Antidepressants that increase mitochondrial energetics may elevate risk of treatment-emergent mania

Pharmacotherapy for bipolar depression is underdeveloped, and antidepressants are widely used despite limited evidence and risks including treatment-emergent mania (TEM+), cycle acceleration, and nonresponse. Antidepressant use in BD has increased substantially over recent decades. Meta-analytic and registry data indicate elevated switch risk with antidepressant exposure, especially early in treatment, with risk influenced by younger age, female sex, mixed features, and BD subtype. Bipolar disorder is linked to mitochondrial dysfunction, including reduced expression of electron transport chain (ETC) genes and impaired ATP production, potentially contributing to illness vulnerability. Preclinical models suggest that antidepressants can differentially modulate mitochondrial activity, with acute treatment often increasing mitochondrial respiration. Given these observations, the study aimed to test whether antidepressants that increase mitochondrial energetics are associated with higher rates of TEM+ compared with those that decrease mitochondrial energetics.

Prior clinical evidence shows increased mania switch rates with antidepressants in BD: a meta-analysis reported a 12.5% switch rate with antidepressants vs 7.5% without; a Swedish registry found increased risk with antidepressant monotherapy early in treatment (HR=2.83). Known clinical risk factors for TEM+ include younger age, female sex, mixed symptoms, and BD subtype. Mitochondrial dysfunction is implicated in BD pathophysiology, with reduced ETC gene expression (frontal cortex, hippocampus) and a shift toward glycolysis with reduced ATP production. Preclinical data indicate heterogeneity in antidepressant effects on ETC complexes, with acute antidepressant exposure often upregulating mitochondrial respiration, while chronic effects may diminish or normalize. Specific agents like venlafaxine and paroxetine have been clinically linked to higher switch risk, aligning with a hypothesis that Mito+ antidepressants may elevate TEM+ risk. The literature also highlights that ETC activity is sensitive to stressors (trauma, chronic stress), metabolic conditions (obesity, diabetes), and environmental exposures (tobacco), all common in BD and potential confounders for mitochondrial function.

Design: Retrospective analysis of participants from the Mayo Clinic Bipolar Disorder Biobank with documented antidepressant exposure and clinical outcomes. Diagnoses were confirmed by structured interview. Recruitment occurred at multiple sites (Mayo Clinic, Lindner Center for HOPE/University of Cincinnati, University of Minnesota, Universidad Autónoma de Nuevo León, and UCLA Child), with IRB approvals and informed consent. Clinical data were obtained via the Bipolar Biobank Clinical Questionnaire (BiB-CQ). Exposure classification: Antidepressants were categorized based on preclinical evidence of their impact on mitochondrial electron transport chain (ETC) activity into Mito+ (increase activity) and Mito– (decrease activity). Agents identified included nortriptyline, paroxetine, and venlafaxine as Mito+, and amitriptyline and escitalopram as Mito– (based on a preclinical review of psychotropic effects on mitochondrial function). Outcome definition: Treatment-emergent mania (TEM+) was defined as a DSM manic/hypomanic episode occurring within 60 days of starting or increasing an antidepressant dose. TEM– controls had at least 60 days of antidepressant exposure without a manic/hypomanic episode. Statistical analysis: Initial comparisons of TEM+ rates across potential confounders (sex, age, BD type, psychosis history, psychiatric hospitalizations, comorbidities) used chi-squared tests and two-sample t-tests as appropriate. The primary analysis compared TEM+ rates between Mito+ and Mito– exposures using generalized estimating equations (logit link, symmetric correlation) to account for individuals exposed to both categories. Models adjusted for age at biobank enrollment, sex, and BD subtype (BD/schizoaffective vs BD). As a retrospective study, analyses did not adjust for dynamic mitochondrial influences such as lifestyle factors, trauma, chronic stress, exercise, or diet.

Sample: 692 subjects were analyzed (62.7% female, 41.9% White, mean age 43.0 ± 14.0 years), including 176 TEM+ (25.3%) and 516 TEM– (74.7%). TEM+ participants were younger at enrollment than TEM– (40.6 ± 13.8 vs 43.8 ± 13.9 years; p=0.019). No significant group differences in BD subtype frequencies (TEM+ 76.1% vs TEM– 70.9%; p=0.31) or in history of psychosis or hospitalization rates. Antidepressant exposure: Exposure counts indicated Mito+ antidepressants were more commonly used than Mito–. The rates of exposure to each class were not significantly different across TEM status groups. Primary outcome: TEM+ occurred more frequently with Mito+ vs Mito– antidepressants (24.7% vs 13.5%). Adjusted analysis (age, sex, BD type) showed an odds ratio of 2.21 for TEM+ with Mito+ vs Mito– (p=0.000009), indicating approximately double the risk with antidepressants that increase mitochondrial energetics.

Findings support the hypothesis that antidepressants which increase mitochondrial energetics are associated with higher rates of treatment-emergent mania in BD. This aligns with prior clinical signals implicating venlafaxine and, to a lesser extent, paroxetine in mood switching, consistent with their classification as Mito+. The work introduces a mechanistically informed framework for antidepressant classification based on mitochondrial effects, beyond traditional monoaminergic mechanisms, and suggests a biological pathway by which certain antidepressants might precipitate mania despite concurrent mood stabilizers. The results underscore the relevance of mitochondrial dysfunction in BD pathophysiology and highlight the potential for integrating mitochondrial bioenergetics into risk stratification and treatment selection.

This study provides preliminary clinical evidence that antidepressants increasing mitochondrial energetics are linked to a higher risk of treatment-emergent mania in bipolar disorder. Classifying antidepressants by their mitochondrial effects may aid clinical decision-making and inform pharmacogenomic approaches. Future prospective studies should validate these findings, explore specific drug–mitochondria interactions, incorporate dynamic lifestyle and stress-related factors, and assess genetic and molecular markers of mitochondrial function to refine risk prediction and guide personalized treatment.

Retrospective design; age at TEM+ occurrence not available for modeling. Lack of control for concurrent psychotropic medications and comorbidities that affect mitochondrial function (e.g., diet, BMI, type 2 diabetes, tobacco use, trauma, chronic stress). Medication regimens at the time of TEM+ were not uniformly available. Classification of antidepressants as Mito+ or Mito– was based on heterogeneous preclinical studies with variable designs and mixed results for several drugs, limiting the strength of the categorization. High rates of polypharmacy in BD may confound mitochondrial activity and risk assessment. Other dynamic mitochondrial influencers (exercise, diet, stress) were not included as covariates.