Abstract
Purpose/Background
There has long been a clinical belief that bupropion exacerbates anxiety. The purpose of the current retrospective study is to compare anxiety severity over time in those prescribed selective serotonin reuptake inhibitors (SSRIs) versus bupropion.
Methods/Procedures
Archival data (N = 8457) from patients receiving psychiatric care from a national tele-mental health company were used. Propensity matching was used to create SSRI and bupropion groups using 17 covariates. These samples were then compared using repeated measures analysis of variance on Generalized Anxiety Disorder Scale 7 scores at start of treatment, 6 weeks, and 12 weeks.
Findings/Results
The SSRI and bupropion groups were significantly different across a number of variables. In the entire sample, the bupropion group had significantly greater anxiety levels. However, for propensity-matched comparisons, there were no significant interactions between group and time (ie, groups did not differ and improved comparably over time).
Implications/Conclusions
Using propensity matching, there were no differences in anxiety outcome between those prescribed selective serotonin reuptake inhibitor versus bupropion across 12 weeks of treatment.
Key Words: bupropion, antidepressant, anxiety, depression, selective serotonin reuptake inhibitors
The only medications currently Food and Drug Administration–approved for generalized anxiety disorder (GAD) are the selective serotonin reuptake inhibitors (SSRIs) paroxetine and escitalopram, and the selective serotonin-norepinephrine reuptake inhibitors (SNRIs) duloxetine and venlafaxine. In clinical practice, virtually all SSRIs, SNRIs, tricyclic antidepressants, and “atypical” antidepressants are used to treat GAD and other anxiety disorders.1 One exception is the norepinephrine-dopamine reuptake inhibitor bupropion, which has several advantages over other antidepressants: it is less likely to cause weight gain or sexual dysfunction, is not sedating, and is less likely to induce mania.2 However, many clinicians believe that it will exacerbate anxiety3 and avoid it for patients with comorbid depression and anxiety.4
In an analysis of 10 pooled studies, Papakostas et al5 found a small advantage for SSRIs versus bupropion in treating anxious depression (ie, depressed individuals with high concurrent anxiety) for response rates, but not for remission. However, various meta-analyses and studies comparing SSRIs with bupropion have found comparable rates of improvement in anxiety.6,7 The largest relevant meta-analysis to date, which included 2890 patients across 10 studies, found that bupropion provided comparable improvement in anxiety symptoms compared with SSRIs6; all included studies used the Hamilton Depression Rating Scale (HDRS) Anxiety-Somatization Factor.
This article attempts to confirm earlier findings that bupropion is an effective treatment for anxiety using a larger and more homogenous dataset with outcomes measured by the GAD Scale 7 (GAD-7),8 a self-report scale that is more commonly used in clinical practice and which has high validity, sensitivity, and specificity for detecting GAD.9 The goal of this study was to compare anxiety over time in those prescribed SSRIs versus bupropion. Our hypothesis is that both groups will provide similar reductions in anxiety.
MATERIALS AND METHODS
Participants
Participant data used in the current investigation were obtained from a national mental health telehealth company (ie, Brightside Health) and consisted of 8457 US-based adult patients, aged 18 to 82 (mean age, 32.52; SD, 8.75) receiving psychiatric care for anxiety or depression between November 2018 and January 2022. Participants were eligible if they (a) were prescribed either a single SSRI or bupropion and maintained on that same medication for 12 weeks, (b) had complete outcome data, and (c) were not prescribed any other psychiatric medications during the 12-week study period. Patients at high risk for suicide, patients with primary substance use disorder, and patients with psychosis or in need of emergency psychiatric services at the initial evaluation were not eligible.
Procedure
All study procedures were approved by the WCG Institutional Review Board for the retrospective analysis of patient data obtained by Brightside Health as part of routine clinical care. Enrolled Brightside Health patients complete an initial digital intake that includes clinically validated measures of depression and anxiety, as well as questions about clinical presentation, medical history, and demographics. All Brightside Health patients are required to complete baseline and intake questionnaires. During a patient's first session, a licensed professional prescribed psychiatric medication(s) for each patient. Over the course of treatment, patients communicated with their provider both asynchronously via messaging and synchronously via video telehealth sessions. Assessments were completed at baseline/intake and periodically thereafter. Surveys were administered digitally through an email prompt. Survey completion at baseline, 6 weeks, and 12 weeks were required for participation.
Measures
The Patient Health Questionnaire 9 (PHQ-9)10 is a 9-item self-report measure used to assess the severity of depressive symptoms present within the prior 2 weeks as outlined by Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria. Respondents rate items on a 4-point Likert scale (0–3) and total scores range from 0 to 27, with scores of 10 or above falling into the clinical range.11 Because provider diagnoses were not available for most participants, criteria for major depressive disorder was considered met if 5 or more of the 9 items were endorsed at least “more than half the days” in the past 2 weeks and one of the symptoms was depressed mood or anhedonia.10 This approach was intended to approximate DSM criteria—that is, at least 5 or more symptoms and at least one of the symptoms must be either depressed mood or loss of interest/pleasure.
The GAD-78 is a 7-item self-report instrument used to assess the severity of anxiety symptoms present within the prior 2 weeks as outlined by DSM-5 criteria. Respondents rate items on a 4-point Likert scale (0–3) and total scores range from 1 to 21, with scores of 10 and above falling into the clinical range.11 The GAD-7 has shown strong reliability and validity, demonstrating 89% sensitivity and 82% specificity for GAD at a score of 10 or greater.9
Other variables measured at baseline included self-reported age, sex, education, race/ethnicity, employment status, income, prior episodes of depression (zero, one, or multiple), duration of the current episode, prior antidepressant use (yes/no), comorbid anxiety disorders (to include anxiety disorder unspecified, generalized anxiety disorder, obsessive compulsive disorder, social anxiety disorder, panic disorder, acute stress disorder, or posttraumatic stress disorder), and total number of chronic health conditions endorsed (including arrhythmia, asthma, cancer, hypercholesterolemia, diabetes, heart condition, irritable bowel syndrome or Crohn disease, lung disease, obesity, thyroid disease, seizures, and chronic pain/fibromyalgia).
Interventions
Participants in the SSRI group were prescribed one of several SSRIs (47.0% escitalopram, 32.8% sertraline, 16.8% fluoxetine, 3.5% other). Dosage adjustments were made based on participant responses to the GAD-7, PHQ-9, and other assessments, as well as virtual visits between patients and providers, and remained within standard therapeutic ranges. Because this was a naturalistic study, dosages were not controlled and varied to meet individual needs. Patients were not prescribed any other psychiatric medications during the 12-week study period.
Data Analyses
Data analyses were performed via SPSS, Version 28. Anxiety severity as measured by the GAD-7 at baseline, 6 weeks, and 12 weeks were compared between the SSRI group and the bupropion group using repeated measures analysis of variance (ANOVA) in the overall sample. Comparisons between groups were made using t tests for continuous variables and χ2 analyses for categorical and evaluated at P < 0.01, two-tailed. Then, propensity matching was done based on a priori variables collected at baseline that might potentially affect outcome.12 This approach attempts to replicate a randomized trial by obtaining treatment groups with similar distributions of known covariates.13 Propensity score matching was done using nearest neighbor pair matching without replacement with a 0.00001 caliper, which is effective at reducing systematic differences between groups.14 Included variables were as follows: age, sex, race/ethnicity, education level, employment status, income level, census-defined region of the country, past antidepressant use (yes/no), number of chronic medical conditions, current smoker (yes/no), duration of current psychiatric illness, baseline depression and anxiety symptom severity, current participation in psychotherapy, frequency of nonwork use of digital technology on a scale from 0 to 4, and current depression and/or anxiety diagnosis. Standardized mean differences were examined to examine balance after matching, with differences of 10% or less indicative of successful matching.15 Mauchly test was used to test the sphericity assumption, with the Greenhouse-Geisser correction16 used for violations.
RESULTS
There were 6528 (77.2%) individuals in the SSRI group and 1929 (22.8%) in the bupropion group. As can be seen in Table 1, these groups differed significantly on several variables, including age, baseline PHQ-9 and GAD-7 scores, sex, prior antidepressant use, education level, race/ethnicity, annual income, depression and anxiety diagnoses, and concurrent psychotherapy treatment. A repeated measures ANOVA comparing the groups on anxiety severity across time revealed that GAD-7 scores significantly decreased over time within each treatment group, F (2, 8454) = 4209.22, P < 0.000, η2 = 0.50. As can be seen in Figure 1, the bupropion group had significantly lower anxiety severity at each time point, P < 0.001, F (2, 8454) = 191.21, P < 0.001, η2 = 0.04.
TABLE 1.
Characteristics of SSRI and Bupropion Groups
| Overall Cohort | Propensity-Matched Cohort | |||||
|---|---|---|---|---|---|---|
| Characteristic | SSRI (n = 6528) | Bupropion (n = 1,929) | d* | SSRI (n = 346) | Bupropion (n = 346) | d* |
| Age | 32.1 (8.6) | 33.8 (8.9) | 0.19 | 32.3 (8.7) | 33.0 (8.5) | 0.08 |
| Baseline PHQ-9 | 16.5 (5.5) | 17.3 (4.7) | 0.14 | 17.7 (5.3) | 17.4 (5.0) | 0.06 |
| Baseline GAD-7 | 14.9 (4.5) | 11.5 (5.1) | 0.75 | 14.9 (4.1) | 14.8 (4.2) | 0.03 |
| No. chronic medical conditions | 0.5 (0.8) | 0.5 (0.8) | 0.06 | 0.5 (0.8) | 0.6 (0.8) | 0.09 |
| Frequency (%) | Frequency (%) | |||||
| Sex | 0.04 | 0.00 | ||||
| Male | 2032 (31.1) | 690 (35.8) | 114 (32.9) | 114 (32.9) | ||
| Female | 4496 (68.9) | 1239 (64.2) | 232 (67.1) | 232 (67.1) | ||
| Region of the country | 0.0 | 0.08 | ||||
| Midwest | 1219 (18.7) | 366 (19.0) | 66 (19.1) | 55 (15.9) | ||
| Northeast | 1348 (20.6) | 419 (21.7) | 80 (23.1) | 66 (19.1) | ||
| South | 2212 (33.9) | 661 (34.3) | 116 (33.5) | 128 (37.0) | ||
| West | 1749 (26.8) | 483 (25.0) | 84 (24.3) | 97 (28.0) | ||
| Prior antidepressant treatment | 3734 (57.2) | 981 (50.9) | 0.1 | 144 (41.6) | 170 (49.1) | 0.08 |
| Prior depressive | 0.01 | 0.04 | ||||
| Episodes | ||||||
| None | 478 (7.3) | 135 (7.0) | 27 (7.8) | 20 (5.8) | ||
| One | 760 (11.6) | 234 (12.1) | 36 (10.4) | 36 (10.4) | ||
| >One | 3222 (49.4) | 932 (48.3) | 179 (51.7) | 187 (54.0) | ||
| Nonepisodic | 2068 (31.7) | 628 (32.6) | 104 (30.1) | 103 (29.8) | ||
| Education: | 0.05 | 0.05 | ||||
| No high school | 89 (1.4) | 18 (0.9) | 4 (1.1) | 6 (1.7) | ||
| High school | 1972 (30.2) | 503 (26.1) | 107 (30.9) | 105 (30.3) | ||
| Some college | 837 (12.8) | 241 (12.5) | 47 (13.6) | 41 (11.8) | ||
| College degree | 2473 (37.9) | 776 (40.2) | 132 (38.2) | 128 (37.0) | ||
| Graduate degree | 1157 (17.7) | 391 (20.3) | 56 (16.2) | 66 (19.1) | ||
| Race/ethnicity | 0.05 | 0.1 | ||||
| White/Caucasian | 5070 (77.7) | 1534 (79.5) | 257 (74.3) | 274 (79.2) | ||
| Asian | 280 (4.3) | 67 (3.5) | 18 (5.2) | 11 (3.2) | ||
| Hispanic | 541 (8.3) | 131 (6.8) | 26 (7.5) | 20 (5.8) | ||
| Black/African American | 297 (4.5) | 75 (3.9) | 16 (4.6) 6 (1.7) | 16 (4.6) 0 (0.0) | ||
| Native American/Pacific Islander | 38 (0.6) | 17 (0.9) | 1 (0.3) | 4 (1.2) | ||
| Other | 302 (4.6) | 105 (5.4) | 22 (6.4) | 21 (6.1) | ||
| Employed | 0.01 | 0.06 | ||||
| Full time | 4512 (69.1) | 1357 (70.3) | 240 (69.4) | 247 (71.4) | ||
| Part time | 758 (11.6) | 209 (10.8) | 34 (9.8) | 41 (11.8) | ||
| Unemployed | 1258 (19.3) | 363 (18.8) | 72 (20.8) | 58 (16.8) | ||
| Annual income | 0.06 | 0.06 | ||||
| <$30,000 | 1965 (30.1) | 511 (26.5) | 103 (29.8) | 100 (28.9) | ||
| $30–60,000 | 2015 (30.9) | 537 (27.8) | 115 (33.2) | 103 (29.8) | ||
| $60–100,000 | 1318 (20.2) | 449 (23.3) | 60 (17.3) | 75 (21.7) | ||
| >$100,000 | 1230 (18.8) | 432 (22.4) | 68 (19.7) | 68 (19.7) | ||
| Depression diagnosis | 6204 (95.0) | 1889 (97.9) | 0.06 | 339 (98.0) | 334 (96.5) | 0.04 |
| Anxiety diagnosis | 5422 (83.1) | 1086 (56.3) | 0.27 | 297 (85.8) | 288 (83.2) | 0.34 |
| Duration of illness | 0.04 | 0.07 | ||||
| <2 wk | 71 (1.1) | 13 (0.7) | 5 (1.4) | 5 (1.4) | ||
| 2 wk to 2 mo | 715 (11.0) | 203 (10.5) | 31 (9.0) | 32 (9.2) | ||
| 2 mo to 1 y | 1789 (27.4) | 570 (29.5) | 90 (26.0) | 104 (30.1) | ||
| 1 to 2 y | 1140 (17.5) | 369 (19.1) | 62 (17.9) | 71 (20.5) | ||
| >2 y | 2813 (43.1) | 774 (40.1) | 158 (45.7) | 134 (38.7) | ||
| Current smoker | 749 (11.5) | 203 (10.5) | 0.01 | 35 (10.1) | 40 (11.6) | 0.02 |
| Current treatment | 0.04 | 0.01 | ||||
| Medication | 5226 (80.1) | 1613 (83.6) | 283 (81.8) | 286 (82.7) | ||
| Medication + therapy | 1302 (19.9) | 316 (16.4) | 63 (18.2) | 60 (17.3) | ||
| Frequency of technology use, 0–4 | 0.01 | 0.08 | ||||
| Seldom, never | 141 (2.2) | 38 (2.0) | 10 (2.9) | 4 (1.2) | ||
| Rarely | 570 (8.7) | 183 (9.5) | 32 (9.2) | 34 (9.8) | ||
| Few times/week | 1192 (18.3) | 348 (18.0) | 64 (18.5) | 71 (20.5) | ||
| Once/day | 1347 (20.6) | 398 (20.6) | 66 (19.1) | 76 (22.0) | ||
| Multiple times/day | 3278 (50.2) | 962 (49.9) | 174 (50.3) | 161 (46.5) |
Depression and anxiety diagnoses are based on PHQ-9 and GAD-7 criteria, respectively.
*d = standardized difference.
FIGURE 1.
Anxiety severity over time for each group, entire sample (N = 8457).
Because of the differences between groups at baseline, particularly in terms of significantly different baseline GAD-7 scores, propensity matching was used to create matched groups (Fig. 2). Each had 346 participants that did not significantly differ on any of the variables in Table 1. Repeated measures ANOVA comparing the groups on anxiety severity across time revealed that GAD-7 scores decreased significantly over time in both groups, F (2, 689) = 684.61, P < 0.001, η2 = 0.67. There was no significant difference in anxiety at any time point, F (2, 689) = 1.20, P = 0.30, η2 = 0.00.
FIGURE 2.
Anxiety severity over time for each group, propensity-matched sample (N = 692).
DISCUSSION
To confirm earlier findings that bupropion provides similar relief of anxiety symptoms to SSRIs, we conducted an iterative analysis of 8457 adult patients receiving psychiatric care through a national mental telehealth company for depression and/or anxiety who were engaged with treatment and prescribed the same agent for 12 consecutive weeks. Symptoms were monitored with the GAD-7, a self-report scale of anxiety symptoms with high diagnostic validity for generalized anxiety disorder and numerous advantages over the HDRS Anxiety-Somatization Factor used in earlier studies. We performed 2 analyses:
1) the entire 8,457 participant sample,
2) propensity-matched groups from the whole sample, with 346 participants in both the bupropion and SSRI treatment groups,
In each analysis, both bupropion and pooled SSRIs were associated with a statistically significant reduction in self-reported anxiety symptoms as measured by the GAD-7 over the 12-week period. In the propensity-matched groups, there was no statistically significant difference in symptom reduction between the SSRI and bupropion group.
In the whole sample analysis, the bupropion group had a lower GAD-7 score than the SSRI group at each time point, which is most likely due to selection bias; prescribers tend to avoid prescribing bupropion for patients with anxiety, which is reflected in the pronounced difference in initial GAD-7 scores between the 2 groups. There were also likely numerous other reasons for each prescriber's clinical gestalt when choosing what agent to prescribe, as reflected in the demographic differences between the bupropion and SSRI groups as reported in Table 1.
Strengths and Limitations
This study has several notable strengths. The sample size examined here is far larger than the largest meta-analysis to date comparing the effect of bupropion to SSRIs on anxiety.6,17 Our propensity-matched sample is also far larger than any individual study included in the analysis. All patients were followed using the same protocol, which minimizes potential differences caused by different methods of measurement and provides a more accurate assessment. Finally, the present analysis extends over 12 weeks, whereas most previously reported studies end at 8 weeks, about the amount of time SSRIs take to reach their full effect. Ending at 12 weeks minimizes the risk of bias in favor of bupropion due to lagging medication responses.
The primary unresolved limitation of this study is selection bias—patients were not randomly assigned to treatment. This analysis also excluded patients who either changed agents or were lost to follow-up. Therefore, we can only conclude that bupropion is as effective as SSRIs for the treatment of anxiety in patients who are able to tolerate treatment. The meta-analysis by Papakostas et al6 found that bupropion, SSRIs, and placebos all had similar discontinuation rates of about one quarter of participants, although there was no analysis reported to indicate whether bupropion's advantage in this respect was statistically significant. In addition, the analyses with propensity-matched groups excluded a large number of patients, thereby limiting the generalizability of the findings. Because there was no control group, this study was unable to separate the benefit of medication from the benefit of being “in treatment.” This question is beyond the scope of this particular study, but given the controversy in recent years over whether antidepressants are overprescribed or even effective, it is an important question to investigate systematically.18 Another limitation is that there was no examination of other interventions besides SSRI or bupropion for treating depression, which again limits generalizability. Finally, the lack of statistical difference in anxiety scores between groups does not mean they do not exist; in other words, type II error cannot be ruled out.
CONCLUSIONS
Bupropion is just as effective a treatment as SSRIs for anxiety symptoms in patients with comorbid major depressive disorder. The common concern among clinicians that bupropion will worsen anxiety in this population is unfounded. Given the many advantages in terms of tolerability bupropion has over SSRIs, clinicians should consider using bupropion more often.
AUTHOR DISCLOSURE INFORMATION
Z. Poliacoff has no conflicts of interest. H.G. Belanger and M. Winsberg are employees and stockholders of Brightside Health.
DATA AVAILABILITY STATEMENT
The data set analyzed during the current study is not available.
Footnotes
Funding for the current research was provided by Brightside Health, Inc, and the University of South Florida. H.G.B. and M.W. are employees of and own stock in Brightside Health.
Contributor Information
Zachary Poliacoff, Email: zpoliacoff@usf.edu.
Mirène Winsberg, Email: mimi.winsberg@brightside.com.
REFERENCES
- 1.Stahl S. Prescriber's Guide: Stahl's Essential Psychopharmacology. 7th ed. Cambridge: Cambridge University Press; 2020. [Google Scholar]
- 2.Stahl SM Pradko JF Haight BR, et al. A review of the neuropharmacology of bupropion, a dual norepinephrine and dopamine reuptake inhibitor. Prim Care Companion J Clin Psychiatry. 2004;6:159–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Naguy A, Badr BHM. Bupropion-myth-busting! CNS Spectr. 2022;27:545–546. [DOI] [PubMed] [Google Scholar]
- 4.Zimmerman M Posternak MA Attiullah N, et al. Why isn't bupropion the most frequently prescribed antidepressant? J Clin Psychopharmacol. 2005;66:603–610. [DOI] [PubMed] [Google Scholar]
- 5.Papakostas GI Stahl SM Krishen A, et al. Efficacy of bupropion and the selective serotonin reuptake inhibitors in the treatment of major depressive disorder with high levels of anxiety (anxious depression): a pooled analysis of 10 studies. J Clin Psychiatry. 2008;69:1287–1292. [DOI] [PubMed] [Google Scholar]
- 6.Papakostas GI Trivedi MH Alpert JE, et al. Efficacy of bupropion and the selective serotonin reuptake inhibitors in the treatment of anxiety symptoms in major depressive disorder: a meta-analysis of individual patient data from 10 double-blind, randomized clinical trials. J Psychiatr Res. 2008;42:134–140. [DOI] [PubMed] [Google Scholar]
- 7.Rush AJ Trivedi MH Carmody TJ, et al. Response in relation to baseline anxiety levels in major depressive disorder treated with bupropion sustained release or sertraline. Neuropsychopharmacology. 2001;25:131–138. [DOI] [PubMed] [Google Scholar]
- 8.Plummer F Manea L Trepel D, et al. Screening for anxiety disorders with the GAD-7 and GAD-2: a systematic review and diagnostic metaanalysis. Gen Hosp Psychiatry. 2016;39:24–31. [DOI] [PubMed] [Google Scholar]
- 9.Spitzer RL Kroenke K Williams JB, et al. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166:1092–1097. [DOI] [PubMed] [Google Scholar]
- 10.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16:606–613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kroenke K Spitzer RL Williams JBW, et al. The Patient Health Questionnaire Somatic, Anxiety, and Depressive Symptom Scales: a systematic review. Gen Hosp Psychiatry. 2010;32:345–359. [DOI] [PubMed] [Google Scholar]
- 12.Rassen JA Shelat AA Franklin JM, et al. Matching by propensity score in cohort studies with three treatment groups. Epidemiology. 2013;24:401–409. [DOI] [PubMed] [Google Scholar]
- 13.Seeger JD, Kurth T, Walker AM. Use of propensity score technique to account for exposure-related covariates: an example and lesson. Med Care. 2007;45:S143–S148. [DOI] [PubMed] [Google Scholar]
- 14.Austin PC. The relative ability of different propensity score methods to balance measured covariates between treated and untreated subjects in observational studies. Med Decis Making. 2009;29:661–677. [DOI] [PubMed] [Google Scholar]
- 15.Zhang Z Kim HJ Lonjon G, et al. Balance diagnostics after propensity score matching. Ann Transl Med. 2019;7:16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Greenhouse SW, Geisser S. On methods in the analysis ofprofile data. Psychometrika. 1959;24:95–112. [Google Scholar]
- 17.Trivedi MH Rush AJ Carmody TJ, et al. Do bupropion SR and sertraline differ in their effects on anxiety in depressed patients? J Clin Psychiatry. 2001;62:776–781. [DOI] [PubMed] [Google Scholar]
- 18.Cipriani A Furukawa TA Salanti G, et al. Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis. Lancet. 2018;391:1357–1366. [DOI] [PMC free article] [PubMed] [Google Scholar]