Background

Type 2 diabetes (T2DM) in young people is an aggressive disease with a greater risk of complications leading to increased morbidity and mortality during the most productive years of life [1]. Evidence and experience are limited in this age group, with much of the data coming from observational studies such as SEARCH [2], the TODAY randomised control study and more recently GLP1 drug studies [3].

These guidelines aim to improve the care of children and young people (CYP) in the UK with type 2 diabetes. Most paediatric diabetes multi-disciplinary teams have good experience in managing type 1 diabetes and complex conditions yet relatively little experience managing type 2 diabetes and associated co-morbidities.

Methods

The Association of Children’s Diabetes Clinicians (ACDC) and National Type 2 Diabetes Working Group co-established a guideline development group, which included a multidisciplinary group of health professionals including general paediatricians, paediatric endocrinologists, diabetes specialist nurses, dietitians and psychologists. The group members agreed on the scope and subdivided topics based on areas of interest and expertise. The subgroups met independently to review the evidence base. The whole group met via teleconference over 1 year to review the evidence and make recommendations. Grading was undertaken using GRADE (Grading Of Recommendations, Assessment, Development and Evaluation).

The guideline was reviewed and endorsed by the Association of British Clinical Diabetologists (ABCD), ACDC and the British Society of Paediatric Endocrinology and Diabetes (BSPED) (Table 1).

Table 1 Summary of the recommendations with GRADE methodology
Full size table

Results and discussion

Diagnosis and investigations

The main risk factors for developing T2DM are excess weight, first or second-degree relative with T2DM, maternal gestational diabetes, high-risk race/ethnicity and insulin resistance. Insulin resistance is hard to measure but associated with acanthosis nigricans, early poor growth (small for gestational age) and other associated comorbidities such as hypertension, hyperlipidaemia and polycystic ovary syndrome (PCOS).

We recommend HbA1c testing as the primary test for screening for type 2 diabetes. The choice of screening test is controversial with HbA1c increasingly recommended over the gold standard oral glucose tolerance test despite limited evidence in CYP. HbA1c does not require a fasting sample and is a predictor of vascular disease across a wider glycaemic range than just the diabetic one. However, it lacks sensitivity and would miss some people with diabetes. Testing fasting blood glucose in addition to HbA1c may diagnose some additional patients but may not be pragmatic in those unlikely to return for a fasted sample. The oral glucose tolerance test is more sensitive, but is inconvenient, more costly, has imperfect reproducibility and is less popular, contributing to poorer uptake [38, 39].

Identification of diabetes type can be challenging with all types having overlapping features, particularly in those with excess weight [1]. Detection of diabetes antibodies identifies those at the highest risk of developing auto-immune driven loss of insulin production and can aid with the correct identification of diabetes subtype. There is insufficient evidence for the use of insulin or C-peptide levels at diagnosis.

Glycaemic targets

Intensive glucose control in adults improved HbA1c and any diabetes-related end point but not mortality [40, 41]. No similar robust studies exist for CYP.

TODAY showed that an HbA1c of 6.3% (45 mmol/mol) or more after initiation of metformin was a predictor of eventual loss of glycaemic control irrespective of treatment arm (defined as HbA1c > 8% (64 mmol/mol)). with every 0.1% increase in HbA1c increasing risk by 16%, with a median time of approximately 11 months to loss of control [42].

We recommend that HbA1c is tested every 3 months with an overall target of < 6.5% (48 mmol/mol) which should be individualised based on circumstances.

Self-monitoring of blood glucose (SMBG)

There is no high-quality evidence that self-monitoring of blood glucose (SMBG) in CYP improves outcomes or that more intense monitoring is better. Overall adherence with twice daily SMBG was low in TODAY (59% initially, < 50% by 12 m) with greater use associated with lower HbA1c [43]. Similar findings were seen in those using insulin in SEARCH [44]. No association was found between SMBG and quality of life (QOL) or depression; however, there remains a significant financial and personal burden in monitoring without clear evidence of benefit [43].

We recommend that self-monitoring of blood glucose is undertaken in all patients in line with other international guidelines [45,46,47]. Monitoring can guide management where the diabetes subtype is not clear and where disease progression is more aggressive or rapid. Frequency should be individually tailored based on hypoglycaemia risk, availability, burden and need for treatment adjustment.

CGM and FGM

There is inadequate evidence to guide the use of continuous glucose monitoring (CGM) or flash glucose monitoring (FGM) in CYP. Inconsistent results are seen in adults [48, 49] and there is a lack of data in CYP with type 2 diabetes. Further research is needed to identify which patients may benefit from the use of CGM and FGM and how they should be used to maximise patient benefit and cost-effectiveness.

There may be a role for the use of CGM/FGM in certain circumstances such as learning difficulties, insulin usage or as a short-term intervention during treatment intensification and education [46].

Structured education

Diabetes-structured education programmes first started in the 1970s. The focus has changed over time from a knowledge-based approach to one supporting self-empowerment. Meta-analysis of structured programmes for CYP with type 1 diabetes found no benefit of structured education versus over informal unstructured education in improving glycaemic control [50]. Similar findings were seen in adults with type 2 diabetes [51].

Two structured education programmes have been developed for CYP with type 2 diabetes but neither have undergone RCT evaluation: the TODAY Standard Education Diabetes Education (TSDE) and iCAN [52, 53]. The iCAN programme is a bespoke group intervention delivered over four 2-h workshops with a focus on food, activity and emotional well-being [53].

Multidisciplinary team (MDT) approach

Multi-disciplinary diabetes team structure and approach recommendations are based on observational studies and expert opinion [45, 54, 55]. We recommend that CYP should be managed by multi-disciplinary secondary care teams with close integration with primary care and collaboration with adult diabetes teams. Teams should include dietitians, paediatric nurses, diabetes educators, psychologists, social workers (sometimes called community/cultural workers) and medical doctors. Multi-agency working is important, including links with youth, family and social workers, and schools.

A unified approach with consistent team messaging likely improves outcomes. Telemedicine may improve access but access can be challenging (including digital poverty).

Lifestyle interventions in type 2 diabetes management in children

Lifestyle change is the cornerstone of type 2 diabetes management.

Weight loss targets

Weight loss in adults improves glycaemic control, with clinical benefits seen from 5% loss and further improvements with additional loss [56,57,58]. There is no direct evidence in young people with type 2 diabetes to recommend a target weight loss as no lifestyle RCT has aimed for sufficient energy restriction to result in 5–10% weight loss seen in adult studies.

Good diabetes outcomes are not dependent on weight loss. The highest improvement of glycaemic control in the TODAY study was seen in the group receiving rosiglitazone even though that group had the greatest increase in body mass index (BMI). However, weight loss of > 7% across all treatment groups was associated with small benefits in cardiometabolic risk factors [59].

Dietary modifications

Dietary recommendations are based on healthy eating principles for all young people and families. A healthy balanced diet is considered to be rich in wholegrains, vegetables, fruit, dairy, nuts and seeds and limits fats, oils and sugary foods. Carbohydrates should provide 40–50% of energy requirements, fat < 35% and protein 15–25%. There is little evidence to suggest other macronutrient balances at the population level; however, the ratio of carbohydrates to fat can be individualised if standard advice does not promote weight loss. Some may benefit from lower carbohydrate intake (and higher fat) whilst others benefit from lower fat.

A whole-family approach should be used to enable change across the whole family with individualisation for each family. Support from specialist diabetes dietitians is likely to be needed and should be offered to all families.

Physical activity

Physical activity is likely to have multiple beneficial health benefits however the evidence for improvement in glycaemic control in youth is limited [60]. A 2010 systematic review found no evidence to guide physical activity modification [61] and the addition of intensive lifestyle support in TODAY improved weight but not glycaemic control [62].

Multiple cohorts show that young people with type 2 diabetes undertake insufficient physical activity: in one cohort 55.7% undertook no regular physical activity, [63] and in another only 6.5 min of moderate/vigorous activity was undertaken each day [64].

Most national guidelines recommend 60 min per day of moderate to vigorous exercise and these should also apply to those with type 2 diabetes [65].

Psychological management

Youth with type 2 diabetes have a higher prevalence of moderate or severe depression than those with type 1 (18 vs. 5% in boys and 20 vs. 9% in girls, respectively), with higher mean HbA1c and frequency of emergency department visits associated with depressed mood [66].

Over 25% reported symptoms of disordered eating behaviours in SEARCH and TODAY, such as skipping insulin, vomiting, and using diet pills or laxatives, and these behaviours were associated with poorer glycaemic control in females [67] and more severe obesity, psychological symptoms of disordered eating, and symptoms of depression [68].

Addressing psychological needs should be incorporated within everyday practice. The PedsQL (paediatric quality of life) questionnaire can be used to evaluate and monitor ongoing issues [69]. There is limited research to determine which behavioural interventions are most effective [4] and approaches could include personalised tailored interventions to their interests, health coaching and behaviour therapy and the use of peer counsellors. These are likely to need to address cultural differences and home lifestyles in order to achieve better outcomes [5].

Pharmacotherapy

Initial treatment

Initial treatment of a CYP with obesity and diabetes should take into account that diabetes type is relatively uncertain for the first few weeks, due to overlap in presenting symptoms and a significant number presenting with ketosis [6].

Metformin should be started once any ketosis has resolved, starting at a low dose to minimise side-effects and titrated to a maximum of 2 g per day, or the maximum tolerated dose. Sustained release preparations are available and should be considered if there are GI side effects or compliance issues that could improve with reduced dose frequency [7].

Patients who present with ketosis, polydipsia, polyuria, weight loss, or have HbA1c > 8.5% (69.4 mmol/mol) should also be treated with insulin. This should be with intravenous insulin if unwell (e.g. systemic illness, not tolerating food or drink), and then subcutaneous basal insulin at 0.25–0.5 units/kg/day titrated to a maximum of 1.5 units/kg/day, pending antibody results and improved glycaemia.

There is some evidence that insulin can be successfully weaned off after initially starting it [8]. Basal insulin can be tapered over 2–6 weeks by decreasing the insulin dose by 10–30% each time the metformin is titrated up and further once metformin is at the maximum dose [9]. There is no evidence that continuing insulin will preserve B-cell function [10].

Treatment Intensification

The goal of initial treatment should be to attain an HbA1c of less than 6.5% (48 mmol/mol).

Liraglutide

Failure to achieve goal HbA1c with metformin monotherapy should prompt consideration of second-line treatment with liraglutide. Liraglutide is licensed for those aged 10 years and above with a BMI > 85th centile.

Liraglutide is given as daily subcutaneous injections and can be started at 0.6 mg daily and can be increased with 0.6 mg increments every 1–2 weeks up to 1.8 mg daily based on fasting capillary blood glucose > 6 mmol/L and tolerability.

Side effects are mainly gastrointestinal (nausea and diarrhoea). Acute pancreatitis is a rare and theoretical risk and patients should be counselled and monitored for signs including persistent, severe abdominal pain. Monitoring of pancreatic amylase and lipase should occur at baseline, after the first clinical review and yearly thereafter. If pancreatitis is suspected, liraglutide should be stopped and reported to the medicines and health products regulatory agency (MHRA) via the Yellow Card scheme.

Insulin

Basal insulin therapy is the only remaining licensed treatment option if metformin and liraglutide are insufficient or not tolerated. Insulin resistance is characteristic in CYP going through mid-late puberty and might require higher doses of basal insulin up to 1.5 units/kg/day to achieve glycaemic control. A single daily long-acting insulin analogue (e.g. glargine, detemir or degludec) is preferred as studies show that compliance with insulin can be a challenge in CYP [11, 12]. Higher concentrations of basal insulin (U-300 glargine, U-200 degludec) may be required to avoid large-volume injections that may further diminish medication adherence [46].

Additional meal-time rapid-acting insulin (e.g. aspart) may be needed if basal insulin cannot achieve glycaemic control. Doses should be titrated based on pre- and post-prandial readings and should be done in discussion with a dietitian who is supporting the young person on lifestyle changes, including diet and weight management.

The main side effects of insulin treatment are hypoglycaemia and weight gain. The incidence of hypoglycaemia is low however all patients should be educated on its treatment, including the use of glucagon [42].

SGLT-2 inhibitors

There is insufficient evidence to recommend the routine use of SGLT 2 inhibitors in CYP [13,14,15,16]. They can be considered for post-pubertal youth not achieving adequate control with licensed medications in collaboration with adult diabetologists.

There is a risk of euglycaemic diabetic ketoacidosis(DKA) and patients should be counselled on the symptoms and advised to seek immediate medical advice if these develop. Baseline C-peptide should be measured in those not on insulin to ensure adequate endogenous insulin production to protect against DKA. Treatment should be discontinued if DKA is suspected or confirmed and not restarted unless the cause of DKA is proven to be unrelated.

Other agents

There was insufficient evidence to make recommendations on sulphonylureas, DPP-4 inhibitors and orlistat (Appendix).

Bariatric surgery

Bariatric surgery in adolescents leads to successful weight loss in those with severe obesity (BMI > 35 kg/m2) and demonstrates greater weight loss outcomes when compared with lifestyle and liraglutide [17].

Bariatric surgery is the most effective current treatment available for reducing metabolic comorbidities in adolescents with remission rates of 85% for type 2 diabetes, 85% for hypertension, 75% for dyslipidaemia and 78% for musculoskeletal problems [18]. Young people are more likely to have diabetes remission than adults (86%) despite having similar weight loss [19].

Nutritional deficiencies are common in adolescents with up to 70% exhibiting some form of micronutrient deficiency [18].

The impact on pregnancy in adults is mixed, with some obesity-related problems improved and others worsened [20, 21].

Complications and comorbidities

Glycaemic control improves microvascular risk. However, the increased risk of a major cardiovascular event (MACE) in adults is not reduced by improved glycaemic control. Additional measures including reduction in excess adiposity, smoking prevention, increased physical activity and reduction of hypertension and dyslipidaemia are essential to reduce MACE risk.

Hypertension in youth with type 2 diabetes

Treatment of hypertension in adults improves microvascular and macrovascular outcomes at least as much as improvement in glycaemic control [41]. It would be reasonable to suggest a similar improvement in outcomes could be achieved in CYP.

Blood pressure in children should be interpreted in relation to age, sex and height using appropriate centile charts [22]. Hypertension in children under 13 years of age is defined as a systolic and/or diastolic blood pressure that is greater than the 95th centile on three or more occasions and > 130/80 in those over 13 [23]. Ambulatory blood pressure monitoring can assist in the diagnosis and exclude those with white coat hypertension.

Weight loss, salt reduction and increased physical activity can improve blood pressure. ACE inhibitors should be used where blood pressure is not responsive to lifestyle change. Electrolytes should be measured at baseline and 4–6 weeks and adequate contraception is in place to avoid teratogenic side-effects.

Dyslipidaemia

Dyslipidaemia is an important modifiable cardiovascular disease (CVD) risk factor. The classic lipid profile seen with obesity, insulin resistance and type 2 diabetes is raised triglycerides and decreased high-density lipoprotein cholesterol (HDL-C) levels. Earlier identification and control of dyslipidaemia reduces the risk of atherosclerosis in early adult life. Treatment of children with familial hypercholesterolaemia shows a reduction in subclinical atherosclerosis [51].

Dietary changes can improve dyslipidaemia and should be attempted for 6 months. Statins should be initiated where LDL-cholesterol remains > 3.4 mmol/L. Liver enzymes and CK should be monitored to detect rare liver side-effects and avoidance of pregnancy with adequate contraception is advised to avoid teratogenic side-effects. Fibrates should be used to avoid pancreatitis where triglyceride levels are very elevated.

Non-alcoholic fatty liver disease (NAFLD)

NAFLD is a histopathological spectrum of liver disease, from early benign steatosis to non-alcoholic steatohepatitis (NASH) and cirrhosis. It contributes to the development of T2D, cardiovascular disease and chronic kidney disease [24]. T2D is one of the strongest clinical predictors of disease progression [25,26,27]. Steatosis is present in 25% to 50% of adolescents with type 2 diabetes and more advanced disease is increasingly common.

Screening method choice is controversial, and all are poor at differentiating disease severity. Liver enzymes should be tested at baseline and yearly thereafter. Abnormal liver function tests (LFTs) should not be attributed to NAFLD without exclusion of alternative diagnoses. Alanine transaminase (ALT) can be normal in severe disease and clinicians should have a high index of suspicion [28, 29]. Referral to a hepatologist or gastroenterologist is recommended where liver enzymes remain over 3 times the upper limit of normal despite the treatments listed below [28, 29].

Ultrasound should be performed at diagnosis and up to 3-yearly thereafter depending on risk factors. Ultrasound can identify steatosis but doesn’t adequately identify the fibrosis associated with severe disease.

Improvements in weight, diet, physical activity, insulin resistance and obstructive sleep apnoea all improve biomarkers and disease severity. Optimising vitamin D may minimise fibrosis [30]. There is insufficient evidence to recommend any one dietary approach, with limited benefit seen with low fructose, low fat and low glycaemic index diets [29, 70,71,72,73]. Psychological support (as part of multidisciplinary management) may help to improve clinical outcomes [74].

There is currently no specific pharmacological treatment for paediatric NAFLD. Metformin, vitamin E, antioxidants, fatty acid supplements and probiotics show no clear benefit on histologic outcomes or sustained reduction in ALT in CYP. In adults, pioglitazone has been recommended for advanced liver fibrosis [33,34,35,36, 75,76,77,78,79,80].

Retinopathy

Diabetic retinopathy (DR) is a major cause of visual loss in working age adults and increases the risk of cataracts and glaucoma [37, 81, 82]. It is mainly associated with long-duration diabetes [83,84,85]. TODAY study showed that high HbA1c was associated with an increased risk of developing retinopathy regardless of disease duration [86].

At 20 years post diagnosis nearly all UK adults with type 1 diabetes and 60% of adults with type 2 diabetes have some degree of retinopathy [87].

Severe disease is unlikely before 12 years [37, 88] and UK guidelines recommend screening from this age regardless of disease duration.

Albuminuria and chronic kidney disease

Kidney disease is an important complication of diabetes and is one of the most common causes of chronic kidney disease (CKD). It is characterised by persistently raised albuminuria beyond 3 months with a low estimated glomerular filtration rate (eGFR) [89].

Albuminuria is strongly associated with the progression of CKD and cardiovascular morbidity and mortality [90,91,92,93]. Albuminuria occurs more commonly in type 2 than type 1 diabetes and is likely independent of differences in body mass index and hypertension [2].

A random urine sample is the preferred outpatient screening and correlates with 24-h collection [94, 95]. Micral (dipstick) is the easiest and most cost-effective screening test [96]. An early morning sample should be tested where random urine albumin creatinine ratio (ACR) is > 3 mg/mmol [97] to minimise false-positive results due to hyperglycaemia, exercise, smoking, menstruation, recent intercourse or sample contamination [46]. Urine ACR should be raised on two of three consecutive tests obtained on different days within a 3- to 6-month period before the diagnosis is confirmed [46].

The severity of the disease should be characterised using increased ACR (3–30 mg/mmol) and severely increased ACR (> 30 mg/mmol) rather than the terms microalbuminuria and macroalbuminuria [98].

Other causes of CKD should be considered, especially if there is reduced eGRF without albuminuria or the presence of both retinopathy and albuminuria > 30 mg/mmol creatinine.

Improved blood glucose and blood pressure prevent and slow the progression of nephropathy. Therapeutic options include the use of angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (see the “Hypertension” section) [31, 32, 46, 99, 100].

We recommend referral to specialist renal/kidney care services where GFR < 30 ml/min/1.73 m2, consistent significant albuminuria > 30 mg/mmol, and significant haematuria or structural abnormalities are detected on ultrasound (1B) [46, 89].

Obstructive sleep apnoea (OSA)

Obstructive sleep apnoea (OSA) is a sleep disorder characterized by repetitive episodes of upper-airway obstruction which result in intermittent hypoxemia and transient arousals leading to sleep fragmentation and poor sleep quality. Sleep disturbance and OSA are increasingly recognised as being associated with obesity, insulin resistance in adults and children and type 2 diabetes in adults. Additionally, it is a risk factor for future cardiovascular disease [101,102,103,104,105,106,107].

Obesity, male sex, and advancing age are the strongest risk factors for OSA [108, 109]. It affects an estimated 1–2% of normal children [110, 111] and is more common in obese youth [45].

Evaluate symptoms of obstructive sleep apnoea using questions about snoring, apnoea, nocturia, enuresis, sleep quality, morning headaches and daytime sleepiness at every visit after diagnosis [45, 112]. Questionnaires alone do not provide a high enough sensitivity or specificity and clinicians should have a high index of suspicion [113, 114].

The gold standard diagnostic test is overnight laboratory polysomnography, which quantifies episodes of apnoea and hypopnoea, with an apnoea–hypopnoea index (AHI) of > 5 being diagnostic [103]. Treatment with continuous airway pressure has been associated with improvement in the glycaemic profile, HbA1c, insulin sensitivity and inflammation in some studies [115,116,117,118,119].

Conclusions

This guideline is the first national guideline on managing type 2 diabetes in children and young people. It gives practical advice on managing this challenging and aggressive condition. Paediatricians do not have extensive experience in managing type 2 diabetes due to a limited number of patients so can struggle with confidence in managing these patients. The authors hope that this guideline will provide paediatricians with the necessary information about managing type 2 diabetes and associated complications.