Lancet Diabetes & Endocrinology June 2013

Special Preview June 2013
www.thelancet.com/diabetes-endocrinology
Review Comment
Title of the thyroid Subclinical article disease
See page 5 ???
Review Articles
Whitehall Title of the II: article defining subtypes of type 2 diabetes See page ???
See page 13
Review
Cystic Title of fibrosis-related the article diabetes in children See page ???
See page 30
Presenting The Lancet Diabetes & Endocrinology, the latest title in The Lancet portfolio dedicated to covering diabetes, endocrinology, and metabolism. Following in the The Lancet tradition, each monthly issue will include original research, expert reviews, informative commentary, and breaking news all of which will provide a clear, independent perspective and keep you well-informed about the clinical advancements and practicechanging research shaping your specialty. Why not acquaint yourself with the journal today?
I
Browse a selection of freely available papers published online in advance of our rst print issue in September 2013 Read more about the scope and aim of the journal and consider us for publication of your original research, review, or personal view. Stay informed about the journals developments in the lead up to launch by registering for email alerts of newly published articles and other relevant freely available online content.
Learn more at www.thelancet.com/diabetes-endocrinology, or submit a manuscript at http://ees.elsevier.com/thelancetde.
in delivering science for better health

THE LANCET is a registered trademark of Reed Elsevier Properties SA, used under licence
The Lancet Diabetes & Endocrinology 32 Jamestown Road, London NW1 7BY, UK T +44 (0)20 7424 4296 F +44 (0)1865 853021 The LancetNew York 360 Park Avenue South, New York, NY 100101710, USA T +1 212 633 3810 F +1 212 633 3853 The LancetBeijing Unit 16, 7F, Tower W1, Oriental Plaza, Beijing 100738 China T + 86 10 85208872 F + 86 10 85189297 diabetes-endocrinology@lancet. com Editor Justine Davies Deputy Editor Sarah Allan Senior Editor Fiona Mitchell Managing Editor Hannah Jones Deputy Managing Editor Laura Benham Senior Assistant Editors Olaya Astudillo Stephanie Bartlett Abi Cantor Sean Cleghorn Tim Dehnel Dara Mohammadi Odhran ODonoghue Helen Penny Frances Whinder Farhat Yaqub Assistant Editors Neil Bennet Hannah Cagney Stephanie Clague Katherine Gourd Natalie Harrison Rebecca Heald Zena Nyakoojo Louise Rishton Web Editors Richard Lane Erika Niesner Senior Assistant Web Editor Katherine Rolfe North America Editor Rebecca Cooney (New York) Asia Editor Helena Hui Wang (Beijing) Conference Editors Laura Hart Nicolai Humphreys Media Relations Manager Daisy Barton Publisher/Editorial Director Richard Horton The Lancet is a registered trademark of Reed Elsevier Properties SA, used under licence.
Special Preview June 2013
Editorial
1 Diabetes and endocrinology: extending The Lancet family
Comment
2

Towards a personalised diagnosis of type 2 diabetes

S H Wild, C D Byrne B Wilcken
3 5

Congenital adrenal hyperplasia: one hundred years of data Subclinical thyroid disease: where is the evidence?
J A Franklyn, K Boelaert
News
7

Gut microbe composition and metabolic syndrome

B Furlow J Jouret
9 10
Interlinks between sleep and metabolism New hormone stimulates pancreatic -cell proliferation
J A Gilbert
In Focus
11 Profile Jean-Pierre Chanoine: taking paediatric endocrinology global
D Holmes
12 Book Adrenaline
T K Burki
Articles
13

Trajectories of cardiometabolic risk factors before diagnosis of three subtypes of type 2 diabetes: a post-hoc analysis of the longitudinal Whitehall II cohort study
K Frch and others
22

One hundred years of congenital adrenal hyperplasia in Sweden: a retrospective, population-based cohort study
S Gidlf and others
Review
30

New insights into cystic fibrosis-related diabetes in children

K L Ode, A Moran
Previously published online See www.thelancet.com for supplementary material See www.thelancet.com for podcast Version verified by CrossMark
Cover International Advisory Board Ariel Barkan (USA) Anthony H Barnett (UK) Richard Bergenstal (USA) John B Buse (USA) Francesco Chiarelli (Italy) Miriam Cnop (Belgium) Professor Annamaria Colao (Italy) Cyrus Cooper (UK) David S Cooper (USA) Andrea Dunaif (USA) Stephen Franks (UK) Edwin Gale (UK) Hertzel C Gerstein (Canada) Lorenz C Hofbauer (Germany) Frank B Hu (USA) William J Jeffcoate (UK) Weiping Jia (China) Ursula B Kaiser (USA) Mikael Knip (Finland) Robert T Lindsay (USA) Glenn Matfin (USA) Jean Claude Mbanya (Cameroon) David M Nathan (USA) Barry Popkin (USA) Richard E Pratley (USA) Guy Rutten (Netherlands) Martin J Schlumberger (France) Paul M Stewart (UK) Cesar G Victora (Brazil) Theo J Visser (Netherlands) Paul Zimmet (Australia) Bernard Zinman (Canada)
David Marck/Skyline Imaging Ltd
Dont just get published. Get read.

Facts:
Immediate and inuential. All original research is fast-track peer-reviewed for quick dissemination of practice-changing research and policy. Others cannot come close to replicating our results, which on average from submission to publication is only 8 weeks. Global media attention. The Lancets family of journals maintain a dedicated press oce, which facilitates global media coverage for selected research and other articles, thereby providing your work with the worldwide media attention it deserves. Personalised service. Authors have direct contact with in-house Editors to benet from one-to-one advice and attention to detail to ensure papers have maximum impact and quality. Comprehensive coverage. Dedicated to providing comprehensive coverage of the eld, the journal will consider for publication the specic topics of diabetes, obesity, nutrition and metabolism, osteoporosis, adrenal disorders, bone metabolism, growth disorders, lipid disorders, neuroendocrinology, paediatric endocrinology, pituitary disorders, reproductive endocrinology, and thyroid disorders. Accessibility. The Lancet Diabetes & Endocrinology is available in print and online via platforms such as thelancet.com, ScienceDirect, and Clinical Key. Gold or Green Open Access is available for research supported by selected funding agencies with whom Elsevier has an agreement. The journal is also freely available to readers in low-to-middle-income countries via the HINARI programme or via IP access at thelancet.com.
Make a real impact, publish in

www.ees.elsevier.com/thelancetde
in delivering science for better health

THE LANCET is a registered trademark of Reed Elsevier Properties SA, used under licence
www.thelancet.com/diabetes-endocrinology
Editorial
Diabetes and endocrinology: extending The Lancet family

The 20th century saw increased understanding and dissemination of information about disorders affecting metabolism and endocrine function. A debate on the nomenclatureincluding use of the indefensible word endocrinetook place within the pages of The Lancet in 1917. Nearly 100 years on, the study of endocrine systems and metabolic disorders has become a diverse and colourful specialty. Disorders of hormone secretion affect all parts of the body and influence health and wellbeing by targeting reproduction, homoeostasis, metabolism, growth, and development. Endocrine and metabolic disorders, and associated conditions, contribute substantially to the increasing burden of non-communicable disease. Around 20% of American adults have an endocrine disorder. Globally, an estimated 475 million adults and 4050 million children are obese, 200 million adults have osteoporosis, and 200 million have a thyroid disorder. Diabetes affects more than 360 million people worldwide, with people in low-income and middle-income countries accounting for 80% of cases. Diabetology, in particular, has emerged as a front runner in cutting-edge medical research in the 21st century as strides have been made in disease prevention, screening and monitoring, pharmaceutical development, and understanding of pathophysiology. Endocrine and metabolic disorders have a substantial effect on society, and to facilitate the dissemination of information that has the potential to change lives, The Lancet Diabetes & Endocrinology will launch in September, 2013. This preview issue is published to coincide with the Endocrine Societys 95th Annual Meeting in San Francisco, CA, USA (June 1518, 2013), and the American Diabetes Association 73rd Scientific Sessions in Chicago, IL, USA (June 2125, 2013), to give an impression of the type of content and scope readers can expect when the journal launches formally later this year. Topics considered by the journal include: diabetes, obesity, nutrition and metabolism, osteoporosis, adrenal disorders, bone metabolism, growth disorders, lipid disorders, neuroendocrinology, paediatric endocrinology, pituitary disorders, reproductive endocrinology, and thyroid disorders. In keeping with the other specialty journals from The Lancet family, The Lancet Diabetes & Endocrinology will publish a range of timely news, views, research, and reviews that advocate change in or illuminate clinical practice, and will offer a fast-track peer-review service for all research articles (submission to online publication within 8 weeks). Expedited priority peer-review will also be available for other selected articles as deemed appropriate by the editorial team. Substantial progress in understanding has been made since use of the term endocrine was first debated, but more is needed to better tackle endocrine disorders. The Lancet Diabetes & Endocrinology hopes to expedite this journey. We hope you enjoy this preview issue and look forward to having you join us on the road ahead. The Lancet Diabetes & Endocrinology
For the latest content and to read more about The Lancet Diabetes & Endocrinology see http://www.thelancet.com/ diabetes-endocrinology To submit a paper to The Lancet Diabetes & Endocrinology see www.ees.elsevier.com/ thelancetde To contact The Lancet Diabetes & Endocrinology email diabetes-endocrinology@lancet. com
www.thelancet.com/diabetes-endocrinology June 2013
Science Photo Library
Comment
Towards a personalised diagnosis of type 2 diabetes

Diagnostic criteria for type 2 diabetes are based on thresholds of biochemical tests, which in turn are based on associations with microvascular disease (particularly retinopathy). The most frequently used criteria are based on random glucose, fasting glucose, and 2 h glucose concentrations after a 75 g oral glucose challenge and, more recently, glycated haemoglobin (HbA1c) concentrations.1,2 One measured concentration greater than the threshold is sufficient to diagnose type 2 diabetes in people with polyuria, polydipsia, or unexplained weight loss; two increased concentrations are needed for diagnosis in asymptomatic individuals. In The Lancet Diabetes & Endocrinology, Kristine Frch and colleagues3 use data from the Whitehall II study to investigate the heterogeneity of type 2 diabetes on the basis of patterns of fasting and 2 h glucose concentrations at diagnosis of screen-detected diabetes. These patterns might show different roles for pancreatic -cell function and hepatic and muscular insulin resistance in the pathogenesis of type 2 diabetes in different people. Evidence that bariatric surgery or severe calorie restriction can achieve normoglycaemia in individuals with type 2 diabetes, at least temporarily, suggests that type 2 diabetes is not necessarily a disease caused by or implicate permanent pancreatic -cell failure.4,5 Genetic studies and insights from incretin biology have shown the complex pathogenesis of type 2 diabetes.6,7 It is now evident that failure to maintain blood glucose concentrations within a physiological range might occur because of a heterogeneous mixture of changed physiological functions, including insulin secretion and sensitivity within several tissue typeseg, pancreas, liver, intestine, adipose, skeletal muscle, brain. Even though the main pathological changes that result in hyperglycaemia can occur in one or all of these tissues, treatment recommendations for type 2 diabetes are the same irrespective of the very different pathogenesis between individuals. Further understanding of the heterogeneity of the pathogenesis of type 2 diabetes could improve treatment choices and subsequent outcomes. The Whitehall II study provides a valuable opportunity to describe the heterogeneity of type 2 diabetes because it has a large, well characterised sample for which data from repeated oral glucose tolerance tests are available. Frch and coworkers analysed data for 6843 participants, including 274 people with screen-detected type 2 diabetes defined on the basis of fasting or 2 h glucose concentrations, or both. 55 (201%) patients met the fasting criterion only, 148 (540%) met the 2 h criterion only, and 71 (259%) met both criteria during a median follow-up of 142 years. Differences in patterns of glucose and insulin concentrations at repeated measurements during follow-up were noted between the three subgroups (ie, fasting, 2 h, and fasting and 2 h) with screen-detected diabetes as early as 18 years before diagnosis. A distinct pattern of increasing insulin secretion (as shown by 2 h insulin concentrations) from 1015 years before diagnosis to a peak 5 years before diagnosis, with subsequent falling concentrations (accompanied by declining -cell function) was reported only in the subgroup with increased fasting and 2 h glucose concentrations. In the fasting subgroup and the 2 h subgroup, 2 h insulin concentrations continued to rise until diagnosis. Additionally, the fasting and 2 h subgroup had the greatest cardiovascular risks and bodymass indices, and the lowest adiponectin concentrations. Frch and colleagues acknowledge that 2 h glucose concentrations are rarely measured in clinical practice, and used classification tree methods to attempt to identify factors that distinguish diabetes identified on the basis of fasting glucose concentrations alone from that identified on the basis of fasting and 2 h concentrations. They concluded that 55 of 65 (85%) people with fasting and 2 h hyperglycaemia would be identified by the combined cutoffs of fasting glucose concentrations of greater than 72 mmol/L and HbA1c greater than 63%. However, they did not describe what factors could be used to identify the large subgroup of people with abnormal 2 h concentrations only, and did not appropriately show that their finding needs confirmation in other datasets. Furthermore, the Whitehall II data presumably provide an opportunity to study the natural history of isolated increased 2 h glucose concentrations, and it would be interesting to know what proportion of people in this group subsequently developed increased fasting glucose concentrations. Frch and colleagues recommend that strategies to prevent loss of -cell function might be beneficial for people with fasting hyperglycaemia, but do not suggest
David Mack/Science Photo Library
Published Online February 21, 2013 http://dx.doi.org/10.1016/ S2213-8587(13)70010-X See Online/Articles http://dx.doi.org/10.1016/ S2213-8587(13)70008-1
Comment
what these strategies might be. They report a subgroup analysis from the Diabetes Prevention Program,8 which suggested that metformin seemed to be more effective at preventing diabetes in people with high fasting glucose concentrations (6269 mmol/L) at baseline than in those with low fasting concentrations (5361 mmol/L) in a population that also had 2 h glucose concentrations in the range of 78110 mmol/L. Frch and coworkers appropriately conclude that further research is needed to establish whether the different subgroups of type 2 diabetes (as defined by glucose or HbA1c criteria) respond differently to lifestyle interventions or drugs, and whether complications differ between subgroups. A better understanding of the pathogenesis of type 2 diabetes in an individual will probably lead to improved targeting of available effective interventions for hyperglycaemia. Sarah H Wild*, Christopher D Byrne
Centre for Population Health Sciences, University of Edinburgh, Edinburgh EH8 9AG, UK (SHW); Nutrition and Metabolism, Faculty of Medicine, University of Southampton, Southampton, UK (CDB); and Southampton National Institute for Health Research Biomedical Research Centre, Southampton, UK (CDB) sarah.wild@ed.ac.uk
We declare that we have no conflicts of interest. 1 2 3 WHO, International Diabetes Federation. Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia. Geneva: World Health Organization, 2006. WHO. Use of glycated haemoglobin (HbA1c) in the diagnosis of diabetes mellitus. Geneva: World Health Organization, 2011. Frch K, Witte DR, Tabk AG, et al. Trajectories of cardiometabolic risk factors before diagnosis of three subtypes of type 2 diabetes: a post-hoc analysis of the longitudinal Whitehall II cohort study. Lancet Diabetes Endocrinol 2013; published online Feb 21. http://dx.doi. org/10.1016/S2213-8587(13)70008-1. Arterburn DE, Bogart A, Sherwood NE, et al. A multisite study of long-term remission and relapse of type 2 diabetes mellitus following gastric bypass. Obes Surg 2013; 23: 93102. Lim EL, Hollingsworth KG, Aribisala BS, Chen MJ, Mathers JC, Taylor R. Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia 2011, 54: 250614. Morris AP, Voight BF, Teslovich TM, et al. Large-scale association analysis provides insights into the genetic architecture and pathophysiology of type 2 diabetes. Nat Genet 2012; 44: 98190. Baggio LL, Drucker DJ. Biology of incretins: GLP-1 and GIP. Gastroenterology 2007; 132: 213157. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002; 346: 393403.
4 5
6 7 8
Congenital adrenal hyperplasia: one hundred years of data

It is rare to find a data-rich review of the diagnosis of only one disorder that spans over a hundred years. But, in The Lancet Diabetes & Endocrinology, Sebastian Gidlf and colleagues1 describe the known cases of congenital adrenal hyperplasia in Sweden between 1910 and 2011, a period that encompasses the discovery and implementation of effective treatment in 1950, the gradual development of better diagnostic methods, and the introduction of early diagnosis by neonatal screening in the 1980s. Congenital adrenal hyperplasia is the most common adrenal disorder in children. Indeed, it is a group of disorders, the most common type being 21-hydroxylase deficiency, associated with low cortisol and aldosterone production. Clinical presentation includes potentially fatal salt-wasting crises, female genital virilisation, and premature pubarche. Gidlf and colleagues Article contributes to our understanding of the disorder in interesting and surprising ways. Most high-income countries
have introduced neonatal screening for congenital adrenal hyperplasia, some as much as 30 years ago,2 and indeed it has recently been introduced in Laos, a country with no previous neonatal screening programme.3 However, congenital adrenal hyperplasia screening has not been implemented in either the UK4 or Australia.5 Early detection of the disorder is mainly aimed at the prevention of salt-wasting crises, wrong sex-assignment, and premature pubarche or accelerated growth. Gidlf and colleagues data clearly show the apparent increase in incidence over time, which they attribute to poor diagnosis before the 1960s and to the fact that the availability of treatment and increased awareness of the disorder increases the likelihood of physicians identifying and diagnosing the disease. Importantly, the investigators postulate that the frequency of this global disease is likely to have remained steady over the period, and thus they are able to calculate the probable number of missed cases over time,
Published Online February 26, 2013 http://dx.doi.org/10.1016/ S2213-8587(13)70009-3 See Online/Articles http://dx.doi.org/10.1016/ S2213-8587(13)70007-X
Sidney Moulds/Science Photo Library
Comment
assuming that almost all severe cases (the saltwasting phenotype) are now effectively diagnosed. Their postulations of the numbers of missed cases was enlightening. Male neonates with the severe salt-wasting phenotype are thought to have a higher risk of death than their female counterparts, because diagnosis of female neonates is easier owing to their virilised genitalia, leading to the possibility that they receive treatment earlier and more often than do boys. Findings from Gidlf and colleagues study, however, showed that in people with the salt-wasting form of the disorder, the risk of death was substantialsome five to ten patients died every year before 19701and much the same in both female and male patients without early diagnosis. Evidently, both male and female babies die undiagnosed, and not, as previously thought, only male babies. Other important information provided by Gidlf and colleaguesinformation usually unavailable in reports of screeningis an estimated false-negative rate of was almost 16%. Such estimates are not usually given in reports of screening, but one report that did, from Minnesota, USA,6 showed a false-negative rate of 22% (15 of 67 cases) for people with the classic form of the disorder during 12 years of neonatal screening between January 1999 and 2010. Of 15 missed cases, five had the salt-wasting form, but four of these five were girls who were diagnosed on the basis of their virilised genitalia. In Gidlf and colleagues Swedish study, more than half of patients had late-onset, non-classic forms of the disorder that were missed by screening.1 However, the detection of late-onset cases is, arguably, not the main target of neonatal screening. Publication of more detailed data from existing screening programmes would be welcome. It is surprising that after 30 years of neonatal screening worldwide there is still a need for additional screening data and, importantly, followup, so that the benefits of screening can be accurately assessed, and screening efficiency can be maximised. Uncertainty still exists about outcomes, and how screening can improve outcomes. An adrenal crisis with accompanying hyponatraemia is thought to cause brain damage, but available evidence does not
lend support to the suggestion that patients with congenital adrenal hyperplasia have any intellectual deficits compared with otherwise healthy individuals.7,8 A clear need exists for more research in this area to be sure that more subtle learning difficulties are not present. However, there is little doubt that screening for the disorder fulfils the essential criteria for screeningit is, after all, a potentially lethal disorderand a 2010 study in the UK concluded that a case can be made for screening.4 Certainly paediatric endocrinologists from Australia agree.5 The Swedish study underlines what can be learned from long-term follow-up, good record keeping, and registers. This type of activity should not only be encouraged but also funded if we are to make best use of our accumulated experience. At the same time, we should remember that for any long-term study, data collected at the beginning might not be entirely comparable with those collected towards the end. Medical diagnostic and therapeutic expertise moves on, so we need to draw conclusions with care. Bridget Wilcken
Sydney Childrens Hospitals Network and University of Sydney, Sydney, Australia bridget.wilcken@health.nsw.gov.au
I declare that I have no conflicts of interest. 1 Gidlf S, Falhammar H, Thiln A, et al. One hundred years of congenital adrenal hyperplasia in Sweden: a retrospective, population-based cohort study. Lancet Diabetes Endocrinol 2013; published online Feb 26. http:// dx.doi.org/10.1016/S2213-8587(13)70007-X. White PC. Neonatal screening for congenital adrenal hyperplasia. Nat Rev Endocrinol2009; 5: 490498. Hoehn T, Lukacs Z, Stehn M, Mayatapek E, Philavan K, Bounnack S. Establishment of the first newborn screening program in the Peoples Democratic Republic of Laos. J Trop Pediatr 2012; published online Oct 26. DOI:10.1093/tropej/fms057. Khalid JM, Oerton JM, Dezateux C, Hindmarsh PC, Kelnar CJ, Knowles RL. Incidence and clinical features of congenital adrenal hyperplasia in Great Britain. Arch Dis Child. 2012; 97: 10106 Warne GL, Armstrong KL, Faunce TA, et al. The case for newborn screening for congenital adrenal hyperplasia. Med J Aust 2010; 192: 107. Sarafoglou K, Banks K, Kyllo J, Pittock S, Thomas W. Cases of congenital adrenal hyperplasia missed by newborn screening in Minnesota. JAMA 2012; 307: 237174. Berenbaum S. Cognitive function in congenital adrenal hyperplasia. Endocrinol Metab Clin North Am 2001; 30: 17392 Berenbaum SA, Bryk KK, Duck SC. Normal intelligence in female and male patients with congenital adrenal hyperplasia. Int J Pediatr Endocrinol 2010; 2010: 853103.
2 3
4 5 6 7 8
Comment
Subclinical thyroid disease: where is the evidence?

Subclinical thyroid disease is very common, particularly in elderly people. Recognition of this endocrine disorder is increasing, partly due to a large increase in thyroid function testing, especially in primary care. Many crosssectional studies have investigated whether subclinical hyperthyroidism or subclinical hypothyroidism are associated with specific symptoms, signs, or comorbidities, and a smaller number of prospective studies have examined whether subclinical thyroid disease predicts specific adverse outcomes.1 What is the latest evidence driving the need, or otherwise, for therapeutic intervention in these common, and largely asymptomatic, biochemically defined disorders? A large and seemingly irrefutable body of evidence exists supporting the association of subclinical hyperthyroidism with atrial fibrillation risk, especially when thyroid-stimulating hormone (TSH) is at undetectable concentrations.2,3 Subclinical hyperthyroidism is also associated with other adverse cardiac outcomes, such as coronary heart disease events and mortality, and heart failure. Evidence linking cardiovascular disorders with tests of thyroid function within the reference range, including higher circulating free thyroxine concentrations,4 suggests that the cardiovascular system is particularly sensitive to subtle changes in thyroid status. Thus, the cardiovascular system is the most important physiological system for which to consider risk, and, in turn, with the potential to benefit from treatment. If the evidence for risk association with cardiovascular endpoints is strong, why is there controversy about intervention? Several crucial reasons exist. First, association does not prove causation, and many studies have not fully considered potential confounders for comorbidities in conditions such as coronary heart disease and heart failure. Second, nearly all studies have been based on one or two TSH measurements in individual subjects, but low TSH is often transient, especially when only slightly low, and frequently reflects non-thyroidal illnesses or drugs, rather than underlying thyroid disease such as mild Graves disease or toxic nodular hyperthyroidism. Third, intervention for subclinical hyperthyroidism means radioiodine therapy or antithyroid drugs, neither of which is trivial. Radioiodine is generally considered
the treatment of choice for toxic nodular disease (the most common underlying thyroid diagnosis, in view of the typical age when subclinical hyperthyroidism is diagnosed). Radioiodine treatment often results in hypothyroidism and the need for permanent thyroxine replacement. Since up to half of patients taking thyroxine have subclinical hyperthyroidism or hypothyroidism biochemically, subclinical hyper thyroidism can be perpetuated or replaced with subclinical hypothyroidism. Finally, there have been no randomised controlled trials of treatment of subclinical hyperthyroidism with meaningful clinical endpoints. Two trials were stopped because of low recruitment and a third has recruitment that is lower than planned, although it is continuing. These issues were described in a recent article about the problems encountered with such trials.4 Despite this absence of evidence, expert groups recommend that treatment should be strongly considered, especially in elderly patients;5 surveys of practice show this is occurring. It seems extraordinary that evidence that the benefit of treatment outweighs the risk does not exist in the 21st century for such a common disorder. We can only hope that evidence will accrue in the next few years. The situation regarding subclinical hypothyroidism is probably more complex and controversial than that for subclinical hyperthyroidism. The most relevant physiological system is again cardiovascular; the largest meta-analysis performed so far reports an association with cardiovascular mortality in more severe cases (ie, serum TSH >10 mIU/L).6 Again, raised TSH is frequently transient, although a persistent increase is a more specific indicator of underlying thyroid disease. However, the upper limit of the TSH reference range rises with age,7 leading to controversy about the definition of disease, especially if TSH is in the 510 mIU/L range in elderly people. Randomised controlled trials of treatment (thyroxine replacement) have been done, but these are largely small, heterogeneous, and underpowered, and their findings are unsurprisingly negative or conflicting. A Cochrane review indicated insufficient evidence to recommend for, or against, treatment, including in those with a TSH greater than 10 mIU/L and in very elderly patients.8 However, new evidence9 exists for improved outcomes
Published Online May 23, 2013 http://dx.doi.org/10.1016/ S2213-8587(13)70030-5
Comment
of coronary heart disease in younger, but not older, patients treated with thyroxine, and there are new data10 showing that thyroxine treatment helps to preserve renal function in people with subclinical hypothyroidism and chronic kidney disease. Fortunately, several clinically relevant trials are underwayincluding one EU-funded multicentre study of patients older than 80 years that will examine cardiovascular and qualityof-life outcomesso the evidence base for subclinical hypothyroidism should increase and better guide us in our therapeutic approach. *Jayne A Franklyn, Kristien Boelaert
Centre for Diabetes, Endocrinology and Metabolism, School of Clinical and Experimental Medicine, College of Medical and Dental Sciences, University of Birmingham, Birmingham, B15 2TT, UK (JAF and KA) j.a.franklyn@bham.ac.uk
We declare that we have no conflicts of interest.
1 2
Cooper DS, Biondi B. Subclinical thyroid disease. Lancet 2012; 379: 114254. Collet TH, Gussekloo J, Bauer DC, et al. Subclinical hyperthyroidism and the risk of coronary heart disease and mortality. Arch Intern Med 2012; 172: 799809. 3 Gammage MD, Parle JV, Holder RL, et al. Association between serum free thyroxine concentration and atrial fibrillation. Arch Intern Med 2007; 167: 92834. 4 Goichot B, Pearce SH. Subclinical thyroid disease: time to enter the age of evidence-based medicine. Thyroid 2012; 22: 76568. 5 Bahn RS, Burch HB, Cooper DS, et al. Hyperthyroidism and other causes of thyrotoxicosis: management guidelines of the American Thyroid Association and American Association of Clinical Endocrinologists. Endocr Pract 2011; 17: 456520. 6 Rodondi N, den Elzen WP, Bauer DC, et al. Subclinical hypothyroidism and the risk of coronary heart disease and mortality. JAMA 2010; 304: 136574. 7 Waring AC, Arnold AM, Newman AB, Buzkova P, Hirsch C, Cappola AR. Longitudinal changes in thyroid function in the oldest old and survival: the cardiovascular health study all-stars study. J Clin Endocrinol Metab 2012; 97: 394450. 8 Villar HC, Saconato H, Valente O, Atallah AN. Thyroid hormone replacement for subclinical hypothyroidism. Cochrane Database Syst Rev 2007; 3: CD003419. 9 Razvi S, Weaver JU, Butler TJ, Pearce SH. Levothyroxine treatment of subclinical hypothyroidism, fatal and nonfatal cardiovascular events, and mortality. Arch Intern Med 2012; 172: 81117. 10 Shin DH, Lee MJ, Kim SJ, et al. Preservation of renal function by thyroid hormone replacement therapy in chronic kidney disease patients with subclinical hypothyroidism. J Clin Endocrinol Metab 2012; 97: 273240.
News
Gut microbe composition and metabolic syndrome

Links between gut microbiota and metabolic diseases have been suggested for many years, but results of studies in human beings and animals are conflicting. Bryant Furlow discusses new research in this area as a new study shows that promoting growth of bifidobacteria reduces risk factors for metabolic syndrome.
The interactions between gut microbial communities, obesity, and human health are complex, and researchers have only begun to thoroughly describe and disentangle them, experts tell The Lancet Diabetes & Endocrinology. The evidence base for therapeutic strategies based on manipulation of the microbiome is still largely descriptive and preclinical, they caution. But a new clinical study offers reason for cautious optimism that manipulating the microbiomes composition in bifidobacterias favour just might help reduce physiological risk factors for metabolic syndrome in overweight adults. The role of microbial fermentation of polysaccharides in facilitating the guts energy harvest is widely appreciated, as is the fact that gut mucosa represents the primary site for microbial interactions with the human immune system, which can trigger local and systemic inflammation, and possibly islet destruction. Mounting preclinical evidence also implicates the microbiome in the release of hormones in the gut, insulin resistance, lipid metabolism, and adiposityand even satiety and mood. Some beneficial gut bacteria do seem to be protective. Bifidobacteria, for example, seem to provide a barrier to mucosal absorption and circulation of antigens from other bacteriaantigens believed to underlie obesity-associated metabolic endotoxaemia. The gut microbiota population is altered towards a less beneficial composition in overweight adults and this change can be accompanied by inflammation, notes lead author Jelena Vulevic (University of Reading, Reading, UK). Bifidobacteria are an exclusively beneficial group of gut bacteria; all bifidobacterial species possess beneficial properties, Vulevic notes. However, for some species, these beneficial effects have been better characterised than for the others. For example, Bifidobacterium bifidum is known to be one of the main species present in healthy breast-fed infants and is absent in infants suffering from allergies. The study is significant since it is one of the first studies of human prebiotic supplementation that has examined the effect of such a potential therapeutic on the gastrointestinal microbiome, which is increasingly being recognised as a key component in defining metabolic and immune status in humans, says Susan Lynch, Director of the Colitis and Crohns Disease Microbiome Research Core at University of California San Francisco, San Francisco, CA, USA. In Vulevics double-blind study of 45 overweight adults with three or more risk factors for metabolic syndrome, participants were randomly assigned to receive either maltodextrin placebo or a mixture of trans-galactooligosaccharides (B-GOS, a prebiotic designed and developed using enzymes from gut-dwelling B bifidum), during a 12-week intervention. B-GOS supplementation was associated at 12 weeks (but not at 6 weeks) with increased numbers of bifidobacteria in participants faecal samples, increased secretory IgA, and decreased faecal calprotectin, and plasma levels of C-reactive protein, insulin, triglyceride, and total cholesterol (TC), and TC-to-HDL cholesterol ratios. Administration of B-GOS to overweight adults resulted in positive effects on the composition of the gut microbiota, the immune response, and insulin, TC, and triglyceride concentrations, Vulevics team reports. B-GOS may be a useful candidate for the enhancement of gastrointestinal health, immune function, and the reduction of metabolic syndrome risk factors. This is an important study in a human population, demonstrating the utility of non-digestible prebiotics as a means to modulate gastrointestinal microbiome com position and significantly improve clinical outcomes in obese subjects, even over a relatively short-term intervention period, says Lynch, who was not involved in the study. The study suggests that supplementation with a mixture of trans-galactooligosaccharides, per haps in combination with other lifestyle changes, may represent an attractive and relatively inexpensive therapeutic strategy for management of obesity-associated metabolic syndrome. The study also bolsters the case that bifidobacteria has a pivotal role in clinical improvements, Lynch notes. Vulevic was surprised that the effects of B-GOS supplementation were not evident before 6 weeks, and suspects the stability of participants microbiota is related to food habits, large calorie intake, and overeating among the volunteers. As a
Published Online January 2013 For more on the gut microbiota, obesity and insulin resistance see Molec Aspects Med 2013; 34: 3958. DOI:10.1016/j. mam.2012.11.001 For more on the potential for dietary manipulation of the gut microbiome and regulatory issues see Curr Opin Biotechnol 2012; 23: 192201. DOI:10.1016/j. copbio.2011.11.004 For the Vulevic and colleagues study see J Nutrit 2013; published online Jan 9. DOI:10.3945/jn/112/166132
News
For the newly-reported pilot study see J Dairy Sci 2013; 96: 8995. DOI:10.3168/ jds.2012-5863
consequence, I think any dietary intervention intended to alter gut microbiota will be in part lost due to these dietary habits and the bioavailability of the substrates, Vulevic cautions, adding that such interventions might fare better if introduced at earlier stages of metabolic syndrome. However, studies of other dietary interventions, such as supplementation with lactobacillus, have yielded mixed results in human and animal studies. In a newly-reported pilot study of 30 human patients with metabolic syndrome, Lactobacillus casei Shirota did not affect insulin sensitivity, -cell function, or biomarkers of inflammation and gut endothelial dysfunction over a 12-week intervention. Another suspected risk factor for metabolic syndrome is chronic stressand this, too, may interact importantly with the microbiome. Evidence is emerging that cortisol levels are affected by the microbiota, Vulevic notes. Mark Lyte (Texas Tech University Health Sciences Center, Abilene, TX, USA) proposes that gut bacteria secrete neuroendocrine hormones to communicate with the enteric nervous system and CNS, and that an intersection exists between neuroendocrinology and
Bifidobacteria colonise the gut
microbiology, which he has termed microbial endocrinology. The bacteria in the intestinal tract are communicating with our brains, says Lyte. Thats a real game changer. Anna Tagliabue from the Human Nutrition and Eating Disorders Research Centre (University of Pavia, Pavia, Italy) cautions that although the study is promising, it is premature to speak of a diet or supplement that can influence metabolic syndrome. Its still early days for gut microbiome research, Vulevic and others agree. The species composition of the human gut microbiome is not well understood, which is not surprising since the gut is a profoundly complex ecosystem, comprised of trillions of bacteria. Early-life exposures to antibiotics probably disrupt normal development of gut microbiota in lasting ways. Marie-France de La Cochetire (INSERMInstitute, Nantes, France), who specialises in molecular analysis of the intestinal microbiota, says that although antenatal antibiotics cannot be totally avoided, each course in early life is like an elephant in a porcelain store. Which bacteria may be proin flam matory or conversely antiinflammatory in type 1 diabetes is not yet known, notes Danny Zipris, (University of Colorado, Denver, CO, USA). We are in the process of analysing the intestinal microbiota of children at risk for type 1 diabetes and hope to identify bacterial strains associated with the disease. Until recently, the composition of the adult gut microbiome was believed to be stable, a steady-state ecosystem. But some suspect that this vision might have been an artifact of simple and relatively insensitive sampling methods. Most gut bacteria cannot be cultureda problem now being remedied with sophisticated next-generation genome sequencing.
Gary Gaugler/Science Photo Library
Sampling techniques remain a challenge. Vulevics team, like most researchers, used faecal sampling, which avoids invasive colonoscopy and the alterations to the gut microbiome that can follow fasting and bowel preparation before colonoscopic procedures. But faecal sampling is far from ideal because the microbes in direct contact with human tissues are not as well represented in faeces as are the lumen-dwelling species associated with food wastes. Establishing causality remains the central conceptual challenge for the field and effectively deciphering microbiome interactions with human immune and endocrine systems and their influence on health will be an interdisciplinary endeavour, Lyte and de La Cochetire agree. Already, however, the field is fostering a reconceptualisation of the human body: To me, the most important advances are in the concept of human as supraorganism: the microbiome ecosystem within the human one, says de La Cochetire. Vulevics team is now studying the effects of B-GOS in diabetic and overweight adults, and its effects on the gutbrain axis and modulation by beneficial gut microbes of cortisol levels, anxiety, and cognition. I think we are currently right in the middle of very exciting times in which human gut microbiota is becoming better understood and implicated in various conditions, Vulevic said. However, we have very little understanding of the mechanisms that underpin some of the observed effects of the human gut microbiota. This is perhaps not surprising, given the diversity and size of the human microbiome. Lots and lots of work has still to be done, agrees de La Cochetire. And each step reveals a new world.
Bryant Furlow
News
Interlinks between sleep and metabolism

Lack of sleep is increasingly associated with weight gain and metabolic problems. Interfaces between the pathways that regulate circadian timing and metabolism might underlie these adverse health effects. Jill Jouret reports.
Getting a good nights sleep is a basic, but often eluded, prescription for good health. Modern lifestyles provide opportunities for 24 h activity, and minimising sleep is often thought to be a harmless, efficient, or merely necessary means to accommodate schedules. However, feeling tired at night is more than an instruction to rest. Behaviour and physiology are intricately linked to light and dark cycles, and an internal timing mechanism has evolved to ensure that physiological processes occur at optimum times in a 24 h cycle. Maintaining the synchrony of this endogenous circadian clock seems to have wide-ranging health implications. Although the mechanisms are not fully clear, evidence is mounting that insufficient sleep and disruption in circadian rhythms contribute to pathogenesis of metabolic disorders, cardiovascular disease, and cancer. Worldwide, meta bolic syndrome is on the rise, as is the introduction of artificial light and activity into nighttime hours. Epidemiological and clinical studies have shown that shortduration and poor-quality sleep predict development of type 2 diabetes and obesity, suggesting that sleep, circadian rhythms, and metabolic systems are interconnected. In mammals, circadian rhythms are generated centrally by the suprachiasmatic nuclei in the anterior hypothalamus. Light perception by the retina synchronises these single-celled oscillators, generating rhythmic outputs that regulate sleep and wakefulness, feeding and energy expenditure, and glucose homoeostasis. This central clock also sends signals via direct innervation and humoral factors to clock components in peripheral tissues, thus maintaining the circadian timing of an array of physiological processes. Transcriptiontranslation feedback loops implicating specific clock genes lead to a roughly 24 h cycle. Molecular links between circadian and metabolic pathways have been identified and many hormones implicated in metabolism and energy balance exhibit circadian oscillation eg, expression and secretion of leptin, a hormone that signals satiety, peaks at night. The complex signalling systems that govern glucose homoeostasis and metabolism of fatty acids, cholesterol, bile acids, and toxins receive inputs from the local and central circadian clocks, allowing cells to anticipate metabolic reactions in a 24 h period. Invitro studies show that metabolic cues can be transmitted to core components of the circadian clock. Such crosstalk suggests a mechanism by which eating (and possibly sleeping) patterns could shift innate circadian timing. A study published in March, 2013, by a group at the University of Surrey (Guildford, UK) highlighted the interconnection between sleep, circadian rhythmicity, and metabolism. Whole-blood RNA samples were taken from participants after a week of restricted nightly sleep (57 h) and also after a week of adequate sleep (85 h). Transcriptome analysis showed that 711 genes were upregulated or downregulated by insufficient sleep, including genes associated with circadian rhythms and metabolism. Sleep restriction also reduced the total number of genes with circadian expression profiles, implying that even a week of poor sleep can disrupt the bodys intricate physiological timing. Melatonin, a key regulator of sleep, could be an important link connecting circadian timing and insulin signalling. Melatonin production is suppressed by light, and peaks around 35 h after sleep onset; it regulates the sleepwake cycle by lowering body temperature and causing drowsiness, and also inhibits insulin secretion by pancreatic cells. A 2013 case-control study within the Nurses Health Study cohort showed that, compared with women in the highest category of melatonin secretion, women in the lowest category had about a twice the risk of developing type 2 diabetes (after controlling for demographic, lifestyle, and other risk factors). Previous studies have shown that single nucleotide polymorphisms of the melatonin receptor are also associated with an increased risk of type 2 diabetes. More clinical research is needed to characterise this association between sleep, melatonin concentrations, and type 2 diabetes, and to elucidate, for example, whether melatonin supplementation has a role in treatment. Irregular and extended working hours are a reality for many industries, and epidemiological studies have shown lower melatonin concentrations in night-shift workers than in day-shift workers and an
Published Online April 2013 For the study of insufficient sleep and transcriptome analysis see Proc Natl Acad Sci USA 2013; published online March 19. DOI:10.1073/pnas.1217154110 For the study of melatonin secretion and incidence of type 2 diabetes see JAMA 2013; published online April 3. DOI:10.1001/jama.2013.2710
Oscar Burriel/Science Photo Library
News
For the study on sleep, food intake, and weight gain see Proc Natl Acad Sci USA 2013; published online April 2. DOI:10.1073/pnas.1216951110
increasing risk of type 2 diabetes with number of years of shift work. For this substantial proportion of the workforce, more solutions are needed to prevent people from falling into economically driven health traps. Insufficient sleep is a risk factor for weight gain and obesity, in addition to type 2 diabetes, and understanding the underlying mechanisms could help to guide novel weight-loss strategies. A study published on April 2, 2013, showed that eating behaviours, particularly night-time eating, con tributed to weight gain during sleep loss. Whole-room calorimetry measured daily energy expenditure in adults undergoing 5-day cycles of inadequate (5 h) or adequate (9 h) nightly sleep. Energy expenditure was about 5% higher with insufficient sleep, but increased food intake more than compensated for this energetic cost. In the sleep-loss condition, participants ate a smaller breakfast but consumed 42% more calories as afterdinner snacks, leading to weight gain. The study investigators suggested
that participants eating patterns during sleep loss resulted from a delayed circadian phaseie, a later onset of melatonin secretion at night, assessed by hourly blood samples from an intravenous catheterwhich might have led to a circadian drive for more food intake. Furthermore, the time between waking and melatonin offset was longer in the 5 h sleep condition; thus, participants awoke during an earlier circadian phase (while still in biological night) and might have been less hungry for breakfast. Previous studies have suggested that disrupted signalling of satiety and hunger hormones leads to the overeating associated with insufficient sleep; however, in both the 5 h and 9 h conditions, excessive food intake was accompanied by appropriate increases in the satiety hormones leptin and peptide YY and decreases in ghrelin, which stimulates hunger. Future studies should examine how sleep deprivation leads to delays in circadian phase and how circadian timing of meals affects
energy metabolism. For the millions of people whose working week necessitates a disrupted sleep schedule, a physiological drive for more food intake, the availability of high-calorie foods, and exhaustion leading to less physical activity overall could be a potent formula for weight gain. Whether for work, play, or travel, voluntary sleep curtailment has become endemic; however, restricted sleep seems to interfere with the crosstalk between complex physiological and circadian networks that have evolved to couple our bodily functions with the Earths 24 h rotation. Many more issues deserve investigation, such as the differential effects on health of acute versus chronic sleep deprivation, and how light exposure mediates the effects of sleep loss. As more evidence emerges of the circadian orchestration of metabolism, perhaps the time has come for sleep to figure more prominently in treatment and public health guidelines.
Jill Jouret
New hormone stimulates pancreatic -cell proliferation

Diabetes affects more than 360 million people worldwide and its prevalence is increasing, with 552 million diabetics predicted worldwide by 2030. Scientists recently discovered a hormone that could improve future diabetes management by stimulating replenishment of insulin-producing cells in the pancreas. The hormone, which has been named covered in studies betatrophin, was dis of a mouse model of severe insulin resistance in which chemical blockade of insulin receptors induced pancreatic -cell proliferation. Betatrophin was identified in murine liver and fat, and its stimulatory effect on cellular replication was limited to cells. Its expression was also reported in human liver tissue. Betatrophin treatment of mice in creased proliferation of pancreatic -cells by an average of 17-fold within a few days, causing an expansion of -cell mass and increased insulin concentrations in the pancreas. Betatrophins discovery is a very exciting new development, and is only the beginning of the story, says C Ronald Kahn (Joslin Diabetes Center, Boston, MA, USA). He adds that unanswered questions include whether action on islets is direct or indirect. We dont know how betatrophin works; is it only one growth factor or one of many? There is a lot of future work to be done. Senior author Douglas Melton (Harvard University, Cambridge, MA, USA) said: Its not often that one finds a new hormone, so it opens up all kinds of possibilities for new treatments. The most immediate application for betatrophin is for the millions of prediabetics who are on their way to getting type 2 diabetes. If these individuals still have cells, this hormone could give them more cells and alleviate the need for insulin injections, Melton continued. Betatrophin may also prove beneficial in type 1 diabetes, which is initiated by an autoimmune process. If the disease is just starting, one could give an immunosuppressant and this hormone to forestall the onset of type1 diabetes. Melton cautions that results from human studies should not be expected quickly. We are currently working with our collaborators Evotec and Janssen to make the human betatrophin protein. This will take more than a year. Results from studies in humans might be available 23 years from now, if all goes well.
John Bavosi/Science Photo Library
Published Online May 2013 For the study see Cell 2013; 153: 74758. http://www.cell. com/abstract/S00928674(13)00449-2
Judith A Gilbert
10
In Focus
Profile Jean-Pierre Chanoine: taking paediatric endocrinology global

Living green, healthy, and thrifty is an ethos we could probably all benefit from sticking to a little more, but for Jean-Pierre Chanoines patients the stakes are considerably higher. One of Chanoines primary interests is the prevention and management of childhood obesity, and the LiGHT programme is one of his newest weapons in the battle to turn the tide against an ever-increasing epidemic. That the fundamental philosophy of the programme seems inimical to the interests of western capitalism only serves to illustrate the scale of the task ahead, Chanoine explains with a wry laugh. There is an antagonism between what we say and what the patient is exposed to when they get out of the door of the clinic, but we wont be able to solve obesity by looking at just the medical point of view because its so engrained into the way we live. If doing battle with the modern world sounds like serious business, Chanoine certainly doesnt take himself too seriously, peppering our conversation with dry, deadpan humour. The immediate past President of the Canadian Pediatric Endocrine Group, Chanoine has called the Endocrinology and Diabetes Unit in British Columbias Childrens Hospital home for the past 13 years, and is the units director. Not bad for someone who, as a child, was adamant he didnt want to go into the family business. Growing up in Belgium, Chanoines father was a physician an internist and a diabetologist in factand Chanoine remembers seeing him only fleetingly as he struggled with an impossible schedule. At the time I said well never, Im never doing this stupid job because I want to have a life, he recalls. But at the time in BelgiumI finished high school in 1975the career choices were way more limited than today. Chanoine went into medicine at the Free University of Brussels, Belgium, in the full knowledge, he says, that he would become a paediatrician. And according to Arya Sharma, Chair in Obesity Research and Management at the University of Alberta, Canada, Chanoines humble, kind hearted, knowledgeable, and soft spoken nature makes him everything one could expect from someone who works with children. It was at medical school that Chanoine got his first taste of what would later become his second major interest. I spent 2 months in the Belgian Congo, back when it was called Zaire. And I really loved that; I always wanted to get into global health, but unfortunately it did not unfold until recently. Compulsory military service intervened, and although there was an option to do social service for 2 years in a developing country, the money to fund it was impossible to come by. I ended up in a cavern in Germany for 13 months, where I had the choice between drinking beer or writing grants. I took the second one, got a grant and was happy to
go to another developing country: the USA, he laughs. It was the beginning of a peripatetic phase of his life. Although he loved his time as a fellow at the University of Massachusetts, and describes his mentor Lewis Braverman as a marvellous man, Chanoine says he always knew that, in the long run, staying in the USA was not an option. The culture is very different; the superficial aspects, the confessional aspects, the guns, he says. With a family its tough out there, and I like the quality of life of Europe. Instead he returned to Belgium, and after a brief spell in what he describes as a depressing and sterile academic setting he turned to industry, taking up a post as Medical Director for Novo Nordisks Belgian operation and founding the medical department there. It was a very interesting experience because Id never worked in a company like that, and these guys think differently. I mean these guys would sell anything to anybody. But the concept was very interesting and actually it was a good company and I learned a lot, but I knew I didnt want to do it forever, he recalls. So I was very lucky to find the position as division head for paediatric endocrinology in Vancouver, BC, Canada. Ive been here since 1997, and they never told me how long it would be for. I never asked, and Im still stuck here in 2013. Perhaps stuck is the wrong word, because as global health increasingly becomes a primary focus for Chanoine, he is spending more time abroad training others. Until 5 years ago, there were just no paediatric endocrinologists in central Africa, he says. Global health was directed, understandably, to the main killers like HIV, maternal health, nutrition. But there is a case for bringing training for specialty and even subspecialty care to developing countries, and diabetes is a great example. Its not an uncommon disease, but in the past, children were dying because there was no care, and they were never even diagnosed with diabetes. Now we can diagnose them, and although the quality of care probably remains quite poor, patients are still living 20 years more than they would have been. Together with colleagues in the Global Paediatric Endocrinology and Diabetes group, Chanoine is absolutely superb at coming up with innovative ideas, sharing a vast canvas of knowledge and experience, and endless sensible enthusiasm for projects that would advance the future, says Margaret Zacharin of the Royal Childrens Hospital in Melbourne, VIC, Australia. In May, Chanoine chairs the first symposium at the Pediatric Academic Societies meeting in Washington, DC to discuss paediatric endocrinology in lowincome settings. Things are really coming along, he says, and Im very enthusiastic about that.
Published Online January 2013
David Holmes
11
In Focus
Book Adrenaline
Adrenaline begins with an elegant couplet of questionable relevance: A little learning is a dangerous thing; drink deep, or taste not the Pierian Spring (Alexander Pope). It is not an accusation that one could make of the author, Brian Hoffman, Professor of Medicine at Harvard Medical School, Boston, MA, USA. But he wears his learning lightly, and briskly outlines his purpose. Adrenaline has enormous cultural significance as a molecule associated with medically important stress and with excitement, anger and terror, he explains. Adrenaline flows in an adventurer ascending a sheer cliff or in a young violinist debuting at Carnegie Hall. A shot of adrenaline empowers superhuman feats in emergencies. On the other hand, an adrenaline surge during a bitter argument may precipitate a heart attack or scare us to death. Hoffmans slender but detailed history of adrenaline, which he regards as the first discovered hormone, despite secretins pretensions to the title, begins, as medical biographies invariably do, in classical antiquity. But this is a rare subject on which the Greeks and Romans stand silent. Herophilus (335280 BC), the father of anatomy, did not mark the adrenal glands, nor did Galen of Pergamon (circa 129210). They go unmentioned in Andreas Vesalius (151464) opus of the human body, De humani corporis fabrica. It was Bartholomaeus Eustachius (circa 150074) who eventually described them. I consider it indicated to say something of the glands, diligently overlooked by other anatomists, he wrote waspishly. Soon enough, the same glands were identified in fish and mammals. The widespread existence of the adrenals in many species spoke to their importance, declares Hoffman, but only in the softest whisper. The authors studious research takes us next to eighteenth century France and to a competition hosted by Bordeauxs Academy of Sciences, which asked what is the use of the adrenal glands?, but found no satisfactory answer. The competitions adjudicator, incidentally, was Charles de Montesquieu (16891755). Hoffman points out, with solemn exactitude, that Montesquieu took an early interest in biology before becoming a great political essayist, which makes the great mans life sound like a question in Trivial Pursuit. Our author does have a dry sense of humour, and it is disappointing that he gives it free rein so infrequently; for the most part Adrenaline is a strikingly earnest endeavour. Witness the critique of Bob Dylans The Lonesome Death of Hattie Carroll, early in the book, which gently upbraids Dylan for ignoring the fact that adrenaline and the closely related substance noradrenaline almost certainly played a significant role in her death, a consequence of the actions of the sympathetic nervous system, all of which may well be true, but it is difficult to write a protest song about the iniquities of the sympathetic nervous system. The chapter A Country Doctors Remarkable Discovery recounts the experiments of English physiologist George Oliver (18411915). His work, measuring the effects of extracts from the adrenal glands on blood pressure, along with Edward Schfer (18501935), who came up with the term endocrine, was of crucial importance in advancing our understanding of adrenaline. An entertaining aside with slightly wistful undertones, View Through a Retrospectoscope applies todays exacting safety standards to Olivers era. Oliver did not have to obtain informed consent from volunteers, nor did he have to deal with government agencies concerned with safe manufacturing practices for experimental substances, writes Hoffman, sounding for all the world like a man whose experiences with government agencies have seldom quickened him with joy. On the other hand, he adds judiciously, in his day, investigators felt a duty to test new chemicals on themselves first, an activity not generally acceptable now. Olivers might have been a simpler time, but it had its own complications. This book is predominantly a tale of the twentieth century. Adrenaline was first named in a scientific journal in 1901; 100 years later, musicians were using -receptor antagonists to quash performance nerves. The chapter and subchapter headingsMolecular Basis for the Activation of Glycogen Phosphorylase; The Perils of an Insufficient Sympathetic Nervous Systemtestify that this is not really a book for the layman, and there are hints that Adrenaline may have started life as a series of lectures. Still, the biochemistry is explained with clarity and Hoffman is a convivial enough host. The text is studded with lively anecdotes, including the amusing story of how the author managed to escape from an enraged motorist. A stocky man leaped out of the car and started towards me on foot; he called out once, youre going down. Somewhat alarmed, I balanced precariously on my bicycle, feet clipped into the pedals. The terrified academic managed to take refuge in a refreshment tent. Not built for fighting, then, but at least in Adrenaline he makes wrestling with ideas look easy.
Adrenaline Brian D Hoffman, Harvard University Press 298 pp. $24.95 ISBN 978-0-674-05088-4 Published Online April 2013
Talha Khan Burki
12
Articles
Trajectories of cardiometabolic risk factors before diagnosis of three subtypes of type 2 diabetes: a post-hoc analysis of the longitudinal Whitehall II cohort study
Kristine Frch, Daniel R Witte, Adam G Tabk, Leigh Perreault, Christian Herder, Eric J Brunner, Mika Kivimki, Dorte Vistisen
Summary
Background Most clinicians acknowledge that type 2 diabetes is multifactorial and has heterogeneous characteristics, but neither prevention nor treatment is systematically stratified. To address the heterogeneity of the disease, we examined whether patients diagnosed on the basis of fasting glucose concentrations, those diagnosed on the basis of 2 h concentrations, and those diagnosed on the basis of both criteria differed in terms of pathogenesis or cardiovascular risks. Methods Retrospectively, we analysed trajectories of cardiometabolic risk factors and 10 year cardiovascular risks in the prospective Whitehall II study cohort by use of multilevel longitudinal modelling. Participants were diagnosed by 75 g oral glucose-tolerance tests. We classified those diagnosed with type 2 diabetes into three subgroups: diagnosed on the basis of fasting glucose concentrations, diagnosed on the basis of 2 h glucose concentrations, and diagnosed on the basis of both concentrations. We also developed a classification tree for identification of individuals who are likely to have high fasting and 2 h glucose concentrations, but for whom only fasting concentrations are available. Results Median follow-up was 142 years with 15826 person-examinations (19912009). Of 10308 individuals, 6843 were included and 6569 remained diabetes free. 274 cases of type 2 diabetes were identified: 55 had high fasting glucose concentrations only, 148 had high 2 h concentrations only, and 71 had high fasting and 2 h concentrations. At diagnosis, participants with high fasting and 2 h glucose concentrations had higher mean body-mass indices (309 kg/m2 [SD 57]) than did those with high fasting concentrations (284 kg/m2 [44]; p=00009) or 2 h concentrations (279 kg/m2 [49]; <00001). Mean glycated haemoglobin A1c concentrations were also higher in the fasting and 2 h subgroup (74% [16]) than in the fasting (59% [05]; <00001) or 2 h (59% [06]; <00001) sugroups. Additionally, the fasting and 2 h subgroup had a higher proportion of individuals with moderate or high risk of cardiovascular disease than did the fasting subgroup (p=002). A classic pattern of -cell decompensation before diagnosis was noted only in the fasting and 2 h subgroup. Additionally, glucose concentrations and insulin resistance accelerated more substantially before diagnosis in the fasting and 2 h subgroup than in the fasting subgroup or the 2 h subgroup. Interpretation Patients with type 2 diabetes diagnosed on the basis of increased fasting glucose concentrations or 2 h glucose concentrations, or both, have distinct cardiometabolic risk development before diagnosis. Funding UK Medical Research Council, UK Economic and Social Research Council, British Heart Foundation, UK Health and Safety Executive, UK Department of Health, US National Heart Lung and Blood Institute, US National Institute on Aging, US Agency for Health Care Policy Research, and John D and Catherine T MacArthur Foundation.
Published Online February 21, 2013 http://dx.doi.org/10.1016/ S2213-8587(13)70008-1 See Online/Comment http://dx.doi.org/10.1016/ S2213-8587(13)70010-X Steno Diabetes Center, Gentofte, Denmark (K Frch PhD, D R Witte PhD, D Vistisen PhD); Centre de Recherche Public de la Sant, Strassen, Luxembourg (D R Witte); Department of Epidemiology and Public Health, University College London, London, UK (A G Tabk PhD, E J Brunner PhD, Prof M Kivimki PhD); First Department of Medicine, Semmelweis University, Budapest, Hungary (A G Tabk); University of Colorado, Aurora, CO, USA (L Perreault MD); and Institute for Clinical Diabetology, German Diabetes Center, Leibniz Center for Diabetes Research, Heinrich Heine University Dsseldorf, Dsseldorf, Germany (C Herder PhD) Correspondence to: Dr Kristine Frch, Steno Diabetes Center A/S, Niels Steensens Vej 2, DK-2820 Gentofte, Denmark. krif@steno.dk
Introduction
Type 2 diabetes is defined as a single disease entity irrespective of the way in which it is diagnosed. Even though most clinicians recognise that the disease is multifactorial and has heterogeneous characteristics, stratification of prevention or treatment is not done systematically. As type 2 diabetes develops, the blood glucose rises from normoglycaemia to a slightly increased concentration (ie, prediabetic state) and finally to a hyperglycaemic state that exceeds the diagnostic criteria for the illness. Previous longitudinal analyses1 from the Whitehall II study have shown that fasting and 2 h postload glucose concentrations increase at different timepoints in the years preceding diagnosis of type 2 diabetes, suggesting possible
differences in disease development. Individuals with prediabetes in whom only fasting plasma glucose concentrations are raised are phenotypically different from those with increased 2 h glucose concentrations only.24 Individuals in whom both fasting and 2 h glucose concentrations are raised have a worse risk profile3,4 and a higher risk of progression to type 2 diabetes5 than do those with isolated fasting or 2 h hyperglycaemia. The major pathophysiological drivers in the different prediabetic states probably continue to operate when fasting or 2 h glucose concentrations further rise into the diabetic range, implying that stratified prevention and possibly stratified treatment might be beneficial. In this longitudinal cohort study, we established 18 year trajectories of traditional cardio metabolic risk factors preceding diagnosis of type 2
13
Articles
diabetes based on fasting glucose or 2 h glucose concentrations, or both. We also developed a model to identify the individuals who are likely to have high fasting and 2 h glucose concentrations, but for whom only fasting concentrations are available.
Methods
Study design and participants

Whitehall II was a longitudinal study of a cohort of nonindustrial UK civil servants and has been previously described in detail.6 Its original aim was to investigate the importance of stress and social class for cardiovascular health. 10 308 participants (6896 [669%] of whom were male) who worked in the London offices of 20 departments
Diabetes based on fasting glucose (n=55) Men Age (years) White ethnic origin Current smoker Family history of diabetes mellitus 10 year cardiovascular disease risk 15% Antihypertensive treatment Lipid-lowering treatment Fasting plasma glucose (mmol/L) 2 h plasma glucose (mmol/L) HbA1c % mmol/mol Body-mass index (kg/m2) Waist circumference (cm) Hip circumference (cm) Waisthip ratio Height (m) Total cholesterol (mmol/L) HDL cholesterol (mmol/L) LDL cholesterol (mmol/L) Triglycerides (mmol/L) Adiponectin (g/mL)|| Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Fasting insulin (pmol/L) 2 h insulin (pmol/L) HOMA-insulin resistance HOMA--cell function Insulin sensitivity index (mg l2/mmol mU min) 59 (05) 41 (57) 284 (44) 1000 (125) 1028 (89) 097 (007) 1722 (99) 58 (12) 14 (03) 37 (12) 16 (08) 65 (64113) 1279 (175) 753 (111) 147 (86221) 575 (352937) 47 (3068) 771 (4831050) 478 (369656) 44 (800%) 600 (69) 43 (782%) 2 (36%) 13 (236%) 22 (400%) 19 (345%) 11 (200%) 73 (05) 82 (20)
and were aged 3555 years were recruited between August, 1985, and April, 1988 (phase 1), and followed up in eight subsequent phases roughly 25 years apart. A questionnaire was administered in all phases, and every second phase (ie, phases 1, 3, 5, 7, and 9) additionally included a clinical health examination. The numbers of participants in these subsequent clinical phases, which were defined as the numbers who either returned a questionnaire or attended a clinical examination, were 6057 men and 2758 women in phase 3 (199193), 5473 men and 2397 women in phase 5 (199799), 4893 men and 2074 women in phase 7 (200204), and 4759 men and 2002 women in phase 9 (200709). The Whitehall II study was reviewed and approved by the University College London
Diabetes based on 2 h glucose (n=148) 107 (723%) 630 (66)* 125 (845%) 11 (74%) 33 (223%) 70 (473%) 58 (392%) 34 (230%) 58 (07)* 122 (10)* 59 (06) 41 (65) 279 (49) 975 (121) 1016 (90) 096 (007) 1689 (93)* 57 (11) 15 (04) 35 (10) 15 (08) 64 (5188) 1326 (184) 776 (122) 106 (67163)* 887 (6011316) 29 (1846)* 957 (6651511)* 339 (283414)* Diabetes based on fasting and 2 glucose (n=71) 57 (803%) 608 (60) 53 (746%) 10 (141%) 17 (239%) 43 (606%)* 22 (310%) 11 (155%) 91 (27) 155 (40) 74 (16) 57 (177) 309 (57) 1044 (114) 1052 (113) 100 (007) 1714 (80) 59 (10) 12 (03) 37 (09) 20 (10) 49 (3387) 1383 (179) 831 (129) 170 (105221) 504 (356908) 62 (3985) 663 (3701000) 318 (255440)* Participants without diabetes (n=6569) 4618 (703%) 604 (79) 6063 (923%) 650 (99%) 617 (94%) 434 (66%) 1478 (225%) 1031 (157%) 52 (05) 60 (16) 55 (04) 37 (46) 262 (41) 918 (122) 998 (79) 092 (008) 1714 (92) 57 (12) 16 (04) 36 (11) 13 (08) 87 (63132)** 1240 (156)** 742 (108)** 66 (4499) 379 (232629) 16 (1124) 863 (6031246) 914 (617150)
Data are n (%), mean (SD), or median (IQR). HOMA=homoeostasis model of assessment. To test differences in characteristics between subgroups, the test was used for categorical variables, t test for normal continuous data, and Mann-Whitney U test for non-normal continuous data. Two-sided 5% level of significance adjusted for multiple testing with Benjamini et als method.13 *Significantly different from fasting subgroup. Significantly different from 2 h subgroup. Significantly different from fasting, 2 h, and fasting and 2 h subgroups. Signicantly different from fasting subgroup and 2 h subgroup. Significantly different from fasting subgroup and fasting and 2 h subgroup. ||n=24 in the fasting subgroup, n=70 in the 2 h subgroup, n=36 in the fasting and 2 h subgroup, and n=2336 in the diabetes-free group. **Significantly different from 2 h subgroup and fasting and 2 h subgroup.
Table: Characteristics of patients with type 2 diabetes diagnosed on the basis of fasting glucose concentrations or 2 h concentrations, or both, at time of diagnosis, and of those without diabetes at last clinical examination
14
Articles
Ethics Committee (85/0938), and written informed consent was obtained from each participant at each phase. The study was done according to the principles of the Helsinki Declaration.
Friedewald formula, and serum adiponectin concentrations measured in 2466 (360%) participants with the Quantikine ELISA kit (R&D Systems, Wiesbaden, Germany).9
Procedures
We used phase 3ie, when the oral glucose tolerance test (OGTT) was first doneas the baseline for our analysis, and final follow-up was the phase 9 examination. We excluded 1032 (100%) participants who were lost to follow-up before phase 3, 153 (15%) participants with prevalent diabetes before phase 3, and 1759 (171%) participants who fasted for less than 8 h before the clinical examination in all study phases. When fasting duration was less than 8 h in one phase, the participant was excluded from that phase only. 795 of the remaining 7364 (108%) participants developed type 2 diabetes during the study. However, because diagnosis with a full and valid OGTT was necessary,7 we further excluded 358 (49%) patients whose type 2 diabetes was diagnosed by a doctor outside the study and 163 (23%) participants in whom the disorder was diagnosed by screening, but for whom data for both a valid fasting and 2 h glucose measurement were not available. Thus, our final sample was 6843 participants (664% of the original sample); median follow-up was 142 years (IQR 87162), 15 826 person-examinations were done, and 274 cases of type 2 diabetes were diagnosed by screen detection by phase 9. We classified these 274 cases into three subgroupsnamely, cases diagnosed on the basis of fasting glucose only (ie, fasting glucose 70 mmol/L, 2 h glucose <111 mmol/L), those diagnosed on the basis of 2 h glucose only (ie, fasting glucose <70 mmol/L, 2 h glucose 111 mmol/L), and those diagnosed on the basis of both criteria (ie, fasting glucose 70 mmol/L, 2 h glucose 111 mmol/L). A flow diagram of the number of participants included at each phase is shown in the appendix.
Statistical analysis
We used the homoeostasis model assessment (HOMA) to estimate insulin resistance and -cell function,10 the insulin sensitivity index to estimate whole-body insulin sensitivity,11 and the Framingham cardiovascular disease risk score to calculate absolute 10 year risk of developing general cardiovascular disease.12 We classified a risk score of 15% or greater as a moderate or high risk of cardiovascular disease. The observation period for retrospective trajectories started at the date of diagnosis by OGTT for patients who developed type 2 diabetes, and at the last screening or questionnaire phase for those who did not develop the disorder (year 0). Trajectories of the following outcomes were followed back in time to the first clinical exam ination: fasting and 2 h plasma glucose concentrations; fasting and 2 h serum insulin concentrations; HOMA-cell function, HOMA-insulin resistance, and insulin sensitivity index; body-mass index and waisthip ratio; systolic and diastolic blood pressure; total, HDL, and
A
10 9
Concentration (mmol/L)
See Online for appendix
Diabetes based on fasting glucose Diabetes based on 2 h glucose Diabetes based on fasting +2 h glucose Participants without diabetes
Concentration (mmol/L) Concentration (pmol/L)
15
8 7 6 5 4
10
Measurements
In phases 3, 5, 7, and 9 of Whitehall II, standard 2 h 75 g OGTTs were done in the morning after overnight fasting (8 h, estimated by the difference between self-reported time of most recent meal and the OGTT), or in the afternoon after no more than a light fat-free breakfast eaten before 0800 h (5 h of fasting). Anthropometry was done and blood pressure measured according to standard protocols.6 Data for ethnic origin, smoking status, and family history of diabetes (types 1 and 2) were gathered via a questionnaire. During all phases, blood samples were handled according to standardised procedures. Blood glucose was measured by the glucose oxidase method;1 serum insulin by in-house radioimmunoassays;8 and serum triglycerides, total cholesterol, and HDL cholesterol concentrations by automated enzymatic colorimetric methods. LDL cholesterol was calculated with the
C
15
Concentration (pmol/L)
100 80 60 40 20 15 10 5 0
10
15
10
Time (years)
Time (years)
Figure 1: Trajectories of fasting plasma glucose (A), 2 h plasma glucose (B), fasting serum insulin (C), and 2 h serum insulin (D) concentrations for a hypothetical, white, 60-year-old man (in year 0) from 18 years before diagnosis of type 2 diabetes or last clinical examination Time 0=diagnosis or last clinical examination. Solid lines are estimated trajectories and dashed lines are 95% CIs. Black bars show data distribution during follow-up.
15
Articles
LDL cholesterol concentrations; triglyceride concen trations; adiponectin concentrations; and the Framingham cardiovascular disease risk score. Before analysis, we log-transformed outcomes that did not follow a normal distribution (fasting and 2 h serum insulin concentrations, adiponectin concentrations, HOMA--cell function, HOMA-insulin resistance, and insulin sensitivity index). We used mixed-effects models to account for the correlation of repeated measurements within participants. Time dependence was allowed to vary across the subgroups, and quadratic and cubic terms for time were included in the three subgroups when significant (twosided 5% significance level). For individuals who did not develop diabetes, year 0 was merely a timepoint in a normal life course, and thus we fitted the trajectories by linear models. We adjusted all analyses for age, sex, ethnic origin, and study phase. Pairwise differences in growth curves between the three subgroups were tested with the F test by comparing the curve of contrasts by diabetes subgroups with a straight line with zero slope that passes through the origin (two-sided 5% significance level). Accordingly, provided p values relate to curve differences in slope or intercept, or both. We did statistical analyses in R (version 9.15.0) and SAS (version 9.2).
A
8 Diabetes based on fasting glucose Diabetes based on 2 h glucose Diabetes based on fasting +2 h glucose Participants without diabetes HOMA--cell function (%)
To acknowledge that OGTTs are rarely done in clinical practice, we developed a classification tree for identi fication of individuals with type 2 diabetes who are likely to have the fasting and 2 h phenotype but for whom 2 h glucose concentrations are not available. We derived a classification tree for type 2 diabetes by both criteria based on a two-step approachan initial screening of the explanatory variables listed in the table (not including 2 h glucose and drugs) by random forests analysis with fasting and 2 h concentrations as the outcome, and derivation of a classification tree based on the five highest ranking variables (appendix).
Role of the funding source

The sponsors of the study did not have roles in study design; data collection, analysis, and interpretation; or writing of the report. All authors accept full responsibility for the study, had full access to all the data, and take responsibility for the integrity of the data and the accuracy of the analysis. The corresponding author had the final responsibility to submit for publication.
Results
The Table shows characteristics of participants at the date of diagnosis for the three subgroups of type 2 diabetes or the last clinical examination for those without type 2 diabetes. Of the 274 people who developed type 2 diabetes, 55 (201%) had high fasting glucose concentrations, 148 (540%) had high 2 h concentrations, and 71 (259%) had high fasting and 2 h concentrations. At diagnosis (year 0), individuals with 2 h hyperglycaemia were older and had lower fasting insulin concentrations, HOMAinsulin resistance, and insulin sensitivity (as measured by the insulin senstivity index), but higher 2 h insulin concentrations and -cell function than did those with fasting hyperglycaemia. Individuals with high fasting and 2 h glucose concentrations were generally more obese, had lower HDL cholesterol concentrations and -cell function, and higher triglyceride and glycated haemoglobin A1c (HbA1c) concentrations and blood pressure than did the subgroups with high fasting concentrations or high 2 h concentrations only. Furthermore, the proportion of individuals with moderate or high cardiovascular risk was higher in the fasting and 2 h subgroup than in the fasting subgroup (table; p=002). For most variables, less than 5% of data were missing. However, for insulin resistance, -cell function, insulin sensitivity index, waisthip ratio, cardiovascular risk, and HDL and LDL cholesterol concentrations, 510% of data were missing. Trajectories of fasting and 2 h glucose differed between the subgroups in accordance with the criteria by which type 2 diabetes was diagnosed (figure 1A, 1B; p<00001 for all comparisons). However, the curves did not diverge until roughly 6 years before diagnosis, when both fasting and 2 h glucose concentrations rose steeply in the fasting and 2 h subgroup. Fasting insulin concentrations rose slightly but steadily in the three diabetes subgroups from
B
120
HOMA-insulin resistance
100
80
60
40
Insulin sensitivity (mg I2/mmol mU min)
200
D
10
150
Concentration (g/mL) 15 10 5 Time (years) 0
100
50
4 15 10 5 Time (years) 0
Figure 2: Trajectories of HOMA-insulin resistance (A), HOMA--cell function (B), insulin sensitivity index (C), and serum adiponectin concentrations (D) for a hypothetical, white, 60-year-old man (in year 0) from 18 years before diagnosis of type 2 diabetes or last clinical examination Time 0=diagnosis or last clinical examination. Solid lines are estimated trajectories and dashed lines are 95% CIs. Black bars show data distribution during follow-up. Adiponectin concentrations are based on a subset of data. HOMA=homoeostasis model of assessment.
16
Articles
15 years before diagnosis compared with those in participants without type 2 diabetes (figure 1C); this increase was less pronounced in the 2 h subgroup (p=0037 compared with fasting subgroup and 00001 compared with fasting and 2 h subgroup). Heterogeneity between the subgroups was greater for the trajectories for 2 h insulin concentrations (figure 1D, p=00033 for fasting subgroup vs 2 h subgroup, 00326 for fasting subgroup vs fasting and 2 h subgroup, and <00001 for 2 h subgroup vs fasting and 2 h subgroup). In the fasting and 2 h subgroup, 2 h insulin concentrations peaked around 5 years before diagnosis and then started to fall, whereas 2 h insulin concentrations increased linearly in patients who developed type 2 diabetes according to fasting criteria or 2 h criteria, but at different rates (figure 1D; p=0003 for fasting subgroup vs 2 h subgroup). Trajectories of insulin resistance broadly resembled those of fasting glucose concentrations and differed significantly between all three subgroups (figure 2A; p<00001 for fasting subgroup vs 2 h subgroup and for 2 h subgroup vs fasting and 2 h subgroup, 00002 for fasting subgroup vs fasting and 2 h subgroup). Trajectories of -cell function also differed significantly between the three subgroups (figure 2B; p=00002 for fasting subgroup vs 2 h subgroup, 00140 for fasting subgroup vs fasting and 2 h subgroup, and <00001 for 2 h subgroup vs fasting and 2 h subgroup). In the fasting and 2 h subgroup, percentage -cell function increased until roughly 810 years before diagnosis and fell progressively thereafter (figure 2B). By contrast, patients in the 2 h subgroup had moderately stable values throughout follow-up (figure 2B). Individuals in the fasting subgroup had significantly lower HOMA--cell function from 18 years before diagnosis until diagnosis than did those who did not have type 2 diabetes (p<00001). Scores on the insulin sensitivity index were lower in the three subgroups of type 2 diabetes before and at diagnosis than they were in the participants who did not develop type 2 diabetes (p<00001 for all comparisons); scores decreased slightly faster in the 2 h subgroup and fasting and 2 h subgroup than in the fasting subgroup (figure 2C, p=00006 for fasting subgroup vs 2 h subgroup, and 00093 for fasting subgroup vs fasting and 2 h subgroup). Individuals diagnosed by fasting and 2 h concentrations had lower serum adiponectin concentrations before diagnosis than did those diagnosed by either fasting or 2 h concentrations (figure 2D, p=00203 for fasting subgroup vs fasting and 2 h subgroup and 00504 for 2 h subgroup vs fasting and 2 h subgroup). Body-mass indices and waisthip ratios increased linearly with time in the three subgroups (figure 3A, 3B). However, obesity was more common in patients diagnosed by fasting and 2 h glucose concentrations than in those diagnosed by either fasting or 2 h concentrations (figure 3A, B; p=00029 for fasting subgroup vs fasting and 2 h subgroup and <00001 for 2 h subgroup vs fasting and 2 h subgroup).
A
32 Body-mass index (kg/m2)
Diabetes based on fasting glucose Diabetes based on 2 h glucose Diabetes based on fasting +2 h glucose Participants without diabetes Waisthip ratio Cardiovascular risk (%) 15 10 5 Time (years) 0
105
30
100
28
095
26 090
24
C
150 Blood pressure (mm Hg)
25 20
140
15 10 5 0 15 10 5 Time (years) 0
130
120
110
Figure 3: Trajectories of body-mass index (A), waisthip ratio (B), systolic blood pressure (C), and 10 year absolute risk of cardiovascular disease (Framingham) (D) for a hypothetical, white, 60-year-old man (in year 0) from 18 years before diagnosis of type 2 diabetes or last clinical examination Time 0=diagnosis or last clinical examination. Solid lines are estimated trajectories and dashed lines are 95% CIs. Black bars show data distribution during follow-up.
The increase in both diastolic (data not shown) and systolic blood pressure with time was steepest in the group diagnosed on the basis of fasting and 2 h sub glucose concentrations (figure 3C; p=00009 for fasting subgroup vs fasting and 2 h subgroup and 00150 for 2 h subgroup vs fasting and 2 h subgroup). Trajectories of estimated risk of cardiovascular disease differed significantly between the three subgroups (figure 3D; p=00024 for fasting subgroup vs 2 h subgroup and <00001 for fasting subgroup vs fasting and 2 h subgroup and 2 h subgroup vs fasting and 2 h subgroup); risk was highest in the fasting and 2 h subgroup (p<00001). The trajectories of total and LDL cholesterol concentrations did not differ significantly between subgroups (figure 4A, 4B; p0334 for all pairwise comparisons). However, HDL cholesterol concentrations decreased significantly more in the subgroup diagnosed on the basis of fasting and 2 h glucose concentrations than in that diagnosed on the basis of 2 h concentrations (figure 4C; p=00005), and patients diagnosed on the basis of both criteria had higher increases in concentrations of plasma triglycerides than did those diagnosed by either fasting or 2 h concentrations (figure 4D, p=00056 for fasting subgroup vs fasting and
17
Articles
A
75 Concentration (mmol/L)
Diabetes based on fasting glucose Diabetes based on 2 h glucose Diabetes based on fasting +2 h glucose Participants without diabetes Concentration (mmol/L)
70
50
65
45
60
40
15 14 Concentration (mmol/L) 13 12 11 10 09
C
Concentration (mmol/L)
D
25
126 patients were used to derive the classification tree. The random forest analysis showed that the five highest ranking explanatory variables that discriminate between people diagnosed on the basis of fasting glucose concentrations only and those diagnosed on the basis of fasting and 2 h concentrations were fasting plasma glucose concentrations, HbA1c concentrations, diastolic blood pressure, systolic blood pressure, and HOMAinsulin resistance. The resulting classification tree was based on these variables, but only fasting glucose and HbA1c concentrations were significant in the model. 55 of 65 (85%) individuals with fasting plasma glucose concentrations greater than 72 mmol/L and HbA1c greater than 63% (45 mmol/mol) had the fasting and 2 h phenotype (figure 5, appendix).
Discussion
We hypothesised that type 2 diabetes is not a single disease entity, but rather a heterogeneous disease with different underlying mechanisms preceding diagnosis in different groups of individuals. We did a simple stratification based on the common diagnostic glucose criteria of the Whitehall II cohort and showed that underlying pathogenesis differed as much as 18 years before diagnosis between patients with type 2 diabetes diagnosed on the basis of increased fasting glucose concentrations, those diagnosed on the basis of 2 h concentrations, and those diagnosed on the basis of both criteria (panel). In the fasting and 2 h subgroup, HbA1c concentrations surpassed the diagnostic threshold for type 2 diabetes (65%) at time of diagnosis, whereas, in the fasting subgroup and the 2 h subgroup, mean HbA1c was less than the cutoff (60%) for high-risk type 2 diabetes.14 Despite moderately normal HbA1c concentrations, both the fasting subgroup and the 2 h subgroup had significantly increased estimated 10 year risk of cardiovascular disease compared with people without type 2 diabetes, suggesting that HbA1c might not be a good marker of cardiovascular risk in all patients with type 2 diabetes. Over the past two decades, the results of several studies2,15,16 have suggested that fasting and post-OGTT hyperglycaemia represent phenotypes with distinct natural histories, a notion supported by our findings. We noted a long-standing reduction in basal -cell function and a progressive increase in insulin resistance in patients developing type 2 diabetes diagnosed on the basis of fasting glucose concentrations. Thus, the increased fasting insulin concentration before onset (figure 1C) shows inadequate compensation for the increased insulin resistance.17 Because of the low -cell function in this subgroup before and at diagnosis, individuals with fasting hyperglycaemia might benefit from early prevention strategies focusing on prevention of further loss of -cell function rather than strategies targeting peripheral insulin sensitivity. This line of thinking is supported by observational and interventional
20
15
15
10 5 Time (years)
15
10 5 Time (years)
Figure 4: Trajectories for total cholesterol (A), LDL cholesterol (B), HDL cholesterol (C), and triglyceride (D) concentrations for a hypothetical, white, 60-year-old man (in year 0) from 18 years before diagnosis of type 2 diabetes or last clinical examination Time 0=diagnosis or last clinical examination. Solid lines are estimated trajectories and dashed lines are 95% CIs. Black bars show data distribution during follow-up.
Fasting glucose
7072 mmol/L
n=41 195% had fasting and 2 h hyperglycaemia
>72 mmol/L n=20 400% had fasting and 2 h hyperglycaemia
HbA1c
63%
>63% n=65 846% had fasting and 2 h hyperglycaemia
Figure 5: Classification tree for identification of patients with fasting and 2 h hyperglycaemia without results from an oral glucose tolerance test Sensitivity=78% (95% CI 6887); specificity=82% (7191); positive predictive value=85% (appendix). HbA1c=glycated haemoglobin A1c.
2 h subgroup and 00004 for 2 h subgroup vs fasting and 2 h subgroup). In our study, 71 of the 126 (563%) individuals with fasting glucose concentrations in the diabetic range also had 2 h concentrations in the diabetic range; the
18
Articles
studies18,19 showing that physical activity, which mainly improves peripheral insulin sensitivity, has little effect on progression to type 2 diabetes in individuals with isolated impaired fasting glucose concentrations (ie, 2 h concentrations are normal), but is protective in those with impaired glucose tolerance (who have prediabetic 2 h glucose concentrations). By contrast, metformin, which mainly works by improving hepatic insulin sensitivity, seemed to be more effective in people with high fasting glucose concentrations than in those with low fasting concentrations in the Diabetes Prevention Program.20 To further disentangle the underlying causes of differences in patterns of insulin resistance and -cell function between subgroups of type 2 diabetes, a detailed examination of the contributing genetic and non-genetic factors is needed. The effects of genes associated with 2 h glucose concentrations increase with age, whereas those of genes associated with fasting glucose concentrations are more stable with time,21 suggesting different genetic effects in the different disease subgroups of our study. In patients who developed type 2 diabetes diagnosed on the basis of 2 h concentrations, 2 h insulin concentrations increased exponentially, whole-body insulin sensitivity fell, and fasting -cell function was stable and almost normal. However, the increase in insulin secretion 2 h after the OGTT was not sufficient to maintain 2 h plasma glucose concentrations within the non-diabetic range, suggesting peripheral insulin resistance with insufficient -cell compensation.17,22,23 In general, type 2 diabetes diagnosed on the basis of 2 h glucose concentrations accounts for 3050% of all cases.24 However, in clinical practice OGTTs are seldom done and thus illness often remains undiagnosed and untreated in individuals with the 2 h phenotype. Because 2 h glucose concentrations are more closely related to cardiovascular risk than are fasting glucose concentrations,25,26 we expected that progression of blood pressure, concentrations of plasma lipids and adiponectin, and cardiovascular risk would be worse in patients in the 2 h subgroup than in those in the fasting subgroup. However, although the 2 h subgroup had a slightly higher estimated cardiovascular risk at time of diagnosis than did the fasting subgroup, they had a slower worsening of plasma lipids. A previous study27 showed that a rapid rise in triglyceride concentrations precedes type 2 diabetes diagnosed on the basis of fasting glucose concentrations, but we did not confirm this finding. Rather, patients with increased fasting and 2 h glucose concentrations had higher triglyceride and lower adiponectin concentrations than did those in the other subgroups, independent of lipid-lowering treatment (data not shown). The pattern of insulin secretion in the fasting and 2 h subgroup (increased -cell function 1015 years before diagnosis and subsequent decline thereafter) was not noted in the fasting subgroup or the 2 h subgroup,
Panel: Research in context Systematic review We searched PubMed for studies of adults (19 years) published in English before Dec 18, 2012, with the terms type 2 diabetes and heterogeneity in the title or abstract. Of the 157 references returned, 72 were focused on genetics of type 2 diabetes. The remaining 85 publications included 34 reviews or meta-analyses, eight clinical trials, and 43 observational studies. 35 publications concerned the pathogenesis of type 2 diabetes. We searched the reference lists in these publications and selected citations that we judged relevant. We noted no previous studies investigating metabolic features preceding development of subtypes of type 2 diabetes diagnosed by fasting versus 2 h glucose concentrations. Interpretation Individuals with prediabetes differ in their natural history and underlying pathogenesis dependent on whether they have increased fasting glucose concentrations or increased 2 h glucose concentrations, or both. We showed that such differences are also present in individuals with incident type 2 diabetes. To distinguish between patients with isolated fasting hyperglycaemia and those with fasting and 2 h hyperglycaemia is important, because these subgroups have distinct natural histories of -cell dysfunction and cardiovascular risk. More research is needed to understand how the specific mechanisms that trigger type 2 diabetes can be used to optimise treatment and prevention of complications.
showing that the common notion that early -cell compensation followed by a progressive loss of -cell function is typical of the development of type 2 diabetes1,28 seems only to apply to the fasting and 2 h subgroup, constituting roughly 25% of all screen-detected patients. This finding also supports the idea that increases of both fasting and 2 h glucose concentrations are not merely a mixture of the underlying conditions of isolated fasting or 2 h hyperglycaemia, but rather a distinct disease entity.3 To identify patients with progressively declining -cell function, accelerated increase of glucose concentrations, and high cardiovascular risk, knowledge of the patients subtype can be useful. Because OGTTs are not done on a regular basis in primary care, other methods to identify individuals in the high-risk subgroup might be needed. We noted that 85% of individuals with a fasting glucose concentration greater than 72 mmol/L and HbA1c concentration greater than 63% (45 mmol/mol) also had 2 h glucose concentrations in the diabetic range. Thus, the combination of fasting glucose and HbA1c concentrations might be useful to identify high-risk individuals who have accelerated -cell dysfunction and are especially likely to benefit from early glucoselowering treatment and intensified cardiovascular disease prevention. However, this finding needs to be confirmed in independent datasets.
19
Articles
For the Whitehall II data sharing policy see http://www. ucl.ac.uk/whitehallII/data_ sharing/index.htm
Our study has some limitations. Estimates of insulin sensitivity and -cell function were based on HOMA and the insulin sensitivity index, which are calculated from fasting and 2 h glucose and insulin concentrations and thus might overlap the classification of the three subgroups of type 2 diabetes at diagnosis. However, individuals in the fasting subgroup had very different trajectories of HOMA--cell function than did those in the fasting and 2 h subgroup, despite diagnoses based on fasting glucose concentrations in both. Whitehall II was an occupational cohort, which might not be representative for a general European population. Nonetheless, no evidence suggests that the trajectories in our study would not apply to other populations, although the distribution of the different subgroups of disease might vary between populations. For example, Asian populations have a higher prevalence of diabetes based on fasting glucose concentrations and they develop type 2 diabetes at lower body-mass indices than do white populations.24,29,30 We were not able to describe sex-specific trajectories because of the low proportion of women in the study. Such findings would have been highly relevant because men have higher fasting glucose concentrations than do women.31 Furthermore, with the acceptance of HbA1c for use in diagnosis, trajectories of HbA1c concentrations in the different subgroups would have been relevant, but concentrations were measured only in the last two clinical phases. Another limitation is that we had to exclude a substantial proportion of incident cases diagnosed outside the Whitehall II study because of unknown or missing OGTT results, which could limit the generalisability of our findings. However, that participants with doctor-diagnosed diabetes would represent a different disease group from those diagnosed on the basis of an OGTT seems unlikely. All trajectories for doctor-diagnosed patients were within the range of the three diabetes subgroups, supporting this suggestion (data not shown). In conclusion, type 2 diabetes seems to have different natural history and pathogenesis dependent on whether it is diagnosed on the basis of increased fasting glucose concentrations or increased 2 h glucose concentrations, or both. Future studies should establish whether glycaemic control, drug needs, and the incidence of cardiovascular disease and microvascular complications differ between patients with different subgroups of disease. Additionally, a better understanding of the genetic and non-genetic causative factors associated with different subgroups is needed. Ultimately, assessment of the need for stratified prevention and treatment strategies is important.
Contributors KF conceived the idea and wrote the first draft, with major input from DV. DV planned and did the statistical analysis. All authors critically assessed and reviewed the paper. MK obtained funding for the Whitehall II study.
Conflicts of interest Steno Diabetes Center, where KF and DV are employed, receives part of its core funding from unrestricted grants from the Novo Nordisk Foundation and Novo Nordisk, and is owned by Novo Nordisk. KF, DV, and DRW own shares in Novo Nordisk. All other authors declare that they have no conflicts of interest. Acknowledgments The Whitehall II study is supported by the UK Medical Research Council, UK Economic and Social Research Council, British Heart Foundation, UK Health and Safety Executive, UK Department of Health, US National Heart Lung and Blood Institute, US National Institute on Aging, US Agency for Health Care Policy Research, and John D and Catherine T MacArthur Foundation. MK is an Economic and Social Research Council professor. Serum adiponectin concentrations were measured at the German Diabetes Center, which is funded by the German Federal Ministry of Health and the Ministry of School, Science and Research of the State of North RhineWestphalia. This study was supported partly by a grant from the German Federal Ministry of Education and Research to the German Center for Diabetes Research. We thank all participating civil service departments and their welfare, personnel, and establishment officers; the UK Occupational Health and Safety Agency; the UK Council of Civil Service Unions; all participating civil servants in the Whitehall II study; and all members of the Whitehall II study team. Whitehall II data, protocols, and other metadata are available to the scientific community. Please refer to the Whitehall II data sharing policy. References 1 Tabk AG, Jokela M, Akbaraly TN, Brunner EJ, Kivimki M, Witte DR. Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: an analysis from the Whitehall II study. Lancet 2009; 373: 221521. 2 Frch K, Vaag A, Holst JJ, Hansen T, Jrgensen T, Borch-Johnsen K. Natural history of insulin sensitivity and insulin secretion in the progression from normal glucose tolerance to impaired fasting gycemia and impaired glucose tolerancethe Inter99 study. Diabetes Care 2009; 32: 43944. 3 Perreault L, Bergman BC, Playdon MC, Dalla Man C, Cobelli C, Eckel RH. Impaired fasting glucose with or without impaired glucose tolerance: progressive or parallel states of prediabetes? Am J Physiol Endocrinol Metab 2008; 295: E42835. 4 Weyer C, Bogardus C, Pratley RE. Metabolic characteristics of individuals with impaired fasting glucose and/or impaired glucose tolerance. Diabetes 1999; 48: 2197203. 5 Engberg S, Vistisen D, Lau C, et al. Progression to impaired glucose regulation and diabetes in the population-based Inter99 study. Diabetes Care 2009; 32: 60611. 6 Marmot M, Brunner E. Cohort profile: the Whitehall II study. Int J Epidemiol 2005; 34: 25156. 7 Alberti KGMM, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus. Report of a WHO consultation. Diabetic Med 1998; 15: 53953. 8 Andersen L, Dinesen B, Jrgensen PN, Poulsen F, Rder ME. Enzyme immunoassay for intact human insulin in serum or plasma. Clin Chem 1993; 39: 57882. 9 Tabk AG, Carstensen M, Witte DR, et al. Adiponectin trajectories before type 2 diabetes diagnosis. Diabetes Care 2012; 35: 254047. 10 Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 41219. 11 Gutt M, Davis CL, Spitzer SB, et al. Validation of the insulin sensitivity index (ISI0,120): comparison with other measures. Diabetes Res Clin Pract 2000; 47: 17784. 12 DAgostino RB, Vasan RS, Pencina MJ, et al. General cardiovascular risk profile for use in primary care. Circulation 2008; 117: 74353. 13 Benjamini Y, Hochberg Y. Controlling the false discovery ratea practical and powerful approach to multiple testing. J R Stat B 1995; 57: 289300.
20
Articles
14 International Expert Committee. International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes. Diabetes Care 2009; 32: 132734. 15 Meigs JB, Muller DC, Nathan DM, Blake DR, Andres R. The natural history of progression from normal glucose tolerance to type 2 diabetes in the Baltimore longitudinal study of aging. Diabetes 2003; 52: 147584. 16 Van Haeften TW, Pimenta W, Mitrakou A, et al. Disturbances in beta-cell function in impaired fasting glycemia. Diabetes 2002; 51 (suppl): S26570. 17 Frch K, Borch-Johnsen K, Holst JJ, Vaag A. Pathophysiology and aetiology of impaired fasting glycaemia and impaired glucose tolerance: does it matter for prevention and treatment of type 2 diabetes? Diabetologia 2009; 52: 171423. 18 Engberg S, Glmer C, Witte D, Jrgensen T, Borch-Johnsen K. Differential relationship between physical activity and progression to diabetes by glucose tolerance status: the Inter99 Study. Diabetologia 2010; 53: 7078. 19 Saito T, Watanabe M, Nishida J, et al. Lifestyle modification and prevention of type 2 diabetes in overweight Japanese with impaired fasting glucose levels: a randomized controlled trial. Arch Intern Med 2011; 171: 135260. 20 Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002; 346: 393403. 21 Jensen AC, Barker A, Kumari M, et al. Associations of common genetic variants with age-related changes in fasting and postload glucose. Diabetes 2011; 60: 161723. 22 Burns SF, Bacha F, Lee SJ, Tfayli H, Gungor N, Arslanian SA. Declining beta-cell function relative to insulin sensitivity with escalating OGTT 2-h glucose concentrations in the nondiabetic through the diabetic range in overweight youth. Diabetes Care 2011; 34: 203340.
23 Bock G, Man CD, Campioni M, et al. Pathogenesis of pre-diabetes: mechanisms of fasting and postprandial hyperglycemia in people with impaired fasting glucose and/or impaired glucose tolerance. Diabetes 2006; 55: 353649. 24 DECODE-study group on behalf of the European Diabetes Epidemiology Group. Is fasting glucose sufficient to define diabetes? Epidemiological data from 20 European studies. Diabetologia 1999; 42: 64754. 25 Blake DR, Meigs JB, Muller DC, Najjar SS, Andres R, Nathan DM. Impaired glucose tolerance, but not impaired fasting glucose, is associated with increased levels of coronary heart disease risk factors: results from the Baltimore longitudinal study on aging. Diabetes 2004; 53: 2095100. 26 Bongaerts BWC, Rathmann W, Kowall B, et al. Postchallenge hyperglycemia is positively associated with diabetic polyneuropathy. Diabetes Care 2012; 35: 189193. 27 Sattar N, McConnachie A, Ford I, et al. Serial metabolic measurements and conversion to type 2 diabetes in the West of Scotland Coronary Prevention Study. Diabetes 2007; 56: 98491. 28 DeFronzo RA. Banting Lecture. From the triumvirate to the ominous octet: a new paradigm for the treatment of type 2 diabetes mellitus. Diabetes 2009; 58: 77395. 29 Nyamdorj R, Pitkaniemi J, Tuomilehto J, et al. Ethnic comparison of the association of undiagnosed diabetes with obesity. Int J Obes (Lond) 2010; 34: 33239. 30 Qiao Q, Nakagami T, Tuomilehto J, et al. Comparison of the fasting and the 2-h glucose criteria for diabetes in different Asian cohorts. Diabetologia 2000; 43: 147075. 31 Frch K, Borch-Johnsen K, Vaag A, Jrgensen T, Witte D. Sex differences in glucose levels: a consequence of physiology or methodological convenience? The Inter99 study. Diabetologia 2010; 53: 85865.
21
Articles
One hundred years of congenital adrenal hyperplasia in Sweden: a retrospective, population-based cohort study
Sebastian Gidlf , Henrik Falhammar, Astrid Thiln, Ulrika von Dbeln, Martin Ritzn, Anna Wedell, Anna Nordenstrm
Summary
Published Online February 26, 2013 http://dx.doi.org/10.1016/ S2213-8587(13)70007-X This online publication has been corrected. The corrected version first appeared at thelancet.com/ diabetes-endocrinology on March 14, 2013 See Online/Comment http://dx.doi.org/10.1016/ S2213-8587(13)70009-3 Department of Molecular Medicine and Surgery, Centre for Molecular Medicine (S Gidlf MD, H Falhammar MD, A Wedell MD, A Nordenstrm MD) and Department of Womens and Childrens Health (Prof M Ritzn MD), Karolinska Institutet, Stockholm, Sweden; Department of Obstetrics and Gynaecology, Karolinska University Hospital Huddinge, Stockholm, Sweden (S Gidlf); Department of Endocrinology, Metabolism and Diabetes (H Falhammar), Centre for Inherited Metabolic Diseases (U von Dbeln MD, Prof A Wedell), and Paediatric Endocrinology (A Nordenstrm), Karolinska University Hospital Solna, Stockholm, Sweden; and Department of Paediatrics, Jnkping Hospital, Jnkping, Sweden (A Thiln MD) Correspondence to: Dr Sebastian Gidlf, Department of Obstetrics and Gynaecology, K57, Karolinska University Hospital Huddinge, S-141 86 Stockholm, Sweden sebastian.gidlof@karolinska.se
Background Congenital adrenal hyperplasia due to 21-hydroxylase deficiency results in cortisol and aldosterone deficiency and is, in its most severe form, lethal. We aimed to assess the effect of historical medical improvements in the care of patients with this disorder over time and to assess the effects of neonatal screening in Sweden. Methods For this retrospective, population-based cohort study, we collected data for all known patients with congenital adrenal hyperplasia in Sweden between 1910 and 2011. Data sources included the registry at the Swedish national screening laboratory, patients identified via the Swedish neonatal screening programme, late-diagnosed patients reported to the laboratory, and patients who underwent genetic diagnostics or became known to us through clinical contacts. All known patients were included in a population-based cohort study of the distribution of clinical severity, genotype, sex, and the effect of nationwide neonatal screening. Findings We identified 606 patients with the disorder, born between 1915 and 2011. The CYP21A2 genotype (conferring deficiency of 21-hydroxylase) was known in 490 patients (81%). The female-to-male ratio was 125 in the whole cohort, but close to 1 in patients detected by the screening. We noted a sharp increase in the number of patients diagnosed in the 1960s and 1970s, and after the introduction of neonatal screening in 1986 the proportion of patients with the salt-wasting form of congenital adrenal hyperplasia increased in both sexes, from 114 (47%) of 242 individuals between 1950 and 1985 to 165 (57%) of 292 individuals between 1986 and 2011 (p=0038). On average, five to ten children were missed every year before 1970. The non-classic form of the disorder was diagnosed more often in women than in men, which accounts for the female preponderance in our cohort. Interpretation Our findings suggest that, contrary to current belief, boys and girls with salt-wasting congenital adrenal hyperplasia were equally missed clinically. Neonatal screening improved detection of the salt-wasting form in girls as well as boys, saving lives in both sexes. The non-classic form was diagnosed more often in women than it was in men, leading to the female preponderance in this cohort. Funding The Swedish Research Council, the Centre of Gender Medicine at Karolinska Institutet, the Stockholm County Council, the Sllskapet Barnavrd Foundation, the Stiftelsen Samariten Foundation, the Stiftelsen Frimurare Barnhuset Foundation, and the Novo Nordisk Foundation.
Introduction
Congenital adrenal hyperplasia due to 21-hydroxylase deficiency is the most common monogenic autosomal recessive endocrine disorder. 21-hydroxylase deficiency due to CYP21A2 genotype results in deficient and potentially fatally low cortisol and aldosterone production, and a concurrent increase in androgen production.1,2 Severity of clinical symptoms varies widely, depending on the patients underlying CYP21A2 genotype.3 The deficiency is present during fetal development and leads to varying degrees of prenatal virilisation of the external genitalia in affected girls. The prognosis for affected patients has changed substantially over the past century. In 1950, treatment with hydrocortisone was reported to be effective in ameliorating androgen overproduction and enabling survival in patients with the salt-wasting form of the disorder.4,5 Later on in the same decade, mineralo corticoid substitution treatment was introduced, further improving outcomes.6 The develop ment of paediatric
endocrinology as an area of expertise started during the 1950s and was pioneered in Sweden by C-G Bergstrand. This development contributed to increased awareness of the symptoms and led to improved diagnostics. The elucidation of the molecular background of this monogenic disease in the 1980s and 1990s and the development of strategies for typing of disease-causing mutations have been instrumental in diagnostics as well as in increasing our knowledge of the patho physiology of the disease. There are few recurring mutations, and reliable genotypephenotype relations have been docu mented that can be used for prognostic assessments.7,8 Congenital adrenal hyperplasia as an endocrine disorder has also proved to be important from a psychological perspective, for the patients as well as for the understanding of endocrine effects on the brain. In the late 1960s, Ehrhardt and Money noted that girls with the disorder had a more boy-like behaviour than did girls without the disorder. This observation pioneered
22
Articles
the interest in endocrine effects on brain and behaviour, which are central to understanding sex differences.9 Structured play studies have shown that affected girls play more with toys designed for boys than do girls without the disorder.10 This type of play behaviour is associated with the CYP21A2 genotypeie, the degree of androgen exposure.11 Also, other features of behaviour and personality such as activity levels, aggression patterns, career choices,10,12,13 and sexual orientation13,14 have been shown to be affected. Corrective surgery is done in female patients with virilised genitalia.15 Reports on the quality of life of people with CAH (irrespective of whether or not they have surgery) have been contradictory.1619 Neonatal screening for congenital adrenal hyperplasia was first done in 1977, in a pilot programme in Alaska, USA.20 Since then many countries have introduced neonatal screening.21 However, the benefit of screening has not always been readily demonstrated. An increased female-to-male ratio has historically been seen; and it probably results in part from the difference in ease of clinical identification of congenital adrenal hyperplasia between newborn boys and newborn girls who have virilised genitalia.22 A resulting higher mortality rate in male patients has been used as an argument for neonatal screening. However, a female preponderance has not always been seen.23 National neonatal screening for congenital adrenal hyper plasia was started in Sweden in 1986. To our know ledge, no other country has published more than 25 years of experience with nationwide screening for this disease. The aim of the screening programme has been to prevent salt crisis and neonatal death from salt crisis in both sexes, as well as to shorten the time to diagnosis in female neonates with ambiguous genitalia.24 The programme does not try to detect milder forms of the disorder. The aim of our study was to assess the effect of historical medical improvements in the care of this group of patients over time and to assess the effects of neonatal screening in this population-based cohort, taking advantage of the exceptionally high coverage of CYP21A2 mutation analyses done in Sweden.
Salt-wasting form Male 19101949 19501985 19862011* Total 2 (67%) 46 (41%) 73 (44%) 121 (43%) Female 1 (33%) 65 (59%) 92 (56%) 158 (57%) Simple virilising form Male 9 (60%) 32 (41%) 41 (49%) 82 (46%) Female 6 (40%) 47 (59%) 43 (51%) 96 (54%)
Methods
Study population
The study population consisted of all patients with congenital adrenal hyperplasia in Sweden included in a registry at the national screening laboratory. We also included data for all patients identified via the Swedish neonatal screening programme, late-diagnosed patients (those diagnosed after 5 years of age) reported to the laboratory, and patients who underwent genetic diagnostics or became known to us through clinical contacts. Additionally, patients known to us through previous and continuing studies were included. Furthermore, laboratories doing steroid analyses in Sweden had been contacted in previous studies to identify patients. The registry comprised 612 patients, 606 of whom had clinically proven 21-hydroxylase deficiency. We used registered data for all patients diagnosed between Jan 1, 1910, and Dec 31, 2011. Patients with a known CYP21A2 genotype were classified into genotype groups depending on the severity of the mildest allele: null, I2 splice, I172N, or V281L genotype groups, as described elsewhere.7,25 Additionally, we categorised patients with known clinical severity of the disease as salt-wasting, simple-virilising, or non-classic. In relation to clinical severity, the null and I2 splice genotype groups were included in the saltwasting group and the I172N and P30L genotype groups in the simple-virilising group. Patients with genetically verified or clinically diagnosed non-classic disease were combined and categorised as the non-classic group. Patients with unknown severity of 21-hydroxylase deficiency were denoted as unknown (table 1). Six of the identified patients had rare forms of congenital adrenal hyperplasia: one had 11-hydroxylase deficiency, three had 3-b-hydroxysteroid dehydrogenase deficiency, and two were diagnosed with P450 oxido reductase deficiency. These patients were excluded from further analysis. We obtained data for the number of livebirths from Statistics Sweden, which holds information between 1749 and 2011.26 The female-to-male ratio in the general population in Sweden was available at Statistics Sweden between 1968 and 2011.27
Non-classic form Male 0 6 (21%) 15 (35%) 21 (27%) Female 5 (100%) 23 (79%) 28 (65%) 56 (73%) Unknown Male 0 32 (62%) 13 (65%) 45 (63%) Female 0 20 (38%) 7 (35%) 27 (37%) 23 (4%) 271 (45%) 312 (51%) 606 (100%) Total
Data are n (%). The salt-wasting form of the disorder is defined clinically as a sodium concentration of less than 125 mmol/L, or is defined genetically depending on the severity of the mildest allele as either null genotype group (deletion, R356W, Q318X, R483GGtoC, ClusterE6, L308F, L307insT+Q318X, G291S, I7 splice, W405X, R356P, or V139E) or I2 splice genotype group (I2 splice, T52P, or R356Q). The simple virilising form is defined clinically as prenatal virilisation of external genitalia in girls or symptoms before 5 years of age in both sexes, or genetically with I172N, P105L+P453S, H62L+P453S, P30L, or G424S on the mildest allele. The non-classic form is defined clinically as onset of symptoms after 5 years of age or genetically with V281L, P453S, R233G, or R341W. Unknown denotes patients with no information about severity of disease. *Includes screened and unscreened individuals.
Table 1: Number of patients with congenital adrenal hyperplasia, by sex and year of birth
23
Articles
80
Male Female 12
The Committee on Ethics in Research at Karolinska Institutet approved the study; informed consent was not required in this registry-based study.
Procedures
The earliest genetic testing for the disorder was first done in the 1980s and was based on linkage to HLA genes.28 In the 1990s, rapid genotyping of common mutations and complete CYP21A2 sequence analysis were introduced.7,29,30 In this study, analyses of CYP21A2 mutations were done by allele-specific PCR and DNA sequencing from genomic DNA prepared from venous blood samples as described elsewhere.31 17-hydroxyprogesterone was used as a marker for disease and analysed by radioimmunoassay between 1986 and 1990, and from 1991 onwards by a dissociation-enhanced, lanthanide fluorescence immunoassay (Delfia; Wallac Oy Corporation, Turku, Finland). The blood samples were taken when individuals were aged 35 days until Nov 15, 2010, and as soon as possible after 48 h thereafter. Results were available at a median age of 87 days (SD 30). Cutoffs related to gestational age were used.24
60 10
Number of individuals Incidence per 100 000 livebirths
8 40 6
20
1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Unknown Salt-wasting Simple-virilising Non-classic
B
125
100
Statistical analysis
We compared the distribution of patients, grouped according to decade of birth, sex, and genotype or clinical severity, using the test. Statistical significance was set at p<005. We used SPSS (version 19) for all statistical analyses.
Number of individuals
75
50
Role of the funding source

The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. SG and AN had full access to all the data in the study. The corresponding author had final responsibility for the decision to submit for publication.
1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
25
C
125 Unknown Null Clinical salt-wasting I2 splice I172N Clinical simple-virilising P30L Clinical non-classic
Results
The study population consisted of 606 patients, born between 1915 and 2011. Most patients (n=583 [96%]) were born after 1950, when treatment with glucocorticoids was introduced. Only 23 patients were born before 1950 (table 1 and appendix). Our data show that the apparent incidence of congenital adrenal hyperplasia in Sweden increased during the 20th century (figure 1), from fewer than one individual per million livebirths between 1910 and 1920, to more than one individual per 9000 livebirths between 1990 and 2000. We detected a major increase in the apparent incidence per decade beginning in the 1950s. During the 1960s and 1970s, the apparent incidence increased substantially from one in 37 000 to the present one in 8800. By the end of 2011, 27 million infants had been screened, representing 998% coverage of all newborn babies in Sweden (population 9 million). We identified 312 individuals with the disorder born after the introduction of neonatal screening (born
See Online for appendix

Number of individuals
100
75
50
Figure 1: Distribution of congenital adrenal hyperplasia patients by sex (A), clinical severity of disease (B), and CYP21A2 genotype or phenotype when genotype was not available (C)
25
1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Decade
24
Articles
180 160 140 Number of individuals 120 100 80 60 40 20 0
Female Male
66
81
92
71
48
47
73
56
Salt-wasting Non-salt-wasting Before screening (195085)
Salt-wasting Non-salt-wasting With screening (19862011)
between 1986 and 2011). Of these, 274 were born in Sweden and hence screened for congenital adrenal hyperplasia: 231 individuals were detected during neo natal screening and 43 detected clinically at a later date. The total cohort consisted of 337 female and 269 male patients. The female-to-male ratio before 1950 was 12 to 11. However, four of the patients diagnosed before 1950 were 46XX male, giving a 46XX to 46XY ratio of 16 to seven. The overall female-to-male ratio in individuals with congenital adrenal hyperplasia detected through screening was close to one (115 vs 116), which did not differ from that of the general population. The proportion of individuals with the salt-wasting form of the disorder increased in both sexes after the introduction of screening (p=0038; figure 2). 25 (58%) of the 43 patients who were identified late, after false-negative results of neonatal screening, were female. 24 of the 38 infants with non-classic disease were
Number of patients (N=490) (Continued from previous column) Non-pseudogene-derived mutations R483GGtoC R483P R341W R233G R354H W22X I1 splice G291S I7 splice R356P W405X R356Q T295N G424S C147R V139E R426H P482S R444X F404C T52P M283V R408C Unknown despite gene sequencing: Total 11 (1%) 2 (<05%) 2 (<05%) 2 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 2 (<05%) 800
Figure 2: Number of patients with congenital adrenal hyperplasia, by sex and screening status The proportion of individuals with salt-wasting and non-salt-wasting forms of the disorder are given for before (19501985) and after (19862011) the introduction of screening for both sexes. The sex ratio did not differ after the introduction of screening, but the proportion of individuals diagnosed with the salt-wasting form increased (p=0038).
Number of patients (N=490) CYP21A2 deletion or large gene conversion Pseudogene-derived mutations I2 splice I172N V281L Q318X R356W P30L Cluster E6 P453S L307insT Combinations of pseudogene-derived mutations Q318X+R356W I2 splice and Q318X (two genes on one chromosome) I172N+P453S L307insT+Q318X Cluster E6+V281L V281L+L307insT I172N+ClusterE6+V281L+L307insT+Q318X+R356W I172N+ClusterE6+L307insT+Q318X I2 splice+I172N I172N+ClusterE6+V281L+L307insT P30L+Q318X H62L+P453S I2 splice+P453S V281L+R356W+A487P 5 (1%) 4 (1%) 4 (1%) 3 (<05%) 2 (<05%) 2 (<05%) 2 (<05%) 2 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) 1 (<05%) (Continues in next column) 218 (27%) 135 (17%) 62 (8%) 31 (4%) 25 (3%) 21 (3%) 7 (1%) 7 (1%) 6 (1%) 220 (28%)
Data are n (%). A total of 490 individuals were genotyped, and because all patients carry two alleles, they can have two different mutations (compound heterozygosity). The genotyped patients thus represent 980 alleles. However, many patients were siblings. In these cases, the families two alleles were counted only once so as not to skew the allele frequencies, and to generate the relative frequencies of unrelated alleles.
Table 2: Frequency of underlying mutations in the congenital adrenal hyperplasia population in Sweden
25
Articles
80 Number of undetected cases 70 60 50 40 30 20 10 0 1910 1920 1930 1940 1950 1960 1970
Salt-wasting Non-salt-wasting
diagnosed, we have calculated an incidence for saltwasting and non-salt-wasting form of the disease (figure 3). Assuming a constant true incidence of this genetic disease we calculated the number of presumed missed cases as presented in figure 3. Hence, five to ten children were missed every year in Sweden before 1970.
Discussion
198085 198689
Figure 3: Estimated number of undetected cases of congenital adrenal hyperplasia, by decade and type Estimations assume a constant incidence and that all patients with non-salt-wasting congenital adrenal hyperplasia born in the 1970s and all patients with a lethal salt-wasting form of the disease born after well-established screening (1990 onwards) have been diagnosed.
not detected by neonatal screening. No patients with a null genotype were missed in the screening. However, in 1989, three children with the I2 splice genotype, which typically presents as the salt-wasting form of the disease, were not detected by screening. Two of these were boys who had 17-hydroxy progesterone concentrations that were above the present cutoff. These patients were diagnosed with signs of pseudopuberty. The other patient was a girl who had a mild clinical phenotype and did not need treatment. This patient is thus a rare case in which genotype and phenotype are discordant, as has been reported previously for the I2 splice genotype group.32 In the whole cohort, 490 (81%) of 606 patients had been genotyped (table 2). All but one of the identified patients born before 1940 were CYP21A2 genotyped. In our cohort, the first patient with salt-wasting congenital adrenal hyperplasia was born in the 1940s. The distribution of different clinical and genotype groups is presented in figure 1. Of the 38 patients with non-classic congenital adrenal hyperplasia who had been screened in the neonatal period, only 14 (37%) tested positive, which suggests that many patients with this form are not identified in neonatal screening. The proportion of identified patients with non-classic congenital adrenal hyperplasia in creased over time until they reached adult age (figure 1). Between 2010 and 2011, we observed no substantial difference in the numbers of female and male patients detected, and the number of patients with non-classic congenital adrenal hyperplasia was low, as expected for the late-onset form (figure 1). After the start of the screening programme, two patients with a known diagnosis had died in the neonatal period. They were born preterm in gestational weeks 25 and 34. We are aware of four patients who died in the neonatal period before screening began. However, the apparent incidence increased from fewer than one individual per million to more than one individual per 9000 during the century, which indicates tionally. Assuming that mortality has decreased propor that all patients with salt-wasting congenital adrenal hyperplasia born in the 1990s and all patients with the non-salt-wasting form born in the 1970s were correctly
26
Our findings show a substantial increase in the apparent incidence of congenital adrenal hyperplasia during the past century, with increases in line with improvements in diagnosis and treatment over time. The absence of a female preponderance has previously been interpreted as a sign of good medical care and diagnosis. However, our data show that both male and female patients with the salt-wasting form are missed clinicallyeven in a country such as Sweden with a developed health-care systemand that neonatal screening improves survival in both sexes. A large proportion of patients were not detected by neonatal screening. Non-classic congenital adrenal hyperplasia is diagnosed more often in women than it is in men, explaining the overall female preponderance despite the fact that male and female patients with salt-wasting congenital adrenal hyperplasia are diagnosed equally via screening. To our knowledge, this study is the most extensive description of the changing apparent incidence of the different clinical forms of the disorder over time (panel). The sharp rise in the apparent incidence during the 1960s (figure 1) can be interpreted as the effect of not only the introduction of treatment with glucocorticoids in the 1950s, but also the concomitant increasing awareness of the disorder and its symptoms. Availability of a treatment generally results in increasing motivation among physicians to identify and diagnose a disease. However, in 1950, when glucocorticoids were introduced, there was only one active paediatric endocrinologist in Sweden, and little knowledge and awareness of congenital adrenal hyperplasia. Despite the obstacles, a few patients in the country were identified and started on glucocorticoid supplementation in the early 1950s. Laboratory diagnosis was difficult in both the 1950s and the 1960s, relying on measurements of urinary steroids. The amount of 17-ketosteroids was used as a measure of the level of androgens, and 17-hydroxysteroids as a measure of cortisol. Gonadotropins were measured with a mouse bio assay:extract from urine was given to prepubertal female mice and the uterine weight was subsequently measured as an indirect indication of the amount of gonadotropins.33 Since the 1950s, once diagnosis was made, treatment was still difficult. Treatment for congenital adrenal hyperplasia has changed little over time, with the exception of the introduction of mineralocorticoid treatment. The first mineralocorticoid, 11-desoxycorticosterone acetate, was given as sublingual tablets or subcutaneous implants that lasted for 3 months, at which point patients were at risk of adrenal crisis. Improved surgical techniques and
Articles
more centralised surgical care have led to better surgical outcomes. However, these modern techniques also have shortcomings, such as strictures, reduced sensitivity, and impaired sexual function.15,34,35 We detected no substantial decrease in the female-tomale ratio after the introduction of screening in our popu lation. However, and more importantly, the proportion of salt-wasting forms of congenital adrenal hyperplasia increased substantially after the introduction of the nationwide neonatal screening programme. This finding should be taken into consideration when assessing the need for neonatal screening in a population. Specifically, the incidence of the salt-wasting form seems to be a more important and adequate measure for improvement of the situation for patients with congenital adrenal hyperplasia compared with the sex ratio alone. In the cohort of patients identified through the neonatal screening programme the female-to-male ratio was close to one. Mild forms of the disorder, presenting later in life and with no risk of salt loss, are detected more often in women. Hence, the female preponderance among the late-diagnosed patients is largely the reason for the high proportion of female patients in the whole cohort (figure 1), and there is a time lag before these patients contribute to the incidence. This occurrence explains the drop in apparent incidence as well as the equal sex ratio after the year 2000 (figure 1). The classification of patients by genotyping81% had a known genotypein com bination with the nationwide coverage since 1986 made these conclusions possible. The carrier frequency, and therefore the actual inci dence of congenital adrenal hyperplasia has probably been stable over time because available data do not suggest a large global variance in the incidence of the salt-wasting form, with the exception of a few populations.3 On the assumption that all patients with salt-wasting congenital adrenal hyperplasia born after screening began have been diagnosed, the number of patients not identified in the pre-screening era can be reliably calculated because the birth rate in Sweden since 18th century is known (figure 3). Patients with the saltwasting form of the disorder most likely did not survive the neonatal period whereas patients with the simple virilising form might have had adrenal crises during later infections,35 or might not have been diagnosed at all, as would also have been the case with patients with the nonclassic form of the disorder. However, the estimated number of non-salt-wasting patients might be less reliable, because these individuals might have been missed in the screening and not all patients with a nonsalt-wasting form of the disorder might have been reported to the registry. Additionally, the proportion of non-classic cases could have increased with the in creasing immigration to Sweden since the late 1960s. The apparent increase in incidence in the 1970s might suggest, to some extent, increased interest in the dis
Panel: Research in context Systematic review We searched PubMed and Web of Science for research articles and reviews written in English with no restriction to year of publication. We used the search terms congenital adrenal hyperplasia, 21-hydroxylase deficiency, and CYP21A2. We also searched the reference lists of retrieved articles. We identified no other studies describing the incidence of congenital adrenal hyperplasia related to the progress of diagnostics and treatment covering the past 100 years. The first effective treatment with glucocorticoids was introduced in the 1950s and improved patients survival. Mineralocorticoid treatment for patients with salt-wasting forms of the disorder was introduced shortly thereafter. Diagnosis has developed from clinical examination to laboratory markers and genetic analyses. Most publications report a skewed sex ratio, with more girls than boys diagnosed, in the prevalence of congenital adrenal hyperplasia, which is widely interpreted as more girls surviving the neonatal period. Interpretation Our findings show how medical development during the past 100 years has improved the survival of patients with congenital adrenal hyperplasia, first by the introduction of glucocorticoid and mineralocorticoid treatment and later with the introduction of neonatal screening. Our findings contradict previous ideas about the disorder: many have postulated that the female preponderance among patients with congenital adrenal hyperplasia is caused by missed diagnosis and increased mortality in boys. However, our data show that both male and female babies with the severe form of the disorder were missed clinically and that neonatal screening improved survival in both sexes equally. Non-classic congenital adrenal hyperplasia is diagnosed more often in women than in men, which accounts for the female preponderance in this cohort.
order and the fact that one of the investigators of this paper (AT) made a concurrent survey to identify all diagnosed patients. Neonatal screening could, in itself, have increased awareness of the disease and thereby also improved clinical diagnoses in patients with milder forms of the disease. More identified patients and an increased apparent incidence result in better awareness of the disease. With screening in combination with the possibility of doing CYP21A2 mutation analyses, physicians no longer have to await electrolyte disturbances in a newborn baby to be able to assess the severity of their disease. Hence, the patients might escape the possible negative effects on brain development. The oldest patients mostly underwent CYP21A2 genotyping, which might be indicative of the fact that they are followed up more often at specialised clinics at university hospitals. Some of the patients born before 1950 with congenital adrenal hyperplasia were initially assigned a male sex, with or without a hypospadias diagnosis. A physically detectable symptom, the virilisation of ex ternal genitalia, aided diagnosis and increased survival. In the 1950s, our knowledge of chromosomes and chromo somal determination of sex increased, and the analysis of Barr bodies became routine clinical practice in the 1950s.36 This development led to the general idea that all 46XX individuals with congenital adrenal hyperplasia should be raised as girls when the patients were diagnosed early.
27
Articles
In the middle of the 20th century, the focus was on the survival of the patients. With improved treatment, other aspects of the patients wellbeing can be addressed. The many patients with a known CYP21A2 genotype has enabled retrospective analyses of treatment outcomes in relation to disease severityie, cortisol deficiency and extent of androgen exposure.11,13,34 The increased awareness of the psychological effects of prenatal exposure to androgens and disappointing surgical results in the more virilised patients, even using improved, modern techniques, have led to discussions about how treatment should be optimised and what sex the more virilised patients should be brought up as. The discussion about sex assignment has just begun.37 The integration of clinical work with molecular genetics and improved specialised care has been instrumental in improving the medical competence and the outcome for patients with congenital adrenal hyperplasia. Improve ments in surgical and psychological care are some of the main challenges for the future.
Contributors SG and AN designed the study and wrote the paper. SG, HF, AT, AW, UvD, MR, and AN collected the data. SG, HF, AT, AW, UvD, MR, and AN interpreted the data and critically revised the paper. Conflicts of interest We declare that we have no conflicts of interest. Acknowledgments This work was supported by the Swedish Research Council (grant number 12198), the Centre of Gender Medicine at Karolinska Institutet, the Stockholm County Council, the Sllskapet Barnavrd Foundation, the Stiftelsen Samariten Foundation, the Stiftelsen Frimurare Barnhuset Foundation, and the Novo Nordisk Foundation. References 1 Nimkarn S, Lin-Su K, New MI. Steroid 21 hydroxylase deficiency congenital adrenal hyperplasia. Pediatr Clin North Am 2011; 58: 1281300. 2 Trapp CM, Speiser PW, Oberfield SE. Congenital adrenal hyperplasia: an update in children. Curr Opin Endocrinol Diabetes Obes 2011; 18: 16670. 3 White PC, Speiser PW. Congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Endocr Rev 2000; 21: 24591. 4 Bartter FC, Forbes AP, Leaf A. Congenital adrenal hyperplasia associated with the adrenogenital syndrome: an attempt to correct its disordered hormonal pattern. J Clin Invest 1950; 29: 797. 5 Wilkins L, Lewis RA, Klein R, Rosemberg E. The suppression of androgen secretion by cortisone in a case of congenital adrenal hyperplasia. Bull Johns Hopkins Hosp 1950; 86: 24952. 6 Engel FL, Owen JA Jr, Wester TB. 9-alpha-Fluorohydrocortisoneinduced hypertension in a male infant with adrenogenitalism, and in 6 adults with Addisons disease. J Clin Endocrinol Metab 1957; 17: 27279. 7 Wedell A, Thilen A, Ritzen EM, Stengler B, Luthman H. Mutational spectrum of the steroid 21-hydroxylase gene in Sweden: implications for genetic diagnosis and association with disease manifestation. J Clin Endocrinol Metab 1994; 78: 114552. 8 Speiser PW, Dupont J, Zhu D, et al. Disease expression and molecular genotype in congenital adrenal hyperplasia due to 21-hydroxylase deficiency. J Clin Invest 1992; 90: 58495. 9 Ehrhardt AA, Epstein R, Money J. Fetal androgens and female gender identity in the early-treated adrenogenital syndrome. Johns Hopkins Med J 1968; 122: 16067. 10 Berenbaum SA, Duck SC, Bryk K. Behavioral effects of prenatal versus postnatal androgen excess in children with 21-hydroxylasedeficient congenital adrenal hyperplasia. J Clin Endocrinol Metab 2000; 85: 72733.
11 Nordenstrom A, Servin A, Bohlin G, Larsson A, Wedell A. Sex-typed toy play behavior correlates with the degree of prenatal androgen exposure assessed by CYP21 genotype in girls with congenital adrenal hyperplasia. J Clin Endocrinol Metab 2002; 87: 511924. 12 Pasterski V, Hindmarsh P, Geffner M, Brook C, Brain C, Hines M. Increased aggression and activity level in 3- to 11-year-old girls with congenital adrenal hyperplasia (CAH). Horm Behav 2007; 52: 36874. 13 Frisen L, Nordenstrom A, Falhammar H, et al. Gender role behavior, sexuality, and psychosocial adaptation in women with congenital adrenal hyperplasia due to CYP21A2 deficiency. J Clin Endocrinol Metab 2009; 94: 343239. 14 Meyer-Bahlburg HF, Dolezal C, Baker SW, New MI. Sexual orientation in women with classical or non-classical congenital adrenal hyperplasia as a function of degree of prenatal androgen excess. Arch Sex Behav 2008; 37: 8599. 15 Braga LH, Pippi Salle JL. Congenital adrenal hyperplasia: a critical appraisal of the evolution of feminizing genitoplasty and the controversies surrounding gender reassignment. Eur J Pediatr Surg 2009; 19: 20310. 16 Jaaskelainen J, Tiitinen A, Voutilainen R. Sexual function and fertility in adult females and males with congenital adrenal hyperplasia. Horm Res 2001; 56: 7380. 17 Johannsen TH, Ripa CP, Mortensen EL, Main KM. Quality of life in 70 women with disorders of sex development. Eur J Endocrinol 2006; 155: 87785. 18 Arlt W, Willis DS, Wild SH, et al. Health status of adults with congenital adrenal hyperplasia: a cohort study of 203 patients. J Clin Endocrinol Metab 2010; 95: 511021. 19 Nermoen I, Husebye ES, Svartberg J, Lovas K. Subjective health status in men and women with congenital adrenal hyperplasia: a population-based survey in Norway. Eur J Endocrinol 2010; 163: 45359. 20 Pang S, Hotchkiss J, Drash AL, Levine LS, New MI. Microfilter paper method for 17 alpha-hydroxyprogesterone radioimmunoassay: its application for rapid screening for congenital adrenal hyperplasia. J Clin Endocrinol Metab 1977; 45: 100308. 21 Speiser PW, Azziz R, Baskin LS, et al. A summary of the endocrine society clinical practice guidelines on congenital adrenal hyperplasia due to steroid 21-hydroxylase deficiency. Int J Pediatr Endocrinol 2010; 2010: 494173. 22 Nordenstrom A, Ahmed S, Jones J, et al. Female preponderance in congenital adrenal hyperplasia due to CYP21 deficiency in England: implications for neonatal screening. Horm Res 2005; 63: 2228. 23 Perry R, Kecha O, Paquette J, Huot C, Van Vliet G, Deal C. Primary adrenal insufficiency in children: twenty years experience at the Sainte-Justine Hospital, Montreal. J Clin Endocrinol Metab 2005; 90: 324350. 24 Thilen A, Nordenstrom A, Hagenfeldt L, von Dobeln U, Guthenberg C, Larsson A. Benefits of neonatal screening for congenital adrenal hyperplasia (21-hydroxylase deficiency) in Sweden. Pediatrics 1998; 101: E11. 25 Krone N, Braun A, Roscher AA, Knorr D, Schwarz HP. Predicting phenotype in steroid 21-hydroxylase deficiency? Comprehensive genotyping in 155 unrelated, well defined patients from southern Germany. J Clin Endocrinol Metab 2000; 85: 105965. 26 Statistics Sweden. Population and population changes 17492011. http://www.scb.se/Pages/TableAndChart____26047.aspx (accessed Dec 23, 2012). 27 Statistics Sweden. Live births by region, sex and age of mother: year 19682011. http://www.scb.se/Pages/SSD/SSD_ SelectVariables____340507.aspx?rxid=1c9e086b-4b7f-4823-a0bd885120416892&px_tableid=ssd_extern%3aFoddaK (accessed Dec 23, 2012). 28 Dupont B, Oberfield SE, Smithwick EM, Lee TD, Levine LS. Close genetic linkage between HLA and congenital adrenal hyperplasia (21-hydroxylase deficiency). Lancet 1977; 2: 130912. 29 Wedell A, Ritzen EM, Haglund-Stengler B, Luthman H. Steroid 21-hydroxylase deficiency: three additional mutated alleles and establishment of phenotype-genotype relationships of common mutations. Proc Natl Acad Sci USA 1992; 89: 723236. 30 Sim SC. CYP21A2 allele nomenclature. http://www.cypalleles.ki.se/ cyp21.htm2012 (accessed Dec 23, 2012).
28
Articles
31 Wedell A. Molecular genetics of 21-hydroxylase deficiency. Endocr Dev 2011; 20: 8087. 32 Wilson RC, Mercado AB, Cheng KC, New MI. Steroid 21-hydroxylase deficiency: genotype may not predict phenotype. J Clin Endocrinol Metab 1995; 80: 232229. 33 Rosemberg E, Lewis W, Russell E. Performance control of preparations which could be used as reference material for the estimation of urinary gonadotropins in the mouse uterine weight method. J Clin Endocrinol Metab 1962; 22: 95358. 34 Nordenstrom A, Frisen L, Falhammar H, et al. Sexual function and surgical outcome in women with congenital adrenal hyperplasia due to CYP21A2 deficiency: clinical perspective and the patients perception. J Clin Endocrinol Metab 2010; 95: 363340.
35 Reisch N, Willige M, Kohn D, et al. Frequency and causes of adrenal crises over lifetime in patients with 21-hydroxylase deficiency. Eur J Endocrinol 2012; 167: 3542. 36 Barr ML, Bertram EG. A morphological distinction between neurones of the male and female, and the behaviour of the nucleolar satellite during accelerated nucleoprotein synthesis. Nature 1949; 163: 676. 37 Lee PA, Houk CP. Review of outcome information in 46,XX patients with congenital adrenal hyperplasia assigned/reared male: what does it say about gender assignment. Int J Pediatr Endocrinol 2010; 2010: 982025
29
Review
New insights into cystic fibrosis-related diabetes in children

Katie L Ode, Antoinette Moran
Published Online May 23, 2013 http://dx.doi.org/10.1016/ S2213-8587(13)70015-9 Department of Pediatrics, University of Iowa Childrens Hospital, University of Iowa, Iowa City, IA, USA (K L Ode MD); and Department of Pediatrics, Amplatz Childrens Hospital, University of Minnesota, Minneapolis, MN, USA (Prof A Moran MD) Correspondence to: Dr Katie L Ode, Department of Pediatrics, Division of Endocrinology and Diabetes, 200 Hawkins Dr, 2861 JPP, Iowa City, IA 52242-1083, USA katie-larsonode@uiowa.edu
Cystic fibrosis-related diabetes (CFRD) is the most common complication of cystic fibrosis. It is associated with significantly increased morbidity and mortality in adults and children. Adolescents with cystic fibrosis have a much higher prevalence of diabetes than any other similar age population. Glucose abnormalities that precede diabetes are even more common, especially in children younger than 10 years. The pathophysiology of glucose metabolic abnormalities is poorly understood, but insulin insufficiency is clearly the main component. Findings from animal studies have provided insight into the pathophysiology of CFRD, and imply that carbohydrate metabolic abnormalities might begin at much younger ages than was previously thought in patients with cystic fibrosis, and might be related to the basic cystic fibrosis chloride channel defect. In this Review we explore present knowledge of CFRD in children and adolescents, and new data that indicate that the pathophysiology of CFRD begins in very young patients.
Introduction
Cystic fibrosis is the most common fatal recessive genetic disorder in white populations. The disorder is caused by mutations leading to abnormalities in the cystic fibrosis transmembrane conductance regulator (CFTR)a chloride channel involved in electrolyte and water movement across cell membranes. The most common mutation of CFTR is F508, in which a phenylalanine is lost at position 508.1 CFTR dysfunction results in multisystem disease, the most important aspect of which is obstructive and inflammatory lung disease caused by thick secretions. The average lifespan of people with cystic fibrosis has increased greatly because of improved treatments for lung disease; thus, other complications of cystic fibrosis have become much more important. Cystic fibrosis-related diabetes (CFRD) is the most common secondary complication of cystic fibrosis, and in some paediatric diabetes practices patients with CFRD are the second largest group of patients after children with type 1 diabetes mellitus.2 CFRD is a unique form of diabetes that does not have the same pathophysiology or clinical picture as either type 1 or type 2 diabetes, although it has features in common with both. Although lung disease is not a major component of type 1 or type 2 diabetes pathophysiology, CFRD increases the risk of pulmonary death in patients with cystic fibrosis.3 CFRD has been associated with rapid decline in lung function and reduced nutritional status, both of which increase morbidity and mortality.4 Furthermore, clinical decline seems to happen years before patients develop frank diabetes, when they have abnormal glucose tolerance.4 Authors of several reviews of CFRD have investigated the disease primarily from the perspective of the adult patient, or by combining paediatric and adult patients.2,510 Stalvey11 discussed the basic science of CFRD, and the Health Technology Assessment from the National Health Service in Great Britain contains a thorough review of screening for CFRD.12 In this Review we focus on what is and is not known about glucose metabolism abnormalities in children and adolescents with cystic fibrosis, with particular attention to questions that have
arisen from new animal models of infant and paediatric glucose metabolism, and what they might imply about the pathophysiology of CFRD.
Pathophysiology
Inadequate insulin secretion is the main pathological feature of CFRD. Patients with cystic fibrosis, even those with normal glucose tolerance, tend to have decreasing insulin secretion over time.13,14 Patients with cystic fibrosis who are 8 years and older have delayed and eventually diminished insulin secretion in response to oral and intravenous stimuli, and increased glucose con centrations compared with controls.15 Insulin sensitivity is generally normal or only slightly decreased, except in the settings of acute illness or glucocorticoid treatment when insulin resistance can be severe.13,14,16,17 The basis of our understanding of the pathophysiology of CFRD and related glucose abnormalities in cystic fibrosis has evolved since diabetes was first described in patients with cystic fibrosis in the 1950s.18 Initially, the disorder was thought to be due to collateral damage to the endocrine pancreas after scarring and fibrosis of the exocrine pancreas. Concentrations of islet hormones other than insulin, such as glucagon, are also abnormal in adults with CFRD, which supports the theory that total islets are destroyed by fibrosis.19 However, although all patients with cystic fibrosis seem to have substantial -cell loss by this mechanism, it cannot be the sole cause of CFRD since results of histopathology studies have shown that patients with cystic fibrosis with CFRD do not necessarily have more pancreatic fibrosis at autopsy than do patients with cycstic fibrosis without CFRD.19 Blackman and colleagues20,21 have shown that susceptibility genes for type 2 diabetes are more common in patients with CFRD compared with patients with cystic fibrosis without diabetes, which suggests that CFRD is associated with similar intrinsic -cell defects as are associated with type 2 diabetes. Furthermore, evidence suggests that the CFTR defect itself might be involved in altered insulin secretion. CFTR is expressed in -cells and -cells in the rat pancreas, although its role in these cells is unknown.22
30
Review
35
Non-CF-1 Non-CF-2 Non-CF-3
CF-1 CF-2 CF-3
30
25 Blood glucose (mmol/L)
20
15
10
10
20
30
40
Age (days)
50
60
70
80
90
Figure 1: Impaired glycaemic regulation in 13-month-old ferrets Blood glucose measured on non-fasted, randomly fed CFTR-null and wild-type ferret sibling pairs during the first 3 months of life. Blood draws were taken simultaneously from each animal in the pair. Three independent cystic fibrosis and non-cystic fibrosis sibling pairs are shown. Non-CF=non-cystic fibrosis. CF=cystic fibrosis. Adapted from Olivier and colleagues,25 by permission of the American Society for Clinical Investigation.
Plasma Insulin concentration (pmol/L)
Advances in the development of new animal models for cystic fibrosis have included CFTR-null and F508 pigs, the CFTR-null ferret, and CFTR-null and F508 mice. CFTR-null ferrets develop multi-system disease, including pancreas disease and lung disease that has very similar pathology to that in people.23,24 These ferrets develop glucose tolerance abnormalities and frank diabetes.25 Endocrine pancreas function in the CFTRnull ferret is abnormal from birth, which suggests that it is an intrinsic defect. Newborn ferrets with cystic fibrosis have abnormal glucose tolerance and reduced first-phase insulin secretion (figures 1 and 2);25 the most interesting aspect of this finding is that these abnormalities seem to precede overt pathology in the pancreas.25 In the same way as in people, the glucose abnormalities progress with age in the CFTR-null animals as pancreatic fibrosis worsens.25 We postulate that a mild intrinsic defect in insulin secretion related to CFTR exists, which then becomes of greater clinical importance as islets are lost through scarring and fibrosis of the pancreas. Other animal models suggest intrinsic defects in insulin secretion in cystic fibrosis. After low-dose streptozocin, CFTR-mutant mice had greater impairment of insulin secretion than did control mice despite similar loss of islets, and increased immune reactivity in response to hyperglycaemia.26,27
Blood glucose concentration (mmol/L)
Animal models and the earliest defects in glucose metabolism
2084
p<00120
111 89 67 44 22 0
p=03884
1389
695
Non-CF
CF
Non-CF
CF
Figure 2: Decreased acute insulin secretion in CFTR-null ferrets Plasma insulin concentrations 2 min after intra-peritoneal injection of 3 g/kg of glucose into newborn CFTR-null and wild-type ferrets who had been fasted for 3 h (A). Blood glucose concentrations from the same blood draw (B). Non-CF=non-cystic fibrosis; CF=cystic fibrosis. Adapted from Olivier and colleagues,25 by permission of the American Society for Clinical Investigation.
New data from CFTR correctors

In a small study of patients with cystic fibrosis aged 652 years given ivacaftora so-called CFTR corrector approved by the US Federal Drug Association restoration of CFTR function for 1 month improved the insulin response to oral glucose by 66178%, apart from in one patient who had had diabetes for more than 16 years. All patients but one showed improvement in first phase insulin secretion after ivacaftor.28
31
Review
60
Prevalence
CFRD total FH+ FH
50
40 Prevalence (%)
30
20
10
<10
1019
70
2029 Age (years)
3039
40
Males Females *
60
Although cases have been reported of diabetes in very young children, including infants, with cystic fibrosis,2932 overt diabetes is rare before the age of 10 years, with a prevalence of roughly 15%.33 In these cases, autoimmune type 1 diabetes must be considered. CFRD prevalence increases with age, with diabetes occurring in about 15% of adolescents and 50% of adults with cystic fibrosis34 (figure 3). In New South Wales, Australia, the incidence of CFRD is 22 per 1000 person-years, and the disorder affects one in five young people with cystic fibrosis.35 45% of French patients with cystic fibrosis had CFRD by the age of 20 years.36 Milder abnormalities of glucose tolerance are even more common than CFRD.19 Bismuth and colleagues37 noted that 20% of children with cystic fibrosis had impaired glucose tolerance at 10 years of age. In our own series,37 39 of 94 children (41%) with cystic fibrosis aged 69 years had abnormal glucose tolerance (or indeterminate glycaemia, defined as a mid-OGTT [oral glucose tolerance test] value >111 mmol/L in patients with normal fasting and 2 h levels), although none had diabetes. These data are the youngest reported systematic OGTT data; nothing is known about glucose tolerance in children younger than 6 years.
50
Implications of diabetes and abnormal glucose tolerance in children

The main acute complications of diabetes are diabetic ketoacidosis and insulin-induced hypoglycaemia. Diabetic ketoacidosis is rare in children and adolescents with CFRD, possibly because they still have endogenous insulin secretion.38 Insulin-induced hypoglycaemia can occur, but in this population it should be distinguished from fasting and reactive hypoglycaemia, which are very common in children with cystic fibrosis even in the absence of diabetes and insulin treatment.39,40 In clinical practice, hypoglycaemia is seldom a major problem. Generally, adults with CFRD do not die from the macrovascular complications typically associated with death in other diabetes populations, although they can have microvascular complications (retinopathy, neuropathy, and nephropathy) and should be screened for these disorders.4144 The primary concern, however, is that CFRD increases mortality from respiratory failure and lung disease.4 This increase is thought to be due to a combination of the catabolic effects of insulin insufficiency and the inflammatory and pathogenpromoting effects of hyperglycaemia.7 In children with cystic fibrosis, diabetes and even abnormal glucose tolerance are reported to be associated with decreased lung function and nutritional status, which contribute to the increased morbidity and mortality from CFRD recorded at later ages. Alicandro and colleagues45 reported that in non-diabetic patients with cystic fibrosis younger than 20 years, those with the lowest fasting insulin concentrations had decreased
Prevalence (%)
40
30
20
10
<10 (n=63)
1019 (n=126)
2029 (n=150) Age (years)
3039 (n=115)
40 (n=73)
Figure 3: CFRD prevalence with and without fasting hyperglycaemia (A) and in males and females (B) by age group CFRD total=total prevalence of CFRD. FH+=CFRD with fasting hyperglycaemia. FH=CFRD without fasting hyperglycaemia. Adapted from Moran and colleagues,34 by permission of The American Diabetes Association.
These data indicate that CFTR itself is probably an essential part of -cell pathophysiology in cystic fibrosis and that glucose and insulin abnormalities are not only attributable to destruction of the endocrine pancreas as a bystander effect of loss of the exocrine pancreas. Additionally, findings from these studies suggest that the effect on insulin secretion and -cell function probably starts at a very young age in patients with cystic fibrosis.
32
Review
nutritional status, supporting the important anabolic role of insulin in growth and development. Brodsky and colleagues46 noted that lung function tests had a strong negative association with the presence of mid-OGTT glucose elevation in children, indicating a potential effect of even intermittent hyperglycaemia on lung function; they strongly argue that OGTT intermediate timepoints are clinically important and that omission of the 1 h timepoint could cause clinically relevant information to be missed.46 An association is also found between increased haemoglobin A1C and decreased nocturnal oxygen saturation even when controlled for body-mass index, lung function, and other sleep parameters, again supporting the theory that even mild hyperglycaemia negatively affects lung function.47 Our group reported that the presence of abnormal glucose tolerance in children aged 610 years strongly predicts development of early CFRD over the next few years,37 and thus this is a marker of more severe metabolic disease with high risk of progression. Curiously, impaired fasting glucose, unlike other glucose abnormalities in cystic fibrosis, is not associated with increased morbidity or mortality.48
guidelines are followed. Table 3 shows glucose tolerance categories based on OGTT, and the panel details diagnostic recommendations from the consensus guidelines.58 All patients with CFRD, including those with normal fasting glucose concentrations, need insulin therapy since insulin has been shown to improve outcomes.34,59 Fewer data exist to guide treatment of abnormal glucose tolerance in cystic fibrosis.
Treatment of diabetes
Treatment of CFRD in children who have fasting hyperglycaemia is similar to that of children with type 1
Method Routine 2 h 75 g (175 g/kg) OGTT Timing Yearly at a state of basal health; beginning at age 10 years Every pulmonary exacerbation requiring intravenous antibiotics or steroids Before pregnancy, 1216 weeks, 2428 weeks, 612 weeks post partum Within 6 months before transplant; After transplant: from critical care period to hospital discharge
Acute illness
Fasting and 2 h postprandial blood sugars for 48 h
Pregnancy
2 h 75 g (175 g/kg) OGTT
Screening of children and adolescents with cystic fibrosis

North American and European consensus guidelines recommend the 2 h 175 g/kg OGTT test for CFRD screening, done annually as an outpatient procedure during a period of stable baseline health (table 1).49,50 Other methods of testing for diabetes are not recommended by consensus guidelines for various reasons, mostly related to inadequate sensitivity (table 2). Continuous glucose monitoring (CGM) is much more sensitive to glucose abnormalities than is any other method of screening in adults and children.5156 CGM correlates with OGTT values and is reproducible.57 However, as not enough data exist to link CGM results to long-term outcomes, its use is not recommended for screening in consensus guidelines. Most patients with cystic fibrosis can have intermittent glucose elevation by CGM, although whether all these patients need therapy for diabetes is not clear. Consensus guidelines recommend starting screening at age 10 years because diabetes is rare before this age. We have advocated for earlier screening, however, beginning at age 6 years. We have shown that children with abnormal OGTT results at 69 years of age progress to CFRD much sooner than do children with normal glucose tolerance.37 Furthermore, in large epidemiological studies, the greatest relative risk for death from CFRD was in patients younger than 10 years at baseline assessment.3
Transplant
Before transplant: 2 h 75 g (175 g/kg) OGTT; After transplant: plasma glucose concentrations
Table 1: Current screening guidelines, by patient group58
Reason not recommended Haemoglobin A1C Insufficient sensitivity in cystic fibrosis, patients might have spuriously low A1C resulting in false-negative screen Insufficient sensitivity in cystic fibrosis Insufficient sensitivity in cystic fibrosis Insufficient sensitivity in cystic fibrosis Shows promise, but insufficient data to correlate with outcomes
Fructosamine Fasting glucose Glycosuria Continuous glucose monitoring
Table 2: Screening methods not recommended by the guidelines
Fasting glucose Intermediate timepoints 2 h timepoint mmol/L (mg/dL ) mmol/L (mg/dL ) mmol/L (mg/dL ) Normal glucose tolerance Impaired fasting glucose Impaired glucose tolerance Indeterminate glycaemia <56 (100) >56 (100) <70 (126) <70 (126) <70 (126 ) <111 (200) <111 (200) <111 (200) >111 (200) NA NA <78 (140) <78 (140) >78 (140) <111 (200) <78 (140) >111 (200) >111 (200)
Cystic fibrosis related diabetes mellitus <70 (126) without fasting hyperglycaemia Cystic fibrosis related diabetes mellitus 70 (126) with fasting hyperglycaemia
NA=not applicable.
Diagnosis and treatment of glucose abnormality in children and adolescents with cystic fibrosis
The diagnosis of glucose tolerance abnormalities in children and adolescents does not differ from that of adultsthe same categories are used and the same
Table 3: Glucose categories from oral glucose tolerance test
33
Review
diabetes. The recommendation from consensus guidelines is to use a basal-bolus regimen with a longacting insulin plus several daily doses of rapid-acting insulin, or to use an insulin pump. Patients are taught how to use insulin-tocarbohydrate ratios to cover carbohydrate intake in meals and snacks.58 One difference in management from that of other paediatric patients with diabetes is that children with cystic fibrosis have very specific dietary recommendations because of the well established need for maintenance of weight and body-mass index. Patients should follow the usual cystic fibrosis dietary guidelines for a high-calorie, high-salt, and high-fat diet. No specific data exist for children with cystic fibrosis to support one type of insulin regimen over another. The average adolescent with CFRD receives roughly 04 units per kg per day of insulin,60 which is less than most adolescents of similar age with type 1 diabetes. These
Panel: Diagnosis recommendations from the guidelines 1 During a period of stable baseline health, CFRD can be diagnosed in patients with cystic fibrosis according to standard ADA criteria: 2 h OGTT plasma glucose >111 mmol/L (200 mg/dL) FPG 70 mmol/L (126 mg/dL) Haemoglobin A1C >65% (<65% does not rule out CFRD because this value is often spuriously low in patients with cystic fibrosis) Classic diabetes symptoms (polyuria and polydipsia) in the presence of a casual glucose concentration >111 mmol/L (200 mg/dL) Testing should be done on 2 separate days to rule out laboratory error, unless there are unequivocal symptoms of hyperglycaemia (polyuria and polydipsia); a positive FPG or A1C can be used as a confirmatory test, but if either is normal the OGTT should be done or repeated. If diagnosis of diabetes is not confirmed, the patient resumes routine annual testing 2 Diagnosis of CFRD can be made in patients with cystic fibrosis with acute illness (intravenous antibiotics in the hospital or at home, systemic glucocorticoid therapy) when the following persist for more than 48 h 2 h postprandial plasma glucose >111 mmol/L (200 mg/dL) FPG 70 mmol/L (126 mg/dL) 3 Diagnosis of CFRD can be made in patients with cystic fibrosis on enteral continuous drip feedings when mid-feeding or post-feeding plasma glucose concentrations exceed 111 mmol/L (200 mg/dL) on 2 separate days 4 Diagnosis of gestational diabetes mellitus should be made on the basis of the recommendations of the International Association of the Diabetes and Pregnancy Study Groups,by which diabetes is diagnosed on the basis of 0, 1, and 2 h glucose concentrations with a 75 g OGTT if any of the following are present: FPG >51 mmol/L (92 mg/dL) 1 h plasma glucose >100 mmol/L (180 mg/dL) 2 h plasma glucose >85 mmol/L (153 mg/dL) (ADA; Consensus) Patients with cystic fibrosis with gestational diabetes mellitus are not deemed to have CFRD, but need CFRD screening 612 weeks after the end of the pregnancy 5 To distinguish between CFRD with and without fasting hyperglycaemia is not necessary 6 The onset of CFRD should be defined as the date that a person with cystic fibrosis first meets diagnostic criteria, even if hyperglycaemia subsequently abates
Adapted from Moran and colleagues,58 by permission of the American Diabetes Association. CFRD=cystic fibrosis-related diabetes; ADA=American Diabetes Association; OGTT=oral glucose tolerance test; FPG=fasting plasma glucose.
modest insulin requirements suggest the persistence of endogenous insulin secretion. This secretion seems to be sufficient to compensate for the usual insulin resistance of puberty, since adolescent doses are similar to insulin doses in adults with CFRD.60 In our experience, we usually recommend starting meal coverage with an insulin-to-carbohydrate ratio of roughly half of what we would recommend for a child with type 1 diabetes of similar age and weight, and then adjust on the basis of 2 h postprandial glucose concentrations. We generally start at 0510 units per 15 g. In the presence of fasting hyperglycaemia, we recommend initiation of longacting basal insulin at roughly 025 units per kg per day, with subsequent adjustment of the dose on the basis of fasting blood glucose concentrations. Patients who are acutely ill need much higher doses, which can then be rapidly reduced as the illness resolves. No treatment guidelines exist to recommend selection of one insulin regimen over another for CFRD treatment when fasting hyperglycaemia is not present. The CFRDT trial59 used pre-meal rapid-acting insulin in adults with CFRD without fasting hyperglycaemia, which resulted in an increase in body-mass index; however, no paediatric patients were included in that trial. On the basis of these data, however, we use pre-meal rapid-acting insulin in children with CFRD without fasting hyperglycaemia. Findings from small series61 have suggested a beneficial effect of once-daily longacting insulin.
Treatment of abnormal glucose tolerance

Promising preliminary results from five of six small, uncontrolled studies6165 suggest that insulin also improves weight and clinical status in non-diabetic adolescents with cystic fibrosis with abnormal glucose tolerance. In these studies, adolescents with cystic fibrosis with abnormal glucose tolerance gained weight in response to insulin therapy of about 025 units per kg per day, without experiencing hypoglycaemia or other adverse events. Adherence was good and the injections were well tolerated. The only study that did not show weight gain in patients used a lower dose of insulin than all but one of the other studies (roughly 015 units per kg per day).66 In cystic fibrosis, treatment of fat-soluble vitamin deficiency and pancreatic exocrine deficiency is routinely begun before evidence of clinical compromise due to such deficiency. Because of early indications that insulin treatment of non-diabetic patients with cystic fibrosis improves nutritional and pulmonary status, and evidence that even healthy children with cystic fibrosis are often insulinopenic, treatment of insulin deficiency in children with cystic fibrosis should be similar to how other known deficiencies in cystic fibrosis are treated.65,66 In infants with cystic fibrosis, aggressive screening for pancreatic insufficiency begins at diagnosis, and supplementation is begun as soon as there is evidence of insufficiency, even though faecal elastase production can vary in the first years of life.67 In the future, we might manage insulin
34
Review
Search strategy and selection criteria We searched PubMed for articles published in English and other languages between Nov 15, 2006, and Nov 15, 2012. We used the search term cystic fibrosis with the term diabetes. Abstracts were reviewed for age ranges of included patients; articles referring solely to patients older than 18 years were excluded, unless they could be considered relevant to the paediatric population (ie, they evaluated long term outcomes or complications). We did not exclude commonly referenced and highly regarded older publications. We also searched the reference lists of articles identified by this search strategy and selected articles that we deemed relevant to this topic. We have included review articles to provide further details and references than could be included in this Review.
Conflicts of interest We declare that we have no conflicts of interest. Acknowledgments We thank John Englehardt, University of Iowa, Iowa City, IA, USA, for the use of his figures. References 1 Kerem B, Rommens JM, Buchanan JA, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989; 245: 107380. 2 Gregory JW. What are the main research findings during the last 5 years that have changed my clinical practice in diabetes medicine? Arch Dis Child 2012; 97: 43639. 3 Chamnan P, Shine BSF, Haworth CS, Bilton D, Adler AI. Diabetes as a determinant of mortality in cystic fibrosis. Diabetes Care 2010; 33: 31116. 4 Milla CE, Billings J, Moran A. Diabetes is associated with dramatically decreased survival in female but not male subjects with cystic fibrosis. Diabetes Care 2005; 28: 214144. 5 Nathan BM, Laguna T, Moran A. Recent trends in cystic fibrosis-related diabetes. CurrOpin Endocrin Diabetes Obes 2010; 17: 33541. 6 Laguna TA, Nathan BM, Moran A. Managing diabetes in cystic fibrosis. Diabetes Obes Metab 2010; 12: 85864. 7 Moran A, Becker D, Casella SJ, et al. Epidemiology, pathophysiology, and prognostic implications of cystic fibrosis-related diabetes: a technical review. Diabetes Care 2010; 33: 267783. 8 Lek N, Acerini CL. Cystic fibrosis related diabetes mellitusdiagnostic and management challenges. Curr Diabetes Rev 2010; 6: 916. 9 Hameed S, Jaff A, Verge CF. Cystic fibrosis related diabetes (CFRD)the end stage of progressive insulin deficiency. Paediatr Pulmonol 2011; 46: 74760. 10 Stecenko AA, Moran A. Update on cystic fibrosis-related diabetes. Curr Opin Pulmon Med 2010; 16: 61115. 11 Stalvey MS, Flotte TR. Endocrine parameters of cystic fibrosis: back to basics. J Cell Biochem 2009; 108: 35361. 12 Waugh N, Royle P, Craigie I, et al. Screening for cystic fibrosis-related diabetes: a systematic review. Health Technol Assess 2012; 16: iiiiv, 1179. 13 Lombardo F, De Luca F, Rosano M, et al. Natural history of glucose tolerance, -cell function and peripheral insulin sensitivity in cystic fibrosis patients with fasting euglycemia. Eur J Endocrinol 2003; 149: 5359. 14 Street ME, Spaggiari C, Ziveri MA, et al. Insulin production and resistance in cystic fibrosis: effect of age, disease activity, and genotype. J Endocrinol Invest 2012; 35: 24653. 15 Battezzati A, Mari A, Zazzeron L, et al. Identification of insulin secretory defects and insulin resistance during oral glucose tolerance test in a cohort of cystic fibrosis patients. Eur J Endocrinol 2011; 165: 6976. 16 Yung B, Noormohamed FH, Kemp M, Hooper J, Lant AF, Hodson ME. Cystic fibrosis-related diabetes: the role of peripheral insulin resistance and -cell dysfunction. Diabet Med 2002; 19: 22126. 17 Elder DA, Wooldridge JL, Dolan LM, DAlessio DA. Glucose tolerance, insulin secretion, and insulin sensitivity in children and adolescents with cystic fibrosis and no prior history of diabetes. J Pediatr 2007; 151: 65358. 18 Caldwell DM, Mcnamara DH. Fibrocystic disease of the pancreas and diabetes in an adult with unusual pulmonary manifestations. Calif Med 1958; 89: 28084. 19 Moran A, Doherty L, Wang X, Thomas W. Abnormal glucose metabolism in cystic fibrosis. J Pediatr 1998; 133: 1017. 20 Blackman SM, Hsu S, Vanscoy LL, et al. Genetic modifiers play a substantial role in diabetes complicating cystic fibrosis. J Clin Endocrinol Metab 2009; 94: 130209. 21 Blackman SM, Hsu S, Ritter SE, et al. A susceptibility gene for type 2 diabetes confers substantial risk for diabetes complicating cystic fibrosis. Diabetologia 2009; 52: 185865. 22 Boom A, Lybaert P, Pollet J-F, et al. Expression and localization of cystic fibrosis transmembrane conductance regulator in the rat endocrine pancreas. Endocrine 2007; 32: 197205. 23 Fisher JT, Liu X, Yan Z, et al. Comparative processing and function of human and ferret cystic fibrosis transmembrane conductance regulator. J Biol Chem 2012; 287: 2167385.
insufficiency similarly, beginning treatment as soon as screening shows insufficiency, rather than waiting for other symptoms and signs to develop. However, further research is needed to confirm this approach.
Future directions
Studies of the CFTR-null ferret and the CFTR-null and F508 pigs are assessing the effects of CFTR mutations on young animal models to further examine the pathophysiology of CFRD. Cell culture studies funded by the US National Institutes of Health are also underway involving isolated islets from human beings, ferrets, and pigs, including CFTR knockout, F508, and wild type. These studies will help to delineate the role of CFTR in insulin secretion. Clinically, because CFTR dysfunction itself might be part of CFRD pathophysiology, determination of whether glucose metabolic abnormalities are present in human infants and very young children with cystic fibrosis is of interest. Our group has begun a study to assess glucose metabolism in children aged 3 months to 5 years with cystic fibrosis compared with that of control children. The new era of clinically relevant CFTR correction offers the unique opportunity to ascertain whether restoration of CFTR function affects development of CFRD. Studies must be designed to follow glucose metabolism in children who begin these medications. Development of CFRD could be slowed or stopped with these new treatments if started in children at a young age. Use of these medications might also allow greater insight into the pathophysiology of CFRD itself. In view of data for milder glucose abnormalities and clinical decline in children,56 further investigation of the therapy of insulinopenia and early glucose abnormalities in children is needed, which would be best addressed by large, collaborative, multi-centre trials.
Contributors KLO did the initial literature search and wrote the initial Review. AM contributed to the content outline and edited the Review. Both approved the final draft.
35
Review
24 Keiser NW, Engelhardt JF. New animal models of cystic fibrosis: what are they teaching us? Curr Opin Pulm Med 2011; 17: 47883. 25 Olivier AK, Yi Y, Sun X, et al. Abnormal endocrine pancreas function at birth in cystic fibrosis ferrets. J Clin Invest 2012; 122: 375568. 26 Stalvey MS, Brusko TM, Mueller C, et al. CFTR mutations impart elevated immune reactivity in a murine model of cystic fibrosis related diabetes. Cytokine 2008; 44: 15459. 27 Stalvey MS, Muller C, Schatz DA, et al. Cystic fibrosis transmembrane conductance regulator deficiency exacerbates islet cell dysfunction after -cell injury. Diabetes 2006; 55: 193945. 28 Bellin MD, Laguna T, Leschyshyn J, et al. Proof-of-concept pilot: insulin secretion improves in cystic fibrosis following ivacaftor correction of CFTR. Pediatr Diabetes 2013; (in press). 29 Casas L, Berry DR, Logan K, Copeland KC, Royall JA. Cystic fibrosis related diabetes in an extremely young patient. J Cyst Fibrosy 2007; 6: 24749. 30 Gelfand IM, Eugster EA, Haddad NG. Infancy-onset cystic fibrosis-related diabetes. Diabetes Care 2005; 28: 259394. 31 Lombardi F, Raia V, Spagnuolo MI, et al. Diabetes in an infant with cystic fibrosis. Pediatr Diabetes 2004; 5: 199201. 32 Siahanidou T, Mandyla H, Doudounakis S, Anagnostakis D. Hyperglycaemia and insulinopenia in a neonate with cystic fibrosis. Acta Paediatr 2005; 94: 183740. 33 Koch C, Rainisio M, Madessani U, et al. Presence of cystic fibrosis-related diabetes mellitus is tightly linked to poor lung function in patients with cystic fibrosis: data from the European Epidemiologic Registry of Cystic Fibrosis. Pediatr Pulmonol 2001; 32: 34350. 34 Moran A, Dunitz J, Nathan B, Saeed A, Holme B, Thomas W. Cystic fibrosis-related diabetes: current trends in prevalence, incidence, and mortality. Diabetes Care 2009; 32: 162631. 35 Rana M, Munns CF, Selvadurai HC, et al. Increased detection of cystic-fibrosis-related diabetes in Australia. Arch Dis Child 2011; 96: 82326. 36 Bismuth E, Laborde K, Taupin P, et al. Glucose tolerance and insulin secretion, morbidity, and death in patients with cystic fibrosis. J Pediatr 2008; 152: 54045, 545.e1. 37 Ode KL, Frohnert B, Laguna T, et al. Oral glucose tolerance testing in children with cystic fibrosis. Pediatr Diabetes 2010; 11: 48792. 38 Swartz LM, Laffel LM. A teenage girl with cystic fibrosis-related diabetes, diabetic ketoacidosis, and cerebral edema. Pediatr Diabetes 2008; 9: 42630. 39 Battezzati A, Battezzati PM, Costantini D, et al. Spontaneous hypoglycemia in patients with cystic fibrosis. Eur J Endocrinol 2007; 156: 36976. 40 Radike K, Molz K, Holl RW, Poeter B, Hebestreit H, Ballmann M. Prognostic relevance of hypoglycemia following an oral glucose challenge for cystic fibrosis-related diabetes. Diabetes Care 2011; 34: e43. 41 Andersen HU, Lanng S, Pressler T, Laugesen CS, Mathiesen ER. Cystic fibrosis-related diabetes: the presence of microvascular diabetes complications. Diabetes Care 2006; 29: 266063. 42 Schwarzenberg SJ, Thomas W, Olsen TW, et al. Microvascular complications in cystic fibrosis-related diabetes. Diabetes Care 2007; 30: 105661. 43 Van den Berg JMW, Morton AM, Kok SW, Pijl H, Conway SP, Heijerman HGM. Microvascular complications in patients with cystic fibrosis-related diabetes (CFRD). J Cyst Fibr 2008; 7: 51519. 44 Lind-Ayres M, Thomas W, Holme B, Mauer M, Caramori ML, Moran A. Microalbuminuria in patients with cystic fibrosis. Diabetes Care 2011; 34: 152628. 45 Alicandro G, Battezzati PM, Battezzati A, et al. Insulin secretion, nutritional status and respiratory function in cystic fibrosis patients with normal glucose tolerance. Clin Nutr 2012; 31: 11823. 46 Brodsky J, Dougherty S, Makani R, Rubenstein RC, Kelly A. Elevation of 1-hour plasma glucose during oral glucose tolerance testing is associated with worse pulmonary function in cystic fibrosis. Diabetes Care 2011; 34: 29295. 47 Suratwala D, Chan JSH, Kelly A, et al. Nocturnal saturation and glucose tolerance in children with cystic fibrosis. Thorax 2011; 66: 57478.
48 Frohnert BI, Ode KL, Moran A, et al. Impaired fasting glucose in cystic fibrosis. Diabetes Care 2010; 33: 266064. 49 Moran A, Brunzell C, Cohen RC, et al. Clinical care guidelines for cystic fibrosis-related diabetes: a position statement of the American Diabetes Association and a clinical practice guideline of the Cystic Fibrosis Foundation, endorsed by the Pediatric Endocrine Society. Diabetes Care 2010; 33: 2697708. 50 ORiordan SMP, Robinson PD, Donaghue KC, Moran A. Management of cystic fibrosis-related diabetes. Pediatr Diabetes 2008; 9: 33844. 51 Dobson L, Sheldon CD, Hattersley AT. Conventional measures underestimate glycaemia in cystic fibrosis patients. Diabet Med 2004; 21: 69196. 52 Martn-Fras M, Lamas Ferreiro A, Colino Alcol E, Alvarez Gmez MA, Yelmo Valverde R, Barrio Castellanos R. Continuous glucose monitoring system in the screening of glucose disorders in cystic fibrosis. An Pediatr (Barc) 2009; 70: 12025. 53 Jefferies C, Solomon M, Perlman K, Sweezey N, Daneman D. Continuous glucose monitoring in adolescents with cystic fibrosis. J Pediatr 2005; 147: 39698. 54 Khammar A, Stremler N, Dubus J-C, Gross G, Sarles J, Reynaud R. Value of continuous glucose monitoring in screening for diabetes in cystic fibrosis. Arch Pediatri 2009; 16: 154046. 55 Moreau F, Weiller MA, Rosner V, et al. Continuous glucose monitoring in cystic fibrosis patients according to the glucose tolerance. Horm Metab Res 2008; 40: 50206. 56 Hameed S, Morton JR, Jaff A, et al. Early glucose abnormalities in cystic fibrosis are preceded by poor weight gain. Diabetes Care 2010; 33: 22126. 57 ORiordan SMP, Hindmarsh P, Hill NR, et al. Validation of continuous glucose monitoring in children and adolescents with cystic fibrosis: a prospective cohort study. Diabetes Care 2009; 32: 102022. 58 Moran A, Brunzell C, Cohen RC, et al. Clinical care guidelines for cystic fibrosis-related diabetes: a position statement of the American Diabetes Association and a clinical practice guideline of the Cystic Fibrosis Foundation, endorsed by the Pediatric Endocrine Society. Diabetes Care 2010; 33: 2697708. 59 Moran A, Pekow P, Grover P, et al. Insulin therapy to improve BMI in cystic fibrosis-related diabetes without fasting hyperglycemia: results of the cystic fibrosis related diabetes therapy trial. Diabetes Care 2009; 32: 178388. 60 Sunni M, Bellin MD, Moran A. Exogenous insulin requirements do not differ between youth and adults with Cystic Fibrosis Related Diabetes. Pediatr Diabetes 2013; published online Jan 28. DOI:10.1111/pedi.12014 (in press). 61 Mozzillo E, Franzese A, Valerio G, et al. One-year glargine treatment can improve the course of lung disease in children and adolescents with cystic fibrosis and early glucose derangements. Pediatr Diabetes 2009; 10: 16267. 62 Dobson L, Hattersley a T, Tiley S, Elworthy S, Oades PJ, Sheldon CD. Clinical improvement in cystic fibrosis with early insulin treatment. Arch Disease Child 2002; 87: 43031. 63 Hameed S, Morton JR, Field PI, et al. Once daily insulin detemir in cystic fibrosis with insulin deficiency. Arch Disease Childhood 2012; 97: 46467. 64 Koloukov S, Zemkov D, Bartoov J, et al. Low-dose insulin therapy in patients with cystic fibrosis and early-stage insulinopenia prevents deterioration of lung function: a 3-year prospective study. J Pediatr Endocrinol Metab 2011; 24: 44954. 65 Bizzarri C, Lucidi V, Ciampalini P, Bella S, Russo B, Cappa M. Clinical effects of early treatment with insulin glargine in patients with cystic fibrosis and impaired glucose tolerance. J Endocrinol Invest 2006; 29: RC14. 66 Minicucci L, Haupt M, Casciaro R, et al. Slow-release insulin in cystic fibrosis patients with glucose intolerance: a randomized clinical trial. Pediatr Diabetes 2012; 13: 197202. 67 OSullivan BP, Baker D, Leung KG, Reed G, Baker SS, Borowitz D. Evolution of pancreatic function during the first year in infants with cystic fibrosis. J Pediatr 2013; 162: 80812.
36
Conference Diary
June 2013
1518 ENDO 201: The Endocrine Societys 95th Annual Meeting and Expo
San Francisco, CA, USA
Website: http://www.endo-society.org/endo2013/
2327 The European Association for the Study of Diabetes 49th Annual Meeting
Barcelona, Spain
Website: http://www.easd2013.com/
October 2013
47 American Society of Bone and Mineral Research Annual Meeting
Baltimore, MD, USA
Website: http://www.asbmr.org/meetings/annualmeeting.aspx
2125 73rd Scientific Sessions of the American Diabetes Association

Chicago, IL, USA
Website: http://professional.diabetes.org/Congress_Display. aspx?TYP=9&CID=91271
July 2013
79 Bristish Society of Neuroendocrinology Annual Meeting
Manchester, UK
Website: http://www.neuroendo.org.uk/
1620 83rd Annual Meeting of the American Thyroid Association

San Juan, Puerto Rico
Website: http://www.thyroid.org/thyroid-events-education-media/83rdannual-meeting-of-the-ata/
November 2013
1116 Obesity Week 2013
Atlanta, GA, USA
Website: http://www.obesityweek.com/
710 ESHRE 29th Annual Meeting

London, UK
Website: http://www.eshre2013.eu/
2226 Society for the Study of Reproduction 46th Annual Meeting

Montreal, QC, Canada
Website: http://www.ssr.org/Meetings.shtml
December 2013
26 IDF 2013 Melbourne World Diabetes Congress
Melbourne, VIC, Australia
Website: http://www.idf.org/worlddiabetescongress/
September 2013
711 37th Annual Meeting of the European Thyroid Association
Leiden, Netherlands
Website: http://www.eurothyroid.com/futureevents.html
1922 9th Joint Meeting of Paediatric Endocrinology

Milan, Italy
Website: http://www.jointmeeting2013.org/
To Advertise in The Lancet Diabetes & Endocrinology Contact: T +44 (0)20 7424 4277 F +44 (0)845 6185 046
lancet-ads@elsevier.com
Your research published worldwide in 8 weeks
The research we publish has the potential to change medical practice. That is why The Lancet Diabetes & Endocrinology is committed to ensuring that breakthrough research is published quickly and disseminated globally. Through our fast-track service, qualified original research is peer-reviewed and published online in advance of the print issue in as little as 8 weeks. A linked commentary by an opinion leader will contextualise your research findings, and a press release to the global media as appropriate will give your research the worldwide attention it deserves. To learn more about our fast-track review and for information on how to submit an article for publication visit www.ees.elsevier.com/thelancetde
In delivering science for better health

The Lancet is a registered trademark of Reed Elsevier Properties S.A., used under licence.
Information for Readers and Advertisers
Press Office
Media Relations Manager Daisy Barton pressoffice@lancet.com T +44 (0)20 7424 4949
Electronic Production
Senior Electronic Production Engineers Colin Brown, Garry Griffin Electronic Production Engineer Dennis Chu
International Advertising Sales

advertisinghssales@elsevier.com Tel: +44 20 7424 4538 North America Account Manager Aileen Rivera a.rivera@elsevier.com
International Reprints Sales

Europe/ROW Senior Business Development Executive Greg Davies gr.davies@elsevier.com North America Reprints Sales Representative Derrick Imasa d.imasa@elsevier.com
Production
Head of Production Dan Lewsley Origination Manager Bente Brattland Art & Logistics Manager Clive Message Production Controller Diandra Coles Senior Production Editors Helen Budd, Matthew Judd, Kate Spencer Production Editors Hilary Lavin, Jules Morgan Illustrators John Higginbottom, Adrian Roots Quality Controller Colin Lake Production Assistant Helen Ng
Journal Office
Lancet Operations Manager Rahila Esposito Senior Administrator Sonia Prescod
Recruitment and Classified advertising

lancet-ads@elsevier.com T +44 (0)1425 462736 F +44 (0)8456 185046
Marketing
Strategic Marketing Manager Nancy Burns
Subscription Customer Service

Customer Service
Eur/ROW T +44 (0)1865 843 077 F +44 (0)1865 843 970 custserv@lancet.com The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK N America T +1 (800) 462-6198 (Toll Free) T +1 (314) 447-8057 (Direct) F +1 (314) 447-8029 (Direct) USLancetCS@elsevier.com The Lancet, 3251 Riverport Ln, Maryland Heights, MO 63043, USA
Printing by Cadmus Professional Communications
: call for papers

We are seeking original papers covering the topics of diabetes, endocrinology, and metabolism for publication in our inaugural September 2013 issue. The journal will consider original research, reviews, expert commentaries, and clinical pictures. Manuscripts in response to this call for papers can be submitted now via our dedicated online system at http://ees.elsevier.com/thelancetde.
* Sign up for email updates at http://www.thelancet.com/diabetes-endocrinology

facebook.com/thelancetmedicaljournal twitter.com/@thelancet THE LANCET is a registered trademark of Reed Elsevier Properties SA, used under licence.

Lancet Diabetes &amp; Endocrinology June 2013

Caricato da

Informazioni sul documento

Descrizione originale:

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Lancet Diabetes &amp; Endocrinology June 2013

Caricato da

Copyright:

Formati disponibili

Special Preview June 2013

Learn more at www.thelancet.com/diabetes-endocrinology, or submit a manuscript at http://ees.elsevier.com/thelancetde.

in delivering science for better health

Special Preview June 2013

Towards a personalised diagnosis of type 2 diabetes

Gut microbe composition and metabolic syndrome

New insights into cystic fibrosis-related diabetes in children

David Marck/Skyline Imaging Ltd

Dont just get published. Get read.

Make a real impact, publish in

in delivering science for better health

Diabetes and endocrinology: extending The Lancet family

www.thelancet.com/diabetes-endocrinology June 2013

Science Photo Library

Towards a personalised diagnosis of type 2 diabetes

David Mack/Science Photo Library

Congenital adrenal hyperplasia: one hundred years of data

Sidney Moulds/Science Photo Library

www.thelancet.com/diabetes-endocrinology June 2013

Subclinical thyroid disease: where is the evidence?

Published Online May 23, 2013 http://dx.doi.org/10.1016/ S2213-8587(13)70030-5

www.thelancet.com/diabetes-endocrinology June 2013

Gut microbe composition and metabolic syndrome

www.thelancet.com/diabetes-endocrinology June 2013

Bifidobacteria colonise the gut

Gary Gaugler/Science Photo Library

www.thelancet.com/diabetes-endocrinology June 2013

Interlinks between sleep and metabolism

www.thelancet.com/diabetes-endocrinology June 2013

Oscar Burriel/Science Photo Library

New hormone stimulates pancreatic -cell proliferation

John Bavosi/Science Photo Library

www.thelancet.com/diabetes-endocrinology June 2013

Profile Jean-Pierre Chanoine: taking paediatric endocrinology global

Published Online January 2013

Talha Khan Burki

www.thelancet.com/diabetes-endocrinology June 2013

Study design and participants

www.thelancet.com/diabetes-endocrinology June 2013

See Online for appendix

Role of the funding source

Insulin sensitivity (mg I2/mmol mU min)

Concentration (g/mL) 15 10 5 Time (years) 0

n=41 195% had fasting and 2 h hyperglycaemia

>72 mmol/L n=20 400% had fasting and 2 h hyperglycaemia

>63% n=65 846% had fasting and 2 h hyperglycaemia

www.thelancet.com/diabetes-endocrinology June 2013

www.thelancet.com/diabetes-endocrinology June 2013

www.thelancet.com/diabetes-endocrinology June 2013

Role of the funding source

See Online for appendix

180 160 140 Number of individuals 120 100 80 60 40 20 0

Salt-wasting Non-salt-wasting Before screening (195085)

Salt-wasting Non-salt-wasting With screening (19862011)

www.thelancet.com/diabetes-endocrinology June 2013

www.thelancet.com/diabetes-endocrinology June 2013

www.thelancet.com/diabetes-endocrinology June 2013

New insights into cystic fibrosis-related diabetes in children

Non-CF-1 Non-CF-2 Non-CF-3

CF-1 CF-2 CF-3

25 Blood glucose (mmol/L)

Plasma Insulin concentration (pmol/L)

Blood glucose concentration (mmol/L)

Lancet Diabetes & Endocrinology June 2013

Lancet Diabetes & Endocrinology June 2013

Special Preview June 2013

www.thelancet.com/diabetes-endocrinology June 2013