- Top of page
- Research Methods and Procedures
Objective: To study the safety, tolerability, and potential efficacy of orlistat in adolescents with obesity and its comorbid conditions.
Research Methods and Procedures: We studied 20 adolescents (age, 14.6 ± 2.0 years; body mass index, 44.1 ± 12.6 kg/m2). Subjects were evaluated before and after taking orlistat (120 mg three times daily) and a multivitamin for 3 months. Subjects were simultaneously enrolled in a 12-week program emphasizing diet, exercise, and strategies for behavior change.
Results: Participants who completed treatment (85%) reported taking 80% of prescribed medication. Adverse effects were generally mild, limited to gastrointestinal effects observed in adults, and decreased with time. Three subjects required additional vitamin D supplementation despite the prescription of a daily multivitamin containing vitamin D. Weight decreased significantly (−4.4 ± 4.6 kg, p < 0.001; −3.8 ± 4.1% of initial weight), as did body mass index (−1.9 ± 2.5 kg/m2; p < 0.0002). Total cholesterol (−21.3 ± 24.7 mg/dL; p < 0.001), low-density lipoprotein-cholesterol (−17.3 ± 15.8 mg/dL; p < 0.0001), fasting insulin (−13.7 ± 19.0 μU/mL; p < 0.02), and fasting glucose (−15.4 ± 7.4 mg/dL; p < 0.003) were also significantly lower after orlistat. Insulin sensitivity, assessed by a frequently sampled intravenous glucose-tolerance test, improved significantly (p < 0.02).
Discussion: We conclude that, in adolescents, short-term treatment with orlistat, in the context of a behavioral program, is well-tolerated and has a side-effect profile similar to that observed in adults, but its true benefit versus conventional therapy remains to be determined in placebo-controlled trials.
- Top of page
- Research Methods and Procedures
Obesity among adolescents in the United States is a problem of increasing concern. The results of National Health and Nutrition Examination Survey III suggest that the prevalence of overweight, defined as the age- and sex-specific 95th percentile for body mass index (BMI), has risen from 6% to 14% in adolescents in the last quarter century (1,2). With the increasing prevalence of overweight and obesity, the comorbid conditions associated with excess weight have also become more common among adolescents (3). As with adults, conventional means of weight loss that involve encouraging healthy diet and exercise, are not uniformly successful in adolescents (4,5). Even for comprehensive programs emphasizing moderate calorie restriction, exercise, and a specific behavioral component (6), weight loss results are disappointing. A review of nine such programs for adolescents with an average BMI of ∼35 kg/m2 (7,8,9,10,11,12,13,14) found weight loss ranged from 2 to 11 kg (mean, 6.6 kg). Given that obesity is a chronic disease with serious long-term health consequences, it is important to explore the safety and efficacy of pharmacotherapy to ameliorate overweight and obesity in heavier adolescents who already manifest the comorbid conditions associated with obesity.
Orlistat is a gastric and pancreatic lipase inhibitor that alters energy balance by reducing the absorption of triglyceride and cholesterol from the gastrointestinal tract (15). In adults, orlistat, at the standard dose of 120 mg three times daily (TID), inhibits ∼30% of triglyceride absorption (16). In placebo-controlled studies, adults treated with orlistat for periods as long as 2 years exhibited greater average weight loss, better weight maintenance, lower total and low density lipoprotein-cholesterol (LDL-cholesterol), and improved glycemic control for patients with type 2 diabetes (17,18,19,20,21,22,23,24,25,26,27,28). Orlistat is minimally absorbed (27) and has few known adverse effects of consequence in adults (18); therefore, it has the potential for a favorable risk-benefit ratio in children and adolescents. Use of orlistat in overweight children <16 years of age has not been studied previously. We therefore examined the effects of orlistat treatment in adolescents to preliminarily estimate its safety, tolerability, and efficacy in this population.
- Top of page
- Research Methods and Procedures
Eighty-five percent of subjects completed 3 months of treatment and reported taking 80% of prescribed medication doses. Because some subjects did not attend every session, weekly reports of the adverse effects experienced were obtained for 79% of the 240 patient-weeks comprising the 3 months of the study period. Adverse effects related to increased fat excretion were generally mild and transient, resolving within the first 6 weeks of treatment (see Table 2). All but one subject reported two adverse effects on at least one occasion. Five of the twelve adverse effects queried, increased defecation, soft stools, fatty or oily stools, oily spotting on clothes, and increased flatus, were reported by >50% of subjects. Only one subject (5%) cited intolerance of adverse effects as the reason for withdrawing from the study.
Table 2. Adverse effects
|Effect||Subjects reporting one or more episodes (%)||Weeks during which an episode was reported (%)|
|Pellets in stool||10||1.1|
|Flatus with discharge||40||6.9|
There were no significant changes in the serum levels of the fat-soluble vitamins A, E (after correcting tocopherol levels for the change in lipid concentrations), or K. A small, but significant, drop in 25-hydroxy vitamin D levels was seen at 1 month (14.9 ± 6.8 vs. 10.6 ± 3.9 ng/mL; p < 0.02). This decrease was ameliorated after extra vitamin D supplementation (50,000 U/d for 1 month) was given to three subjects with vitamin D levels below the normal range (<9 ng/mL). These three subjects had an average vitamin D concentration of 7.5 ± 1.6 ng/mL at baseline, at which time they were placed on a multivitamin containing 400 IU of vitamin D. Levels decreased further, to 6.5 ± 1.3 ng/mL at 1 month, and rebounded to 10.1 ± 1.7 ng/mL by 3 months. Fecal-fat excretion increased 11.8 ± 10.9 g/24 hours (p = 0.003) with orlistat treatment (Table 3). Thyroid-stimulating hormone, free thyroxine, apolipoproteins A1 and B, glycosylated hemoglobin, calcium, phosphorous, magnesium, zinc, measures of iron stores, pulse, and blood pressure did not change significantly during the study.
Table 3. Baseline and 3-month data
|Weight (kg)||123.4 ± 43.0||119.0 ± 43.1*|
|BMI (kg/m2)||44.1 ± 12.4||42.2 ± 13.0*|
|Body fat (%)||47.9 ± 5.4||46.9 ± 6.8|
|Lean body mass (kg)||62.4 ± 19.8||61.6 ± 16.4|
|Waist circumference (cm)||114.3 ± 19.6||112.1 ± 21.4†|
|Triceps skinfold (mm)||33.3 ± 10.2||30.3 ± 8.3|
|Cholesterol (mg/dL)||177.8 ± 41.2||156.6 ± 34.2*|
|LDL-cholesterol (mg/dL)||120.2 ± 30.1||102.9 ± 29.5‡|
|HDL-cholesterol (mg/dL)||46.5 ± 10.9||43.0 ± 8.3†|
|HDL/LDL ratio||0.42 ± 0.21||0.45 ± 0.18|
|Triglycerides (mg/dL)||110.2 ± 55.0||113.8 ± 51.8|
|Leptin (ng/mL)||35.6 ± 16.2||26.4 ± 18.0§|
|Fecal fat (g/24 hours)||4.9 ± 1.5||16.7 ± 11.0§|
Subjects’ weight decreased on average by 4.4 ± 4.6 kg (range +4.0 kg to −17.7 kg, p < 0.001; 3.8 ± 4.1% of original body weight), waist circumference decreased 2.4 ± 4.0 cm (p < 0.03) and mean BMI decreased 1.9 ± 2.5 kg/m2 (p < 0.0002), although percentage of body fat and lean body mass as determined by air displacement plethysmography remained unchanged (Table 3). Overall, the weight loss among white subjects was greater than the weight loss among African-American subjects. At the end of 3 months, seven subjects (35% of total; 86% white; 57% girls) lost >5% of initial body weight (Figure 1). Three of these seven subjects (15% of total; 100% white; 67% female) lost >10% of initial body weight.
Figure 1. Individual weight change after three months of study. Subjects who withdrew from the study are represented by ○; those remaining in treatment for the entire 3-month study interval are represented by •.
Download figure to PowerPoint
Lipid levels changed significantly during weight-reduction treatment using orlistat (Table 3). Total cholesterol (−21.3 ± 24.7 mg/dL; p < 0.001), LDL-cholesterol (17.3 ± 15.8 mg/dL; p < 0.0001), and HDL-cholesterol (3.4 ± 5.7 mg/dL; p < 0.05) decreased, whereas the HDL/LDL ratio and serum triglycerides remained unchanged during orlistat treatment. Changes in total- or LDL-cholesterol measurements were not significantly correlated with the amount of weight lost (r ≥ 0.38, p ≥ 0.13). Fasting baseline levels were correlated with the decreases in total-cholesterol (r = 0.54, p = 0.02), but not LDL-cholesterol (r = 0.29, p = 0.26), obtained after three months of orlistat treatment. Fasting serum leptin concentrations decreased by 9.2 ± 10.6 ng/mL (p = 0.01).
Indicators of glycemic control also improved (Table 4). Fasting glucose (−15.4 ± 7.4 mg/dL; p < 0.003) and insulin (−13.7 ± 19.0 μU/mL; p < 0.02) decreased after 3 months of orlistat treatment. The decrease in fasting insulin was significantly correlated with the degree of weight loss (r = 0.68, p = 0.006; Figure 2), but the decrease in fasting glucose was not (r = 0.34, p = 0.23). However, decreases in both 0-hour oral glucose-tolerance test insulin (r = 0.84, p < 0.001) and glucose (r = 0.61, p = 0.01) after 3 months of orlistat treatment were correlated with baseline 0-hour levels. At the 2-hour point of an oral glucose-tolerance test, glucose and insulin values, although somewhat lower, were not significantly different from baseline values. There was no correlation between baseline values and decreases in the 2-hour values after 3 months of orlistat treatment. Insulin sensitivity (SI), derived from the frequently sampled intravenous glucose tolerance test, increased an average of 28% (p < 0.0003); whereas the acute insulin response to glucose (AIRG) did not change significantly. Conversely, the decrease in AIRG was correlated with baseline AIRG (r = 0.64, p = 0.02), whereas the increase in insulin sensitivity was not correlated with the baseline level. Calculated indicators of endocrine pancreatic function also improved. With the homeostatic method of analysis, the insulin resistance index fell significantly (Table 4, p < 0.04). The degree of improvement in the insulin resistance index (r = 0.70, p < 0.002) was correlated with the baseline level as well. However, the change in percentage of β-cell function was not significantly correlated with the baseline value or with weight change.
Table 4. Glucose metabolism
|Glycohemoglobin (%)||5.9 ± 0.9||6.0 ± 1.1|
|SI (mU/min × 10−4)||0.91 ± 0.67||1.17 ± 0.77*|
|AIRG (pmol/L)||1545 ± 693||1536 ± 592|
|IRI||6.55 ± 4.08||5.08 ± 2.91†|
|β-cell function (%)||402 ± 321||323 ± 229|
|Oral glucose tolerance test||Fasting||2 Hours||Fasting||2 Hours|
|Glucose (mg/dL)||101.7 ± 38.3||123.1 ± 64.2||86.3 ± 6.4‡||106.4 ± 21.5|
|Insulin (μU/mL)||34.0 ± 16.3||117.0 ± 71.7||20.3 ± 10.6†||94.3 ± 85.4|
Figure 2. Individual weight loss (•) vs. individual decreases in fasting insulin after 3 months of orlistat treatment were significantly correlated (r = 0.647; p < 0.004).
Download figure to PowerPoint
- Top of page
- Research Methods and Procedures
This open-label pilot trial of the gastrointestinal lipase inhibitor, orlistat, is the first study to explore orlistat's safety, tolerability, and efficacy in an adolescent population. In general, orlistat was well tolerated and no unanticipated adverse events were observed. More adverse effects were reported in this adolescent sample than have generally been reported in the literature regarding studies in adults. This might be attributable to adolescents’ higher sensitivity to the effects of orlistat, the use of a questionnaire specifically designed to probe for the occurrence of specific adverse effects, or it may be related to adolescents’ greater difficulties in adhering to the recommended moderate dietary-fat intake. Anecdotally, it appears that some subjects may have learned to skip their orlistat dose when they consumed a high-fat meal to avoid adverse effects. However, adverse effects did not limit participation or decrease adherence to the medication regimen. Subject retention after 3 months was 85%, and only one subject withdrew from the study because of adverse effects. In addition, pill-count data and self-reports suggested a 75% to 80% medication compliance rate.
The vitamin D deficiency observed in three African Americans was easily corrected through additional supplementation. Because African Americans may be at risk for low vitamin D even before administration of orlistat (46,47,48), it is possible that it will be necessary to monitor for hypovitaminosis D in African-American adolescents who are taking orlistat. Randomized controlled studies in larger numbers of African-American adolescents will be required to determine definitively the need for such monitoring. Although we did not find other fat-soluble vitamin deficiencies over the short-term, it remains unknown whether clinically significant decreases in vitamins A, E, or K will be observed with long-term use.
Because this is the first clinical trial of orlistat in an adolescent population and because no subjects were enrolled in the behavioral weight-reduction program without receiving orlistat, this study can provide no definitive evidence for orlistat's effects on body weight in adolescents; however, some comparisons to previous studies in adults can be made. Mean weight reduction over 12 weeks, 4.4 kg (3.8%), was somewhat less than the amount generally reported in adult studies (17,18,20,21,22,23,25,26). Adult clinical trials typically report an average weight loss of 5.9 to 10.3 kg over 1 to 2 years, with an approximate 6- to 7-kg reduction over the first 3 months. However, one study (49) with the same duration as this study, 3 months, but did not use a comprehensive behavioral weight-reduction program, reported a similar weight loss of 4.7 kg. In addition, a random survey of 1000 results from 20,000 orlistat prescriptions found that 43% of patients had lost <5% of body weight in the first 3 months of treatment (50).
Our results fall within the range of weight loss, 2 to 11 kg, achieved in other adolescent programs that did not use a pharmaceutical agent, although our population is dissimilar in that our subjects were much heavier, having a mean BMI of 44 kg/m2 vs. 30 to 35 kg/m2 (7,8,9,11,12,13,14,51). In addition, there are two reports in adolescents of the use of metformin, a dimethylbiguanide, antihyperglycemic agent, which has promoted weight loss in adults. In a randomized, double-blind, placebo-controlled, 8-week trial of metformin in 24 adolescents (age, 14 to 16 years; mean BMI, 41 kg/m2), Kay et al. (52) achieved a weight loss of 6.5 ± 0.8% in the treatment group who took 850 mg of metformin twice a day in conjunction with a 1500 to 1800 kcal/d diet vs. a loss of 3.8 ± 0.4% in the placebo group with diet alone. Conversely, Freemark, et al. (53) conducted a randomized, double-blind, placebo-controlled, 6-month trial of 500 mg metformin taken twice daily vs. placebo without any dietary prescription in a group of 29 adolescents (age, 12 to 19 years; BMI, >30 kg/m2). In this study, the differences between groups were very small. The treatment group lost 1.3% of body weight vs. a gain of 2.3% of body weight in the control group. However, both studies obtained significant decreases in fasting glucose and insulin (52,53). It remains to be seen under which conditions, and at what dosages, pharmaceutical agents will function best as adjunctive weight-loss agents in severely obese adolescents.
There are several potential explanations why weight loss in this study was somewhat lower than typically found in adult studies. First, the average initial BMI in adult studies was ∼36 kg/m2, whereas the average BMI in this adolescent study was 44.1 kg/m2. Individuals with greater BMI may have more difficulty making the dietary and activity changes necessary to lose weight successfully. Second, adult studies have generally been designed with a 4- to 6-week lead-in period before randomization when all of the subjects were following the recommended diet and exercise program while taking placebo. These lead-in periods allowed a specific level of medication compliance to be an inclusion criterion. Thus, adult studies may have eliminated subjects anticipated not to have success with weight reduction and may have increased, to some extent, the magnitude of weight loss observed in both placebo and orlistat study arms. The present study did not exclude subjects for poor study compliance because the purpose was to examine orlistat use among adolescents with obesity-related comorbid conditions. Our results may, therefore, be more representative of orlistat's effectiveness among the overweight adolescents for whom pharmacotherapy would be recommended (i.e., those who are heavy enough to suffer from an obesity-related comorbid condition and who have also previously failed more conventional methods used alone).
The modest reductions in adolescents’ body weight observed in the present study improved obesity-related comorbid conditions. As seen in adult studies (24,54,55), there were improvements in several cardiovascular risk factors including LDL- and total-cholesterol levels. The improvements in fasting glucose and insulin we observed also agree with those reported in the literature for adults (22,26). Improvements in cholesterol were not significantly correlated with the amount of weight loss, supporting the growing evidence that there are weight-independent, pharmacological, lipid-lowering effects of orlistat (55,56).
This study is limited by its open-label design, short duration, and small sample size. We cannot separate the effects of the behavioral weight-loss program from the effects of orlistat without a placebo-controlled trial. We also cannot determine the significance of the difference in response between whites and African Americans. Although these limitations prevent delineation of the magnitude of orlistat's effects on obese adolescents’ body weight, they do not affect the purposes of this study, which were to determine if orlistat was well tolerated and safe to be taken by an adolescent population still growing and advancing through puberty. We believe the data of the present study support the contention that orlistat has a safety and tolerability profile for adolescents that is similar to its profile for adults.
In conclusion, orlistat appears to be a relatively safe and well-tolerated adjuvant weight loss therapy for use in an adolescent population. However, there may be a somewhat greater incidence of initial gastrointestinal adverse effects in adolescents, and hypovitaminosis D may be observed despite vitamin supplementation in some Africa-American adolescents taking orlistat. Our findings support the need for further randomized, double-blind, placebo-controlled trials of this drug in a large, multiracial adolescent population. Therefore, the use of orlistat in individuals <16 years of age continues to be experimental, and there is no FDA approval of the drug for use in children and adolescents at this time.