| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institute of Endocrinology and Diabetes (S.S., G.R.A., S.P.G., M.T., C.T.C.), The Children’s Hospital at Westmead, Westmead, New South Wales 2145, Australia; Discipline of Paediatrics and Child Health (S.S., L.A.B., S.P.G., C.T.C.), University of Sydney, Sydney 2006, Australia; Department of Paediatrics (F.Y.), KK Women’s and Children’s Hospital Singapore, Singapore 229899; and Department of Endocrinology (G.M.W.), St. Vincent’s Hospital Melbourne, Fitzroy, Victoria 3065, Australia
Address all correspondence and requests for reprints to: Dr. Shubha Srinivasan, Institute of Endocrinology and Diabetes, The Children’s Hospital at Westmead, Locked Bag 4001 Westmead, New South Wales 2145, Australia. E-mail: shubhas{at}chw.edu.au.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: We assessed the effect of metformin on body composition and insulin sensitivity in pediatric subjects with exogenous obesity.
Design and Setting: Patients referred to a pediatric endocrine clinic were enrolled in a randomized, double-blind, crossover trial.
Patients: Twenty-eight patients (13 males) aged 9–18 yr participated in the study.
Intervention: Patients received metformin (1 g twice daily) and placebo for 6 months, each with a 2-wk washout period.
Main Outcome Measures: Body composition (anthropometry, dual-energy x-ray absorptiometry, and abdominal magnetic resonance imaging), and insulin sensitivity (Si; minimal model, fasting insulin and glucose) were measured at baseline and 6 and 12 months.
Results: Mean age of subjects at baseline was 12.5 ± 2.2 yr, median body mass index z-score 2.54 (range, 1.93–2.85). Metformin had a greater treatment effect over placebo for weight (–4.35 kg, P = 0.02), body mass index (–1.26 kg/m2, P = 0.002), waist circumference (–2.8 cm, P = 0.003), sc abdominal adipose tissue (–52.5 cm2, P = 0.002), and fasting insulin (–2.2 mU/liter, P = 0.011). Si improved in 45% of subjects while on metformin and 27% of subjects while on placebo (P = 0.21).
Conclusions: Metformin therapy for obese insulin-resistant pediatric patients results in significant improvement in body composition and fasting insulin. Although improvement in Si was noted in many individuals, Si was a less useful parameter for analysis of group data, possibly because of effects of variable compliance and changing Si during puberty.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
There are few data on the role of metformin in insulin resistance associated with obesity before the development of type 2 diabetes in children. The potential clinical application of metformin in the pediatric population was first described in a small study in the 1970s with a beneficial effect on weight and insulin concentrations in 8- to 14-yr-old obese children (10). Subsequent pediatric randomized, controlled trial data have shown improvement in body mass index (BMI), fasting serum glucose, and insulin and improved lipid profile in patients on metformin therapy for exogenous obesity with insulin resistance (11, 12) as well as psychotropic drug-induced weight gain (13). However, insulin sensitivity, as measured by minimal model, did not significantly improve in adolescents receiving metformin, compared with placebo, in a case-controlled, randomized, controlled trial (12), raising the question of whether metformin specifically improves peripheral insulin sensitivity.
By conducting a crossover, randomized, controlled trial, we sought to clarify the role of metformin therapy in pediatric patients with obesity, specifically addressing the effect on anthropometry, body fat compartments, and insulin sensitivity parameters.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Participants were 9 to 18 yr olds referred to the endocrine clinic at The Children’s Hospital at Westmead between March 2002 and March 2003 with obesity, as defined by the International Obesity Task Force (14), and clinical suspicion of insulin resistance, as defined by either a fasting insulin (milliunits per liter) to glucose (millimoles per liter) ratio greater than 4.5 (15) or the presence of acanthosis nigricans. Exclusion criteria were known type 1 or type 2 diabetes mellitus, contraindications to metformin therapy, and/or magnetic resonance imaging (MRI) scanning and weight greater than 120 kg due to technical difficulties with dual-energy x-ray absorptiometry (DXA) scans. All parents and patients were given verbal and written information about the study before providing written consent. All participants were invited to give verbal and written feedback of individual results at the end of the study. This study was approved by The Children’s Hospital at Westmead Ethics Committee and registered with the International Standard Randomized Controlled Trial scheme (ISRCTN43267711).
Study design
Participants were randomized to receive metformin and placebo for 6 months each in a crossover design, with a 2-wk washout period in between. Block randomization (blocks of four) with stratification by pubertal stage (Tanner 1–2 or Tanner 3–5) was performed by computer-generated random number allocation, and placebo or metformin was dispensed by the hospital pharmacy. All participants and investigators were blinded to the intervention. Unblinding occurred after final data analysis. Both metformin and placebo doses were gradually built up over a 3-wk period to a final dose of 1 g twice daily. Standardized information on healthy eating and exercise was given to all patients.
Investigations
The time line for investigations is illustrated in Fig. 1
. At baseline and 6 and 12 months, participants attended The Children’s Hospital at Westmead for clinical assessment including anthropometry, frequently sampled iv glucose tolerance test, DXA imaging, and MRI of the abdomen as detailed below. At 3 and 9 months, participants underwent clinical assessment and fasting biochemical profile. Liver function tests, serum creatinine, and serum lactate levels were measured every 3 months to assess metformin safety profile. Pill counts were conducted every 3 months by the hospital pharmacy to calculate percent adherence to therapy based on number of capsules consumed vs. anticipated capsule consumption for each 3-month period.
|
Height was measured to the nearest 0.1 cm and weight with minimal clothing using electronic scales to the nearest 0.1 kg. Waist circumference was calculated from the average of three measures at the level of the umbilicus. BMI was calculated using the following formula: BMI = weight/height (kilograms per square meter). BMI for age z-scores were calculated from the U.S. Centers for Disease Control and Prevention reference data 2000 (16). Waist circumference z-scores were calculated from recent multiracial American reference data (17). Pubertal stage was assessed using the standards of Tanner and Whitehouse (18). Blood pressure was measured on the right arm with an appropriately sized cuff using a DynaMap machine with the subject seated. The lowest of three measures was recorded. Routine physical examination was performed before each set of investigations to rule out significant intercurrent illness. Acanthosis nigricans was assessed for severity at the neck by a validated scale ranging from grade 0 (not present) to grade 4 (severe: extending anteriorly, visible when the participant is viewed from the front) (19). This was performed clinically by the principal investigator (S.S.) and based on clinical photographs by an independent observer.
Frequently sampled iv glucose tolerance test
After an overnight fast, subjects underwent a 180-min iv glucose tolerance test for minimal model analysis of parameters of insulin sensitivity (20). An iv cannula was inserted into each arm, one for sampling and the other for glucose and insulin boluses. After taking baseline samples, 0.3 g/kg dextrose (25% solution) was infused over 90 sec at time 0. At 20 min 0.03 U/kg Actrapid insulin (1:10 dilution) was given over 90 sec. Paired insulin and glucose samples were taken at 2, 3, 4, 5, 6, 8, 10, 12, 14, 16, 19, 22, 23, 24, 25, 27, 30, 40, 50, 60, 70, 80, 90, 100, 120, 140, 160, and 180 min. Glucose was analyzed immediately on the Dade Dimension ARX using hexokinase-glucose-6-phosphate dehydrogenase method. The insulin assay was performed on the Immulite analyzer (Diagnostics Products Corp., Los Angeles, CA) using an immunometric assay. Insulin and glucose values were entered into the MINMOD Millennium computer program (40) to calculate insulin sensitivity, glucose effectiveness (Sg), acute insulin response (AIR) disposition index (DI) and glucose disposal (Kg). All analyses were performed by the same investigator (S.S.).
DXA
DXA scans were performed on the GE Lunar Prodigy machine (GE Lunar Corp., Madison, WI) in the Department of Medical Imaging at The Children’s Hospital at Westmead. Participants were positioned on the scanner table using standard procedures, and total body cuts were positioned as per standard manufacturer specifications. Data obtained from the DXA scans were processed using GE Lunar enCore software (version 6.10.029; GE Lunar Corp.) to calculate percent total body fat.
MRI of the abdomen
All subjects were scanned on a 1.5 Tesla Philips (Best, The Netherlands) ACS-NT whole-body MRI unit. Five cross-sectional images, each 10 mm thick, of the abdomen were acquired. The center image (slice 3) was positioned at mid-L-4 with two images acquired above (slices 1 and 2) and below (slices 4 and 5). Analyze software (version 4.0; Mayo Clinic, Rochester, MN) was used to quantify the surface area (square centimeters) of visceral abdominal adipose tissue (VAAT) and sc abdominal adipose tissue in each of the five slices. The mean of the five slices was used in the final analysis. All analyses were performed by the same investigator (S.S.).
Statistical analysis
For all outcomes, data were analyzed as a simple two-period crossover trial using Statistical Package for the Social Sciences (SPSS; version 11.5.1; Chicago, IL). Normally distributed data are reported as mean ± SD and nonparametric data as median (range). To assess the effect of metformin vs. placebo, the paired sample t test was used to compare means for normally distributed data and the Wilcoxon signed-ranks test to compare paired medians for nonparametric data. These tests were applied to period 1 and period 2 differences for groups A and B. The difference between the means of the two groups was taken as twice the size of the treatment effect (21). Linear mixed model analysis was performed to assess possible confounding effect of change in pubertal stage and poor adherence to therapy on insulin sensitivity.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
.
|
The mean age of the participants was 12.5 ± 2.2 yr with 14 of 28 (50%) being in Tanner stage 1–2 and 14 of 28 in Tanner stage 3–5. There were significantly more girls than boys in Tanner stage 3–5 puberty (P = 0.02); however, other characteristics were similar for both males and females (Table 1). Eighteen participants (64%) were from ethnic backgrounds with high prevalence of insulin resistance and the metabolic syndrome (e.g. Indian subcontinent, Pacific islands), seven (25%) were from a northern European background, and three participants (11%) were from a mixed background. A family history of features of the metabolic syndrome in either first- or second-degree relatives was noted in 25 participants (89%). Twenty-five of 28 (89%) participants had acanthosis nigricans. There was no difference in baseline characteristics between groups A and B.
|
Metformin therapy had a significant beneficial treatment effect over placebo for weight, BMI (Fig. 3A), and waist circumference, both as raw measures and z-scores (Table 2). Whereas metformin therapy resulted in a tendency to reduction in total body fat percentage from DXA measurements (treatment effect 0.67%), this was not significant (P = 0.062). A beneficial treatment effect of metformin over placebo was found for sc abdominal adipose tissue (treatment effect 52.5 cm2; P = 0.002) but not VAAT (treatment effect –6.3 cm2; P = 0.231), suggesting that the weight loss was primarily sc, rather than visceral, fat. Median acanthosis nigricans neck severity score on metformin was 3.0 (0–4) and placebo was 4.0 (0–4), P = 0.304.
|
|
Metformin therapy had a beneficial treatment effect over placebo for fasting insulin (Fig. 3B
) and a small but significant beneficial effect for fasting glucose (Table 2). Insulin sensitivity measured from the minimal model improved in 10 of 22 (45%) patients on metformin and six of 22 (27%) patients on placebo (P = 0.21). There was no significant beneficial treatment effect of metformin over placebo for insulin sensitivity parameters (insulin sensitivity, Sg, AIR, DI, Kg) measured by minimal model analysis (Table 2 and Fig. 3C).
Side effects, adherence to therapy, and safety profile
Both metformin and placebo were well tolerated with only two participants (aged 9 yr) unable to tolerate 1 g metformin twice daily due to nausea. Both of these participants tolerated 750 mg twice daily with slower dose increments. Adherence to therapy based on pill counts for the whole group was similar for metformin and placebo (metformin median adherence 78%, range 15–99%; placebo median adherence 78%, range 35–98%; P = 0.689). Furthermore, individuals demonstrated consistency with adherence to metformin and placebo (r = 0.566, P = 0.004).
There was no difference in liver function tests, serum creatinine, or lactate levels while on metformin or placebo (serum alanine aminotransferase 49.9 ± 25.1 vs. 55.3 ± 31.5 U/liter, P = 0.100; serum creatinine 60.6 ± 9.7 vs. 62.4 ± 9.0 µmol/liter, P = 0.141; serum lactate 0.29 ± 0.47 vs. 0.40 ± 0.49 mmol/liter, P = 0.437).
Role of adherence to therapy and change in pubertal status in insulin sensitivity measures
Eight patients took less than 75% of their prescribed metformin. In addition, six patients underwent a change in pubertal stage from Tanner 1–2 to Tanner 3–5 over the 1-yr period. Linear mixed-model analysis using the unstructured repeat covariance type (lowest –2 restricted log likelihood, i.e. best estimate of fit) was used to assess the possible confounding effect on insulin sensitivity of poor adherence to therapy and change in pubertal status. After adjusting for change in pubertal stage and adherence to therapy, insulin sensitivity on metformin was 0.172 (milliunits per liter)–1 per minute–1 higher [95% confidence interval –0.146 to 0.491 (milliunits per liter)–1 per minute–1] than insulin sensitivity on placebo (P = 0.273).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our study population was 9 to 18 yr olds referred to the endocrine service for management of obesity and insulin resistance. Many were experiencing relentless weight gain with the development of clinical features of insulin resistance, such as acanthosis nigricans, despite attempts at appropriate lifestyle changes. Obesity and insulin resistance in the peripubertal child and adolescent can be frustrating to manage for the child, their family, and their health care workers because entrenched unhealthy lifestyle patterns are often compounded by the physiological insulin resistance of puberty. However, there is little evidence-based information to guide the clinician in the management of insulin resistance in children and adolescents.
By conducting a double-blind, placebo-controlled, crossover trial, we assessed the role of metformin over placebo for young individuals with insulin resistance and obesity. Each patient acted as their own control, thereby minimizing inaccuracies in case-control matching and variability in adherence between patients on metformin vs. placebo. We safely used a maximum daily dose of 2 g, whereas data from adults with type 2 diabetes suggest that a total daily dose of 3 g may be required to maximize the metabolic benefits of metformin therapy (22). Many patients were poorly adherent to prescribed therapy; however, this is not uncommon in the 9- to 18-yr-old age group (23) and reflects real life challenges in managing young people with obesity. Of note, participants were consistent in their adherence to therapy when on metformin and placebo.
Similar to two previous studies assessing the role of metformin in a small number of pediatric subjects with obesity and insulin resistance (11, 12), we found that 6 months of metformin therapy resulted in improvement in anthropometry, fasting serum glucose, and insulin but not insulin sensitivity. This suggests that simple clinical parameters are useful and sensitive enough to detect changes over relatively short periods. Furthermore, we have shown that metformin-induced weight loss reflects loss of sc rather than VAAT. Visceral abdominal fat is implicated in the development of insulin resistance in adolescents (24, 25), and loss of visceral, rather than sc, fat in adults has greater metabolic benefits (26). Therefore, the lack of improvement in insulin sensitivity after 6 months of metformin therapy in our study may reflect inadequate visceral fat loss. Another explanation is the questionable ability of the minimal model technique to detect small changes in insulin sensitivity in severely obese patients (27). In addition, whereas the physiological and dynamic nature of the minimal model technique make it an attractive tool, these features may contribute to a potentially large coefficient of variation (27), thereby affecting accuracy of longitudinal data.
Puberty is a time of physiological insulin resistance (28, 29, 30, 31, 32, 33), and six of the 22 patients underwent a change in their pubertal stage over the course of the study. Whereas this may have confounded the effect of metformin on insulin sensitivity measures, it is not possible to determine whether insulin resistance may have been worse had the patient not been on metformin. The patient numbers in this study were insufficient to statistically assess the effect of pubertal stage on response to metformin therapy.
The primary mechanism of action of metformin is by suppression of hepatic glucose production through activation of the insulin receptor, preferentially through insulin receptor substrate-2 (34, 35). However, whether metformin specifically improves peripheral insulin sensitivity in addition to suppression of hepatic glucose has not been consistently demonstrated in previous clinical studies involving adult patients. Some studies using the hyperglycemic clamp method have shown increased Kg, implicating muscle as the main site of metformin action (36), although others have not shown improvement in insulin-mediated Kg (34, 37). We used the minimal model technique of assessing insulin sensitivity because it enables determination of insulin sensitivity, both insulin dependent (insulin sensitivity) and insulin independent (Sg), as well as insulin secretion (AIR) from a single 3-h test. However, we did not find significant differences in the effect of 6 months of treatment with metformin over placebo in any of these parameters. Recent studies indicate that metformin may have antiinflammatory and lipolytic effects mediated through adipocytokines (8, 38, 39). We did not address this potential mechanism of action of metformin; however, it would be important to consider in future studies.
The alarming rise in childhood obesity and its metabolic complications is well described, and public health issues in relation to primary prevention are vital. At the same time, the clinician faced with young patients with insulin resistance and obesity needs treatment options that are safe and effective. Metformin therapy has beneficial effects on body composition and fasting insulin. Although 6 months of therapy may not be sufficient to have an impact on visceral adipose tissue loss and insulin sensitivity, ethical considerations and patient participation may make longer-term studies difficult to conduct. We did not specifically address the role of dietary or exercise interventions, and the combination of lifestyle and pharmacological interventions to reduce the morbidity of high-risk patients needs to be assessed in future studies.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online April 4, 2006
Abbreviations: AIR, Acute insulin response; BMI, body mass index; DI, disposition index; DXA, dual-energy x-ray absorptiometry; Kg, glucose disposal; MRI, magnetic resonance imaging; Sg, glucose effectiveness; VAAT, visceral abdominal adipose tissue.
Received February 2, 2006.
Accepted March 29, 2006.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. R Desilets, S. Dhakal-Karki, and K. C Dunican Role of Metformin for Weight Management in Patients Without Type 2 Diabetes Ann. Pharmacother., June 1, 2008; 42(6): 817 - 826. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Velasquez-Mieyer, C. P. Neira, R. Nieto, and P. A. Cowan Review: Obesity and cardiometabolic syndrome in children Therapeutic Advances in Cardiovascular Disease, October 1, 2007; 1(1): 61 - 81. [Abstract] [PDF]
|
|
|
|
|
|
K. C Dunican, A. R Desilets, and J. K Montalbano Pharmacotherapeutic Options for Overweight Adolescents Ann. Pharmacother., September 1, 2007; 41(9): 1445 - 1455. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. Ponder and M. A. Anderson Childhood Obesity: Practical Considerations for Prevention and Management Diabetes Spectr, July 1, 2007; 20(3): 148 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Freemark Pharmacotherapy of Childhood Obesity: An evidence-based, conceptual approach Diabetes Care, February 1, 2007; 30(2): 395 - 402. [Full Text] [PDF]
|
|
|
|
 |
|
 E. T. Beckley Weight-Loss Drugs Appropriate for Some Morbidly Obese Teens DOC News, August 1, 2006; 3(8): 9 - 9. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
