Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Obesity (Silver Spring). Author manuscript; available in PMC 2009 May 1.
Published in final edited form as:
PMCID: PMC2674126

Changes in Symptoms of Depression with Weight Loss

Results of a Randomized Trial


Recent studies of rimonabant have re-awakened interest in the possible adverse psychiatric effects of weight loss, as well as of weight-loss medications. The present study examined changes in symptoms of depression in 194 obese participants (age = 43.7 ± 10.2 yr; BMI = 37.6 ± 4.1 kg/m2) in a 1-yr randomized trial of lifestyle modification and medication. Participants were assigned to: 1) Sibutramine Alone; 2) Lifestyle Modification Alone; 3) Sibutramine Plus Lifestyle Modification (i.e., Combined Therapy) or; 4) Sibutramine Plus Brief Therapy. Participants completed the Beck Depression Inventory (BDI-II) at baseline and weeks 6, 10, 18, 26, 40 and 52. At 1 yr, participants in Combined Therapy lost the most weight and those in Sibutramine Alone the least (12.1 ± 8.8% vs. 5.5 ± 6.5%; p < 0.01). Mean BDI-II scores across all participants declined from 8.1 ± 6.9 to 6.2 ± 7.7 at 1 yr (p < 0.001), with no significant differences among groups. Despite this favorable change, 13.9% of participants (across the 4 groups) reported potentially discernible increases (5 or more points on the BDI-II) in symptoms of depression at week 52. They lost significantly less weight than participants in the rest of the sample (5.4 ± 7.8% vs. 9.0 ± 7.8%, respectively, p < 0.03). The baseline prevalence of suicidal ideation was 3.6%. Seven new cases of suicidal ideation were observed during the yr, with three in Lifestyle Modification Alone. Further research is needed to identify characteristics of obese patients at risk of negative mood changes (and suicidal ideation) in response to behavioral and pharmacologic therapies.

A recent meta-analysis concluded that the weight loss medication rimonabant was associated with an increased risk of adverse psychiatric events (1). In June 2007 a U.S. Food and Drug Administration panel expressed concern about the drug’s association with suicidal ideation and behavior (2); and in November 2008 rimonabant’s manufacturer decided to discontinue development of the medication. The mechanisms by which rimonabant may increase the risk of adverse emotional responses are not well understood. Moreover, studies to date have not isolated the potentially separate effects on mood of medication versus weight loss.

Numerous studies reported that weight loss, achieved with diet and exercise (i.e., lifestyle modification), was associated with mean improvements in symptoms of depression, when mood was measured before and after weight reduction using standardized measures such as the Beck Depression Inventory (BDI) (3-7). Similarly positive findings were reported when weight loss was induced by lifestyle modification in combination with medications including sibutramine (8-11), orlistat (11, 12), and fenfluramine (13, 14). None of these studies, however, reported changes in mood at more frequent intervals. Furthermore, none described potentially adverse changes in psychosocial functioning in individual participants, which may have been overlooked by examining only mean values.

The present study investigated changes in symptoms of depression in obese individuals who participated in a 1-year randomized weight loss trial, described previously (15), which compared the efficacy of lifestyle modification alone, sibutramine alone, and the combination of the two approaches. Mood was assessed seven times during the year using the BDI-II (16), which includes an item on suicidal ideation. The study provided an opportunity to determine whether different weight-loss approaches (i.e. lifestyle modification versus sibutramine) are associated with differential changes in mood and whether some participants experience negative emotional responses to weight loss that are not captured by traditional before- and after-treatment assessments.



The 194 participants (162 women and 32 men) were selected from a larger parent study (N = 224), described previously (15). They were included in the sub-study if they completed the BDI-II at baseline and on at least one subsequent occasion. Thirty participants from the parent study did not meet this requirement. Participants had a mean (± SD) age of 43.7 ± 10.2 years, weight of 105.8 ± 16.4 kg, and body mass index of 37.6 ± 4.1 kg/m2. Additional baseline characteristics are summarized in Table 1. Exclusion criteria included: uncontrolled hypertension; cerebrovascular, cardiovascular, renal or hepatic disease; type 1 or 2 diabetes; pregnancy or lactation; a history of anorexia or bulimia nervosa; the use of weight-loss medications; or a weight loss ≥ 5 kg in the preceding 6 months. Participants with clinically significant psychiatric conditions were excluded, including those taking anti-depressant or other psychiatric medications. Depression was assessed by the BDI-II and an interview with a clinical psychologist (DBS, LGW, TAW). Participants typically were excluded if they had a BDI-II score ≥ 25 or indicated, during the interview or on the BDI-II, that they experienced suicidal ideation. These exclusions could be overridden by the examiners if they believed that the elevated depression score was specifically related to participants’ distress about their obesity or that suicidal ideation was not clinically significant (i.e., in the past, or passive and fleeting, with no intent). Eleven of the 194 participants in the current sample were admitted to the study on these bases. This study was approved by the University of Pennsylvania’s Institutional Review Board.

Table 1
Baseline characteristics of participants in four treatment groups.

Treatment Conditions

Participants were randomly assigned to one of four treatment interventions that have been described previously (15) and are only briefly summarized here. All participants were instructed to eat a 1200-1500 kcal/d balanced diet and to gradually increase their physical activity to 30 minutes a day most days of the week.

Sibutramine Alone (n = 45)

These participants initially received sibutramine (5 mg/d) which was titrated up to 15 mg/d over a 6-week period. They met with a primary care provider (for 10-15 min) at weeks 1, 3, 6, 10, 18, 26, 40 and 52. The provider assessed their blood pressure and pulse and asked about any symptoms. Participants were given a pamphlet on healthy eating and activity habits but did not receive specific instructions for changing their behavior.

Lifestyle Modification Alone (n = 47)

Participants in this intervention attended weekly group meetings from weeks 1-18, every other week sessions from weeks 20-40, and a follow-up visit at week 52. Meetings lasted 90 minutes and were led by psychologists who followed the LEARN Program for Weight Control (17), copies of which were provided to group members. Participants were instructed to keep daily records of their food intake and physical activity for at least the first 18 weeks. Meetings from weeks 20 to 40 were conducted using the Weight Maintenance Survival Guide (18)

Combined Therapy (n = 53)

These participants were provided the combination of sibutramine and group lifestyle modification (i.e., Combined Therapy). Thus, they received sibutramine and attended medical visits on the same schedule as participants treated by Sibutramine Alone, while also attending group sessions on the same schedule as participants in the Lifestyle Modification Alone group. These meetings followed a version of the LEARN Program adapted for use with sibutramine (19), as well as the Weight Loss Maintenance Survival Guide.

Sibutramine Plus Brief Therapy (n = 49)

Participants in this group received sibutramine (with the same titration schedule) and visited a primary care provider for 10-15 min at weeks 1, 3, 6, 10, 18, 26, 40, and 52. They also received the two treatment manuals (18, 19) and, similar to participants in the two group lifestyle modification conditions, were instructed to complete food and activity records, which they reviewed with their providers. This treatment condition was designed to determine whether primary care providers could provide effective lifestyle counseling during brief visits.



Weight was assessed at all treatment visits using a digital scale (model 6800A, Detecto). Participants were dressed in light clothing.


Symptoms of depression were assessed at a screening visit and then at weeks 0, 6, 10, 18, 26, 40, and 52 using the BDI-II, a 21-item scale that assesses mood over the previous two weeks. Total scores range from 0-63, with higher scores indicating greater symptoms of depression. Scores of 0-13 reflect minimal (i.e. subclinical) symptoms, whereas values of 14-19, 20-28 and > 29 indicate mild, moderate, and severe symptoms of depression, respectively (16). Item #9 assesses suicidal ideation, as judged by the following responses: (a) “I do not have any thoughts of killing myself;” (b) ”I have thoughts of killing myself, but I would not carry them out;” (c) ”I would like to kill myself;” and, (d) ”I would kill myself if I had the chance.” For the purposes of this study, respondents who endorsed items b, c, or d on this question were deemed to have suicidal ideation.

Psychiatric history

The Weight and Lifestyle Inventory (20) was administered at baseline to assess psychiatric history. Participants were judged to have a psychiatric history if they responded “yes” to the question, “Have you ever had any problems at anytime with depression, anxiety, or other emotions that disrupted your normal functioning?”

Adverse events

Adverse events, including psychiatric complications, were recorded in response to participants’ spontaneous reports and in response to regular monitoring of blood pressure and pulse rate, as described previously (15). After the screening assessment, the BDI-II was not the principal method used to monitor changes in mood, but instead was employed as an outcome measure. Participants’ affect was monitored during interactions with treatment staff who included clinical psychologists (who lead the group treatment sessions) and primary care providers (who conducted the Sibutramine Alone and Sibutramine Plus Brief Therapy sessions). Research coordinators typically reviewed the results of the BDI-IIs, when collected at each assessment period, and informed treatment staff of participants who acknowledged suicidal ideation. Treatment staff met weekly (and later twice monthly) to discuss any clinical concerns observed in patients.

Statistical Analyses

Changes in weight and total BDI-II scores from baseline to week 52 were examined using a repeated measures analyses of variance (ANOVA; treatment group x time); the last observation carried forward (LOCF) method was used in the case of missing data. Chi-square analyses were used to compare differences among treatment groups in both decreases and increases of 5 or more points on the BDI-II. This degree of change was considered to be potentially discernible to participants (i.e., a recognizable difference in mood state) because it represents more than half a standard deviation and more than half the range separating mild and moderate symptoms of depression. These analyses were conducted comparing baseline scores with those at weeks 6, 10, 18, 26, 40, and 52. Changes in suicidal ideation were examined in a similar manner, using dichotomous criteria (i.e., did or did not affirm any of the three responses suggestive of suicidal ideation). Analyses of covariance (that controlled for the effect of treatment group) were used to compare changes in depression and weight among participants who were categorized at baseline with minimal, mild, and moderate symptoms of depression.


Eighteen of 194 (9.3%) participants did not complete the present study by providing a measured weight at week 52. Characteristics of drop-outs are discussed in a later section.



Figure 1 shows that at 1 year participants treated by Sibutramine Alone lost 5.5 ± 6.5% of initial weight, compared with 7.8 ± 8.0% for the Lifestyle Modification Alone group, 12.1 ± 8.8% for Combined Therapy, and 7.7 ± 6.9% for participants who received Sibutramine Plus Brief Therapy. Mean losses for the Combined Therapy group were significantly greater (p < 0.01) than those for the Sibutramine Alone group at 1 year. There were no other significant differences among the groups.

Figure 1
Mean (± SE) percent of initial weight lost in the four groups, as determined by a last observation carried forward analysis. Note: The weight losses for each group and the statistical differences differ from those reported in the parent study ...

Mean Changes in Depression

Participants (N = 194) had minimal symptoms of depression at baseline (i.e., week 0), with a mean BDI-II value of 8.1 ± 6.9. An ANOVA revealed no significant differences among the four groups at this time. As shown in Table 2, the mean BDI-II score across the four treatment groups declined at week 52 to an average of 6.2 ± 7.7. There were no significant differences in changes among the four groups; however, scores did decline significantly (p < 0.001) as a function of time. Similarly, there were no significant differences among the four groups in changes in BDI-II scores from baseline to weeks 6, 10, 18, 26, or 40, despite significant (p < 0.001) reductions (across groups) at all time points.

Table 2
Mean (± SD) Beck Depression Inventory scores for participants in the four treatment groups at seven assessment periods.

Case Analysis of Symptoms of Depression

Discernible changes in BDI-II scores at week 52

The data were examined further to identify individuals who may have had discernible decreases, as well as increases, in symptoms of depression (as indicated by changes of 5 or more points on the BDI-II). As shown in Figure 2, 28.6 - 43.4% of participants in the four treatment groups (mean = 35.6%, 69 of 194 participants in the full sample) registered potentially discernible decreases on the BDI at the end of treatment, with no significant differences among the groups. By contrast, 8.5 - 22.5% of participants in the four groups (mean = 13.9%, 27 of the full sample) had potentially discernible increases in symptoms of depression at the end of treatment. Significantly (p < 0.05) more participants in the Sibutramine Plus Brief Therapy group (22.5%) reported increases of this magnitude than in the other three groups.

Figure 2
The percentage of participants in each group who registered potentially discernible increases (worsening) or decreases (improvements) of ≥ 5 points on the Beck Depression Inventory at the end of treatment.

Discernible changes in BDI-II scores at any time

The percentage of participants in each group who reported changes of 5 points or more at any time during the year was examined to capture the highest possible incidence of positive or negative changes. As shown in Figure 3, 38.8 – 57.4% of participants in each group (mean = 48.3%, 94 of the full sample), showed improvements in their symptoms of depression at some point during the study. There were no significant differences among the four treatment groups on this measure. By contrast, 15.6 – 30.2% of participants in the four groups (mean = 23.7%, 46 of the full sample) reported worsening of symptoms of depression at some time during the year. There were no significant differences among the four groups on this measure.

Figure 3
The percentage of participants in each group who registered potentially discernible increases (worsening) or decreases (improvements) of ≥ 5 points on the Beck Depression Inventory at any time during treatment.

Characteristics of patients with increased depression

The 13.9% of participants (i.e., 27 of 194) who reported worsening symptoms of depression at the end of treatment did not differ significantly at baseline from the remaining participants in initial BMI, education, gender, ethnicity, or initial BDI-II score. However, 44.4% of participants (i.e., 12 of 27) who reported worsening of symptoms had a psychiatric history, compared with only 26.9% (i.e., 45 of 167) of the rest of the sample (p < 0.05). They also lost significantly less weight at week 18 than remaining participants (6.9 ± 5.1% vs. 9.3 ± 5.1%, p < 0.05), as they did at week 52 (5.4 ± 7.8%, vs. 9.0 ± 7.8% respectively, p < 0.03).

Case Analysis of Suicidal Ideation

Prevalence of suicidal ideation

The prevalence of suicidal ideation at baseline, as judged by the single BDI-II item, was 3.6% of the entire sample (i.e., 7 of 194). All seven of these participants reported passive suicidal ideation; they acknowledged ”I have thoughts of killing myself, but I would not carry them out.” No participant endorsed the more serious options available (i.e., ”I would like to kill myself,” or ”I would kill myself if I had the chance”). The Combined Therapy and the Sibutramine Plus Brief Therapy groups each contained three participants at baseline who expressed suicidal ideation. However, none of the three Combined Therapy participants continued to report suicidal ideation after the baseline assessment. In the Sibutramine Plus Brief Therapy group, one participant remained passively suicidal at week 40. The participant did not complete the BDI-II at week 52. However, she did complete the study, and no clinical concerns were raised about the participant at this time. The other two participants reported no suicidal ideation at week 52. Only one participant in the Sibutramine Alone group reported suicidal ideation at baseline. She continued to do so at every time point, but no clinical concerns were reported by staff. No participants in the Lifestyle Modification Alone group reported suicidal ideation at baseline. (In the full sample of 224 participants described in the parent study, two additional individuals reported suicidal ideation at baseline. These two participants did not complete BDI-II questionnaires at their subsequent visits and, thus, were not eligible for the current analyses. No clinical concerns were reported by either participant, each of whom completed the study.)

Incident cases of suicidal ideation

Figure 4 presents the incidence (i.e., new cases) of suicidal ideation occurring after baseline. In the three treatment groups that received sibutramine, a total of four new cases of passive suicidal ideation emerged during the year. Two were in the Sibutramine Alone group (at weeks 18 and 52, respectively); no clinical concerns were reported with either of these participants. Two other participants (in the Combined Therapy group) acknowledged passive suicidal ideation on the BDI-II at weeks 26 and 52. No clinical concerns were reported by one participant. The other was determined by a staff psychiatrist to be experiencing significant depression. He was referred to a mental health professional and reported, at a follow-up visit at week 65, that his depression had largely remitted. Three new cases of passive suicidal ideation also were observed in the Lifestyle Modification Alone group - one each at weeks 18, 40, and 52. The male participant who acknowledged suicidal thoughts at week 18 subsequently discontinued the study at week 22 because of scheduling conflicts. No mood disturbance was noted at this time when he was contacted by phone by a staff psychologist. A female participant who reported passive suicidal ideation at week 40 was determined by a staff psychologist to be experiencing significant depression and was prescribed an anti-depressant medication by her physician. Her mood was improved at her visit at week 52 and remained so at a follow-up visit at week 65. No clinically significant concerns were observed in the third participant at week 52 (when she acknowledged passive suicidal ideation on the BDI-II) or when she returned for a visit at week 65.

Figure 4
One-year cumulative incidence of suicidal ideation in each of the four treatment groups during the study.

Characteristics of patients with suicidal ideation

The 14 participants who acknowledged suicidal ideation, either at baseline or later in the year, began the study with significantly higher baseline BDI-II scores than those who did not report such ideation (12.2 ± 8.2 vs. 7.9 ± 6.9, p < 0.03). In addition, 42.9% (i.e., 6 of 14) of the former group reported a psychiatric history, compared to 28.3% (i.e., 51 of 180) of the latter group (p < 0.05). No significant differences between the two groups were observed at baseline in age, gender, education or race. There also were no significant differences between these groups in percent weight lost during the study (9.8 ± 9.2% vs. 8.4 ± 7.8%). Participants with suicidal ideation reported increases in symptoms of depression from baseline to the end of treatment (rising from 12.2 ± 8.2 to 14.2 ± 12.3), while those free of such ideation had declining scores (falling from 7.9 ± 6.9 to 5.5 ± 6.7). The difference between groups was highly significant (p < 0.001). There were no attempted or completed suicides during the study.

Relation of Baseline Depression to Changes in Depression

Secondary analyses examined changes in depression and weight in relation to participants’ baseline level of depression, grouped by “minimal” (0-13), “mild” (14-19) and “moderate” (≥ 20) levels. (Five participants with severe depression were included with the moderate group because of the small number of cases in the former category.) An analysis of covariance (that controlled for treatment group) revealed that participants who had mild or moderate symptoms of depression at baseline had significantly (p < 0.002) greater reductions on the BDI-II at week 52 than did participants who reported minimal symptoms at baseline. In the former two groups, BDI-II scores reached their nadir at week 18 and then increased from weeks 40 to 52, in parallel with increases in body weight (see Figure 5).

Figure 5
a. Changes in Beck Depression Inventory score by participants’ baseline level of depression, categorized as minimal, mild and moderate depression.

Relation of Baseline Depression to Changes in Weight

Examination of the total sample (N = 194) revealed no relationship between baseline BDI-II scores and baseline BMI. Partial correlation analyses (which controlled for treatment group, gender, age, race, and education) revealed no relationship between baseline BDI-II scores and percentage of initial weight lost at any assessment period. Similarly, there were no significant differences in weight loss when participants were grouped by the minimal, mild, and moderate categories of depression, described previously. (The percentage reduction in initial weight did not correlate significantly with the reduction in BDI-II scores, r = 0.13, p < 0.08.)

For the 11 participants who were admitted to the study as a result of overriding the usual exclusion criteria for depression and/or suicidal ideation, the mean baseline BDI-II was 22.7 ± 10.2. At week 52, this value decreased by 10.6 ± 10.5, which was a significantly (p < 0.001) greater decline than the 1.4 point reduction observed in the remainder of the sample. Weight losses of these 11 participants did not differ significantly from the remainder of the sample (8.5 ± 8.7% versus 8.5 ± 7.9%).

Adverse Events and Attrition

Increases in systolic or diastolic blood pressure (> 10 mm) or in pulse rate (> 15% above baseline) were the most common adverse events, observed in a total of 30 participants who took sibutramine. These events were expected prior to study initiation and medication doses were reduced (following an algorithm) until vital signs returned to acceptable levels. As described previously, two cases of significant major depression were observed that required treatment. Five participants discontinued the study because of adverse events that included breast cancer, hepatitis C, Lyme disease, back injury, and a metallic taste in the mouth.


Of the 18 participants who dropped out of the present study, 10 did so because of dissatisfaction with treatment and 3 because of scheduling conflicts. Five participants dropped out because of adverse events (described above). Baseline BDI-II scores of the 18 participants who prematurely left the study did not differ from those who completed it (10.3 ± 5.4 vs. 7.9 ± 7.0). Completers and non-completers did not differ on any demographic variables except for age (44.2 vs. 37.8 years, p < 0.02).


The first of this study’s two principal findings was that a mean weight loss of 8.4% was associated with a statistically significant reduction in BDI-II scores. The mean (N = 194) value fell from 8.1 at baseline to 6.2 at week 52. Reductions in symptoms of depression were observed in all four treatment groups, with no statistically significant differences among conditions. The present findings replicate previous studies that reported improvements in mean BDI-II scores, following weight loss, in participants treated by lifestyle modification alone (3-7) or in combination with fenfluramine (13, 14), sibutramine (8-11), or orlistat (11, 12). Approximately 35.6% of participants in the present trial may have experienced a discernible improvement in mood at the end of treatment, as suggested by a decrease of 5 points or more on the BDI-II. As many as 48.3% of participants experienced a reduction of this magnitude at a minimum of one of the assessment periods.

The second principal finding was that, despite the mean reduction in BDI-II score, 13.9% of the total sample registered an increase of 5 points or more on the BDI-II at week 52, suggesting a discernible worsening in mood in a minority of participants. Moreover, 23.7% of participants registered an increase of this magnitude at one or more assessment periods. Stunkard (21) described a “dieting depression” more than 50 years ago, and a later pilot investigation found that increasing the frequency of mood assessment, beyond pre- to post-treatment, increased the likelihood of observing adverse changes in mood during weight reduction (22, 23). In the present study, participants who experienced worsening symptoms of depression (increases of 5 points or more) at week 52 were significantly more likely to have a self-reported history of psychiatric complications than were participants whose BDI-II scores declined or were unchanged at week 52. In addition, participants with increased BDI-II scores at week 52 lost significantly less weight (at weeks 18 and 52) than the remainder of the study sample, suggesting that smaller losses could have resulted in disappointment or dysphoria (detected on the BDI-II). Alternatively, increased symptoms of depression may have prevented these participants from losing weight as successfully as others. Jain et al. (24), in a study of bupropion, found that an end-of-treatment reduction in BDI-II scores of 50% or more was associated with an increased odds of losing 5% or more of initial weight. We did not observe this finding in a secondary analysis of the present study. However, results of the present study and those of Jain et al. should relieve earlier concerns that larger weight losses are associated with an increased risk of depression (23).

Further research is needed to identify baseline patient characteristics, as well as treatment factors, that may be associated with the increased symptoms of depression that appear to occur in a minority of participants who undergo weight reduction. Such studies should include a control group that is not subjected to dieting and weight loss, in order to identify the normal variation in mood — both positive and negative. (The lack of such a placebo-control group is a clear limitation of the present study.) In addition, psychometric testing should be complemented by the use of a structured clinical interview that is designed to assess the potential clinical significance of the changes in mood, both positive and negative, observed on self-report inventories, such as the BDI-II. This includes characterizing the onset (and remission) of conditions that meet criteria, described in the Diagnostic and Statistical Manual on Mental Disorders (IV-TR) (25), for major depressive episode, dysthymic disorder, bipolar disorder, adjustment disorder with depressed mood, and other complications (including various anxiety disorders). Clinical interview may capture a broader range of adverse psychiatric events than can be detected by a psychometric test (such as the BDI-II) that targets a specific population (23).

The present trial was conducted prior to the reports of suicidal ideation in patients treated by the cannabinoid agent, rimonabant (2). Thus, our study was not designed prospectively to assess suicidal ideation or behavior and relied on a single item from the BDI-II to do so. Moreover, our small sample size was not powered to detect significant changes in a relatively infrequent event (i.e., suicidal cognition). With these limitations acknowledged, during the 1-year trial we nonetheless observed seven incident cases of passive suicidal ideation, as judged by the BDI-II item. Three of these cases occurred in participants who received group lifestyle modification alone, while the other four were in participants treated by sibutramine, either alone (N = 2) or in combination with lifestyle modification (N = 2). In five cases, participants did not report clinically significant distress to study staff. However, in the two remaining cases (one each in Lifestyle Modification Alone and Combined Therapy), participants complained of symptoms of depression and were referred for treatment. Mood improved in both individuals when observed at follow-up visits.

Given that our study did not systematically monitor mood and adverse emotional reactions, as now required by the Food and Drug Administration of studies of cannabinoid and other central nervous system (CNS) agents, the findings reported here may well underestimate participants’ experience of adverse emotional events, including depression and suicidal ideation. This underestimation may have occurred despite the study team’s including four mental health professionals who met regularly to discuss any clinical concerns observed with participants.

The FDA has recommended that future trials of experimental weight loss medications assess mood with the Patient Health Questionnaire-9 (PHQ-9) (26), anxiety with the General Anxiety Disorder-7 (GAD-7) (27), and suicidal ideation with the Columbia Suicide Severity Rating Scale (C-SSRS) (28). Critical scores have been identified on the PHQ-9 and C-SSRS that mandate the patient’s referral to a mental health professional for further evaluation. We believe that these or similar procedures are appropriate in randomized trials of new pharmacologic agents that affect the CNS. Such assessment should identify the small minority of participants who may be at risk of adverse psychiatric complications.

Results of the present study also raise the question of whether, for each new medication tested, rates of suicidal ideation and depression should be assessed in at least one study that includes a group of individuals who are treated by a program of lifestyle modification alone designed to induce a weight loss comparable to that achieved with the experimental medication. Inclusion of such a group would allow investigators to separate the possible adverse effects on mood of the medication per se from those potentially associated with weight loss alone or the severity of caloric restriction. As reported in the present study, the lifestyle modification alone group was associated with three incident cases of suicidal ideation, suggesting that this problem may not be limited to weight loss medications.

Results of the present study also underscore how little empirical research has been conducted on the treatment of obesity in individuals who also suffer from major depression (or other psychiatric disorders that impair daily functioning). We traditionally have excluded such individuals from our randomized trials of lifestyle modification because of concerns that they would not have the emotional resources required for rigorous behavioral treatment (i.e., restricting food intake, exercising, etc.) and that weight loss (with its effects on thyroid and other hormones) could exacerbate their depression (29-31). Trials of weight-loss medications also frequently require the exclusion of patients who are taking anti-depressants and other drugs because of concerns of adverse medication interactions. We generally followed these exclusion criteria in the present trial but did admit 11 patients with passive suicidal ideation and/or elevated BDIII scores. In this investigation, participants with moderate levels of depression (N = 14) lost comparable amounts of weight as individuals with minimal (N = 156) or mild (N = 24) depression, and they also reported large reductions in symptoms of depression at all assessments. These findings suggest that careful study is needed of the behavioral, neuroendocrine, and psychological effects of weight loss in obese patients with current depressive disorders, as well as in persons with a history of these and other psychiatric conditions. In addition to examining depressed individuals who seek to reduce their weight using traditional diet and exercise interventions, study is particularly needed of obese individuals who seek bariatric surgery, given the high rates of depression in these individuals (32-34). Thus, while researchers and clinicians must take care to protect the small minority of individuals who may be at risk of experiencing adverse psychiatric complications with weight loss, they must also identify the conditions under which obese individuals with significant psychiatric complications can safely undertake weight reduction.


This research was supported by grants DK 56124 and DK 065018 from the National Institutes of Diabetes Digestive and Kidney Disease. Abbot Laboratories is acknowledged for providing the medication used in this study.


1. Christensen R, Kristensen PK, Bartels EM, Bliddal H, Astrup A. Efficacy and safety of the weight-loss drug rimonabant: a meta-analysis of randomised trials. Lancet. 2007;370:1706–13. [PubMed]
2. US Food and Drug Administration Advisory Committee FDA Briefing Document: Zimulti (rimonabant) Tablets, 20mg 2007. FDA; Rockville: (assessed Aug 6, 2007).
3. Alger SA, Malone M, Cerulli J, Fein S, Howard L. Beneficial effects of pharmacotherapy on weight loss, depressive symptoms, and eating patterns in obese binge eaters and non-binge eaters. Obes Res. 1999;7:469–76. [PubMed]
4. Gladis MM, Wadden TA, Vogt R, Foster G, Kuehnel RH, Bartlett SJ. Behavioral treatment of obese binge eaters: do they need different care? J Psychosom Res. 1998;44:375–84. [PubMed]
5. Wadden TA, Butryn ML, Byrne KJ. Efficacy of lifestyle modification for long-term weight control. Obes Res. 2004;12(Suppl):151S–62S. [PubMed]
6. Wadden TA, Foster GD, Letizia KA. One-year behavioral treatment of obesity: comparison of moderate and severe caloric restriction and the effects of weight maintenance therapy. J Consult Clin Psychol. 1994;62:165–71. [PubMed]
7. Wing RR, Epstein LH, Marcus MD, Kupfer DJ. Mood changes in behavioral weight loss programs. J Psychosom Res. 1984;28:189–96. [PubMed]
8. Appolinario JC, Bacaltchuk J, Sichieri R, Claudino AM, Godoy-Matos A, Morgan C, et al. A randomized, double-blind, placebo-controlled study of sibutramine in the treatment of binge-eating disorder. Arch GenPsychiatry. 2003;60:1109–16. [PubMed]
9. Elfhag K, Rossner S, Barkeling B, Rooth P. Sibutramine treatment in obesity: initial eating behaviour in relation to weight loss results and changes in mood. Pharmacol Res. 2005;51:159–63. [PubMed]
10. Hainer V, Kunesova M, Bellisle F, Hill M, Braunerova R, Wagenknecht M. Psychobehavioral and nutritional predictors of weight loss in obese women treated with sibutramine. Int J Obes. 2005;29:208–16. [PubMed]
11. Kiortsis DN, Tsouli S, Filippatos TD, Konitsiotis S, Elisaf MS. Effects of sibutramine and orlistat on mood in obese and overweight participants: a randomised study. Nutr Metab Cardiovasc Dis. 2008;18:207–10. [PubMed]
12. Rucker D, Padwal R, Li SK, Curioni C, Lau DC. Long term pharmacotherapy for obesity and overweight: updated meta-analysis. BMJ. 2007;335:1194–9. [PMC free article] [PubMed]
13. Brownell KD, Stunkard AJ. The double-blind in danger: untoward consequences of informed consent. Am J Psychiatry. 1982;139:1487–9. [PubMed]
14. Craighead LW, Stunkard AJ, O’Brien RM. Behavior therapy and pharmacotherapy for obesity. Arch Gen Psychiatry. 1981;38:763–8. [PubMed]
15. Wadden TA, Berkowitz RI, Womble LG, et al. Randomized trial of lifestyle modification and pharmacotherapy for obesity. N Engl J Med. 2005;353:2111–20. [PubMed]
16. Beck A, Steer RA, Brown GK. Beck Depression Inventory-II (BDI-II) Psychological Corporation; San Antonio: 1996.
17. Brownell KD, editor. The LEARN Program for Weight Control. American Health Publishing; Dallas: 1998.
18. Brownell KD, Rodin J, editors. The Weight Maintenance Survival Guide. Brownell & Hager; Dallas: 1990.
19. Brownell KD, Wadden TA, editors. The LEARN Program For Weight Control: Special Medication Edition. American Health Publishing; Dallas: 1999.
20. Wadden TA, Foster GD. Weight and Lifestyle Inventory (WALI) Obesity. 2006;14(Suppl 2):99S–118S. [PubMed]
21. Stunkard AJ. The dieting depression; incidence and clinical characteristics of untoward responses to weight reduction regimens. Am J Med. 1957;23:77–86. [PubMed]
22. Wadden TA, Stunkard AJ, Smoller JW. Dieting and depression: a methodological study. J Consult Clin Psychol. 1986;54:869–71. [PubMed]
23. Smoller JW, Wadden TA, Stunkard AJ. Dieting and depression: a critical review. J Psychosom Res. 1987;31:429–40. [PubMed]
24. Jain AK, Kaplan RA, Gadde KM, Wadden TA, Allison DB, Brewer ER, et al. Bupropion SR vs. Placebo for weight loss in obese partients with depressive symptoms. Obes Res. 2002;10:1049–56. [PubMed]
25. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision. American Psychiatric Association; Washington, DC: 2000.
26. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD: the PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA. 1999;282:1737–44. [PubMed]
27. Spitzer RL, Kroenke K, Williams JB, Lowe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166:1092–7. [PubMed]
28. Posner K, Brent D, Lucas C, Gould M, Stanley B, Brown G, et al. The Columbia Suicide Severity Rating Scale. Columbia University; New York, NY: 2008.
29. Chaput JP, Arguin H, Gagnon C, Tremblay A. Increase in depression symptoms with weight loss: association with glucose homeostasis and thyroid function. Appl Physiol Nutr Metab. 2008;33:86–92. [PubMed]
30. Joffe RT, Blank DW, Post RM, Uhde TW. Decreased triiodothyronines in depression: a preliminary report. Biol Psychiatry. 1985;20:922–5. [PubMed]
31. Wadden TA, Mason G, Foster GD, Stunkard AJ, Prange AJ. Effects of a very low calorie diet on weight, thyroid hormones and mood. Int J Obes. 1990;14:249–58. [PubMed]
32. Sarwer DB, Cohn NI, Gibbons LM, Magee L, Crerand CE, Raper SE, et al. Psychiatric diagnoses and psychiatric treatment among bariatric surgery candidates. Obes Surg. 2004;14:1148–56. [PubMed]
33. Sarwer DB, Wadden TA, Fabricatore AN. Psychosocial and behavioral aspects of bariatric surgery. Obes Res. 2005;13:639–48. [PubMed]
34. Wadden TA, Sarwer DB, Fabricatore AN, Jones L, Stack R, Williams NS. Psychosocial and behavioral status of patients undergoing bariatric surgery: what to expect before and after surgery. Med Clin North Am. 2007;91:451–69. [PubMed]