Differential diagnosis of type 1 diabetes: which genetic syndromes need to be considered?


  • Timothy Geoffrey Barrett

    Corresponding author
    1. Diabetes Department, Institute of Child Health, Diabetes Department, University of Birmingham, Birmingham B15 2TT, UK
    Search for more papers by this author

Prof. Timothy G Barrett
Diabetes Unit
Birmingham Children’s Hospital
Steelhouse Lane
Birmingham B4 6NH
Tel: (0) 121 333 9267;
fax: (0) 121 333 9272;
e-mail: t.g.barrett@bham.ac.uk


Abstract:  Recently it has become apparent that not all diabetes presenting in childhood is type 1. Increasingly type 2 diabetes, secondary diabetes, maturity onset diabetes of the young, and rare syndromic forms of diabetes such as Wolfram syndrome and Alstrom syndrome have been identified in children. Although individually rare, collectively they make up about 5% of children seen in diabetes clinics. The importance of these syndromes for children lies in the recognition of treatable complications, and for their parents, the possibility of genetic counselling. The scientific importance is enormous as they are experiments of nature that reveal basic mechanisms of insulin and glucose metabolism. We are now able to offer mutation analysis to correlate the clinical pattern to the genotype, and seek novel therapeutic approaches based on the developing knowledge of gene and protein functions. This review focuses on monogenic syndromes of diabetes, particularly where significant advances have been made in our understanding recently. Neonatal diabetes is a specialist field in its own right and is not included, except to discuss Kir6.2 diabetes which may develop in infancy. This review is written for the paediatric diabetes specialist and aims to provide information on the clinical features, natural history, genetics and management of children with diabetes as part of a syndrome. Finally there is information on useful investigations to aid diagnosis.

Children presenting with thirst, polyuria, polydipsia with weight loss, and a raised venous glucose concentration have traditionally been assumed to have type 1 or autoimmune diabetes; they would eventually develop an absolute insulin deficiency and require insulin replacement. Other forms of diabetes in children were considered to be so rare that they did not need to be considered. The clinical practise of diabetes in children was essentially a management issue; there was only one medical therapeutic option (insulin), and the goals were the avoidance of short-term metabolic complications (hypoglycaemia and ketoacidosis) and long-term microvascular and macrovascular complications. Over the past 15 yr or so, the heterogeneity of childhood diabetes has been increasingly recognized. This has occurred against the background of a rising prevalence of type 1 diabetes mellitus (DM) . The natural history and monogenetic basis of maturity-onset diabetes of the young (MODY) has been described, inherited diabetes syndromes have been characterized clinically and at the molecular level, and the emergence of obesity-related type 2 diabetes in children is considered pandemic. In addition, cases of secondary diabetes are increasing: posttransplant diabetes is an increasingly recognized complication of solid organ and bone marrow transplants. Risk factors for diabetes include the use of tacrolimus immunosuppression and high-dose corticosteroid therapy. Non-type 1 diabetes in children is rare: in 2000, the UK national survey of clinicians caring for children with diabetes found that only 0.7% of children with diabetes had non-type 1 diabetes, although this is likely to be an underestimate (1). A UK case finding survey of newly presenting non-type 1 diabetes over a 13-month period identified 168 cases; 40% had type 2 diabetes, 22% had secondary diabetes, 10% had MODY, and 10% had syndromic diabetes (20% remained unclassified) (2). It is important to distinguish these entities and to make an accurate diagnosis as non-type 1 diabetes can have marked differences in treatment and complications from type 1 DM. There is also diagnostic confusion between diabetes syndromes (e.g. the obesity syndromes Alstrom and Bardet–Biedl). Most importantly, the differing inheritance patterns of syndromes can require differences in the approach to genetic counselling. The purpose of this review is to outline some syndromes for which there have been notable recent advances in our understanding and to provide a clinical update and resource for diagnosis and management. The conditions can be divided broadly into defects of insulin secretion (β-cell defects or apoptosis) and defects of insulin action (insulin resistance) (Table 1 and 2).

Table 1.  Differential diagnosis of the main diabetes syndromes
Presumed mechanismConditionGene (s)Other features
  1. MODY, maturity-onset diabetes of the young; TRMA, thiamine responsive megaloblastic anaemia.

Insulin secretion defectsMODY-3HNF-1αAutosomal dominant family history of diabetes, sulphonylurea responsive
MODY-2GlucokinaseAutosomal dominant family history of diabetes, fasting hyperglycaemia 5.5–8 mmol/L
MODY-5HNF-1βRenal cysts
Kir6.2KCJN11Infancy-onset insulin dependent
WolframWFS-1Optic atrophy, deafness, diabetes insipidus, neuropathic bladder, ataxia, neurodegeneration
Wolcott–RallisonEIF2AK3Acute hepatic failure, spondylo-epiphyseal dysplasia, developmental delay, renal impairment
TRMASLC19A2Megaloblastic anaemia, deafness under 3 yr
MitochondrialtRNALEUDeafness, epilepsy, maternal inheritance
Insulin resistance defectsDonohue, Rabson–MendenhallInsulin receptorIntra-uterine growth retardation, hirsutism, reduced subcutaneous fat
AlstromALMS1Acute cardiomyopathy, retinal dystrophy, obesity
Bardet–BiedlBBS1-11Obesity, retinal dystrophy, polydactyly, renal impairment
LipodystrophyBSCL1, AGPAT2Absent subcutaneous fat, virilization, acanthosis nigricans
Table 2.  Presentations of diabetes syndromes and suggested investigations
PresentationDiabetes typeInvestigations
  1. HNF, hepatocyte nuclear factor; MODY, maturity-onset diabetes of the young; TRMA, thiamine responsive megaloblastic anaemia.

Autosomal dominant family historyMODYGenetic analysis
Fasting glucose >5.5 mmol/LGlucokinase 
Fasting glucose <5.5 mmol/LHNF-1α 
With renal cystsHNF-1β 
Infancy-onset diabetesKir6.2Genetic analysis
Diabetes and deafness maternal family historyMitochondrialGenetic analysis
Diabetes and optic atrophyWolframAudiogram, screen for diabetes insipidus, magnetic resonance imaging for brain atrophy
Diabetes and acute hepatic failureWolcott–RallisonX-ray epiphyses, check liver and renal function, developmental assessment
Diabetes, megaloblastic or sideroblastic anaemia, deafness under 3 yrTRMABone marrow aspirate, ophthalmic assessment, trial of thiamine
Diabetes in slim teenager with hirsutism and acanthosisInsulin receptor mutationFasting insulin, C-peptide, lipids
Acute cardiomyopathy without diabetes, early-onset obesityAlstromFasting insulin, C-peptide, ophthalmic assessment, audiogram
Infancy-onset obesity, polydactylyBardet–BiedlOphthalmic assessment, audiogram, renal function, fasting insulin, C-peptide
Severe loss of subcutaneous fat, acanthosis nigricansLipodystrophyFasting insulin, C-peptide, lipids

Defects in insulin secretion

Maturity-onset diabetes of the young

MODY comprises a heterogeneous group of disorders caused by defects in single genes. MODY is characterized by β-cell dysfunction, young onset (<25 yr), and autosomal dominant inheritance (3) and should be considered in non-obese patients with diabetes developing before 25 yr of age and a two or three generation history of diabetes in one side of the family. MODY accounted for 10% of all non-type 1 diabetes in the UK case finding survey (2).

The first MODY gene was found in 1992, and there are now seven known genes in which defects have been shown to cause MODY (3). These genes encode for glucokinase, which acts as a glucose sensor, carboxyl ester lipase (4), and five transcription factors [hepatocyte nuclear factor (HNF)-1α, HNF-1β, HNF-4α, insulin promoter factor-1 (IPF-1), and neurogenic differentiation-1 (NEURO-D1)]. Transcription factors are proteins that bind to promoter regions of genes and activate transcription into messenger RNA. They initiate the production of proteins, which are important in the development of the pancreas. Each genotype produces a unique phenotype (5). In the UK, HNF-1α mutations (HNF-1α; MODY-3) account for 63% of MODY, glucokinase mutations (MODY-2) account for 20%, and other known genes <10%. Ten to fifteen percent of MODY patients do not have a known MODY mutation and are termed MODY-X.

Heterozygous inactivating glucokinase mutations produce mild fasting hyperglycaemia (5.5–8 mmol/L), which is usually non-progressive (6). This is because of a glucose-sensing defect: these children have their resting glucose homeostatic mechanisms reset at a higher level. Insulin secretion can achieve the same maximum, but for any given glucose level, the insulin response is set lower. Patients do not usually show symptoms, so may be diagnosed coincidentally at any age during routine testing for other reasons. If tested by an oral glucose tolerance test (OGTT), their glucose levels rapidly fall to the baseline level (their fasting glucose is usually more than 5.5 mmol/L, but the increment between their 2-h value and their fasting value is usually less than 3.5 mmol/L) (7). The haemoglobin A1c (HbA1c) level is at or slightly above the upper limit of normal. Affected children are sometimes misdiagnosed as type 1 diabetes and treated with insulin. Clues to this diagnosis include the child with surprisingly good glycaemic control both on HbA1c and capillary glucose monitoring, requiring <0.5 u/kg/d of insulin outside the honeymoon period, and with no tendency to ketonuria with hyperglycaemia. Treatment is almost always unnecessary for children with heterozygous glucokinase mutations. Interestingly, homozygous inactivating glucokinase mutations are a very rare cause of permanent neonatal diabetes.

HNF-1αmutation carriers usually have normal blood glucose levels until the teenage years when diabetes presents. If a child with an HNF-1αmutation is born to a mother who is diabetic, they may present with diabetes considerably earlier. HNF-1α patients have an inexorable β-cell impairment; the progressive nature of the diabetes means that many patients will eventually require insulin and are at risk of microvascular complications. The initial abnormality is postprandial hyperglycaemia as a result of an inadequate rise in insulin in response to a glucose challenge. If these patients are tested by an OGTT, the 2-h glucose level may be more than 6 mmol/L higher than the fasting level, even when fasting glucose levels are in the normal range (7).

Children with HNF-1αmutations may respond to diet, and HNF-1α patients are particularly sensitive to sulphonylureas (8). The response to sulphonylureas and insulin sensitivity in these patients is similar to those seen in non-diabetics. The efficacy of sulphonylurea treatment in HNF-1α MODY was shown in a randomized crossover trial, comparing gliclazide and metformin between patients with HNF-1αmutations and type 2 diabetes (9). In type 2 diabetes, the glycaemic response was similar with either metformin or gliclazide. In contrast, in HNF-1α patients, there was a fivefold greater response to gliclazide than metformin, and the response to sulphonylurea was fourfold greater than that in type 2 patients. Patients with HNF-1α may be initially managed on diet alone; low-dose sulphonylurea (e.g., gliclazide 20 mg, once daily) can be added when HbA1c levels rise. Some of these patients eventually require insulin treatment as pancreatic impairment progresses. The reasons for this inexorable β-cell impairment are not yet understood.

Patients with HNF-1βmutations have associated renal cystic disease. This can vary from single kidney involvement to severe urogenital disease. Renal disease usually precedes the development of diabetes. Other extrapancreatic manifestations include uterine and genital anomalies, abnormal liver function tests, gout, hyperuricaemia, and possibly abnormalities of the gastrointestinal system, such as pyloric stenosis (10). Diabetes is rare below 10 yr of age. The mechanism of diabetes is a combination of hepatic insulin resistance and β-cell dysfunction. The deterioration of β-cell function is more rapid than with HNF-1α diabetes, diabetic ketoacidosis may occur, and patients are not sensitive to sulphonylureas. Many patients eventually require insulin treatment.

Kir6.2 infancy-onset diabetes

Heterozygous, activating mutations in the KCJN11 gene, encoding the Kir6.2 subunit of the inwardly rectifying potassium channel, have recently been shown to be the most common cause of diabetes presenting under 6 months of age (11) (Fig. 1). The β-cell potassium channel is intimately involved in the regulation of insulin secretion. It is an octomeric complex composed of four inwardly rectifying potassium channels (Kir6.2) and four sulphonylurea receptor subunits (SUR1s). Kir6.2 binds ATP to close the channel, and magnesium nucleotides bind to SUR1 causing activation of the channel. In the fed state, high intracellular glucose generates ATP through glycolysis and closes the potassium channel, leading to cell membrane depolarization and insulin release. Fasting lowers the intracellular ATP:ADP ratio, resulting in opening of the potassium channel and inhibition of insulin secretion. Both Kir6.2 and SUR1 are vital for the correct regulation of insulin secretion, and inactivating mutations in the genes encoding both have been shown to cause autosomal recessive persistent hyperinsulinaemic hypoglycaemia of infancy, characterized by complete loss of glucose-stimulated insulin release. Activating mutations in KCJN11 cause diabetes by making the KATP channel less likely to close in the presence of ATP. This results in reduced potassium efflux, which, in turn, hyperpolarizes the β-cell membrane, reducing insulin secretion. Children usually present with insulin-dependent diabetes under 6 months of age, with ketoacidosis, and no measurable C-peptide. About 25% of children with transient neonatal diabetes also have activating mutations in KCJN11, and rarely, children may present after 6 months without a history of neonatal diabetes (12).

Figure 1.

Infancy-onset diabetes because of activating KCJN11 mutation. Potassium channel fails to close in response to increased intracellular ATP. Sulphonylureas bind to the sulphonylurea receptor 1 on the potassium channel, closing it via an ATP-independent route and allowing insulin secretion in response to glucose. ATP, adenosine triphosphate; GCK, glucokinase; GLUT2, glucose transporter.

These children secrete insulin in response to sulphonylureas; the drugs bind to the SUR 1 on the KATP channel and close the channel via an ATP-independent route. A recent follow-up of 49 patients with diabetes because of KCJN11 mutations has shown that 90% of children have successfully switched from insulin to sulphonylureas, with an improvement in glycaemic control that was sustained at 1-yr follow-up (13). Children presenting with infancy-onset diabetes almost all require insulin at diagnosis; genetic testing is relevant for all children diagnosed under 6 months. Additional clues to a KCJN11 mutation include a family history of early-onset diabetes, although some children present with de novo mutations. Once the infant is stabilized on insulin and the diagnosis of Kir6.2 diabetes is made, a trial of sulphonylurea therapy can be attempted. Sulphonylurea preparations are available as suspensions and require gradual introduction while weaning off insulin, so as to avoid hypoglycaemia.

Wolfram syndrome

Wolfram syndrome [also known as diabetes insipidus, diabetes mellitus, optic atrophy, and deafness (DIDMOAD)] is an autosomal recessive syndrome in which the association of insulin-dependent diabetes with progressive optic atrophy under 16 yr of age is diagnostic (14) (Fig. 2). Other features include bilateral progressive sensorineural deafness, cranial diabetes insipidus, autonomic nervous system dysfunction leading to neuropathic bladder, and other signs of neurodegeneration including cerebellar ataxia, myoclonic epilepsy, and brainstem atrophy. The complete phenotype is seen in about 75% of patients. The diabetes is non-autoimmune and insulin deficient and presents at a median age of 6 yr. Patients usually require insulin treatment from diagnosis. The median age of death in Wolfram syndrome is 30 yr, and development of the full phenotype is seen with increasing age.

Figure 2.

Natural history of Wolfram syndrome. D, deafness; DI, diabetes insipidus; DM, diabetes mellitus; Neuro, neurodegeneration; OA, optic atrophy; renal, neuropathic bladder.

The syndrome is caused by loss of function mutations in the Wolfram (WFS-1) gene (15), encoding Wolframin, an endoplasmic reticulum (ER) membrane protein. Pancreatic β-cells are some of the most susceptible cells to ER stress, and ER stress-mediated apoptosis causes diabetes in Wolfram syndrome. Mutations are distributed throughout the entire gene, but with one loss of function mutation at the carboxy-terminal end (c.2648-2651delTCTT; F883fsX950) recurring in white European populations (16). Mutations in the WFS-1 are present in 90% of patients. In general, patients with one or more inactivating mutations will have a severe phenotype, whereas homozygosity or compound heterozygosity for missense mutations is more likely to result in a mild phenotype (17). The most important part of management is in supporting the family with this devastating diagnosis and teaching the child low skills, while there is useful vision remaining.

Wolcott–Rallison syndrome

Wolcott–Rallison syndrome is a rare autosomal recessive condition characterized by extremely early-onset diabetes, epiphyseal dysplasia, renal impairment, acute hepatic failure, and developmental delay. Non-autoimmune insulin-deficient diabetes presents in infancy but has been reported to range from 2 wk to 30 months, with a mean age in our series of 10 months (18). Infancy-onset diabetes mellitus and generalized epiphyseal dysplasia are the characteristic features of this syndrome. Hepatic abnormality is the next most consistent feature. Most children develop acute liver failure with each intercurrent illness. Liver biopsies show cholestasis with steatosis and fibrosis. Global developmental delay has been reported frequently, and two of our three patients had developmental regression after episodes of acute hepatic failure. Hepatic and renal failure are the commonest causes of death in this syndrome, with children not usually surviving beyond childhood.

The gene for Wolcott–Rallison has been isolated as EIF2AK3, a regulator of protein translation (19). EIF2AK3 codes for eukaryotic initiation factor 2α kinase 3 (EIF2AK3) and is also known as the pancreatic endoplasmic reticulum kinase (PERK). PERK regulates the cellular response to ER stress.

Any infant with diabetes mellitus should be screened for epiphyseal dysplasia so that the diagnosis of Wolcott–Rallison syndrome can be made early. Each intercurrent illness can potentially be complicated by acute liver failure and/or renal failure. These children should have a clear care plan prepared so that there is no delay in initiating treatment for acute hepatic failure. Together with implications for management planning, early diagnosis has a role for prenatal diagnosis for affected families.

Thiamine responsive megaloblastic anaemia/Roger’s syndrome

Thiamine responsive megaloblastic anaemia (TRMA) with diabetes and deafness is an autosomal recessive disorder because of defective thiamine transport. It was first described in an 11-yr-old girl in whom treatment with thiamine reduced her insulin requirement by 50%, making this a unique, vitamin-dependent form of diabetes (20). In addition to the cardinal features of megaloblastic anaemia with ringed sideroblasts, diabetes, and sensorineural deafness, other reported features include cardiac and visual defects and trilineage myelodysplasia. Diabetes mellitus usually presents coincidentally with glycosuria or after a short history of osmotic symptoms associated with intercurrent infection. Most children require less than 0.5 u/kg/d of insulin during the first decade of life with good glycaemic control. Insulin requirements increase from the onset of puberty. All reported children have been negative for antibodies to islet cells and glutamic acid decarboxylase (GAD). Children may show an initial improvement in glycated haemoglobin and reduction in insulin dose with oral thiamine treatment, but this improvement is usually lost during puberty when insulin requirements increase. There is little evidence that thiamine treatment above 25 mg/d offers any additional benefit and no evidence that thiamine treatment altered sensorineural deafness in our families. Two Italian children were followed to puberty (21): they were treated with thiamine in childhood, but at puberty, glycaemic control deteriorated and oral hypoglycaemic agents were added. Both patients showed reduced first-phase insulin response and reduced insulin secretory capacity at the end of puberty. It remains unclear whether a lipophilic preparation of thiamine with enhanced bioavailability (benzoyloxy-ethyl-thiamine; Bioindustria, Novi Ligure, Italy) might increase intracellular thiamine levels.

The cause of TRMA syndrome is because of mutations in the SLC19A2 gene (22). SLC19A2 encodes a membrane-bound thiamine transporter protein known as THTR-1. Thiamine uptake is thought to take place via two pathways: active transport by a saturable, high-affinity carrier and passive uptake by a low-affinity carrier. Once taken up by cells, thiamine is converted into its active form, thiamine pyrophosphate, which is required for the proper functioning of four enzymes: the pentose phosphate shunt enzyme transketolase and three multienzyme complexes involved in oxidative decarboxylation (pyruvate dehydrogenase, α-ketoglutarate dehydrogenase, and branched chain acid dehydrogenase). Transketolase is the rate-limiting enzyme in the non-oxidative pentose shunt pathway for ribose synthesis, vital for nucleic acid production. Mutations in the high-affinity thiamine transporter lead directly to an increased rate of apoptosis in cells that have a high rate of nucleic acid synthesis – that is, those with high turnover such as the marrow or those with high translation rates such as the secretory β-cells.

Mitochondrial mutations

In 1992, a mutation in mitochondrial DNA (mtDNA) at nucleotide position 3243 was found to segregate with diabetes (23). This has proved to be the commonest mitochondrial mutation causing diabetes, but overall mitochondrial mutations are a rare cause of childhood diabetes. Factors that are suggestive of mitochondrial diabetes are maternal inheritance (as mitochondria are inherited exclusively from the maternal side) and associated deafness, myopathy, or neurological deficits.

Mitochondria are the intracellular organelles responsible for the generation of energy by the process of oxidative phosphorylation (OXPHOS). The human mtDNA is a 16 565-nucleotide, double-stranded closed circular molecule, which encodes 13 genes of OXPHOS together with the structural ribosomal RNA’s and transfer RNA’s needed for their expression. Mitochondrial diseases may result from either mutations of mtDNA or nuclear DNA as genes from both genomes are needed to encode subunits of OXPHOS.

The commonest cause of diabetes because of mitochondrial disorders is a point mutation at nucleotide pair (np) 3243. The mutation has been described in families with a mild form of type 2 DM of adult onset and sensorineural deafness. The diabetes associated with np3243 is usually diagnosed in the third to fifth decades of life but may present between the late teens and mid-80s. Hyperglycaemia is often mild and controlled by diet at diagnosis but tends to be progressive. The underlying pathogenesis is thought to be β-cell failure with evidence of reduced insulin secretion in the presence of normal insulin sensitivity. The same mitochondrial mutation may result in the less common but severe mitochondrial encephalopathy with lactic acidosis and stoke-like episodes (MELAS) phenotype. MELAS present in childhood with short stature develop bilateral deafness in their teens and then diabetes, seizures, stroke-like episodes, and an encephalopathy in their third or fourth decade. Trials of specific treatments for mitochondrial diabetes have been unpromising, and patients usually require insulin from diagnosis.

Defects in insulin resistance

Insulin receptor mutations

Insulin receptor defects produce a spectrum of insulin-resistant phenotypes (24), ranging from the severe Donohue syndrome to the relatively mild insulin resistance type A syndrome. Patients with Donohue syndrome (previously known as leprechaunism) have severe intra-uterine growth restriction and hyperinsulinaemia with abnormal glucose homeostasis, characterized by fasting hypoglycaemia and postprandial hyperglycaemia, facial dysmorphism, reduced subcutaneous fat, and protuberant abdomen. Patients have <10% of wild-type insulin binding and either premature stop mutations or mutations in the extracellular domain of the receptor. These defects are associated with death occurring before 2 yr of age. In contrast, patients with Rabson–Mendenhall syndrome (RMS) have mutations in the intracellular domain of the receptor and insulin binding at levels of up to 25% of normal. RMS can be distinguished from Donohue syndrome by the presence of dysplastic gums and teeth, thickened nails, and hirsutism. Children with RMS develop progressive ketoacidotic diabetes, most dying before adolescence. Type A syndrome in contrast affects lean, predominantly female adolescents with severe insulin resistance, hyperandrogenism, and acanthosis nigricans. Female patients have androgen excess ranging from mild hirsutism to frank virilization. Diabetes or impaired glucose tolerance is one of the later manifestations of the disorder. There is some evidence that metformin can be of use in type A insulin resistance.

Alstrom syndrome

Alstrom syndrome is the autosomal recessive inherited association of retinal dystrophy, sensorineural deafness, and obesity (25). Other features include diabetes mellitus, hypertriglyceridaemia, cardiomyopathy, hepatic disease, and urological abnormalities. The prevalence is less than 1:100 000.

Progressive cone–rod dystrophy causes nystagmus in infancy and blindness by adulthood; sensorineural deafness presents in childhood. Other patients present with dilated cardiomyopathy in infancy, and this can recur or present de novo later. Obesity also develops in infancy, with diabetes mellitus in 70% of subjects in the second to third decade. Euglycemic hyperinsulinaemic clamps confirm marked insulin resistance. The hyperinsulinaemia is associated with acanthosis nigricans, hypertension, and hypertriglyceridaemia, which can be severe and lead to pancreatitis (26). Liver function abnormalities and non-alcoholic steatohepatitis are common. Other endocrine complications include hypogonadotropic hypogonadism in male patients and hypothyroidism. Urological dysfunction because of detrusor–urethral dyssynergia occurs in 50% of patients in the second to third decade.

Cardiomyopathy accounts for most mortality in the first three decades of life; in the UK, death from cardiac failure occurred in 27% of Alstrom patients in the first or second decade (Alstrom syndrome UK; unpublished data). Renal failure is the commonest cause of death in older patients. Late fibrotic changes have been described at postmortem in several organs including the kidneys, liver, and lungs.

Mutations in the gene ALMS1 on chromosome 2p13.1 were recently identified in patients (27). ALMS1 consists of 23 exons encompassing over 224 kb of genomic DNA, encoding a polypeptide of 4169 amino acids with a predicted molecular mass of 461.2 kDa. The ALMS1 protein is of unknown function, is widely expressed in human and mouse tissues, and localizes to centrosomes and the base of cilia. The basic pathophysiology is thought to involve intracellular trafficking. Management includes encouragement of healthy eating and exercise within the constraints imposed by the sensory deficits; metformin has been of benefit in children if tolerated. Insulin is effective to improve glycaemic control. Raised triglycerides are frequently seen in children and may respond to fibrates or nicotinic acid. Regular echocardiograms have been recommended to detect signs of impending cardiomyopathy, and ace inhibitors have been used to treat the hypertension.

Bardet–Biedl syndrome

The key manifestations of Bardet–Biedl syndrome (BBS) are rod–cone dystrophy (atypical retinitis pigmentosa), postaxial polydactyly, central obesity, mental retardation, hypogonadism, and renal dysfunction. Thirty percent of patients develop insulin-resistant diabetes (28). The syndrome is linked to 11 loci, and inheritance seems to be variable with some pedigrees showing recessive and some multigenic inheritance.

The prevalence of Bardet–Biedl syndrome is about 1 in 125 000, and criteria for diagnosis have been published (29). In this large survey of 109 BBS patients in the UK and their families, the average age at diagnosis was 9 yr, although parents first noticed abnormalities in their children at a mean age of 3 yr. Obesity only began to develop at around 2–3 yr, and retinal degeneration did not become apparent until a mean age of 8.5 yr. Diabetes tended to present in the teenage years or adulthood and showed insulin resistance. Six genes are characterized in BBS, although their functions are still unknown. For the BBS6 locus, positional cloning identified the MKKS gene, which codes for a chaperone protein. Mutations identified in MKKS result in a shortened chaperone protein and are present in 5–7% of BBS cases; however, the links between MKKS, its eventual target proteins, and the BBS clinical traits are largely unknown. A newly identified locus, BBS10, has recently been found to code for C12orf58, a vertebrate-specific chaperone-like protein, and was found to be mutated in 20% of the populations examined from various ethnic backgrounds. Functional studies performed in single-cell organisms have shown that certain BBS genes are specific to ciliated cells. Ciliated cells have a role in mammalian development, contributing to right/left asymmetry, thus enabling the organs (e.g., heart, liver, and lungs) to be correctly positioned within the biological system. Dysfunction in processes affecting ciliated cells may contribute to the alterations in pigmentary epithelia and structural anomalies noted in certain organs in patients with BBS; however, the relationship between cilia and obesity remains enigmatic. Preliminary findings suggest that the BBS genes, like the gene for Alstrom syndrome, may play an important role in intracellular signalling.

Berardinelli–Seip congenital lipodystrophy

Berardinelli–Seip congenital lipodystrophy (BSCL) is a rare disorder characterized by congenital absence of subcutaneous and visceral fat with severe insulin resistance that progresses to diabetes in early adolescence (30). Other features of the condition are acanthosis nigricans, hypertriglyceridaemia, fatty liver, virilization, and cardiomyopathy. The condition is autosomal recessive; two genes have been isolated by positional cloning, which account for 82% of cases.

The first, BSCL1 on chromosome 9q34, encodes 1-acylglycerol-3-phosphate-O-acyltransferase 2 (AGPAT2). AGPAT2 catalyses an important step in the triacylglycerol pathway. AGPAT2 regulates conversion of lysophosphatidic acid to phosphatidic acid. Phosphatidic acid is the substrate for phosphatidylinositol production. Phosphatidylinositol is an important messenger in insulin signalling pathways, and so disruption of its production may be the cause of insulin resistance in BSCL.

The second locus, BSCL2 on 11q13, maps to a gene Gng31g, which encodes a protein, seipin, of unknown function. Unlike AGPAT2, Gng31g is widely expressed in the brain. Patients with Gng31g mutations have learning difficulties (80%), are three times more likely to suffer from cardiomyopathy, have diabetes with onset at a significantly earlier age, and may be at risk of early death when compared with patients with mutations in AGPAT2. There is poor genotype to phenotype correlation in BSCL, suggesting that alternative environmental or genetic influences may be important.


The vast majority of children presenting with diabetes will have type 1 diabetes. The safest form of treatment for a newly presenting diabetic child is insulin, particularly if the child is unwell or has ketonuria. The well child with mild symptoms of diabetes is more problematic, but the overweight child from an ethnic minority is likely to have type 2 diabetes, and the slim white child is likely to have MODY; however, there are non-obese, white type 2 patients and ethnic minority MODY patients. There are clues to a diagnosis of a diabetes syndrome in the history and examination; a family history of diabetes is important, particularly if there is a clear maternal history of diabetes and deafness or epilepsy (mitochondrial) or autosomal dominant history (MODY). Children with autosomal recessive syndromes often have parents who are related and inherit a homozygous mutation by descent from a common ancestor. A history of previous early childhood deaths or miscarriages is relevant, particularly to autosomal recessive disease. A history of rapid infancy-onset weight gain, removal of accessory digits, or cardiomyopathy suggests an obesity syndrome. Pointers in the examination include evidence of sensorineural hearing loss or vision defects or developmental delay. Useful investigations include autoantibodies to GAD, islet cells, and insulin to exclude type 1 diabetes, audiogram and visual evoked responses, and fasting insulin and C-peptide to identify hyperinsulinaemia. Other more specialized investigations to consider include an echocardiogram (Alstrom), bone marrow aspirate (TRMA), and skeletal survey (Wolcott–Rallison). Genetic testing for most of the conditions described above is now available with a useful resource on the International Society for Paediatric and Adolescent Diabetes (ISPAD) website (http://www.ispad.org/). We now have the diagnostic tools to accurately characterize rarer forms of diabetes in children; meticulous phenotyping is likely to identify further genetic subtypes in future. Only by understanding the biochemical pathways that these syndromes illustrate will we be able to develop specific treatments to help individual families.

Conflicts of interest

The author has declared no conflicts of interest.