Rtms Reduces Cue Induced Food Craving in Bulimic Disorders
Total Page:16
File Type:pdf, Size:1020Kb
Repetitive Transcranial Magnetic Stimulation Reduces Cue-Induced Food Craving in Bulimic Disorders Frederique Van den Eynde, Angelica M. Claudino, Andrew Mogg, Linda Horrell, Daniel Stahl, Wagner Ribeiro, Rudolf Uher, Iain Campbell, and Ulrike Schmidt Background: Craving or the “urge to consume” is a characteristic of bulimic eating disorders and addictions. Dysfunction of the dorsolat- eral prefrontal cortex (DLPFC) is associated with craving. We investigated whether stimulation of the DLPFC reduces food craving in people with a bulimic-type eating disorder. Methods: Thirty-eight people with bulimic-type eating disorders were randomly allocated to receive one session of real or sham high- frequency repetitive transcranial magnetic stimulation (rTMS) to the left DLPFC in a double-blind procedure. Outcome measures included self-reported food craving immediately after the stimulation session and frequency of bingeing over a 24-hour follow-up period. Results: Compared with sham control, real rTMS was associated with decreased self-reported urge to eat and fewer binge-eating episodes over the 24 hours following stimulation. Conclusions: High-frequency rTMS of the left DLPFC lowers cue-induced food cravings in people with a bulimic eating disorder and may reduce binge eating. These results provide a rationale for exploring rTMS as a treatment for bulimic eating disorders. ␣ϭ Key Words: Addiction, bingeing, craving, eating disorders, prefron- previous study (13) with 80% power at one-sided .05. tal cortex, repetitive transcranial magnetic stimulation Therefore, 38 participants (5 male participants, 33 female partic- ipants), 18 to 60 years old with BN or Eating Disorder Not raving is a precipitant of bingeing behavior (1,2). Mecha- Otherwise Specified-bulimic type (EDNOS-BN, including Binge nisms proposed to underlie this “irresistible urge to con- Eating Disorder) were recruited from the Eating Disorders out- C sume” include hyperactivity of orbitofrontal and anterior patient department at the Maudsley Hospital, London, and by cingulate brain circuits and impaired inhibitory control from lateral e-mail advertisement in King’s College London. Diagnosis was prefrontal circuits (3,4). Imaging data show similar activation pat- established with the eating disorders module of the Structured terns in addictions, obesity, and bulimia nervosa (BN) (4,5). Clinical Interview for DSM-IV, Axis I Disorders/Patient Version High-frequency repetitive transcranial magnetic stimulation (15). People with EDNOS were enrolled if they had at least six (rTMS) activates the cortex underlying the stimulation site and binges over the previous 28 days as assessed with the Eating inhibits more remote areas (6). We hypothesized that applying Disorder Examination-Questionnaire (16). Contraindications to rTMS to the dorsolateral prefrontal cortex (DLPFC) would coun- rTMS were checked with the Adult Safety Screen Questionnaire teract neural processes that underlie craving by stimulating this (17). Exclusion criteria were left-handedness, being on a dose of area and inhibiting the orbitofrontal and anterior cingulate psychotropic medication that had not been stable for at least 14 cortices. Stimulation of the left DLPFC rTMS reduces cigarette days before enrollment, pregnancy, nicotine use exceeding the craving and consumption (7–10), alcohol craving (11), and cocaine equivalent of Ͼ10 cigarettes/day, and substance dependence. craving (12). In healthy women with high levels of food craving, Local ethical committee approval was obtained. Written in- rTMS to DLPFC prevented an increase in food craving following formed consent was obtained from all participants. cue exposure to food (13). One previous rTMS trial in BN did not For the rTMS, we used a Magstim Rapid device, with real and detect differences between real and sham rTMS effects on sham figure eight coils (Magstim, Whitland, Wales, United King- bulimic and mood symptoms (14). dom) and we applied the same (real and sham) rTMS procedure The aim of the present randomized sham-controlled study and parameters previously described (13). was to establish—across diagnoses—whether the effect of left The Hospital Anxiety and Depression Scale (18) and Food DLPFC rTMS extends to cue-elicited food craving in people with Craving Questionnaire-Trait (19) were administered at baseline. any bulimic eating disorder in which bingeing and craving are Outcome measures, administered before and after the rTMS prominent. session, included a 10-cm visual analogue scale (VAS) of “urge to eat” (primary outcome), the Food Craving Questionnaire-State Methods and Materials (FCQ-S) score, and VAS of “hunger,” “tension,” “mood,” and A power calculation indicated that 18 individuals per group “urge to binge eat” (secondary outcomes). A follow-up phone were needed to detect an effect size of d ϭ .85 reported in our call was made 24 hours after the assessment to check whether binges occurred since the rTMS (secondary outcome). To stimulate food craving, participants were shown a From the Institute of Psychiatry (FVdE, AMC, AM, LH, DS, WR, RU, IC, US), King’s College London, London, United Kingdom; and Department of 2-minute film of four segments showing people eating palatable Psychiatry (AMC, WR), Federal University of São Paulo, São Paulo, Brazil. foods and then presented with a buffet of the same foods and Authors FVdE and AC contributed equally to this article. asked to rate their appearance, smell, and taste. Finally, partici- Address correspondence to Frederique Van den Eynde, M.D., Section of pants were asked to rate the five VAS and complete the FCQ-S. Eating Disorders PO59, De Crespigny Park, SE5 8AF, London, United A randomization procedure, stratifying by diagnosis (BN; Kingdom; E-mail: [email protected]. EDNOS-BN), was used. Two researchers were involved; one Received Aug 7, 2009; revised Nov 4, 2009; accepted Nov 11, 2009. conducted the rTMS and the other supervised the food challenge 0006-3223/10/$36.00 BIOL PSYCHIATRY 2010;67:793–795 doi:10.1016/j.biopsych.2009.11.023 © 2010 Society of Biological Psychiatry 794 BIOL PSYCHIATRY 2010;67:793–795 F. Van den Eynde et al. Table 1. Baseline Characteristics for the Real and Sham rTMS Groups measures between the real and sham groups, with baseline measurements as covariates. Real rTMS Sham rTMS (n ϭ 17) (n ϭ 20) Results ϭ ϭ Diagnosis BN (n 10) BN (n 10) Table 1 shows the baseline characteristics of participants by EDNOS (n ϭ 7) EDNOS (n ϭ 10) group; no significant group differences were observed. Age (mean Ϯ SD) 30.5 Ϯ 11.2 years 29.5 Ϯ 8.4 years Thirty-seven participants completed the procedure. One par- Male/Female Ratio 3/14 2/18 Body Mass Index 25.8 Ϯ 11.5 kg/m2 25.0 Ϯ 8.5 kg/m2 ticipant, randomized to real rTMS, dropped out due to discomfort (mean Ϯ SD) after 4 trains of rTMS. Five participants in each group reported Mean Number of Binges/ .53 Ϯ .30 .70 Ϯ .65 side effects immediately after the rTMS session, i.e., a transient 2 ϭ ϭ Day in the Last 28 Days slight headache [ (1) .907; p .763]. Sixteen in each group (EDE-Q) (mean Ϯ SD) reported that they would agree to take part in a treatment study 2 HADS Total Score 15.1 Ϯ 5.7 16.9 Ϯ 9.7 if this study’s results were positive [ (1) ϭ 1.567; p ϭ .211]. HADS Depression Subscore 5.3 Ϯ 3.2 6.7 Ϯ 5.2 Blinding was only partially successful as 88.2% (15/18) of HADS Anxiety Subscore 9.8 Ϯ 3.8 10.3 Ϯ 5.5 those who received real rTMS guessed correctly. In the sham FCQ-T Score 152.1 Ϯ 33.7 160.4 Ϯ 40.0 group, 55% (11/20) believed that they had been given real rTMS Nonsmokers/Smokers Ratio 13/4 15/5 [2(1) ϭ 4.859; p ϭ .028]. Number on 5/17 4/20 On the primary outcome measure of urge to eat VAS, the real Antidepressants fluoxetine (n ϭ 2) fluoxetine (n ϭ 2) rTMS group showed a significant reduction, compared with the venlafaxine (n ϭ 2) citalopram (n ϭ 1) sham group [F(1,37) ϭ 3.925; p ϭ .028 one tailed; Cohen’s d ϭ escitalopram (n ϭ 1) mirtazepine (n ϭ 1) Ϫ.55, upper limit of 95% confidence interval (CI): .005]. This Number of Meals/Day 2.3 Ϯ .3 2.7 Ϯ .2 finding remains significant when the data for the dropout were ϭ ϭ Duration of Illness 0–5 years: n 6 0–5 years: n 12 included. In contrast, the scores on the VAS scales of hunger, ϭ ϭ 5–10 years: n 5 5–10 years: n 2 tension, mood, and urge to binge eat did not differ between the ϭ ϭ 10–15 years: n 3 10–15 years: n 4 groups. The FCQ-S score in the real rTMS group was not Ͼ 15 years: n ϭ 3 Ͼ 15 years: n ϭ 2 significantly different from that in the sham group [F(1,37) ϭ BN, bulimia nervosa; EDE-Q, Eating Disorder Examination-Question- 1.952; p ϭ .086; Cohen’s d ϭϪ.35, upper limit of 95% CI: .19]. naire; ENDOS, Eating Disorder Not Otherwise Specified; FCQ-T, Food Craving However, in the real group, there was a reduction in the FCQ-S Questionnaire-Trait; HADS, Hamilton Depression and Anxiety Scale; rTMS, score (before vs. after rTMS) in that the prepost difference in the repetitive transcranial magnetic stimulation. FCQ-S in the real rTMS group was nearly twice that observed in the sham group (real rTMS 10.4 Ϯ 14.0; sham 5.5 Ϯ 9.4). Table task and questionnaire completion and conducted follow-up 2 shows the means and SDs of all outcome measures before and calls. The latter and the participants were blind to the rTMS (real after rTMS by group.