Obesity and morbidity

The rapidly increasing prevalence of obesity among children is a major challenge for paediatricians. Reilly et al3 provide a critical appraisal of the literature in their review article, summarising both the short and longer term consequences of childhood obesity. Key short term consequences included psychological effects. Obese children are more likely to experience psychological or psychiatric problems than non-obese children; girls being more at risk than boys and this risk increases with age. Approximately 34% of obese, 13–14-year-old white girls had low self-esteem compared with 8% non-obese white girls.4 In the long term, paediatric obesity tends to persist into adulthood5 ,6 with adverse effects on social and economic outcomes in young adult life,7 increased adult morbidity and increased risk of earlier death.5 Obesity increases the risk of developing dyslipidaemia, hypertension, coronary heart disease, asthma symptoms, type 2 diabetes, cerebrovascular accidents and reduces life expectancy.3

The consequences of obesity are not limited to the direct impact on health; there is a wider impact on the national economy in terms of working days lost and decreased productivity. It is estimated that the treatment of obesity costs approximately round £1.1 billion a year, whereas the overall costs to the economy are projected at around £3.5 billion annually.8

Assessment of obesity

BMI is the most widely used metric by which to assess overweight. BMI is not a direct measure of adiposity and experts have debated whether it is the most appropriate measure or whether alternatives such as waist to hip ratio should be considered. Accepting its limitations, BMI is recommended by the Scottish Intercollegiate Guideline Network (SIGN) and National Institute of Clinical Excellence (NICE) guidance as the most appropriate measure.9 ,10 There is no clear consensus as to exactly which cut-offs should be used to define obesity.11 It is important to differentiate the cut-offs used in clinical practice from those used epidemiologically to assess the degree of overweight and obesity within the population.12 Epidemiologically BMI >85th centile is often defined as overweight and >95th centile obese. In clinical practice SIGN guidance suggests BMI of >91st centile is classed as overweight, >98th centile as obese and >99.6th centile as severe obesity (+2.67 SD).9 Somewhat confusingly the SIGN guidance also uses SD scores (SDS) where there are no appropriate centiles and defines very severe obesity as BMI>3.5 SDs above the mean and extreme obesity as >4.0 SDs above the mean.9 The International Obesity Task Force have drawn up recommendations for use internationally13 and while they broadly correlate with the 91st Centile (overweight) and 98th centile (obese) of the UK 1990 growth standards, they are not completely concordant. It is recommended in the UK that a centile based approach and criteria are used.14

Management of obesity

Lifestyle interventions are the first line treatment of obesity in children. There are a wide range of behavioural interventions and weight management programmes available nationally including MEND, Morelife, Alive N Kicking, Watch IT and HENRY to name a few. There is wide variation in local provision with different localities adopting different programmes. The Cochrane review of interventions to treat obesity in children identified 54 randomised control trials: 12 focused on physical activity and sedentary behaviour, six on dietary interventions and 36 were behaviourally orientated treatment programmes.15 The authors concluded that there was little evidence to recommend one treatment programme over another but that combined behavioural lifestyle interventions produced a significant and clinically meaningful reduction in children's weight.15 Pooled meta-analysis (on only two to four trials in most cases) suggested a reduction in absolute BMI of 3.04 kg/m² with a corresponding reduction in BMI SDS of −0.14. Effect size was greater in older (>12 years) compared with younger children. This contrasts with the systematic review by McGovern et al which included larger numbers of studies in the meta-analysis (23 trials).16 McGovern et al found a more modest, but still significant effect of combined lifestyle interventions on BMI of approximately 1.5–1.6 kg/m².

The SIGN guidelines recommend family based behaviour change programmes that reduce the overall dietary intake, increase physical activity (at least 60 min of moderate activity a day) and decrease time spent on sedentary behaviours.9 NICE guidance recommends aiming for weight loss of 0.5–1.0 kg/month in adolescents based on British Dietetic Association recommendations.10 It is important that lifestyle programmes do not affect the normal growth and development of the child. Therefore, in younger children weight maintenance may be a suitable goal rather than weight loss.

Indications for pharmacological treatment of obesity

Unfortunately individuals are often unable to achieve or maintain significant weight loss with lifestyle interventions alone. In adults, NICE recommends starting antiobesity drugs when life style changes have failed in helping the individual lose weight or when there are associated comorbidities. In children, NICE does not recommend drug therapy, except in specific situations:10

• In children less than 12 years, only when there are severe life-threatening comorbidities like sleep apnoea or raised intracranial pressure.

• In children 12 years and older, only when physical or significant psychological comorbidities are present.

SIGN guidelines suggest pharmacological intervention may be considered with severe obesity (BMI >99.6th centile) when there are comorbidities or with very severe (BMI SDS >3.5) or extreme obesity (BMI SDS >4.0).9 Both recommend referral to and management within specialist clinics—though neither define a ‘specialist’ clinic. Although there are limited data on the use of these drugs in children, a recent study concluded that, in the UK, prescribing of antiobesity medicines for adolescents and children <18 years has risen by 15-fold between 1999 and 2006.17

Therapeutic options are limited. Over recent years only three drugs have been licensed and marketed specifically for weight loss: rimonabant, sibutramine and orlistat. None were licensed for use in children.

Rimonabant selectively blocks the centrally located cannabinoid-1 receptors, decreasing food intake.18 It was withdrawn several years ago due to adverse effects on mental health.

Sibutramine was originally developed as an antidepressant. It worked as a centrally acting serotonin and norepinephrine reuptake inhibitor that increased satiety thus reducing food intake.19 Studies in adolescents had demonstrated its efficacy. Meta-analysis of four randomised controlled trials of sibutramine showed a reduction in BMI of 2.2 kg/m².20 However, based on the findings of the SCOUT (Sibutramine Cardiovascular Outcome Trial) study,21 it was concluded that in patients with pre-existing cardiovascular conditions and taking sibutramine, there was 16% increased risk of non-fatal myocardial infarction and stroke but no increased risk of cardiovascular death. It was therefore concluded that the cardiovascular risks of sibutramine were greater than its benefit and its use was suspended by the European Medicine Agency in January 2010.

Orlistat

Only orlistat remains available with weight loss as a specific licensed indication. Although the Federal Drug Administration has approved orlistat for use in individuals over 12 years of age its use in children in Europe remains off licence.

Orlistat is a synthetic derivative of lipostatin, which is a naturally occurring lipase inhibitor isolated from the bacterium, Streptococcus toxytricini.22 It inhibits intestinal lipase by binding irreversibly with it. It has shown to reduce fat absorption by 30% in patients eating a 30% fat diet, thereby achieving a reduction of approximately 200 calories per day.22 Orlistat has a low systemic absorption and most of the drug is excreted unchanged in stools. It has been extensively studied in adults, has a good safety record and is generally well tolerated. Its principal side effects are gastrointestinal. Cholelithiasis has been reported on rare occasions.23

The usual dose of orlistat is 120 mg three times daily with meals or within an hour of the meal. NICE recommends a 6–12-month trial if orlistat is prescribed in children with regular reviews. Orlistat has been shown to reduce absorption of fat soluble vitamins particularly betacarotene and vitamin E.24 ,25 Multivitamins supplementation is recommended while on orlistat.24 ,25

A number of studies have explored the use of orlistat in adolescent populations.26–28 Many studies are small and of short duration. Several systematic reviews and meta-analyses have been undertaken.16 ,20 ,29 In the analysis performed by Viner and colleagues20 only two studies (573 patients combined) fulfilled the inclusion criteria.23 ,30 In both studies, in addition to orlistat, the patients also underwent behavioural, dietary and exercise counselling. They concluded that orlistat together with behavioural support led to a reduction of 0.83 kg/m² in BMI compared with placebo.20 The total weight loss averaged 2.3 kg over 12 months.20 Side effects reported by patients on orlistat were mainly gastrointestinal including oily spotting, flatulence, faecal incontinence and oily loose stools.23 The gastrointestinal side effects can be minimised by reducing dietary fat. In the larger of the two studies included in the analysis (533 adolescents aged 12–18 years) by 12 months, 19% of individuals in the orlistat group had lost 5% of their body weight compared with 11.7% in the placebo group. Approximately 10% of the individuals in the orlistat group had lost more than 10% of their body weight compared with 3.3% in placebo group.23

A separate meta-analysis by Czernichow included different studies and reported a reduction in BMI of 1.67 kg/m² and an absolute weight loss averaging 6.0 kg.29 Other meta-analyses have reported weight loss more in keeping with Viner and colleagues. Oude et al reported a reduction in BMI of −0.76 kg/m² and McGovern et al −0.7 kg/m².15 ,16

Dropout rates ranging from 6% up to 33.8% of participants have been reported.29 Gastrointestinal side effects have been reported in up to 50% of participants.23 UK prescribing data suggest that 45% of individuals prescribed orlistat take it for less than a month suggesting that gastrointestinal side effects may be more significant than reported in some studies.17

The majority of studies of orlistat have focused on adolescent populations. Only one study of 11 children has looked at the use of orlistat in prepubertal children aged 8–12 years.^w1 It was a ‘before and after’ study. Total weight loss was approximately 4 kg over 12 weeks.

Orlistat appears to support modest weight loss of 2.5–6.0 kg in both prepubertal and adolescent children though gastrointestinal side effects are troublesome for some patients.

Metformin

In obese children, particularly adolescents, concern about insulin resistance and type 2 diabetes has prompted the use of metformin. There is evidence that metformin may delay the progression from impaired glucose tolerance to type 2 diabetes.^w2 Metformin may also promote modest weight loss, although it is not yet specifically licenced for weight loss. Weight loss occurs as a result of metformin's non-insulinotropic action as well as its effects of increasing satiety and gastrointestinal side effects.^w3

Metformin is a biguanide used as an oral hypoglycaemic agent in treatment of type 2 diabetes. Its cellular mechanism is poorly understood. It has three major effects: it reduces hepatic glucose production which is the main source of circulating glucose; it decreases intestinal absorption of glucose; and increases sensitivity to insulin increasing peripheral utilisation of glucose. Lactic acidosis is a very rare complication of metformin, with an incidence of 0.05 per 1000 patient years.^w4

A number of studies have examined its use, principally in adolescents. A systematic review identified five studies for meta-analysis.20 Pooled data suggested that metformin reduced BMI by 1.42 kg/m² (approximately 0.4 SDS) compared with placebo. It was well tolerated with dropout rates of 11.6% in the metformin treated groups compared with 16% in placebo.^w5 The principal side effects were gastrointestinal reported in up to 20%–30%.20 ^{w6 w7} Modest reductions in cholesterol and insulin resistance (measured by homeostatic model assessment) were reported.^w8 Most studies though were relatively small and short-term.

In a study, treating of 24 obese hyperinsulinaemic, but non-diabetic, adolescent patients with metformin and diet produced weight loss of 6.5% compared with 3.8% in the placebo group. Metformin produced greater improvement in body fat, insulin sensitivity, cholesterol and triglycerides.^w9 A further study in 28 overweight, non-diabetic adolescents between 9 and 18 years demonstrated improvements in insulin sensitivity in 45% of subjects on metformin compared with 27% of subjects on placebo.^w10

A study in younger children aged 6–12 years also showed a reduction in absolute BMI of 1.09 kg/m² (BMI SD—0.07) compared with placebo. Total weight loss across children of differing ages was 3.38 kg.^w11

These small, short clinical trials provide preliminary evidence that metformin is associated with weight loss and positive effects on metabolic syndrome but large scale, long term randomised placebo controlled trials are necessary to evaluate its role in weight loss alone.

Other historical treatments

Over the years there have been a variety of drugs trialled for weight management including a range of drugs acting on the monamine system including amphetamines and amphetamine like analogues (eg, phentermine, diethylpropion) and fenfluromines. Some have been formally marketed for weight loss but many have subsequently been withdrawn as adverse side effect profiles emerged.^w12

A number of drugs licensed for other indications have been investigated for their weight loss potential including several antidepressant drugs also active within the monamine system. Fluoxetine, a selective serotonin reuptake inhibitor, and buproprion, an antidepressant which inhibits reuptake of dopamine and norepinephrine, have both been used off licence for weight loss. Fluoxetine produces short-term weight loss but longer term individuals tend to regain weight.^{w13 w14} In a meta-analysis, weight loss of 2.8 kg was reported with buproprion at 6–12 months compared with placebo.^w15 Studies of bupropion (sometimes used for smoking cessation) in combination with the opioid receptor antagonists naltrexone have been shown to be efficacious.^w16

The β adrenergic agonists caffeine and ephedrine, which are postulated to activate lipolysis in adipocytes, enhance thermogenesis and delay gastric emptying (increasing satiety), have been advocated. They were used extensively in the 1990s in North America and Denmark.^w17 However, the FDA revoked the licence in 2004 due to cardiovascular side effects. In a study on 32 obese adolescents, mean weight loss was 5.5 kg and BMI change −1.4 kg/m² compared with placebo with minor side effects.^w18

However, the evidence base to support the use of these antidepressants, caffeine and ephedrine in adolescent obesity is weak and they cannot be advocated outside of formal clinical trials.

Potential future developments

A number of potential therapies are under investigation. Octreotide, a synthetic analogue of somatostatin (which inhibits insulin secretion, secretion of gastric acid and delays intestinal transit time), has been shown to be effective in reducing insulin secretion, weight and BMI in children with severe hypothalamic obesity post brain tumour or cranial irradiation.^w19 However, parenteral administration is a major obstacle to its wider use.

Several drugs currently licensed and prescribed for diabetes in adults have demonstrated potential to also facilitate weight loss. Exanetide is a glucagon-like peptide-1 agonist isolated from the lizard Heloderm suspectum. It stimulates glucose dependant insulin exocytosis which helps in better glycaemic control in patients with type 2 diabetes. It has also shown to delay gastric emptying and causes reduction in food intake, thereby leading to weight reduction. Studies have shown modest improvements in body weight from −2.8 to −1.6 kg in individuals on exanatide compared with placebo (+0.3 to +0.9 kg).^w20 Liraglutide, another glucagon-like peptide-1 analogue, has also been associated with dose dependant weight loss in addition to improvement of HbA1C and triglyceride levels.^w21

Pramlintide, a soluble synthetic analogue of the pancreatic hormone amylin, is used as an adjunct to insulin in type 1 and type 2 diabetics who have poor glycaemic control. Amylin is cosecreted with insulin and acts to mediate satiety effects in the body. In adults, pramlintide used in conjunction with lifestyle interventions produced sustained weight loss of up to 7.2 kg over 12 months.^w22 Another recent drug is metreleptin which is a leptin analogue and works by reducing fat accumulation in organs, thereby increasing insulin sensitivity. Pramlintide combined with metreleptin has also been demonstrated to promote significant weight loss. Minor side effects like nausea and injection sites events were noted.^w23

Other specific weight reduction medications under investigation include cetilistat, another gastrointestinal lipase inhibitor. In a short (12 weeks) phase multicentre randomised placebo controlled study, statistically significant weight loss of up to 4.1 kg was noted.^w24 There were also significantly improved outcomes in terms of total serum and low density lipoprotein cholesterol by 3%–11% in the treatment group. Cetilistat was also found to have fewer adverse side effects as compared with orlistat and hence better tolerability.^w25

Topiramate, an anticonvulsant gamma-Aminobutyric Acid (GABA) agonist, has also been noted to result in weight loss by suppressing appetite.^w12 However, specific studies investigating its role in weight loss were stopped in phase III because of the high incidence of side effects.^w26 Recently, a new combination of topiramate with phentermine has been shown to be effective in improving weight.^w27

None of the above drugs, trialled almost exclusively in adults, has yet received a licence to treat obesity but several may have potential. There are however limited data in adults and none in adolescents.

Bariatric surgery

Bariatric surgery is gaining increasing acceptance in young people. It is beyond the scope of this review on pharmacological treatment of childhood obesity but NICE guidance makes provision for surgery in adolescents in exceptional circumstances10 and there is an increasing body of evidence to support its efficacy.^{w28 w29}