under-recognised;5 and there is little focus on lifestyle and prevention, with most services targeting infertility and costly assisted reproductive technology. Given the prevalence, disease burden, health costs and clear gaps in care, PCOS is highlighted in national policy and has been prioritised by government, with funding for development of a national PCOS evidence-based guideline and translation of evidence into practice. Here, we present a brief general clinical introduction to PCOS, concise guideline algorithms and clinical pathways and a comprehensive summary of the Evidence based guideline for assessment and management of PCOS (available at http://www.managingpcos.org.au/pcos-evidence-based-guidelines).8 PresentationPCOS has significant and diverse implications, including reproductive (hyperandrogenism, hirsutism, anovulation, infertility), metabolic (insulin resistance [IR], impaired glucose tolerance, DM2, adverse cardiovascular risk profiles) and psychological features (increased anxiety and depression and worsened QoL).9 Presentation varies across the life span. Hyperandrogenic features are often most prominent among adolescents,10 fertility issues most prominent among women in their 20s and 30s, and metabolic challenges most notable after this.11 The propensity to weight gain and psychological challenges affect all ages, and metabolic features can occur early, especially among those who are overweight. Variations across ethnic groups should also be noted, such as fewer hyperandrogenic dermatological features and more severe metabolic features in Asian women, even without weight gain. Indigenous women appear to have a higher prevalence and severity of PCOS.1,12 The clinical presentation of PCOS is outlined in Figure 1. Figure 1. The aetiological, hormonal and clinical features of polycystic ovary syndrome Adapted and reproduced from Teede at al with permission from the Royal Australian College of General Practitioners.13 Diagnosis and investigationsDiagnosis of PCOS is now largely based on the Rotterdam criteria,14 which are inclusive of the original National Institutes of Health (NIH) criteria15 and require two of three key features: oligo- or anovulation, clinical and/or biochemical hyperandrogenism and polycystic ovaries on ultrasound (Figure 2). However, as noted, PCOS phenotypes vary widely depending on life stage, genotype, ethnicity and environmental factors, including lifestyle and body weight. Diagnostic investigations must exclude other causes and include thyroid function tests and prolactin and follicle-stimulating hormone (FSH) levels.9 For diagnosis, androgen levels should be measured; however, optimal methodology remains very controversial and is addressed in Section 1. Vaginal ultrasound is often needed for diagnosis where hyperandrogenism and anovulation are not both clearly present. Ultrasound can check for polycystic ovaries and endometrial thickness. However, vaginal ultrasound should be reserved for sexually active women. The role of ultrasound remains controversial for adolescents, among whom a polycystic appearance of the ovaries is very common, potentially leading to overdiagnosis;16 hence, this area is also covered in the guideline. Other diagnostic investigations are based on clinical discretion. Screening is also vital to detect PCOS complications and guide prevention and treatment. Comprehensive cardiovascular risk-factor screening, including family history, ethnic group, body mass index (BMI), waist circumference, smoking status, blood pressure, glycaemic status (oral glucose tolerance test [OGTT]) and lipid profile, is important at diagnosis and should be repeated with a frequency informed by metabolic risk (eg, body weight, age, family history, ethnicity) as outlined in Section 3. Optimal methodology for the routine screening for prediabetes and DM2 has been controversial in PCOS, but because lifestyle change and metformin improve IR in PCOS, and among other at-risk groups these measures have been shown to dramatically reduce progression to diabetes, early detection, including detection of prediabetes, is vital and is addressed in the guideline. Figure 2. The Rotterdam criteria for diagnosis of polycystic ovary syndrome (PCOS)14 The Rotterdam criteria are inclusive of National Institutes of Health (NIH) criteria in that a woman diagnosed with PCOS using the NIH criteria will also meet Rotterdam criteria; however, a woman diagnosed with PCOS using Rotterdam criteria may not meet NIH criteria. AetiologyPCOS is an endocrine disorder, the pathophysiology of which remains unclear. Genetic and environmental contributors combine with obesity, ovarian dysfunction and hormonal drivers to contribute to the aetiology of PCOS.17,18 The underlying hormonal imbalance may include a combination of increased androgens and/or hyperinsulinaemia secondary to IR (Figure 1). Greater understanding of cause has been hampered by a lack of ideal methods to assess either hyperandrogenism or IR. Hyperandrogenism is detected in around 60%–80% of women with PCOS, and IR is a pathophysiological contributor in around 50%–80%.19 Obesity increases reproductive features — hyperandrogenism, hirsutism, infertility and pregnancy complications — both independently and by exacerbating PCOS.20,21 Furthermore, obesity exacerbates the PCOS-related increased risk factors for impaired glucose tolerance, DM2 and CVD,22 while obesity also affects psychological features of PCOS. Clinical featuresPCOS is a chronic condition that manifests across the life course. Women with PCOS present with psychological,5,23 reproductive24 and metabolic implications. In terms of psychosocial implications, challenges to feminine identity and body image due to obesity, acne, excess hair, infertility and long-term health-related concerns compromise QoL and adversely affect mood and psychological wellbeing. With a higher prevalence and greater severity of depression and anxiety, low self-esteem, negative body image, and psychosexual dysfunction,5,25 assessment of psychological functioning in women with PCOS is vital. This is relevant to clinical care as mood disturbance, in turn, impairs QoL and adversely affects ability to self-manage and optimise lifestyle. Optimal approaches to screening and assessment of psychological functioning in PCOS are unknown and recognition is generally poor; hence, this area was prioritised in the guideline (Section 4). If mood disturbance is detected during screening, further assessment and management is required. Reproductive and reproductive hormonal features are often the best-recognised features in PCOS as they form the basis of the diagnostic criteria.14 These include clinical and biochemical hyperandrogenism, anovulation, subfertility and polycystic ovaries on ultrasound. A key point is that fertility is not necessarily impaired in all PCOS cases — some women conceive without medical intervention, depending on the severity of the condition. Age and BMI have a critical role in infertility risk in PCOS; therefore, early family initiation (before the age of 30–35 years) combined with maintaining a BMI < 31 kg/m2 is ideal for increasing the chance of conceiving. Metabolic features of PCOS include an apparent propensity for excess weight gain, an increased prevalence of prediabetes and DM2, a 5–10-fold risk of progression from prediabetes to DM2 and a 4–7-fold risk of DM2.11 Cardiovascular risk factors are increased and CVD appears more prevalent among women with PCOS despite inadequate long-term studies to appropriately address this question.22 In the general population, IR is a predictor of CVD.26,27 Women with PCOS also have an increased prevalence of metabolic syndrome (associated with an increased risk for DM2 and CVD),28 individual risk factors for CVD and clinical signs of atherosclerosis,29,30 which are all exacerbated by obesity. Women with PCOS are therefore a population at high risk of developing DM2 and CVD. As DM2 and subsequent CVD are the primary cause of death in Australian women, any increase in prevalence will have significant public health implications. It is also important to note that relatives of women with PCOS may have increased risk of diabetes and increased CVD risk factors. Metabolic features are often poorly appreciated in PCOS; hence, recommendations on screening and assessment of DM2 and CVD risk factors are covered in the guideline. Obesity or excess weight is a major cause of chronic disease in Western countries. In Australia, 56% of the adult population is overweight (BMI ≥ 25 kg/m2) or obese (BMI ≥ 30 kg/m2). In 2007, 31% of women were overweight and 24% of women were obese. Recent data from the Australian Longitudinal Study on Women’s Health showed that among 26–31-year-old women, 20.4% were overweight and a further 13.9% were obese.31 Overall, the proportion of adults who are obese has doubled in the past 20 years.32 Obesity is now the primary cause of chronic disease among Australian women, with adverse outcomes including DM2 and CVD.31 Obesity has a specific impact on women’s reproductive health, increasing the prevalence and severity of PCOS, infertility, pregnancy complications, gestational diabetes and fetal pregnancy complications, with substantial and escalating economic costs.33,34 Indeed, the adverse impact of obesity on fertility, exacerbated by delay in childbearing, is resulting in a significant social, health and economic burden in Australia.35 Given the dramatic increase in obesity, the guideline addresses weight loss and prevention of weight gain through lifestyle intervention (Section 5). ManagementTherapy should focus on both the short- and long-term reproductive, metabolic and psychological features. It is important to address psychological factors initially to optimise self-efficacy, readiness to change and sustainability of lifestyle interventions as well as to improve QoL. Screening, assessment and treatment of depression and anxiety are vital, and recognition of other aspects of emotional wellbeing, including poor body image, sexual dysfunction, disordered eating and eating disorders — all more common among women with PCOS — is important for improving QoL. Optimal approaches to screening and assessment of emotional wellbeing among women with PCOS remain unclear and are also addressed in this guideline (Section 4). Once recognised, poor emotional wellbeing and mood disorders should be addressed to improve QoL among women with PCOS. PCOS management should focus on support and education, and needs to strongly emphasise healthy lifestyle, with targeted medical therapy as required. Given the putative aetiological role of IR and obesity in PCOS, prevention of weight gain is important across the life span. Furthermore, among those who are already overweight, multidisciplinary lifestyle intervention aimed at improving IR and aiding weight management is recognised as first-line therapy for most women who are overweight.6 Modest weight loss of 5%–10% of initial body weight significantly reduces IR and has been demonstrated to ameliorate many of the features of PCOS.6 Optimal methods for achieving weight loss and prevention of weight gain remain unclear and are a focus of this guideline (Section 5). Short-term diets rarely lead to permanent weight loss, and lifestyle change requires behavioural change. Health-coaching principles can be incorporated to optimise readiness to change, and include education and accurate risk perception, which can assist with motivation through education and tailoring the knowledge relevant to the individual. Once ready to change, support is needed to convert this to action with effective strategies including patient-driven goal setting (eg, 5% of body weight loss, small improvements in exercise), so that these incremental changes are seen as achievements. Multidisciplinary involvement in care is often useful in the early stages to support education and behaviour change and is explored further in this guideline (Section 2). In addition to lifestyle measures, therapy in PCOS can be targeted to specific clinical presentations. Although there is a plethora of options for therapy in PCOS, in this guideline we have focused on the most controversial interventions, where little guidance is currently available. Assessment of mood disorders and emotional wellbeing is prioritised in the guideline, yet treatment is well guided by a range of existing clinical guidance tools.36-43 Hirsutism treatment is also guided by a recent and comprehensive international statement,44 so these areas are not covered in the guideline. Infertility remains a highly controversial area and is covered in detail in the guideline (Sections 6, 7 and 8). DM2 and CVD risk assessment is included (Section 3); yet treatments for these established complications are covered in other specific national evidenced-based guidelines,45,46 and have not been reproduced here. Optimal therapy for infertility is one of the most controversial areas of PCOS management and includes lifestyle interventions, medical and surgical ovulation induction, consideration of bariatric surgery for preconception weight loss, and in-vitro fertilisation (IVF). All these areas are covered in the guideline except for IVF therapy, as this was deemed to be of a lower priority than the first-line lifestyle measures and ovulation induction therapies. Potential targeted treatment options for PCOS are summarised in Box 1. Box 1. Summary of potential targeted treatment options for polycystic ovary syndrome (PCOS) Oligomenorrhoea/amenorrhoea Lifestyle change (5%–10% weight loss + structured exercise) Oral contraceptive pill (OCP) (low oestrogen doses [eg, 20 μg] may have less impact on insulin resistance)47 Cyclic progestins (eg, 10 mg medroxyprogesterone acetate 10–14 days every 2–3 months) Metformin (improves ovulation and menstral cyclicity) Hirsutism Choice of options depends on patient preferences; impact on wellbeing; and access and affordability:44 Self-administered and professional cosmetic therapy are first line (laser recommended) Eflornithine cream can be added and may induce a more rapid response If cosmetic therapy is not adequate, pharmacological therapy can be considered Pharmacological therapy Medical therapy if patient is concerned and cosmetic therapy is ineffective/inaccessible/unaffordable Primary therapy is the OCP (monitor glucose tolerance in those at risk of diabetes) Anti-androgen monotherapy (eg, spironolactone or cyproterone acetate) should not be used without adequate contraception Trial therapies for ≥ 6 months before changing dose or medication Combination therapy — if ≥ 6 months of OCP is ineffective, add anti-androgen to OCP (twice daily spironolactone > 50 mg or cyproterone acetate 25 mg/day, days 1–10 of OCP) Infertility Lifestyle intervention (to optimise preconception health and fertility and reduce pregnancy and long-term complications) Advise on folate, smoking cessation and optimal weight and exercise before conception Given age-related infertility, advise women to optimise family initiation Infertility therapies may include clomiphene citrate, metformin, gonadotrophins, surgery and in-vitro fertilisation. Cardiometabolic risk Lifestyle change: > 5% weight loss in those who are overweight reduces diabetes risk by approximately 50%–60% in high-risk groups48 Optimise cardiovascular risk factors Consider metformin* (reduces the risk of diabetes by ~ 50% in adherent high-risk groups)48 Adapted and reproduced with permission from Teede et al,9 not generated directly from the evidence-based guidelines. Hirsutism therapy is summarised from existing hirsutism clinical practice guidelines.44 * Metformin and the OCP are not currently approved for use to manage PCOS by many regulatory bodies. The OCP is indicated for contraception and metformin for diabetes. However, their use is supported by evidence and is recommended by international and national specialist societies.49 Considerations for Indigenous women with PCOSThe prevalence of PCOS among Indigenous Australian women appears to be as high as 21% by the Rotterdam criteria12 and the NIH criteria,50 and increases with rising BMI.12 In a group of Indigenous women with PCOS, 30.3% were obese and 7.0% had a normal BMI.12 DM2 and obesity are associated with major morbidity among Indigenous women. The National Aboriginal and Torres Strait Islander Health Survey (NATSIHS) found that Indigenous Australians are 1.2 times more likely to be overweight or obese than non-Indigenous Australians, and this disparity is greatest for women.51 DM2 is the second-commonest cause of mortality and disability-adjusted life-years (DALY) among Indigenous women.52 The DALY rate ratios (age-standardised to total Indigenous population) for ischaemic heart disease and DM2 among Indigenous women, compared with all Australian women, are 6.6 and 6.3, respectively; and the mortality rate ratio is 5.0 for ischaemic heart disease and 18.9 for DM2.52 The risk of metabolic complications is already high among Indigenous women, independent of PCOS; therefore, PCOS can amplify metabolic risk in these women. Given that these metabolic complications are largely preventable, it is important to provide early access to care. The leading cause of burden of disease among Indigenous women in the NATSIHS was anxiety and depression, accounting for 10% of the burden.51 Little is known about the prevalence of eating disorders and disordered eating among Indigenous women. Social and cultural factors influence emotional wellbeing, and the challenges facing many Indigenous women are likely to amplify the impact of PCOS on emotional wellbeing. Further research in this area is needed. Access to culturally appropriate care, services and programs is currently not optimal. Access issues are a key barrier for many Indigenous women, as health services generally, and women’s health services in particular, are limited in rural and remote locations. There are many barriers to healthy lifestyles, including the high cost of maintaining a healthy diet in rural and remote locations. Socioeconomic factors, such as poverty and overcrowding, make the use of refrigerators and kitchen equipment to cook healthy food difficult. Lifestyle programs may need to be applied in different ways to engage Indigenous women and incorporate exercise into daily activities, especially in rural and remote locations, due to a lack of service provision and facilities. Other issues for Indigenous women include that the role of ultrasound in the Indigenous setting is questionable due to limitations in care and access to ultrasound facilities and service provision in rural and remote locations; and there may be cultural factors and potential issues around acceptability of bariatric surgery. It is important to encourage and enable Indigenous women with PCOS to access services that are available and address potential barriers presented by cultural and traditional health practices. Work is currently underway to adapt, translate and implement the recommendations outlined here to Indigenous settings. Development of an evidence-based guidelineRationale and methodsGiven its heterogeneous clinical features across the life span, PCOS is a condition that engages many health disciplines. The associated complications are serious yet are often largely preventable; however, there is a lack of awareness of PCOS among consumers and health professionals. It is essential that consumers and health professionals recognise the life-course implications of PCOS, identify the early signs and symptoms and work together to manage PCOS and prevent its complications — especially as the burden and cost of PCOS complications, including infertility, DM2, CVD and emotional wellbeing issues are significant. Currently, there is limited consensus among different medical specialties as to the optimal management of PCOS in Australia.7,53 Diagnosis and treatment of PCOS can therefore differ depending on the health professional consulted (eg, general practitioner, endocrinologist or gynaecologist).7 There are limited clinical guidelines and no evidence-based guidelines, either in Australia or internationally, for assessment or management of women with PCOS; rather, PCOS is briefly mentioned within guidelines for the management of obesity and DM2.45,54 Where international clinical guidelines for the assessment and management of women with PCOS exist, they are informed by expertise and, in some cases, evidence, but are not rigorously developed evidenced-based guidelines; they do not consider psychological issues; and they offer simplistic advice on lifestyle management of PCOS. There is no guidance on the assessment and management of PCOS among Indigenous women, nor any adaptation for the Australian context. Overall, many areas of controversy remain in PCOS. Comprehensive evidence-based guidelines are warranted to optimise diagnosis, assessment and management. Accordingly, the Jean Hailes Foundation for Women’s Health has facilitated the formation of an independent PCOS Australian Alliance, bringing together health professionals, researchers, consumers and policymakers to advance knowledge and quality of care in PCOS. The federal government has funded the PCOS Australian Alliance, under the auspices of the Jean Hailes Foundation, to produce national evidence-based guidelines. The full version of the guideline has been approved by the National Health and Medical Research Council (NHMRC) (for detail about obtaining NHMRC approval, please see the full guideline), is endorsed by the Royal Australian College of General Practitioners and is freely available at http://www.managingpcos.org.au/pcos-evidence-based-guidelines.8 This is a summary version of the full guideline. The Jean Hailes Foundation was funded to translate the guideline into practice, including freely available independent evidence-based information on PCOS for health professionals and women at http://www.managingpcos.org.au. ScopeThe purpose of the guideline is to integrate the best available evidence with clinical expertise and consumer preferences; to provide health professionals, consumers and policymakers with guidance on timely diagnosis, accurate assessment and optimal management of PCOS; and to promote consistency of care and prevention of complications in primary care and specialist settings. The guideline is relevant to the assessment and management of reproductive-age adolescents and women with PCOS, including women with PCOS who are experiencing infertility. The guideline will apply in all health care settings and to a broad audience, including: community care practitioners; Indigenous health care workers; GPs; nurses; endocrinologists; obstetricians and gynaecologists; allied health professionals — psychologists, dietitians, exercise physiologists and physiotherapists; patients; community support groups (eg, the Polycystic Ovary Syndrome Association of Australia [POSAA]); the general public; students; and policymakers. PCOS is a syndrome, and as such, no single diagnostic criterion is sufficient for diagnosis. The 2003 Rotterdam consensus workshop concluded that PCOS diagnosis requires at least two of: oligo- or anovulation, hyperandrogenism (clinical and/or biochemical) and polycystic ovaries on ultrasound (Figure 2).14 The evidence-based guideline development groups and the Alliance agreed to endorse the Rotterdam diagnostic criteria for the guideline, while recognising there are current limitations of all definitions. MethodologyGuidelines are intended to improve patient outcomes, promote standardised care, develop standards to assess the clinical practice of health care professionals, and promote research and translation into practice. Guidelines are developed by drawing from clinician judgement, patient preference and research evidence, and are intended to aid clinical judgement and patient preference, not to replace it (Figure 3). The ultimate decision about clinical management of an individual patient will always depend on the clinical circumstances, patient preferences, and the clinical judgement of the health care team. Although there are many types of guidelines, this NHMRC-approved evidence-based guideline followed a rigorous, systematic process of development, which is briefly outlined below and in detail in the full guideline.8 Figure 3. Evidence-based guidelines are intended as an aid to clinical judgement and patient preference, not to replace it An independent PCOS Australian Alliance was formed in 2008 after a national workshop facilitated by the Jean Hailes Foundation for Women’s Health, which brought together key leaders from the research and multidisciplinary clinical sectors, with consumers providing a driving force through the peak national support group, POSAA. The vision of the Alliance is to improve the lives of Australian women with PCOS through education, research and evidence-based health care. One of the priorities of the Alliance was to develop an evidence-based guideline for PCOS. This guideline was developed as outlined in the NHMRC standards and procedures for externally developed guidelines.55 The Alliance identified key clinical objectives for the guideline based on highest clinical priority, greatest knowledge gaps, factors identified by the Australian government (which funded the guideline), and the expertise of Alliance members. The identified key clinical priorities focused on care of women with PCOS to facilitate early diagnosis of PCOS; early detection and treatment of depression, anxiety and mood disorders; early detection and diagnosis of risk factors for prediabetes, DM2 and CVD; and early detection and treatment of fertility problems and prevention of pregnancy complications. Multidisciplinary guideline development committees included a Project Board, PCOS Australia Alliance Strategic Advisory Group and four guideline development groups. Each guideline development group comprised a chair, professional group members with specific expertise in PCOS and the clinical area of interest (eg, psychologist in the emotional wellbeing guideline development group), a consumer representative from POSAA, evidence officers and, where possible, a representative to provide context for the Indigenous setting. Indigenous representation was present on the PCOS Australian Alliance Strategic Advisory Group, and the guideline development groups comprised clinicians with experience working with Indigenous communities. These multidisciplinary groups determined and prioritised the clinical questions addressed in this guideline and developed the clinical practice and research recommendations from the evidence reviews. For more detail about the development and prioritisation of clinical questions, please see the full guideline.8 To facilitate this process using an evidence-based approach, the chairs of each guideline development group attended a 1-day workshop, facilitated by the Southern Health Centre for Clinical Effectiveness, where the methods of identifying, appraising and synthesising evidence; grading the strength of evidence and its suitability to support evidence-based recommendations; and the process of guideline development overall were described in detail. Evidence reviews were conducted for each of the 22 identified clinical questions. Search strategies were developed according to a-priori selection criteria for each clinical question. Searches were limited to English language articles and there were no limits on year of publication. The literature was searched until November 2010. The following electronic databases were employed to identify relevant evidence: Australasian Medical Index, CINAHL, the Cochrane Library, the Cochrane Database of Systematic Reviews, DARE (Database of Abstracts of Reviews of Effects), the Cochrane Central Register of Controlled Trials, the Cochrane Database of Methodology Reviews, the Cochrane Methodology Register, Health Technology Assessment Database, the United Kingdom National Health Service Economic Evaluation Database, EMBASE, EBMR, MEDLINE and PsycINFO. Bibliographies of relevant studies identified by the search strategy and relevant reviews/meta-analyses were also searched. Included studies were classified according to the NHMRC levels of evidence56 and appraised using a-priori criteria according to study design, using a descriptive component approach to assign a risk of bias rating.57 In accordance with the selection criteria, data were extracted from included studies using a specially developed data extraction form,57 and meta-analyses were performed where appropriate. The guideline development groups were able to develop guideline recommendations from these evidence reviews. The evidence reviews for each question can be found in the supporting document to the full guideline: Evidence report: evidence based guidelines for assessment and management of PCOS, available at http://www.managingpcos.org.au/pcos-evidence-based-guidelines. The guideline contains 38 recommendations, each of which is assigned a grade. In developing the guideline recommendations, the guideline development groups placed emphasis on accurate assessment and management of PCOS. The recommendations in this guideline are strengthened by the use of rigorous methodology for evidence review and guideline development, including use of: study designs least susceptible to bias; a-priori criteria for inclusion and appraisal of studies; extraction of study data; and meta-analysis where appropriate. The recommendations were formulated using a considered judgement process that took into account the amount and quality of available evidence as well as its generalisability and applicability to current practice in Australia. Each evidence-based recommendation was given an overall grading from A to D, according to the NHMRC grades of recommendations for guideline developers (Table 1).56 Evidence grading is provided primarily to inform users about the strength of the evidence underpinning each recommendation. Where there was insufficient high-quality evidence in specific patient groups, lower-quality evidence or data from other patient groups, and where there was consensus among the guideline development group, combined with clinician and patient preferences, clinical consensus recommendations were developed. Clinical practice points have also been included, where important issues (such as safety, side effects or risks) arose from discussion of evidence-based or clinical consensus recommendations. Further points of relevance to the clinical implementation of recommendations were made in “implications of the recommendations” sections, including consideration of resource implications. Table 1. National Health and Medical Research Council grades for recommendations56 A Body of evidence can be trusted to guide practice. B Body of evidence can be trusted to guide practice in most situations. C Body of evidence provides some support for recommendation but care should be taken in its application. D Body of evidence is weak and recommendation must be applied with caution. The words “should”, “could” and “should not” do not directly reflect the grade or classification allocated to a recommendation, and are independent descriptors intended to reflect the judgement of the multidisciplinary guideline development group about the practical application of the recommendation, balancing benefits and harms. Where the word “should” is used in the recommendations, the guideline development group judged that the benefits of the recommendation (whether evidence-based or clinical consensus) clearly exceed the harms, and that the recommendation can be trusted to guide practice. Where the word “could” is used, either the quality of evidence was underpowered, or the available studies demonstrated little clear advantage of one approach over another, or the balance of benefits to harm was unclear. Where the words “should not” are used, there is either a lack of appropriate evidence, or the harms outweigh the benefits. In formulating the recommendations for this guideline, the guideline development groups recognised and took into account several factors and limitations pertaining to the available evidence. For many aspects of PCOS, there is little or no evidence or the evidence is of poor quality, with other potential biases resulting from different methods for diagnosis of PCOS and differing end points. Public and targeted consultation on the draft guideline was conducted for 30 days commencing 5 March 2011, in accordance with the legislative requirements for approval of externally developed guidelines under Section 14A of the National Health and Medical Research Council Act 1992 (Cwlth). All aspects of the guideline were developed as outlined in the NHMRC standards and procedures for externally developed guidelines,55 and accordingly, the guideline was approved by the NHMRC in July 2011. In approving the full version of the guideline, the NHMRC is satisfied that it is based on the systematic identification and synthesis of the best available scientific evidence and makes clear recommendations for health professionals practising in an Australian health care setting. This guideline does not seek to provide full safety and usage information on pharmacological and surgical interventions. The pharmacological and surgical interventions recommended in the guideline should not be applied without consideration of the patient’s clinical profile and personal preferences. It is recommended that the reader consults the Therapeutic Guidelines (http://www.tg.com.au) and the National Prescribing Service (http://www.nps.org.au) for detailed prescribing information, including indications, drug dosages, methods and routes of administration, contraindications, supervision and monitoring, product characteristics, and adverse effects. It is intended that this evidence-based guideline summary be used alongside the full guideline.8 The guideline should be considered according to the limitations outlined within, and used in conjunction with clinical judgement and patient preference. For a detailed description of the methodology used to develop the guideline, please see the full guideline.8 Translation of the guideline, including the production and dissemination of guideline-associated tools and resources, is the responsibility of the Jean Hailes Foundation for Women’s Health as a national not-for-profit women’s health organisation funded by the federal government. The PCOS Alliance and POSAA provided significant contribution to these resource developments.

Helena J Teede MB BS, FRACP, PhD · Marie L Misso PhD, BSc(Hons) · Amanda A Deeks BMed, GradDipPsych, PhD · Lisa J Moran BSc(Hons), BND, PhD · Bronwyn G A Stuckey FRACP · Jennifer L A Wong MB BS(Hons), FRACP, MIH · Robert J Norman MB ChB(Hons), MD, FRANZCOG · Michael F Costello MB BS, FRANZCOG, CREI · on behalf of the Guideline Development Groups

To the Editor: Few reports have been published on bone mineral density (BMD) among randomly sampled populations. Organisations such as Osteoporosis Australia and the Australian and New Zealand Bone and Mineral Society rely on research to supply reliable data that are representative of the Australian community. This information informs practitioners, researchers and policymakers of the size of the problem of osteoporosis in Australia. Our study aimed to document the proportion of individuals who have reduced BMD. The Geelong Osteoporosis Study recruited a random population-based sample of individuals from the Barwon Statistical Division, an area surrounding Geelong, Victoria. This region is well suited to epidemiological research as it is geographically well defined and has a large, stable population (259 000) with sufficient socioeconomic diversity for it to be representative of the Australian population.1,2 Age-stratified random samples of 1494 women (median age 54.0 years; range 20–94 years; recruited 1994–1997)3 and 1467 men (median age 56.0 years; range 20–97 years; recruited 2001–2006)4 were drawn from electoral rolls. Participation rates were 67% for men and 77% for women. We measured BMD at the spine and femoral neck by dual energy x-ray absorptiometry (Lunar; GE Healthcare, Madison, Wis, USA). Reference ranges for BMD in men4 and women3 have been published previously. We categorised BMD as normal (T score, > − 1.0), osteopenia (T score, − 2.5 to − 1.0) or osteoporosis (T score, < − 2.5) using the osteoporosis and osteopenia thresholds developed for postmeno-pausal women. Among those with discordant BMD, the site with the lower BMD was used in this classification. Normal BMD was predominant among men aged < 50 years (Box). A consistent proportion of men older than 50 years, including those aged > 80 years, had osteopenia (range, 49%–64%); 19% of those aged > 80 years had osteoporosis. Most women aged < 55 years had BMD in the normal range. Osteopenia was most prevalent among those aged 55–79 years, and osteoporosis dominated (51%) among those aged > 80 years. The osteoporosis and osteopenia thresholds developed for postmenopausal women may not be the most appropriate cut-points for diagnosis in younger women and in men; however, we used them because they are the currently accepted thresholds. Although sex-specific thresholds for defining osteoporosis might best be defined on the basis of absolute fracture risk, until such data are available, T-score criteria continue to be used. After standardising for age and sex to the 2006 Australian population,5 we found that 5.9% of men and 22.8% of women aged 50 years and over, and 12.9% of men and 42.5% of women aged 70 years and over, would be classified as having osteoporosis. Proportion of participants in the Geelong Osteoporosis Study with osteoporosis or osteopenia, by age group, according to bone mineral density at the spine or femoral neck Men* (no. [%]) Women† (no. [%]) Age group (years) Osteoporosis‡ Osteopenia§ Normal¶ Osteoporosis‡ Osteopenia§ Normal¶ 20–24 0 (0) 14 (15.7%) 75 (84.3%) 1 (1.0%) 21 (20.6%) 80 (78.4%) 25–29 0 (0) 24 (26.1%) 68 (73.9%) 0 (0) 24 (22.2%) 84 (77.8%) 30–34 1 (1.1%) 29 (30.9%) 64 (68.1%) 0 (0) 25 (22.5%) 86 (77.5%) 35–39 0 (0) 31 (30.7%) 70 (69.3%) 2 (1.7%) 41 (35.3%) 73 (62.9%) 40–44 4 (3.8%) 36 (34.3%) 65 (61.9%) 1 (0.9%) 32 (29.9%) 74 (69.2%) 45–49 4 (3.9%) 34 (33.3%) 64 (62.7%) 2 (1.8%) 32 (28.1%) 80 (70.2%) 50–54 3 (2.5%) 60 (50.0%) 57 (47.5%) 5 (4.7%) 43 (40.6%) 58 (54.7%) 55–59 2 (1.9%) 61 (58.7%) 41 (39.4%) 9 (8.9%) 57 (56.4%) 35 (34.7%) 60–64 4 (3.9%) 50 (48.5%) 49 (47.6%) 22 (21.0%) 54 (51.4%) 29 (27.6%) 65–69 6 (5.7%) 67 (63.8%) 32 (30.5%) 24 (24.0%) 48 (48.0%) 28 (28.0%) 70–74 9 (7.5%) 72 (60.0%) 39 (32.5%) 43 (32.6%) 70 (53.0%) 19 (14.4%) 75–79 14 (13.1%) 63 (58.9%) 30 (28.0%) 32 (42.7%) 33 (44.0%) 10 (13.3%) 80+ 34 (18.5%) 107 (58.2%) 43 (23.4%) 105 (51.0%) 88 (42.7%) 13 (6.3%) * Recruited 2001–2006. † Recruited 1994–1997. ‡ T score, < − 2.5. § T score, − 2.5 to − 1.0. ¶ T score, > − 1.0.

Margaret J Henry · Julie A Pasco · Geoff C Nicholson · Mark A Kotowicz

21-Hydroxylase deficiency (21-OHD) is the most common cause of congenital adrenal hyperplasia, with an incidence of 1 : 14 000 live births and equal prevalence among males and females. Newborns with the most severe “salt-wasting” form of 21-OHD are susceptible to salt-wasting crises in the first few weeks of life. This is associated with morbidity and mortality. 21-OHD newborn screening (NBS) is currently performed in many countries. Despite several prominent medical societies recommending 21-OHD NBS, no state in Australia currently screens for this condition. We report a case that illustrates the need to reconsider including 21-OHD in NBS. 21-OHD NBS can be reliable, sensitive and effective in reducing morbidity and mortality.

Joyce Y Wu MB BS, MAACB, FRCPA · Sudeep MB BS, FRACP, DCH · David M Cowley MB ChB, FRCPA, FHGSA · Mark Harris MB BS, FRACP, MD · Ivan N McGown BSc, MIT, MHGSA · Andrew M Cotterill MB BS, FRACP, MD

To the Editor: The Hyperglycemia and Adverse Pregnancy Outcomes (HAPO) study, a large, blinded, multinational study, showed an increased risk of adverse maternal and neonatal outcomes in relation to maternal glycaemia, at glucose levels below the current Australian criteria for diagnosing gestational diabetes mellitus (GDM).1 The International Association of Diabetes and Pregnancy Study Groups (IADPSG), an international consensus group, has proposed new criteria for the diagnosis of GDM.2 As a result, these new criteria have been adopted by the American Diabetes Association, which predicts a significant increase in the prevalence of GDM.3 The new criteria were discussed at the Australasian Diabetes in Pregnancy Society annual scientific meeting in 2010. Moses and colleagues accurately outline the increased prevalence of GDM if IADPSG criteria are adopted in Australia.4 An increased prevalence has implications for resource allocation, and the anticipated increase in workload can be managed by appropriate planning and exploration of alternative models of care. We surveyed attitudes to the management of GDM among general practitioners already involved in antenatal shared care programs in the Liverpool and Fairfield areas of Sydney (GDM is not currently part of the shared care program in this urban area, which has a high prevalence of diabetes). Around 120 GPs are enrolled in the antenatal shared care program in the Liverpool and Fairfield areas. Forty-six of these GPs attended an educational meeting at which the survey was distributed, and of the 46 (who all completed the survey), only seven believed that GDM can always be managed in the antenatal shared care program. Seventeen felt that, due to lack of time or lack of access to appropriate resources, GDM cannot be managed at all by GPs as part of shared antenatal care; eight of these 17 indicated that they never initiated insulin for patients with type 2 diabetes. Only two indicated that no up-skilling was required for them to manage GDM. These attitudes may be limited to GPs in urban practices. Whether the involvement of GPs in the management of GDM is appropriate is unclear, and the provision of supporting resources requires further review. Additionally, as determined by Moses and colleagues,4 the predicted increase would come from older women who are possibly more likely to have other comorbidities that make them less suitable for shared care.

Barbara Depczynski · Vincent W Wong · Hamish D Russell · Nicole Opie

New data from the Health in Men Study raise questions about the role of testosterone supplementation in ageing men Ageing of the “baby boomer” generation foreshadows a future shaped by demographic change, with increasing numbers of older Australians. The large, longitudinal Western Australian Health in Men Study (HIMS) is therefore timely, as it examines the endocrinology of male ageing and predictors of health in community-dwelling older men.1,2 As part of HIMS, my colleagues and I surveyed 3274 men aged 75–95 years in 2008–2009 using a questionnaire that included items on sexual activity.3 Of 2930 men who reported on the importance they attached to sex, 48.8% considered it important, and of the 2783 men who provided data on sexual activity, 30.8% had at least one sexual encounter (defined as any mutually voluntary activity with another person that involves sexual contact, whether or not intercourse or orgasm occurs4) in the previous 12 months.3 Of these older sexually active men, 56.5% were satisfied with the frequency of sex, while 43.0% would have preferred sex more frequently.3 These findings indicate that many older Australian men consider sexual activity important and desirable. In HIMS, factors that predicted reduced sexual activity were increasing age, osteoporosis, prostate cancer, diabetes, antidepressant use, β-blocker use, and partner’s lack of interest or physical limitations.3 Living with a partner and having a non-English-speaking background were associated with increased sexual activity. Interestingly, a 1 SD increase in testosterone level, measured in blood samples collected in 2001–2004, was associated with a 20% increased likelihood of being sexually active in 2008–2009. Therefore, while older men with lower testosterone levels are likely to report symptoms such as reduced frequency of sexual thoughts and erectile difficulties,5 higher testosterone levels predict sexual activity several years into the future. This raises the question of whether giving exogenous testosterone to induce a comparable increase in circulating total testosterone levels (+ 5.6 nmol/L) would increase the frequency of sexual activity for older men. Epidemiological studies such as HIMS show that men with testosterone levels in the low-normal range have poorer health outcomes; for example, those with testosterone levels in the lowest quartile (< 11.7 nmol/L) have increased risk of stroke or transient ischaemic attack.6 Lower testosterone levels are associated with mortality in older men.7 Studies of testosterone therapy in older men show favourable effects on body composition, with increased lean mass and bone mineral density and, to an extent, improved muscle strength.8 However, there is no evidence as yet that testosterone therapy reduces cardiovascular events or mortality, or that it increases sexual activity in older men. In fact, administering higher doses of testosterone to older men with limited mobility might result in an excess of adverse cardiovascular events.9 More data are needed to help design optimal studies to clarify the role of testosterone supplementation in ageing men. In HIMS, the mean serum total testosterone level in 3638 men aged 70–89 years was 15.4 nmol/L (reference range, 8–35 nmol/L), and only a minority would have been classified as having unequivocally low testosterone levels.1 Uncertainty remains around the extent to which lower testosterone levels reflect underlying comorbidity; appropriate testosterone thresholds for the diagnosis of androgen deficiency in older men; and effects of testosterone therapy on cardiovascular risk.8 The current Testosterone Trial (ClinicalTrials.gov identifier NCT00799617) in the United States, due for completion in 2015, is recruiting older men with lower testosterone levels and will examine the effect of transdermal testosterone gel on end points of walking speed, sexual activity, vitality, memory and anaemia correction. So while the question of whether testosterone therapy might protect against cardiovascular events remains unresolved, its impact on sexual activity in the setting of a randomised controlled trial might not be known for another 4 years. Under these circumstances, the clinical approach to ageing men with symptoms of testosterone deficiency must be prudent, taking both known risks and potential benefits into account.8 Testosterone supplementation could be considered in men who are clearly hypogonadal. Symptoms of androgen deficiency should be assessed, and the diagnosis based on at least two unequivocally low early-morning testosterone levels, preferably assayed using a mass spectrometry-based methodology.10 Men should be counselled as to the risks and benefits of testosterone therapy, and treatment should be accompanied by safety monitoring, including prostate evaluation and monitoring of prostate-specific antigen levels and haematocrit. The anticipated effect of testosterone therapy would be to increase libido, and this should be included in the discussion of benefit and risk. While higher testosterone levels are associated with sexual activity in older men, non-hormonal factors are also important. HIMS found that increasing age predicted declining sexual activity; after adjusting for this and other covariates, men were four times more likely to be sexually active if they were living with a partner.3 Conversely, they were much less likely to be sexually active if their partner lacked interest in sex or had physical limitations. Medical comorbidities including diabetes and use of antidepressants were also associated with reduced likelihood of being sexually active. Therefore, social and medical factors are key determinants of whether ageing men remain sexually active. The increasing numbers of men transitioning from middle to older age should be encouraged to maintain their personal health and the health of their relationships to maximise their chances of having sex in future years.

Bu B Yeap MB BS, FRACP, PhD

HbA1c may be a practical alternative to blood glucose for the diagnosis of diabetes For more than 15 years, glycated haemoglobin (HbA1c) has been recommended as the key tool for assessing glycaemic control in people with diabetes. Only in 2009 did the first advice to use HbA1c levels for diabetes diagnosis appear, using a cut-point of ≥ 6.5%.1 To date, no clear argument has been articulated to explain why HbA1c levels have been deemed superior to laboratory-determined blood glucose levels for determining the need for insulin therapy, but not for diagnosing diabetes; however, it is likely that implications of the former are greater than those of the latter, for both individuals and society. Blood glucose values are considered the gold standard for diabetes diagnosis, but they have significant limitations. Day-to-day variability in glucose levels is considerable, and the glucose concentration in a plasma sample falls within a short period, even if the blood has been collected in a fluoride tube. In addition, when stable samples are tested in two different laboratories, the results will differ by at least 14% in more than a third of cases.2 Furthermore, even when using a single laboratory, only 70% of people with a blood glucose value that indicates a diagnosis of diabetes have the diagnosis confirmed by repeat testing 2 weeks later, compared with 83% for HbA1c.2 So, is HbA1c the answer to the challenges of diabetes diagnosis? Until recently, the problem with HbA1c has been the concern that results vary considerably between laboratories. In the 1990s, laboratory differences of more than two percentage points were not uncommon, but the United States National Glycohemoglobin Standardization Program (NGSP) has progressively driven improvements in assay standards. The latest results from the largest global survey of quality of HbA1c measurement show that, for reference samples with HbA1c levels of 4.0%–6.0%, 91% of more than 3000 laboratories could obtain an HbA1cvalue that was within 6.0% of the target.3 In a recent Australian study, whole blood samples were sent to more than 200 laboratories and more than 90% obtained HbA1c values that were within 6% of the median.4 Thus, for a sample with a median value of 5.3%, over 90% of laboratories obtained values within the range 5.0%–5.6%, and for a median value of 7.4%, over 90% obtained values within the range 7.0%–7.8%. In addition, combined data from eight studies conducted between 1988 and 2004 (using assays in eight different laboratories, none of which may have performed as well as those available now) showed that HbA1c levels were at least as strongly correlated with diabetic retinopathy as were blood glucose levels.5 HbA1c is not without limitations. First, an HbA1c test is more expensive than a fasting glucose test, but costs about the same as an oral glucose tolerance test. The extra cost of using HbA1c instead of fasting glucose as the initial blood test needs to be weighed up against the potential for the HbA1c test (which does not require the patient to fast) to be used more widely, to identify more undiagnosed cases of diabetes, and to save money by preventing complications of diabetes. To our knowledge, no cost–benefit analyses comparing the HbA1c test with the fasting glucose test have been published — this should be a high priority. Second, HbA1c can be unreliable in the presence of haemoglobin variants or alterations in red blood cell turnover. Most HbA1c assays are now able to adjust for the most common haemoglobin variants, but where there is uncertainty relating to the reliability of HbA1c, blood glucose will remain the preferred test. If the potential exists to use HbA1c for the diagnosis of diabetes, how can a practitioner know whether a particular laboratory can be relied on? A joint working party of the Australian Diabetes Society, the Royal College of Pathologists of Australasia, and the Australasian Association of Clinical Biochemists is developing a formal laboratory and clinical framework within which the diagnosis of diabetes by HbA1c testing can be undertaken. In the meantime, it would be reasonable to think that a laboratory can be relied on, in the context of using HbA1c as a diagnostic tool, if the routine coefficient of variation is ≤ 3.0% (the 2010 accreditation target used by the NGSP) and the external quality assurance results are consistently within the Royal College of Pathologists of Australasia Quality Assurance Programs method-specific performance targets (allowable limits of performance). This information should be available from laboratories on request. Practical aspects of using HbA1c for the diagnosis of diabetes remain to be finalised. One option may be to request a fasting glucose test and HbA1c test at the same time, with the HbA1c to be performed only if the fasting glucose level is ≥ 5.5 mmol/L. This strategy would limit the additional costs of HbA1c testing while decreasing the number of patients who are lost to follow-up for an oral glucose tolerance test. The cost of an HbA1c test that is used for diagnosis is not currently reimbursed by Medicare. However, when used appropriately, the HbA1c test appears to be at least as useful for diagnosing diabetes as a blood glucose test. Australia may soon be ready to join countries such as the United States and Japan in using HbA1c, a measure of chronic glycaemia, for the diagnosis of diabetes, a disease of chronic glycaemia.

On behalf of the Joint HbA1c Working Party of the Australian Diabetes Society, the Royal College of Pathologists of Australasia, and the Australasian Association of Clinical Biochemists

Haemoglobin A1c (HbA1c — a term that is sometimes used interchangeably with “glycated haemoglobin”) measurements are an indicator of time-averaged blood glucose levels (previous 2–3 months), and are used as the best marker of long-term diabetes control. A recent consensus statement on the worldwide standardisation of HbA1c measurement1 has updated previous international recommendations on the standardisation of HbA1c measurement and reporting.2 Here, we provide the rationale and guidance for implementation of HbA1c reporting in the new Système International (SI) units in Australia. This article represents the views of the Australasian Association of Clinical Biochemists, the Australian Diabetes Educators Association, the Australian Diabetes Society and the Royal College of Pathologists of Australasia, and was prepared by a working party of representatives of these organisations. The International HbA1c Consensus Committee recommends that all HbA1c levels be reported in SI units (mmol/mol, no decimal places) — with results directly traceable to the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) reference method — and in the currently used, National Glycohemoglobin Standardization Program (NGSP) units (percentage, one decimal place). We recommend that dual reporting in Australia begins in July 2011, and that reporting of percentages ceases 2 years later. In New Zealand, dual reporting commenced in August 2009. The key reasons for implementing this recommendation in Australia are that: the SI units relate to a scientifically valid measure of HbA1c; the SI units remove potential confusion between HbA1c values as a percentage and blood glucose values in mmol/L; the change is in keeping with the international consensus statement;1 and the change has already been initiated in New Zealand and a number of countries in the European Union. Until now, all HbA1c measurements performed in Australia have been reported as percentages (HbA1c as a percentage of total haemoglobin) that are aligned with those produced in the Diabetes Control and Complications Trial.3 These units and this standardisation have been promoted by the NGSP in the United States, and the activities of this organisation have produced marked improvement in the accuracy of HbA1c results worldwide. More recently, the IFCC has developed a reference method that is more specific for HbA1c and more analytically robust.4 The IFCC method is now used as the reference system by the NGSP and for all routine methods for measurement of HbA1c, although a numerical conversion is required during the calibration process. The changes recommended here will provide results that are directly aligned with the IFCC method. As the IFCC method is more specific for HbA1c, not measuring several other haemoglobin–sugar complexes, the results are 10% to 40% lower than those from the NGSP system, depending on HbA1c concentration. Because reporting these results as percentages may lead to confusion (eg, producing a result of 5.3% rather than 7.0%), the units are changed to mmol/mol (millimoles HbA1c per mole of total haemoglobin [53 mmol/mol in the previous example]), which is consistent with the SI units recommended for use in Australia. There is a linear relationship between results from the two methods, and the “master equation” is used to convert results between the two methods: HbA1c SI unit (mmol/mol) = 10.93 × HbA1c NGSP unit (%) − 23.50.5 To make the conversion easier for clinicians, it is important to translate current treatment advice to the new units. A general conversion table for clinical use is provided in Box 1. The general HbA1c target of ≤ 7.0% for patients with type 1 and type 2 diabetes mellitus equates to ≤ 53 mmol/mol, although these values need to be individualised for patients. The recently updated diabetes treatment guidelines are shown with SI units in Box 2 and Box 3,6 and recommendations for reporting HbA1c levels in Australia are summarised in Box 4. In addition, supporting material for doctors and patients will be presented in SI units in the future. The routine reporting of an estimated average glucose (eAG) value may be useful for consultations with individual patients. However, the working party strongly agrees with the revised consensus statement that routine reporting of eAG with all requests for HbA1c analysis is not appropriate.1 The reasons for this include variability in the methods used to measure eAG, the risk of confusing a measure of long-term glycaemia (eAG) with a measure of short-term blood glucose control (actual blood glucose level), and its lack of applicability in the majority of patients with type 2 diabetes (in whom blood glucose levels are not measured at frequent intervals).7 Nonetheless, eAG values will be used as an educational tool at the discretion of individual clinicians, who can assist patients to understand the significance and limitations of the result. 1 Conversion table for haemoglobin A1c (HbA1c) values HbA1c as percentage (old units) HbA1c in mmol/mol (new units) 5.0 31 6.0 42 6.5 48 7.0 53 8.0 64 9.0 75 10.0 86 11.0 97 12.0 108 2 Recommended haemoglobin A1c (HbA1c) target ranges for patients with type 1 diabetes6 HbA1c target General target ≤ 53 mmol/mol, ≤ 7.0%* Specific clinical situations Pregnancy or planning pregnancy ≤ 53 mmol/mol, ≤ 7.0%*† Children and adolescents ≤ 58 mmol/mol, ≤ 7.5%* Recurrent severe hypoglycaemia or hypoglycaemia unawareness ≤ 64 mmol/mol, ≤ 8.0% Patients with major comorbidities likely to limit life expectancy Symptomatic therapy of hyperglycaemia‡ and avoidance of ketosis * Achievement of HbA1c targets must be balanced against risk of severe hypoglycaemia. † An HbA1c level of ≤ 42 mmol/mol (≤ 6.0%) is desirable if it can be achieved safely. ‡ Where practical, suggest blood glucose target level < 15 mmol/L to help minimise risk of infection. 3 Recommended haemoglobin A1c (HbA1c) target ranges for patients with type 2 diabetes6 HbA1c target General target ≤ 53 mmol/mol, ≤ 7.0%* Specific clinical situations Diabetes of short duration† and no clinical cardiovascular disease Requiring lifestyle modification ± metformin ≤ 42 mmol/mol, ≤ 6.0%* Requiring any antidiabetic agents other than metformin or insulin ≤ 48 mmol/mol, ≤ 6.5%* Requiring insulin ≤ 53 mmol/mol, ≤ 7.0%* Pregnancy or planning pregnancy ≤ 42 mmol/mol, ≤ 6.0%* Children and adolescents ≤ 53 mmol/mol, ≤ 7.0%* Diabetes of longer duration† or clinical cardiovascular disease (any therapy) ≤ 53 mmol/mol, ≤ 7.0%* Recurrent severe hypoglycaemia or hypoglycaemia unawareness (any therapy) ≤ 64 mmol/mol, ≤ 8.0% Patients with major comorbidities likely to limit life expectancy‡ (any therapy) Symptomatic therapy of hyperglycaemia§ * Achievement of HbA1c targets must be balanced against risk of severe hypoglycaemia, especially among older people. † In an older adult, long duration might be considered to be > 10–20 years, but for a person who develops type 2 diabetes at a young age, it might be considerably longer. ‡ Examples of major comorbidities include chronic medical conditions, such as chronic kidney disease stages 4 or 5; heart failure stages III or IV (New York Heart Association grading); incurable malignancy; and moderate to severe dementia. § Where practical, suggest blood glucose target level < 15 mmol/L to help minimise risk of infection. 4 Recommendations for reporting haemoglobin A1c (HbA1c) levels in Australia From July 2011, HbA1c levels should be reported in both National Glycohemoglobin Standardization Program units (percentage) and the Système International (SI) units (mmol/mol) by all pathology laboratories and, where possible, from point-of-care devices. The period of dual reporting will be 2 years, after which only the SI units will be used. These recommendations are consistent with international recommendations and are already in place in New Zealand.

Graham R D Jones MB BS, DPhil, FRCPA, Chemical Pathologist · George Barker BHSc, CDE-RN, NP · Ian Goodall BSc, FAACB, FFRCPA · Hans-Gerhard Schneider MD, FRACP, FRCPA · Mark D S Shephard MAACB, FFRCPA, PhD · Stephen M Twigg MB BS, PhD, FRACP

To the Editor: Atypical femur fractures appear to be an emerging adverse outcome of long-term use of bisphosphonates. Although analyses of epidemiological data suggest that subtrochanteric and diaphyseal fractures per se are rare,1,2 the true incidence of atypical fractures (a distinct subset of such fractures) is unknown. In a recent 5-year retrospective study, we reviewed individual radiographs of 152 patients with subtrochanteric and diaphyseal femur fractures and identified 20 atypical fractures.3 Seventeen of these 20 atypical fractures had occurred in patients treated with oral bisphosphonates. In light of these findings, we sought to calculate the incidence of atypical femur fractures in the population served by our large tertiary referral hospital in Sydney. According to the Australian Bureau of Statistics 2006 Census, the hospital’s catchment population (as defined by the New South Wales Department of Health) was 174 448. The annual incidence of atypical femur fractures in 2006 was therefore estimated to be 0.23 per 10 000 in the general population, and 1.6 per 10 000 in people aged over 65 years. These estimates, based on a stringent radiological definition of atypical femur fractures, confirm that, on a population basis, these fractures are indeed rare. We also sought to define the mean annual incidence of atypical femur fractures in patients treated with oral bisphosphonates. Thus, we obtained data on the wholesale purchase of alendronate and risedronate by pharmacies within the hospital’s catchment area over the 5 years of the original retrospective study (1 June 2003 to 30 May 2008) from IMS Health Australia (market researchers for the global pharmaceutical and health care industries). A mean number of 2860 patients per year were prescribed alendronate and 1265 patients per year were prescribed risedronate. This corresponded to a mean annual incidence of atypical femur fractures of 10 per 10 000 in patients taking alendronate and three per 10 000 in those taking risedronate. The use of a relatively small number of cases to calculate these incidences is a particular limitation of our research, and is a reflection of the rarity of these events. Furthermore, our estimates need to be considered within the wider context of the established beneficial effects of bisphosphonates in patients with osteoporosis. On the basis of randomised trials, it has been estimated that treating 1000 women with oral bisphosphonates for 3 years prevents 100 fractures.2 Also, several meta-analyses have confirmed significant reductions in the risk of osteoporotic vertebral and non-vertebral fractures in patients treated with oral bisphosphonates.4,5 In spite of two recent large database studies that showed a greater incidence of atypical fractures among long-term bisphosphonate users,6,7 atypical fractures are rare and the risks of using this class of drugs appears to be strongly outweighed by their proven efficacy in preventing fractures.

Christian M Girgis · Markus J Seibel

The Australian Government is planning to pilot a model of prepaid funding for coordinated care of patients with diabetes in general practice. Patients will register with a practice that undertakes to coordinate their care, and practices will manage pre-allocated funds to provide services instead of billing Medicare. Systems to manage prepaid funds in Australian general practice have not yet been developed. In the model that has been proposed, practices with a small register of patients will be at risk of overspending, which may threaten practice viability and patient services. If the initiative is to have integrity, all patient services should be paid from the prepaid funds and patients should only attend the practice with which they have registered. Risks should be delineated and contingency plans made explicit before practices and patients commit to the initiative.

Douglas A Pritchard MB BS, FRACGP, PhD

To the Editor: We read with interest the recent letter by He and colleagues.1 The authors described a case of ischaemic stroke in a patient with scurvy and considered whether there was a connection between the two conditions. They referred to evidence that vitamin C deficiency may be a risk factor for cerebrovascular disease, but acknowledged that a direct causal link is unlikely to be established. We propose adiponectin as the causal link between vitamin C deficiency and stroke. Adiponectin is an adipokine, secreted in multimers by adipose tissue, with insulin-sensitising, antiatherogenic and cardioprotective properties.2 A decrease in adiponectin levels — particularly the more biologically active, high molecular weight (HMW) multimers — is implicated in a number of disease states, such as obesity, type 2 diabetes, heart disease and some cancers. Adiponectin has also been reported to have cerebroprotective properties,3 and there is some evidence that levels may be reduced in patients with cerebrovascular disease.4 A recent study demonstrated that vitamin C supplementation increases the proportion of HMW adiponectin secreted from human adipocytes.5 Vitamin C levels are lower in obese and diabetic patients, and the patient reported by He et al had also been recently diagnosed with type 2 diabetes. While scurvy is admittedly rare, He and colleagues noted that subclinical vitamin C deficiency is not uncommon in the general population (about 10%). We suggest that it would be worthwhile to assay plasma vitamin C levels and total and HMW adiponectin levels in patients presenting with cerebrovascular events, and to undertake prospective studies to determine whether vitamin C supplementation improves patients’ adiponectin levels. Increased HMW adiponectin levels could be explained, at least in part, by the role of vitamin C in the multimerisation of adiponectin. Adiponectin levels are also reduced in other vascular conditions, such as ischaemic heart disease and peripheral vascular disease.4 As plasma vitamin C levels of patients with these diseases are also reduced, it is tempting to posit that vitamin C supplementation could have a role as a treatment, or even a prophylactic, in populations at risk of a range of adiponectin-related vascular diseases.

Felicity J Rose · Jonathan P Whitehead

To the Editor: We read with interest the recent editorial by Nordin,1 which makes several imprecise observations on our recent position statement in the Journal.2 Here, we analyse some of his statements, as we believe that they are misleading to your readers and hazardous to institutionalised older persons. The Consensus Conference on Treatment of Osteoporosis in Residential Aged Care Facilities (RACFs) was organised as a unique type of meeting in which Australian experts in osteoporosis and geriatric medicine, including representatives from the Australian and New Zealand Bone and Mineral Society (ANZBMS) and Osteoporosis Australia, participated in multiple interactive sessions with 50 geriatricians and general practitioners who practise in RACFs. The goals were to appraise current evidence in the field of falls and fracture prevention in RACFs and to define practical, evidence-based recommendations. A similar meeting took place in 2004 in Canada,3 where conclusions played a pivotal role in optimising osteoporosis care in RACFs. Conclusions of the Australian meeting and recommendations subsequently published in the Journal are products of both the general consensus of the participants in the final plenary session and contributions of all the coauthors.2 Our article states, “In adults with a baseline calcium intake of 500–900 mg/day, increasing or supplementing this intake by a further 500–1000 mg/day has a beneficial effect on BMD [bone mineral density]”.2 In light of the evidence that we cited4 and more recent evidence,5,6 it is Nordin’s responsibility to highlight the potential risks of excessive calcium intake, from dietary sources plus supplements, in a population that is at high risk of cardiovascular disease.7 Moreover, we are not alone in raising this concern — a recent statement from the American Society for Bone and Mineral Research has expressed similar concern.8 In addition, Nordin cites the seminal work of Chapuy and colleagues; although this study was carried out in nursing homes and apartments for older people, it only looked at independent, ambulatory older people.9 Furthermore, the statement regarding calcium compliance is again supported by a study performed in ambulatory populations, in which compliance is likely to differ from that in our population of interest. Finally, Nordin asserts that our article promotes the use of bisphosphonates, particularly the intravenous variety, and makes the unfounded suggestion that the recommendation of bisphosphonates was due to the relationship between the sponsor and some of the coauthors. This is inaccurate. The University of Sydney, funded by a medical education grant, organised the Consensus Conference. Also, both the ANZBMS and Osteoporosis Australia endorsed the meeting and the integrity of the funding process. Indeed, maintaining independence from the sponsor was a major goal of the Consensus Conference, which was attested to by the feedback from participants. In conclusion, a basic knowledge of geriatric pharmacology and a good understanding of the current literature on geriatric medicine are enough to value the recommendations presented in our article.

on behalf of Jacqueline J Close, Julien P de Jager, Peter R Ebeling, Charles Inderjeeth, Stephen Lord, Andrew J McLachlan, Ian R Reid, Bruce R Troen and Philip N Sambrook

Screening plus equitable provision of vitamin D supplements could mitigate many adverse outcomes For nothing worthy proving can be proven, Nor yet disproven: wherefore thou be wise, Cleave ever to the sunnier side of doubt. Alfred, Lord Tennyson, The ancient sage Tennyson may not have been alluding to the need for high-level evidence from randomised controlled trials (RCTs) to alter clinical practice, but he would have been aware of children with rickets. Evidence has accumulated linking vitamin D deficiency to adverse outcomes in pregnancy, such as pre-eclampsia, hypertension, higher rates of caesarean section and preterm delivery. Lau and colleagues (→ Serum 25-hydroxyvitamin D and glycated haemoglobin levels in women with gestational diabetes mellitus) contribute to this evidence by demonstrating that, in women with gestational diabetes mellitus (GDM), a lower serum 25-hydroxyvitamin D (25[OH]D) concentration was independently associated with poorer glycaemic control.1 Of 147 women who were studied late in pregnancy (at a mean of 35 weeks’ gestation), about 40% had vitamin D insufficiency or deficiency (serum 25[OH]D concentrations ≤ 50 nmol/L). Most of the women in this study were not white, and ethnicity, occupational status and season, not surprisingly, all influenced 25(OH)D concentrations, while body mass index did not. Perhaps more surprisingly, however, 25(OH)D concentrations were inversely associated with fasting and 2-hour glucose levels measured during an oral glucose tolerance test and with the marker of glycaemic control, glycated haemoglobin. Most importantly, serum 25(OH)D was an independent predictor of glycaemic control. In adults, a number of large cross-sectional studies have shown a consistent, independent and positive relationship between serum 25(OH)D and insulin sensitivity, and an inverse relationship with risk of diabetes.2-5 Serum 25(OH)D levels have been shown to account for 42% of the variation in insulin sensitivity assessed by hyperglycaemic clamp.3 Consistent with Lau et al’s findings, fasting and 2-hour levels of glucose and insulin have been shown to be independently and inversely associated with serum 25(OH)D levels.3-5 In a United States study, the odds ratio for diabetes was 0.25 (95% CI, 0.11–0.60) for non-Hispanic white participants in the highest versus the lowest serum 25(OH)D quartile.2 The highest-level evidence to date comes from a large prospective study with a 17-year follow-up that showed people in the highest quartile of serum 25(OH)D had a 40% decreased risk of type 2 diabetes compared with those in the lowest quartile.6 Based on these data, RCTs of vitamin D supplementation in adults to improve insulin sensitivity and reduce diabetes risk are underway. GDM is becoming increasingly more common, affecting up to 10% of pregnancies. The presence of GDM is not trivial and has long-term implications for the health of mothers and their children. The former have an increased risk of developing type 2 diabetes, while their offspring have an increased risk of obesity and diabetes later in life. Vitamin D deficiency is also very common in pregnancy. The prevalence of inadequate levels of vitamin D in Lau et al’s study is comparable with rates of vitamin D insufficiency of 47.1% and 83.5% in white and black pregnant women, respectively, in the northern US (defined as 25[OH]D < 80 nmol/L)7 and 65.3% in pregnant women in rural Victoria (defined as 25[OH]D < 75 nmol/L).8 RCTs of vitamin D supplementation, initiated early in pregnancy, are now required to demonstrate whether vitamin D supplementation might reduce the incidence or severity of GDM. International debate is currently focused on the optimal level of serum 25(OH)D. Based on meta-analyses using musculoskeletal end points in older individuals, cut points of 60 nmol/L and 75 nmol/L seem appropriate to prevent falls and fractures, respectively.9 However, a recent Institute of Medicine report recommended at least 50 nmol/L,10 which appears overly conservative and does not take season into account. The public health implications of vitamin D deficiency in pregnancy are far broader than glycaemic control. In Australia, there has been a resurgence of rickets — partly owing to an increased refugee population comprising dark-skinned and veiled women with vitamin D deficiency, and also because of decreased exposure of babies to sunlight, lack of supplementation of infant feeds with vitamin D and weaning of infants onto non-milk liquids. Milder forms of bone disease may also occur with vitamin D deficiency. Recently, a study that used three-dimensional ultrasonography in pregnant women showed that vitamin D deficiency was associated with increased femur metaphyseal cross-sectional area and increased femur splaying (the ratio of femoral metaphyseal cross-sectional area to femoral length) at as early as 19 weeks’ gestation.11 In addition, it was previously shown that children born to mothers with vitamin D deficiency (< 50 nmol/L) during pregnancy exhibit deficits in total body bone mineral content as great as 11% at 9 years of age.12 This could lead to an increased risk of osteoporotic fracture later in adult life, but this is unlikely to be evaluated in long-term studies. In addition, maternal or early life vitamin D deficiency has been linked to an increased risk of several other disorders, including neonatal craniotabes, prematurity, type 1 diabetes mellitus, schizophrenia, and childhood respiratory infections and wheeze.13,14 Current evidence strongly supports routine screening for vitamin D deficiency early in pregnancy. Furthermore, vitamin D supplementation to correct deficiency should be initiated early in pregnancy as it might reduce the incidence or severity of GDM and because changes in skeletal morphology of the fetus associated with deficiency are seen as early as 19 weeks’ gestation. The most common recommended daily doses of cholecalciferol are 1000 IU–2000 IU, however, daily doses of up to 4000 IU have recently been shown to be safe in pregnancy (Bruce W Hollis, Professor, Department of Paediatrics, Medical University of South Carolina, USA, personal communication). What is problematic is the equitable provision of vitamin D supplements to pregnant Australian women with deficiency. Pregnant and breastfeeding women who are most at risk of vitamin D deficiency are often the least likely to be able to afford supplements. In the United Kingdom, vitamin D supplements are provided free of charge to such women through the Healthy Start program.15 There is evidence to support more widespread use of vitamin D supplements during pregnancy in Australia, although more research is required. One way to increase access might be to alter the scheduling of higher-dose, lower-cost vitamin D supplements.

Peter R Ebeling MB BS, MD, FRACP

Objective: To test the hypothesis that lower 25-hydroxyvitamin D (25[OH]D) levels in late pregnancy are associated with poorer glucose control in gestational diabetes mellitus (GDM).Design and setting: Retrospective cross-sectional study, in a GDM clinic at a tertiary referral centre.Patients: Women attending the GDM clinic at Westmead Hospital from 1 February 2007 to 1 February 2008, excluding those with prepregnancy glucose intolerance.Main outcome measures: Levels of glycated haemoglobin (HbA1c) and 25(OH)D measured during the third trimester; maternal age, ethnicity, body mass index (BMI) and occupational status; and results of oral glucose tolerance testing (OGTT).Results: 147 women with a mean gestational age of 35 ± 2 weeks were included, of whom 41% had insufficient or deficient levels of 25(OH)D (≤ 50 nmol/L). Ethnicity, occupational status and season significantly influenced 25(OH)D levels (P < 0.01 for all) but BMI did not. 25(OH)D levels were inversely associated with fasting and 2-hour blood glucose levels during OGTT (Spearman r = − 0.16; P = 0.05 for both) and with log[HbA1c] (Spearman r = − 0.32; P < 0.001). BMI and insulin doses were also associated with HbA1c levels. Multivariable analysis identified 25(OH)D and blood glucose levels during the OGTT as independent predictors of HbA1c levels.Conclusions: Lower 25(OH)D levels are independently associated with poorer glycaemic control. Future randomised trials are needed to determine whether vitamin D plays a role in glycaemic control in GDM. Regardless, maternal vitamin D insufficiency has adverse effects including neonatal hypocalcaemia and rickets. The 41% prevalence of inadequate 25(OH)D levels in the women in our study is unacceptably high. We propose routine 25(OH)D testing of all pregnant women at screening for GDM or earlier, and treatment of women who are found to be deficient.

Sue Lynn Lau MB BS, FRACP · Jenny E Gunton MB BS, FRACP, PhD · Neil P Athayde MB BS(Hons), FRANZCOG, CMFM · Karen Byth PhD · N Wah Cheung MB BS, FRACP, PhD

Objective: The International Association of Diabetes and Pregnancy Study Groups (IADPSG) has proposed new criteria for the diagnosis of gestational diabetes mellitus (GDM). The aim of this study was to compare the prevalence of GDM when IADPSG criteria were used with the prevalence when the current Australasian Diabetes in Pregnancy Society (ADIPS) criteria were used.Design, setting and participants: This was a prospective study over a 6-month period, examining the results of all glucose tolerance tests (GTTs) conducted for the diagnosis of GDM in Wollongong, a city using the public and private sectors.Main outcome measures: The prevalence of GDM using the existing (ADIPS) and the proposed (IADPSG) criteria.Results: There were 1275 evaluable GTTs (571 public and 704 private). Using the current ADIPS diagnostic criteria, the prevalence of GDM was 8.6% (public), 10.5% (private) and 9.6% (overall). Using the proposed IADPSG criteria, the prevalence of GDM was 9.1% (public), 16.2% (private) and 13.0% (overall).Conclusions: The proposed IADPSG criteria would increase the prevalence of GDM from 9.6% to 13.0% (P < 0.001). In our study in the Wollongong area, which has a population with a predominantly white background, this increase came mainly from older women attending a private pathology provider. Data from both the public and private sectors need to be included in any discussion on the change in prevalence of GDM.

Robert G Moses MD · Gary J Morris BAppSc · Peter Petocz PhD · Fernando San Gil PhD · Dinesh Garg MD

Androgen deprivation therapy (ADT) in men with prostate cancer increases the risk of osteoporotic fractures, type 2 diabetes and, possibly, cardiovascular events. There is considerable uncertainty about the risk–benefit ratio of ADT in non-palliative treatment; the benefits of ADT in treating non-metastatic prostate cancer need to be carefully weighed against the risks of ADT-induced adverse events. Baseline assessment of bone health at the initiation of ADT should include measurement of bone mineral density (BMD) by dual energy x-ray absorptiometry and, in men with osteopaenia, a thoracolumbar spine x-ray. General measures to prevent bone loss, including regular physical activity, as well as ensuring calcium and vitamin D sufficiency, should be instituted routinely. All men with a previous minimal trauma fracture should receive pharmacological therapy unless contraindicated; for those who have not sustained a minimal trauma fracture, treatment is advised if the BMD T score is ≤ − 2.0, or if the 10-year risk of a major osteoporotic fracture exceeds 20%. Men with prostate cancer who are receiving ADT should be closely monitored for weight gain and diabetes; intensive lifestyle intervention is recommended to prevent ADT-induced weight gain and insulin resistance. Management of the metabolic sequelae of ADT includes optimal reduction of cardiovascular risk factors, with particular attention to weight, blood pressure, lipid profile, smoking cessation, and glycaemic control.

Mathis Grossmann MD, PhD, FRACP · Emma J Hamilton MB BS · Christopher Gilfillan MB BS, PhD, FRACP · Damien Bolton MB BS, FRACS · Daryl Lim Joon MB BS, FRANZCR · Jeffrey D Zajac MB BS, PhD, FRACP

Omission: In “Detecting undiagnosed diabetes using glycated haemoglobin: an automated screening test in hospitalised patients” in the 21 February 2011 issue of the Journal (Med J Aust 2011; 194: 160-164), the following were omitted:

Nyoli A Valentine MB BS · Tariq M Alhawassi BScPharm, MClinPharm · Greg W Roberts BPharm, FSHP, BCPS · Parind P Vora MB BS, MPH · Stephen N Stranks MB BS, FRACP · Matthew P Doogue MB ChB, FRACP

Objective: To assess the utility of glycated haemoglobin (HbA1c) level as an automated screening test for undiagnosed diabetes among hospitalised patients and to estimate the prevalence of undiagnosed diabetes among hospitalised patients.Design, participants and setting: A 3-month prospective study of all adult patients admitted to a tertiary hospital. An HbA1c test was automatically undertaken on admission for all patients with a random plasma glucose (RPG) level ≥ 5.5 mmol/L. Demographic, admission and biochemical data were obtained from hospital databases. A subset of patients was recruited for an oral glucose tolerance test (OGTT) after discharge.Main outcome measures: Prevalence of undiagnosed diabetes (defined as HbA1c ≥ 6.5% in accordance with International Expert Committee and American Diabetes Association recommendations) and utility of automated HbA1c testing.Results: The prevalence of undiagnosed diabetes was 11% (95% CI, 9.8%–12.4%) (262/2360) during the study period. A further 312 patients with known diabetes were admitted. The prevalence of undiagnosed diabetes was highest in the 65–74-years age group. The HbA1c test cost was $152 per new diagnosis of diabetes. Conservatively assuming an annual incidence of undiagnosed diabetes of 0.8%, the ongoing cost of testing hospitalised patients would be $2100 per new diagnosis of diabetes. RPG testing was not sensitive or specific in diagnosing diabetes. Patients were poorly compliant with the post-discharge OGTT (27% completion rate).Conclusions: HbA1c is a simple, inexpensive screening test that can be automated using existing clinical blood samples. Hospital screening for diabetes needs to be coupled with resources for management in the community.

Nyoli A Valentine MB BS · Tariq M Alhawassi BScPharm, MClinPharm · Greg W Roberts BPharm, FSHP, BCPS · Parind P Vora MB BS, MPH · Stephen N Stranks MB BS, FRACP · Matthew P Doogue MB ChB, FRACP

To the Editor: Rahman and colleagues suggest that iodine deficiency in Victoria increased between 2001 and 2006, based on the findings of thyroid-stimulating hormone (TSH) levels in neonates at routine newborn screening.1 Indeed, their data as presented suggest a doubling of the percentage of mothers with iodine deficiency to over 9% during that period. This could be correct. Certainly, as they state, there is much evidence to suggest that there is mild iodine deficiency in Australia. However, there are caveats about the data they report which are not mentioned. Data from New South Wales do not show this trend. While they do suggest a degree of mild iodine deficiency, there was no increase in the percentage of neonates with TSH levels > 5 mIU/L of whole blood from 2002 to 2009 (Box), although the average age at sampling falls slightly (from 2.96 to 2.32 days) over this period. The World Health Organization has defined iodine sufficiency as being indicated, inter alia, when more than 3% of newborns aged 3–4 days have a TSH level > 5 mIU/L of whole blood.2 Factors that affect the TSH level in a newborn screening program include the precise age at sampling, and any changes to the method of TSH analysis used. In a Swiss study assessing the efficacy of iodine supplementation, there was a small but significant decrease in the TSH level from Day 3 to Day 4 of age.3 This is unsurprising: following the TSH surge in the first hours after birth, TSH levels decline gradually to a steady level at about Day 7.4 There could well have been a trend to earlier sampling in Victoria, within the bounds of the 2–4 days of age assay that Rahman and colleagues mention, over the period studied, but these crucial data are not given. The dried blood spot TSH assay method is not described either. A change in any aspect of the methodology; for example, if the manufacturer modified the antibody used, may produce a small, clinically insignificant but numerically significant, change in results. If the data presented by Rahman and colleagues for Victoria do not have these biases, then the situation warrants further investigation, but whatever is happening in Victoria seems not to be replicated over the border. Percentage of newborns with thyroid-stimulating hormone (TSH) level > 5 mIU/L of whole blood, detected by routine newborn screening in New South Wales, by year Year Newborns with TSH level > 5 mIU/L 2002 3.80% 2003 3.68% 2004 3.87% 2005 5.02% 2006 4.48% 2007 3.56% 2008 3.92% 2009 4.00%

Bridget M Wilcken · Veronica C Wiley

In reply: The methods used for blood sample collection and analysis remained unchanged during our data collection period. One source of thyroid-stimulating hormone (TSH) calibrators and reagents was used over the study period by a single laboratory covering all of Victoria. Material from the United States Centers for Disease Control and Prevention was used for external quality assurance, ensuring that the results were in agreement with those of other laboratories. The per cent coefficient of variation over the period ranged from 10% to 20%. The table of neonatal TSH values for New South Wales provided by Wilcken and Wiley further demonstrates the value of using TSH levels as a screening tool for population iodine status, even with a decreasing mean age of sample collection. While we dealt with the effect of sample collection time in our published article,1 here we present a table illustrating analysis of the Victorian neonatal TSH values for samples collected at 48, 72 and 96 hours after birth (Box). The percentage of elevated TSH values increased from 2001 to 2006 at each collection time and, although the percentage of elevated TSH values decreased with increasing age, these values were still indicative of iodine deficiency. Iodine status varies between regions. The National Iodine Nutrition Study (NINS) found both South Australia and Queensland iodine sufficient, while the neighbouring states of NSW and Victoria were iodine deficient.2 The results also indicated that iodine status was worse in Victoria than in NSW; therefore, we might expect similar differences in TSH values. We are now in the process of analysing Victorian TSH values for 2007 to 2010. Percentage of newborns with thyroid-stimulating hormone (TSH) level > 5 mIU/L for blood samples collected at 48, 72 and 96 hours after birth, Victoria, 2001–2006 Sample collection time (h) Percentage of neonates with TSH > 5 mIU/L according to birth year 2001 2002 2003 2004 2005 2006 48 5.73% 6.83% 8.47% 10.58% 11.86% 13.53% 72 4.20% 5.15% 6.87% 7.01% 9.13% 9.38% 96 2.49% 3.21% 4.37% 3.78% 5.98% 5.34%

Ashequr Rahman · Gayle S Savige · Nicholas J Deacon · Ivan Francis · Janice E Chesters

To the Editor: Girgis and Seibel1 are to be congratulated as endocrinologists for raising the important issue of bisphosphonate therapy and femoral fractures. This is a relatively new phenomenon that even a couple of years ago, although already documented, did not seem to be on many endocrinologists’ radar. However, I question the authors’ statement that such fractures are rare. In my small outpost of the orthopaedic world (Northeast Health Wangaratta), three cases have been seen in 12 months. All patients characteristically had a spontaneous non-traumatic fracture, a short oblique or transverse fracture in the subtrochanteric area, and had been on alendronate for more than 5 years. If a journeyman orthopaedic surgeon is seeing a cluster of cases, I suggest that these fractures are not rare. On the other hand, a review of our audit figures in Wangaratta in north-eastern Victoria (where we have a stable rural population of about 17 000) provides an interesting statistic. Over 15 years (1993–2008), Northeast Health Wangaratta admissions for fractured neck of femur fell steadily from 74 to 35 a year. This trend continues. We should not throw out the baby with the bathwater — osteoporosis treatment is likely partly responsible. Also, a deeper level of community aged care support and much better comorbidity management may have helped to reduce falls. The real question is whether all patients who have been on alendronate (or any bisphosphonate) for 5 years need to have a mandatory holiday from the drug. In line with Wolff’s law (bone will adapt to loads under which it is placed), the precursor femoral stress lesions will heal by remodelling, as long as bisphosphonate therapy is suspended. I suggest that the suspension should be for 2 years, which is the time required for full fracture remodelling. I agree with Girgis and Seibel that these drugs have been effective in reducing fracture incidence; however, their long-term use needs further study.

Michael P Falkenberg

In reply: We thank Falkenberg for his comments in reference to our article in the Journal.1 Two recent large-scale population-based studies have suggested that subtrochanteric femur fractures are rare both in the general population and among bisphosphonate users.2,3 While personal experience may often suggest otherwise, a cluster of atypical fractures cannot be used as an indicator of true incidence in the absence of data on the frequency of bisphosphonate use in a particular population. It is certainly safe to say that atypical fractures occur much less frequently than osteoporotic hip fractures. In a 5-year retrospective analysis of femur fractures at our centre, osteoporotic hip fractures outnumbered atypical fractures by a factor of greater than 60.4 Given the body of high-quality evidence on the antifracture efficacy of bisphosphonates, we agree that discarding an effective class of drugs because of a presumed association with an uncommon fracture pattern would be like throwing out the baby with the bathwater. Should we routinely advise patients to take a drug holiday after, say, 5 years of bisphosphonate therapy? There is no good evidence for that either. Although bisphosphonates bind to bone for extended periods, severely suppressed bone turnover or signs of mechanical failure (microcracks) are rarely, if at all, seen in patients chronically treated with bisphosphonates. Similarly, the few bone biopsy studies in patients with atypical fractures do not uniformly support the hypothesis of severely suppressed bone turnover as a cause of these fractures.5,6 Instead of being based on the theoretical assumption of an uncertain risk, the decision for a drug holiday should be made on a case-by-case basis, guided by factors such as the patient’s on-drug fracture history, the presence of other relevant risk factors for osteoporosis, and changes in bone density and bone turnover. Many questions remain unanswered regarding atypical femoral fractures and their biomechanical evolution. Until further research is conducted, the fear of the unknown, namely the impact of bisphosphonates on bone remodelling and microfracture accumulation, should not replace strong evidence in support of their antifracture efficacy.

Christian M Girgis · Markus J Seibel

Reports of adverse events related to calcium supplementation should be supported by rigorous evidence Calcium is an essential nutrient, not only because of its major role in bone, where 99% of it is stored, but because of its central role in neuromuscular function. It is this latter role that explains why ionised calcium in the blood and tissue fluids is one of the most tightly controlled analytes of those that are commonly measured.1 However, maintenance of the calcium level in tissue fluids carries with it the penalty of continuous loss of calcium through the kidneys, bowel and skin, even on a low calcium intake, which is why the recommended daily calcium allowance for adults is relatively high, at 1000 mg.2 Nutritional deficiencies of other minerals, such as magnesium and phosphate, are rare because their tissue fluid levels are not tightly controlled but vary with intake and, therefore, so does their excretion. Calcium is different; reducing calcium intake has a marginal effect on extracellular calcium (and therefore on calcium excretion) because bone is mobilised to maintain the calcium level, which leads sooner or later to the development of osteoporosis. This is the case in laboratory animals3 and, by implication, in humans. Osteoporosis is therefore the index disease for calcium deficiency,4 just as rickets and osteomalacia are the index diseases for vitamin D deficiency; however, there is some overlap between them because the secondary hyperparathyroidism associated with hypovitaminosis D5 increases bone resorption. This is not to suggest that all adult osteoporosis is due to calcium deficiency, but simply to point out that the increase in bone resorption which follows menopause6 can be largely or wholly explained by the fall in calcium absorption and rise in obligatory calcium excretion which occur at this time,7 and also occur in oophorectomised animals.8,9 (The loss of a direct antiresorptive action of oestrogen on bone at menopause cannot be excluded but is probably quantitatively much less important.) For these reasons, it has become standard practice to recommend calcium supplementation to postmenopausal women, increasingly with vitamin D, to prevent or delay bone loss and reduce fracture risk. In the largest meta-analyses, calcium with vitamin D in adequate dosage reduces fracture risk by 25% or more, but vitamin D alone is not effective.10,11 Until very recently, calcium supplementation was not thought to cause any significant side effects. However, a New Zealand team recently reported an increase in the mean rate of mainly self-reported myocardial infarction in participants who were allocated to receive calcium supplements in five prospective trials for which patient-level information was available.12 Although the effect was not significant in any of the trials individually, it was significant at the 5% level in the whole series and has attracted sufficient media attention to endanger the use of calcium in the prevention of osteoporosis in postmenopausal women. An extension of this case against calcium recently appeared in this Journal, in a position statement on fracture prevention in aged-care facilities that was co-authored by one member of the New Zealand team.13 The article not only ignores the seminal work of Chapuy and colleagues on fracture prevention with vitamin D and calcium in aged care homes,14 but specifically opposes the general use of calcium supplementation on four separate grounds, none of which are directly referenced. The first is that long-term compliance with calcium supplementation is very poor, whereas in most trials it is not significantly different from compliance with placebo.15 The second is that the anti-fracture efficacy of calcium is marginal, despite overwhelming evidence to the contrary in the largest meta-analyses.10,11 The third is a bizarre claim that calcium could increase the rate of hip fracture; this is only supported by one trial (by one of the co-authors of the position statement) in which the adverse effect was not remotely significant in participants who complied with calcium supplementation,15 which is widely regarded as an anomaly and is contradicted by a later meta-analysis.16 The final is that calcium supplementation could increase the risk of myocardial infarction, which is highly contentious and negated by the latest meta-analysis of 17 trials.17 These negative statements about calcium (which are not reflected in the article’s abstract) are coupled with the promotion of bisphosphonates — particularly the intravenous variety — despite the fact that virtually all the bisphosphonate trials have incorporated calcium supplements. It may therefore be relevant that this article arose from a meeting financed by a pharmaceutical company that happens to market an intravenous bisphosphonate and gave some form of assistance to six of the 10 authors.13 Since it is clearly stated that this meeting was endorsed by the Royal Australian College of General Practitioners, the Australian and New Zealand Bone and Mineral Society and Osteoporosis Australia, there is a strong implication that these bodies also support the article itself. It is questionable whether such public bodies should lend their authority to a position statement of uneven quality and which runs the risk of being seen as commercially driven.

B E Christopher Nordin MD, FRACP, DSc

Objective: To determine the prevalence and diagnosis rates of Klinefelter syndrome (KS) in Victoria, Australia, and compare these to previous international findings.Design, setting and participants: A Victorian population-based descriptive study of all cytogenetic examinations resulting in a diagnosis of KS, including prenatal diagnoses from 1986 to 2006 and postnatal diagnoses from 1991 to 2006.Main outcome measures: Birth prevalence and diagnosis rates of KS.Results: The birth prevalence of KS in Victoria is estimated to be 223 per 100 000 males (95% CI, 195–254), with about 50% of cases remaining undiagnosed.Conclusions: KS may be occurring more frequently than has been reported previously, yet many cases remain undiagnosed. Our results highlight the need for increased awareness leading to timely detection.

Amy S Herlihy BSc, GradDipGenCounsel · Jane L Halliday BSc(Hons), PhD · Megan L Cock BSc(Hons), PhD · Robert I McLachlan MB BS, PhD

Clinical record A 56-year-old woman, with a history of a right adrenal phaeochromo-cytoma excised in 1974, presented to the emergency department in 2008 with a hypertensive crisis characterised by 3 days of severe headache, malignant hypertension (blood pressure, 200/115 mmHg) and seizures complicated by bilateral humeral fractures. After her surgery in 1974, she had unresolved hypertension and raised urinary catecholamine levels suggestive of residual or metastatic disease. However, no additional tumour could be identified despite venous sampling studies. The patient declined further treatment and was lost to follow-up from 1978. She subsequently had three uncomplicated, successful pregnancies. During the patient’s 2008 hospital admission, her levels of urinary catecholamines and plasma metanephrines were elevated (Box 1). She was treated with phenoxybenzamine and diltiazem (extended release), doses of which were titrated up to 60 mg twice daily and 360 mg daily, respectively, before uneventful surgical repair of both humeral fractures. Localisation studies with a 123 I-metaiodobenzylguanidine scan identified increased tracer uptake within both the left adrenal region and left anterior mediastinum. Computed tomography and magnetic resonance imaging (MRI) scans showed large bilateral renal angiomyolipomas (AMLs), a right renal artery aneurysm, a bulky left adrenal gland with no discrete mass lesion, and a 15 mm calcified lesion within the left anterior mediastinum (not related to the sympathetic chain). Results of a positron emission tomography scan were negative. However, an octreotide scan showed mild tracer uptake in the left adrenal region and marked focal uptake within the left anterior mediastinum (Box 2). It was felt unlikely that the left adrenal gland image represented a phaeochromocytoma and likely that the mediastinal lesion was a phaeo-chromocytoma lymph node metastasis. After the addition of atenolol 100 mg daily to the patient’s drug regimen, the left anterior mediastinal lesion was excised via a cervical approach without complication. Histological examination confirmed a phaeochromo-cytoma metastasis within a lymph node. After the operation, blood pressure improved significantly. Plasma metanephrine levels have remained normal for over a year. Tests for phaeochromocytoma genetic syndromes revealed a missense mutation (Gly144Arg) on exon 2 of the von Hippel–Lindau (VHL) gene, and we are awaiting testing of the patient’s eight siblings to determine whether or not this is a de novo mutation causing VHL disease. Although two of her three children have inherited the mutation, they have had no disease manifestations. Further investigations of the patient have revealed mild sensorineural hearing loss, but no evidence of haemangio-blastomas via fundoscopy or on MRI scanning of the brain and spinal cord. While there is an increased risk of renal cell carcinoma with VHL disease, recent imaging has shown no change in the patient’s presumed bilateral AMLs and she has declined surgery or embolisation. The left adrenal gland and right renal artery aneurysm have also remained stable on serial imaging. Malignant phaeochromocytoma has been reported to be a risk factor for osteoporosis,1 but the patient’s bone mineral densitometry showed only osteopaenia of her lumbar spine, while her left hip bone measurements were within normal limits. Our patient’s history shows the need for long-term periodic follow-up of people treated for phaeochromocytomas. It also illustrates the association between von Hippel–Lindau (VHL) disease and phaeochromocytoma. VHL disease, an autosomal dominant condition, is caused by a mutation of the VHL tumour suppressor gene on chromosome 3p; the mutation rate is one in 36 000 live births.2 The VHL gene is involved in regulating the transcription of vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF) and other hypoxia-inducible proteins. Inactivation promotes tumour angiogenesis and growth through overexpression of VEGF and PDGF receptor agonists, resulting in haemangioblastomas of the central nervous system or retina, pancreatic neuroendocrine tumours and cysts, renal clear cell carcinomas, phaeochromocytoma, endolymphatic sac tumours, and papillary cystadenomas of the epididymis (men) or broad ligament (women).3 In a series report of 246 patients with VHL disease, 26% of patients developed phaeochromocytomas at a mean age of 29 years, 39% had bilateral adrenal phaeochromocytomas, and extra-adrenal disease occurred in up to 30%.4 The adrenal medulla is the main site in the body where the enzyme that converts noradrenaline to adrenaline, phenylethanolamine N-methyltransferase, is localised, and, consequently, adrenal phaeochromocytomas typically have raised levels of both noradrenaline and adrenaline. By contrast, extra-adrenal phaeochromocytomas and phaeochromocytomas in patients with VHL disease do not usually express phenylethanolamine N-methyltransferase and are therefore characterised by raised levels of noradrenaline and normetadrenaline with relatively normal levels of adrenaline and metadrenaline. For our patient, elevations in noradrenaline levels exceeded those of adrenaline both at the time of the original phaeochromocytoma diagnosis and subsequently, with adrenaline levels normalising after the original operation, in keeping with extra-adrenal or metastatic disease. Kindreds with VHL disease can be divided into two types (Types I and II) based upon the incidence of phaeochromocytoma, with Type I generally unaffected and Type II at high risk.5 Type II is further divided — IIA and IIB have a low or high incidence of renal clear cell carcinomas, respectively, while IIC is characterised by the development of phaeochromocytomas without other VHL manifestations. Of the 1548 reported mutations in the VHL gene, 52% are missense mutations and the one previously reported Gly144Arg mutation was associated with polycythaemia.6 Renal angiomyolipomas (AMLs) occur in 0.1% of the population, and although those with bilateral renal AMLs have a greater chance of having tuberous sclerosis, this patient has no other features of the condition.7,8 However, it is possible that she has a new phenotype of Type II VHL disease associated with AMLs. Between 12% and 24% of patients with ostensibly sporadic phaeochromocytomas harbour phaeochromocytoma genetic syndromes, including VHL disease. A study of 271 such patients detected a germline VHL mutation in 11%, increasing to 42% in those who presented at age 18 years or younger.9 Therefore, it is recommended that patients diagnosed with a phaeochromocytoma before the age of 50 years should undergo screening for underlying germline mutations.10 For relatives identified as having predisposing germline mutations, investigations and follow-up are dictated by the underlying disease, but should include clinical genetics and endocrinological input. In the case of VHL disease, biochemical screening for phaeochromocytomas should start at age 4 years, and tumours should be removed if functional, as indicated by elevated catecholamine levels.4 While our patient’s experience highlights the importance of screening for underlying germline mutations, it also reinforces the unpredictable nature of phaeochromocytoma metastases. 1 Patient’s levels of urinary catecholamines and plasma metanephrines at phaeochromocytoma diagnosis and post-treatment, and metastasis diagnosis and post-treatment Investigation 1974 1976 Investigation 2008 April 2009 Feb 2010 Urine noradrenaline (RR, < 80 μg/d) 1681.8* 851.4* Urine noradrenaline / Cr (RR, 0–60 nmol/mmol) 152.3 58.2 48.5 Urine adrenaline (RR, < 20 μg/d) 86.5* 4.7* Urine adrenaline / Cr (RR, 0–10 nmol/mmol) 17 < 1.3 < 2.5 Plasma normetadrenaline Assay not available Assay not available Plasma normetadrenaline (RR, < 900 pmol/L) 3680 658 692 Plasma metadrenaline Assay not available Assay not available Plasma metadrenaline (RR, < 500 pmol/L) 591 113 122 Cr = creatinine. RR = reference range. * Noradrenaline and adrenaline / Cr ratios unavailable. 2 Octreotide scans of patient with phaeochromocytoma metastasis showing abnormal focal tracer uptake in anterior mediastinum (arrows in anterior views) * Scan 24 hours after intravenous injection of radioactive octreotide. † Scan 48 hours after intravenous injection of radioactive octreotide. Lessons from practice Persistent hypertension and elevated catecholamine levels after excision of a phaeochromocytoma are suggestive of residual or metastatic disease. Patients with a history of phaeochromocytoma require long-term surveillance. Between 12% and 24% of patients with ostensibly sporadic phaeochromocytomas harbour predisposing genetic syndromes, including von Hippel–Lindau (VHL) disease, and genetic screening is recommended in patients diagnosed under 50 years of age. VHL disease is a dominantly inherited syndrome, involving mutation of the VHL tumour suppressor gene — the mutation occurs in one in 36 000 live births.

Christopher J Yates MB BS · Sybil A McAuley MB BS · Simon Grodski MB BS, FRACS · Peter Shane Hamblin MB BS, FRACP · Peter R Ebeling MB BS, FRACP, MD

To the Editor: The article by Philip Sambrook et al on the impact of adverse news media on prescriptions for osteoporosis1 contains a number of serious errors and ignores more recent data. Their article quotes the estimated incidence of osteonecrosis of the jaw (ONJ) after oral bisphosphonate treatment for osteoporosis to be between 1 in 10 000 and 1 in 100 000 patient treatment-years. This is in fact a gross underestimate.2 The results of an independent study funded and conducted by the United States Food and Drug Administration (FDA) found an incidence of between 1 in 537 and 1 in 1537.3 This result is similar to that of an independent Australian study, which found an incidence of 1 in 296 to 1 in 1130.4 Thus, there is evidence that bisphosphonate-associated ONJ is much more common in patients treated with oral bisphosphonates for osteoporosis than the authors claim. Bisphosphonate-associated ONJ is also much more serious than presented. Patients with ONJ can be affected for years. The condition causes considerable morbidity, with greater interference in a patient’s life than the condition of osteoporosis and vertebral fractures. It should be noted that the FDA study3 and the Australian study4 were based on documented cases of ONJ, whereas the study by Sambrook et al1 was not based on actual fractures. It is unfortunate that, despite the wording of the Pharmaceutical Benefits Scheme (PBS) regulations, the Pharmaceutical Benefits Advisory Committee has let it be known that osteoporosis diagnosed by bone mineral density is not a requirement for PBS-supported bisphosphonate therapy — all that is required is a minimal trauma fracture, although it is well known that most patients with minimal trauma fracture do not have osteoporosis (however that is defined).5 The issue of adverse effects of bisphosphonate treatment for osteoporosis is currently being tested by a US class action. In the bellwether case of Boles v Merck & Co,6 the issue being tested is whether it was appropriate to prescribe alendronate for a patient with osteopenia but no fractures, and how much this treatment contributed to the patient’s end-stage ONJ. The plaintiff had exposed bone, constant pain, a pathological fracture, and pus dripping from her chin. The New York Supreme Court found in her favour. Despite the claim by Sambrook and colleagues that unbalanced media coverage “has the potential to do more harm than good”,1 the media do have an important role to play in exposing to the public the risks of pharmaceutical products and the actions of some pharmaceutical companies. It should be noted that Professor Sambrook was not only involved in The 7:30 Report current affairs program, but was also subsequently granted the unusual right of reply on that program. Following these two programs, the public made its decision based on facts about the incidence and morbidity of ONJ that have been confirmed by recent independent studies.3,4,6

Paul J Sambrook · B E Christopher Nordin · Alastair N Goss