Abstract
Objective
Childhood obesity is associated with long-term health complications. Liraglutide is approved for use in adolescents for weight loss and has shown beneficial outcomes in clinical trials. Continuous glucose monitoring (CGM) is widely used in type 1 diabetes mellitus. To look at the effect of liraglutide treatment on cardiometabolic variables, glycaemic control (as assessed by CGM), body composition, quality-of-life and satiety levels in adolescents with severe obesity.
Methods
Patients aged 12 to 17.9 years were commenced on liraglutide in addition to lifestyle support. Pediatric Quality of Life 4.0 generic scale and Three-factor Eating Questionnaire R18 were completed at baseline and after 3-months.
Results
Twenty-four subjects (10 male: 14 female) took part. Significant improvements in weight, body mass index (BMI), BMI standard deviation scores, percentage body fat and fat mass following liraglutide treatment. A significant reduction in glycated haemoglobin, triglyceride and cholesterol levels, as well as a reduction in uncontrolled eating behaviour were observed. The time spent within normal glucose range (3.9-7.8 mmol/L; 70.2-140.4 mg/dL) was lower than in healthy peers (91.76% vs. 97.00%) at baseline but improved after liraglutide treatment. The cohort reported lower health-related quality-of-life scores and exhibited more uncontrolled eating and emotional eating behaviours, compared to the healthy population.
Conclusion
We report, for the first time, the role of CGM in identifying glycaemic dysregulation in children and young people with obesity before and after liraglutide treatment. The results have shown significant potential for liraglutide treatment in improving outcomes. Earlier identification of glycaemic dysregulation and targeted therapy could potentially reduce the long-term risk of developing type 2 diabetes mellitus.
What is already known on this topic?
Liraglutide has shown beneficial effects in managing adolescent obesity in clinical trials and has therefore been approved for clinical use. Continuous glucose monitoring (CGM) is routinely used in type 1 diabetes mellitus, but not for type 2 diabetes mellitus or in patients with obesity.
What this study adds?
Significant improvements were observed in adolescents with obesity following liraglutide treatment in a routine clinical setting. These included improvements in weight, body mass index (BMI), BMI-standard deviation score, body fat percentage, fat mass, percentage time glucose levels spent within normal range, levels of glycated haemoglobin, cholesterol, and triglycerides, and less uncontrolled eating behaviour.
Introduction
Childhood obesity is a worldwide concern due to its rising prevalence and the associated complications. Type 2 diabetes mellitus (T2DM) is a chronic condition that is associated with a number of long-term issues, such as micro- and macrovascular diseases (1). It has been shown that children and young people (CYP) with obesity are at a four-fold increased risk of developing T2DM (2, 3).
Treatment options for obesity within the pediatric population remain limited, with lifestyle intervention being the first line therapy. Liraglutide, a glucagon-like peptide 1 analogue, has received European Medicines Agency and FDA approval for adolescents aged 12-17 years with a body weight of at least 60 kg and an initial body mass index (BMI) corresponding to 30 kg/m2 or greater for adults (4, 5). It has been shown to have a significant effect on BMI standard deviation (SD) scores (SDS) and glycated haemoglobin (HbA1c) in adolescents (6, 7, 8, 9). The efficacy and tolerability in adolescents have been reported as similar to those seen in adults, with no unexpected issues found (8, 9, 10). Liraglutide has both central and peripheral mechanisms of action, which result in slower gastric emptying, thus increased satiety, and increased insulin production (11).
In view of liraglutide’s potential to improve glycaemic control and concerns of increasing prevalence of T2DM secondary to childhood obesity, there is a need for further research within this area. The introduction of continuous glucose monitoring (CGM) for patients with type 1 diabetes mellitus has been beneficial and reports in adults show the potential for its use in T2DM to improve glycaemic control and reduce complications (12). The use of CGM to investigate glycaemic dysregulation in childhood obesity is limited, and to the best of our knowledge it has not been used in normoglycaemic CYP with obesity who have received liraglutide treatment.
The aim of the present study was to investigate the effect of three months of liraglutide treatment on glycaemic control, anthropometry, metabolic outcomes, quality of life and satiety levels, in CYP with obesity.
Methods
This was a retrospective study over a period of 18 months. Patients were commenced on liraglutide (Saxenda; NovoNordisk) and followed regularly to assess progress (Figure 1). All patients were under the care of the Tier 3 multi-disciplinary, weight management service at a tertiary children’s hospital and received lifestyle intervention in the outpatient setting. Anthropometric measurements and fasting cardiometabolic variables were performed on a regular basis. The CYP had a CGM device inserted and completed three questionnaires at baseline and at 3-months post liraglutide treatment to investigate the effects of the treatment in a routine clinical setting.
Measurements
The CYP had their weight, BMI and BMI SDS measurements undertaken by the same staff in the medical day case unit. The body composition was obtained using a TANITA: RD-545HR device (Manchester, M1 2HY) which reported body fat percentage (%), fat mass (kg) and free fat mass (kg).
Biochemistry
Fasted blood samples were taken at baseline and at 3 months for HbA1c, insulin, c-peptide, liver enzymes (aspartate aminotransferase and alanine transaminase) and lipid profile. If the patient was not fasted then the insulin, c-peptide and triglyceride levels were omitted from analysis.
Continuous Glucose Monitoring
CGM was used at these time periods to investigate alterations in glycaemic dysfunction following the use of liraglutide. Dexcom G6 CGM devices (Broadway, London, United Kingdom) were used for the majority of the CYP. Two CGM data sets were collected using FreeStyle Libre Pro devices (Abbott House, Vanwall Business Park, Maidenhead, SL6 4XE).
All devices were placed onto the upper arm of the patients and removed on day 10 for the Dexcom wearers and day 14 for Libre wearers. When analysing the data, the readings for the first 12 hours were omitted to ensure accuracy.
Questionnaires
The patients and their guardians were asked to complete the following questionnaires at baseline and 3-months.
1. Pediatric Quality of Life (PedsQL) Inventory, Version 4.0 - English (UK): Child or teenager report (13).
2. PedsQL Inventory, Version 4.0 - English (UK): Parent/Guardian report (13).
Both the above questionnaires assess the quality of life for respondents. They include questions that investigate physical functioning, emotional functioning, social functioning, and school functioning. Transformed scores (0-100) are produced from the raw scores and the mean value is calculated. The higher the score, the better the quality of life reported. From these, the psychosocial health summary score, physical health summary score and total score are also calculated.
3. Three Factor Eating Questionnaire R18 (TFEQ-R18) (14, 15).
This questionnaire was completed by the participants and investigates their levels of uncontrolled eating, cognitive restraint, and emotional eating. Each statement has a score dedicated to one of the sections and the raw score is converted to a transformed score (0-100).
Liraglutide
Liraglutide was commenced following the baseline CGM, at a dose of 0.6 mg once daily as a subcutaneous injection. It was increased in increments of 0.6 mg with at least one-week intervals up to the maximum dose of 3 mg, or highest tolerated dose, with the support of the specialist nurse.
Statistical Analysis
The measurements, blood investigations and questionnaire results were analysed using a paired sample t-test to compare the means of each time point. The average glucose and the SD from the CGM data was also analysed using a paired sample t-test. The percentage time that glucose levels were in and out of range were analysed using the Wilcoxon signed ranks test to compare the medians at baseline and 3-months post-liraglutide treatment. All statistical analysis was performed using IBM Statistical Package for the Social Sciences, version 27 (IBM Inc., Armonk, NY, USA).
Results
A total of 24 patients were included in the study. The median (range) age of the patients was 14.7 (12-17.9) years and 14 (58.3%) were female. Two patients had pre-existing T2DM, and one patient had impaired glucose tolerance. The average measurements at baseline and 3-months are shown in Table 1. A significant mean weight reduction was found (-2.95 kg, p=0.001), along with a significant decrease in the mean BMI and BMI SDS (-1.38 kg/m2; p<0.001 and -0.09; p=0.003, respectively). Body composition was performed, which evaluates body fat percentage, fat mass and free fat mass. A significant reduction was seen over the course of the 3-months of liraglutide treatment in mean body fat percentage (-2.09%; p=0.043) and fat mass (-4.00 kg; p=0.011).
The biochemistry results showed a reduction in HbA1c, insulin, liver function and lipid profile levels (Table 2). Significant reductions in HbA1c [-1.35 mmol/mol; -2.3% (DCCT), p=0.025], triglycerides (-0.122 mmol/L; p=0.010) and cholesterol (-0.28 mmol/L; p=0.029) were noted over the 3-month treatment course. The change in high-density lipoprotein-cholesterol (HDL-C) levels were not significant, but the mean result went from being in the abnormal range to normal range over the course of the three months.
The CGM devices were worn for a median of 9.12 (2.2-13.9) days. The mean glucose level reduced slightly over the treatment course with a mean glucose of 6.56 mmol/L (118.1 mg/dL) at baseline and 6.50 mmol/L (117.0 mg/dL) at 3-months (p=0.828). The SD calculated by the CGM device improved from 1.11 mmol/L to 0.97 mmol/L (20.0-17.5 mg/dL), which was not significant (p=0.066) (Table 3a). The median percentage time that the patient’s glucose levels were within normal range (3.9-7.8 mmol/L; 70.2-140.4 mg/dL) significantly improved with liraglutide treatment and increased from 91.76% to 94.18% (p=0.048). It was also noted that the percentage time that glucose levels were >7.8 mmol/L (140.4 mg/dL) or >10 mmol/L (180 mg/dL) also reduced during the intervention period (Table 3b). These results show that the use of liraglutide does improve glucose variation in CYP with obesity. When compared to 12-18 year olds who are healthy, with a BMI below the obesity threshold and do not have diabetes, the time spent with glucose levels within the normal range remained lower in our cohort (91.76% and 94.18% versus 97% for healthy CYP) (Table 3b) (16).
Quality of life was measured through both child and parent-reported questionnaires and the mean transformed scores were calculated. The child-reported scores showed an improvement in all sections, except for social functioning, which remained similar. None of the improvements were found to be significant (Table 4). The parent-reported scores also showed an increase in quality of life for all areas, except for school functioning. Even though an overall positive change was noted, the results remain much lower than those reported by Upton et al. (17), which represents healthy children and parents and also those with a chronic condition (Table 4). This highlights the association between childhood obesity and lower quality of life levels.
Liraglutide is known to improve satiety (11) and this was shown by the results of the TFEQ-R18 scores. A significant reduction was seen in uncontrolled eating with a mean baseline score of 51.94 and 3-month score of 41.01 (-10.93; p=0.006). Cognitive restraint tended to increase and emotional eating tended to reduce over the 3-months, but the differences were not statistically significant (Table 4).
Discussion
The results showed significant improvements following three months of liraglutide treatment in a routine clinical setting. Adolescents with obesity participants exhibited improvements in weight, BMI, BMI SDS, body fat percentage, fat mass, percentage time glucose within normal range, HbA1c, cholesterol, triglycerides and uncontrolled eating behaviour.
The change in weight, BMI and BMI SDS reported herein support those found in earlier clinical trials (7, 18, 19, 20). Zhao et al. (21) recently published a systematic review and meta-analysis comparing medications for the treatment of obesity. Compared to metformin, orlistat, exenatide and topiramate, liraglutide had shown to be the most beneficial treatment for weight loss in CYP. Our results showed a significant reduction in body fat percentage and fat mass. The effect of liraglutide on body composition has been shown in adults, assessed using dual energy X-ray absorptiometry (22, 23, 24). Feng et al. (22) showed a significant reduction in total fat mass, trunk, limb, android and gynoid fat following a 24-week course of liraglutide treatment (1.8 mg) in adult patients with T2DM and non-alcoholic fatty liver disease. Significant reductions in total body fat, visceral adipose tissue and liver fat were seen following a median course of 36 weeks of liraglutide (3 mg) versus placebo using magnetic resonance imaging (25). These changes are thought to reduce the risk of cardiovascular disease complications (25). There are very limited studies available looking at the effect liraglutide has on body composition in the pediatric population. Our results are promising after 3-months of liraglutide treatment. Liraglutide likely has this positive effect on body composition secondary to its known mechanisms, which include lower plasma glucagon levels, delayed gastric emptying and increased satiety. It has been suggested that there may also be a weight independent effect of liraglutide on the distribution of body fat (25). Further research is needed in this area, especially in CYP.
Kelly et al. (7) did not show significant changes in glycaemic and cardiometabolic variables, but interestingly our results contradict their findings. We showed improvements in all the biochemical variables, except for fasting c-peptide levels. HbA1c, triglyceride and cholesterol levels reduced significantly and, even though the difference was not significant for HDL-C, it had improved from the abnormal to normal range over the three month treatment course. Similar results have also been shown previously (19, 20). These improvements also support the potential cardiovascular benefits that liraglutide might have in the long run.
Due to its positive effect on glycaemic control, liraglutide has been approved for the treatment of T2DM in children aged 10 years and over. Fasting glucose and HbA1c levels have been analysed during liraglutide treatment and it has shown good effect compared to placebo (6). The rising rates of childhood obesity are associated with increased prevalence of T2DM in CYP. T2DM is known to be associated with multiple complications which may have a significant impact on the individual. Identifying glycaemic dysregulation in CYP with obesity, who are at increased risk of T2DM, would result in commencing interventions and potentially reducing the chance of the individual developing T2DM. The use of CGM has been shown to be a feasible and acceptable tool to investigate glycaemic control in both pediatric and adult populations (26, 27). To the best of our knowledge, CYP with obesity, but not T2DM, on liraglutide treatment have not had their glucose levels analysed using CGM devices. Our patients tolerated the monitors well and the results show a significant improvement in the time that their blood glucose was within the normal range following treatment. Shah et al. (16) analysed glycaemic variation in healthy children and adults and this data was used for comparison. Their results from the 12-18-year-old participants were used as this was the same age range as our patients. At best, the median percentage time in range was 94.18%, which was still lower than the results seen in the healthy population (97.00%). Interestingly, Shah et al. (16) did not report any blood glucose levels above 10 mmol/L, whereas for our patients, they had 0.08% and 0.02% of time spent above this level at baseline and 3-months, respectively. This demonstrates that CYP with obesity already show evidence of glycaemic dysregulation, which improved with liraglutide treatment. This highlights the potential to commence at-risk individuals on treatment at the earliest point, thereby preventing them from progressing to pre-diabetes or T2DM.
Lower health-related quality of life (HR-QoL) has been shown to be associated with childhood obesity. Younger children have been reported to score lower for physical functioning compared to psychosocial functioning (28). Whereas other studies have shown that the individuals’ psychosocial QoL is impacted more as they get older into adolescence and with the most severe grade of obesity (29, 30). The total score for HR-QoL improved in our cohort of patients following liraglutide treatment. For both child and parent-reported questionnaires the psychosocial score was lower than the physical health score. Adult studies have shown this association (31, 32), whereas Kelly et al. (7) did not find any difference in overall weight-related quality of life in adolescents. Data from Upton et al. (17) was used for comparison, and the scores from the healthy population were much higher in all sections than those reported by our cohort. This highlights the importance of psychological support within the multidisciplinary team in the management of childhood obesity.
The GLP-1 hormone acts by delaying gastric emptying and appetite suppression (7, 33). We analysed the effect of this in the clinical setting by asking our patients to complete the TFEQ-R18 questionnaire. A significant reduction was seen in uncontrolled eating behaviours and both cognitive restraint and emotional eating showed improvement during the treatment course. The study of De Lauzon et al. (34) was used as the comparison healthy population. As expected, their scores for emotional and uncontrolled eating were lower than our results but interestingly, our patients scored higher for cognitive restraint. Emotional eating was the highest score, which links with the poor psychosocial scores previously mentioned. These results can be used by the psychology team to help improve eating behaviours. This questionnaire can also be used when assessing a patient with suspected hyperphagia.
Study Limitations
The limitations of our study include a small sample size, in comparison to clinical trials previously published. It is important to note that these patients were under the care of a Tier 3 multidisciplinary, weight management service, which has a specific inclusion criterion. This means that all the patients had significantly elevated BMIs (>+3 SDS) and associated complications. Therefore, these results show the effect of liraglutide on adolescents with severe obesity in the routine clinical outpatient setting. Due to the retrospective nature of data collection and patient engagement, some measurements and results were unavailable and the sample size (n) for each section has been highlighted throughout.
Conclusion
The results of the present study showed the positive effect of liraglutide on CYP with obesity following a 3-month treatment course in a routine clinical setting. Significant improvements were shown in anthropometric measurements, cardiometabolic variables and uncontrolled eating behaviour. We have also shown the potential use of CGM in identifying glycaemic dysregulation in CYP at risk of T2DM due to obesity and the significant improvement that liraglutide has on the time that the glucose levels were spent within the normal range. Further research is needed, but there is promise for the use of CGM in identifying glycaemic dysregulation earlier, so that interventions can be targeted with the view of preventing the complications or progression to T2DM.
