In patients with binge eating disorder (BED), lisdexamfetamine appears to be associated with greater abstinence than second-generation antidepressants (RR 1.56, 95 % CI 1.06 to 2.29).
The network meta-analysis «Peat CM, Berkman ND, Lohr KN, ym. Comparative Effe...»1 presents results from the systematic review about the comparative effectiveness of selected treatments for adults with BED. Eligible studies were randomized controlled trials (RCTs) with sample sizes of 10 or more and published in English. Searches identified 4794 potentially relevant citations. The full systematic review included 87 of these publications and this study focused on the 42 studies providing evidence on the comparative effectiveness of treatment among adults with BED.
Participants were individuals of all races, ethnicities, and cultural groups who met DSM-IV or DSM-5 criteria for BED. Interventions included psychological, behavioral, pharmacological, or complementary and alternative treatments or combinations of treatments, and comparators included two or more active treatments.
Four major outcomes were of particular importance: binge-eating outcomes (e.g. abstinence, binge-eating frequency), eating-related psychopathology outcomes (e.g. obsessions, compulsions), weight-related outcomes [e.g. BMI], and general psychological outcomes (e.g. depressive symptoms).
A total of 12 trials provided evidence of comparative effectiveness of pharmacological interventions. A single head-to-head trial compared two second-generation antidepressants: fluoxetine and sertraline. The remaining 11 pharmacological trials contributed to the network meta-analysis of major outcomes. Of these, eight involved second-generation antidepressants (included bupropion, citalopram, duloxetine, escitalopram, fluoxetine, fluvoxamine, and sertraline) and three involved lisdexamfetamine. The evidence of comparative effectiveness of psychological treatments was reported in 18 trials (in 29 articles).
A total of 18 trials examined various forms of CBT (cognitive behavioural therapy), IPT (interpersonal therapy), BWL (behavioral weight loss treatment), dietary approaches, and inpatient interventions for managing BED. These 18 trials presented 26 treatment comparisons. Of these, four were replicated in more than one trial and 22 were confined to a single trial each.
Comparative effectiveness of second-generation antidepressants versus Lisdexamfetamine
A total of 725 participants (517 lisdexamfetamine and 208 second-generation antidepressants) constituted the treatment arms across these trials. Lisdexamfetamine was associated with statistically significant greater abstinence than second-generation antidepressants, relative risk 1.56 (95 % CI, 1.06 to 2.29; moderate strength of evidence for greater benefit). Differences between second-generation antidepressants and lisdexamfetamine were not statistically significant for either binge frequency (mean difference in days per week -0.48; 95 % CI, -1.17 to 0.20; low strength of evidence for no difference) or obsessions and compulsions (mean difference -2.68; 95 % CI, -5.41 to 0.06; low strength of evidence for no difference). Both second-generation antidepressants and lisdexamfetamine improved binge-eating outcomes and eating-related psychopathology when compared with placebo. Network meta-analysis revealed that lisdexamfetamine was superior to second-generation antidepressants only in achieving binge abstinence.
Summary
Network meta-analysis and qualitative results revealed important information on which interventions might be more effective in improving at least one type of outcome for patients with BED; these treatments include lisdexamfetamine, therapist-led CBT, and BWL. These interventions produced reductions in key BED outcomes including binge-eating abstinence and binge-eating frequency, but they did not demonstrate superiority on all relevant outcomes. Lisdexamfetamine was better at increasing binge abstinence than second-generation antidepressants; therapist-led cognitive behavioural therapy was better at reducing binge-eating frequency than behavioural weight loss, but behavioural weight loss was better at reducing weight.
Comment: Most of the patients includen in the studies were owerveight or obese caucasian women. Thus, the applicability of these results to various age groups and ethnic groups, men and people of normal weigt remains unclear. Most of the comparisons were head-to-head. A long follow-up was only occasionally possible.