Bipolar disorder is one of the most disabling psychiatric conditions worldwide, characterized by alternating episodes of depression and mania. Manic episodes involve elevated or irritable mood, inflated self-esteem, decreased need for sleep, racing thoughts, and impulsive behavior, and they frequently require hospitalization. Clinicians currently assess the severity of mania with rating scales such as the Young Mania Rating Scale, or YMRS, and gauge treatment response by observing how those scores decline over weeks of medication. What they lack is a biological marker, something measurable in the body that could predict, early on, whether a given patient is likely to respond well to standard treatment. The new study addresses precisely this gap.

The theoretical backdrop is the oxidative stress hypothesis of psychiatric illness. Oxidative stress arises when reactive molecules, including hydrogen peroxide and hydroxyl radicals, are produced faster than cellular defense systems can neutralize them. The brain is particularly vulnerable because it consumes a disproportionate share of the body’s oxygen and is rich in lipids that are easily damaged by reactive species. Over the past two decades, researchers have reported abnormalities in antioxidant enzymes and molecules in patients with schizophrenia, depression, and bipolar disorder, but results have often been inconsistent, and few studies have tracked these markers longitudinally through an actual episode of illness and its treatment.

To fill that gap, the team led by Yujun Sun of Kunshan Mental Health Centre and colleagues at Suzhou Guangji Hospital of Soochow University recruited 70 inpatients with manic episodes of bipolar disorder from Wutai Mountain Hospital in Yangzhou, along with 70 age- and sex-matched healthy control subjects. All participants underwent standardized assessment of manic symptoms, and peripheral blood samples were collected to measure serum levels of a suite of oxidative stress markers: catalase, glutathione peroxidase, glutathione, hydrogen peroxide, hydroxyl radical, and total antioxidant capacity. The patients were then followed for four weeks of treatment, with symptom severity and blood markers reassessed at the end of that period.

The baseline comparisons revealed a distinctive oxidative profile in acute mania. Before treatment, patients showed significantly lower levels of catalase, glutathione peroxidase, glutathione, hydrogen peroxide, and hydroxyl radical than healthy controls, together with significantly higher total antioxidant capacity. At first glance, the elevated TAC in acutely manic patients may seem paradoxical, since antioxidant capacity might be expected to fall when defenses are overwhelmed. The authors interpret this pattern as evidence of a compensatory response: as the manic state imposes oxidative burden, the body appears to mobilize total antioxidant defenses upward, even as individual enzymatic components such as catalase and glutathione peroxidase are depleted or consumed.

Four weeks of treatment substantially changed the picture. Symptom severity, as measured by YMRS scores, decreased significantly over the follow-up period, a change the authors report with a t statistic of 46.220 and a P value below 0.001. Alongside clinical improvement, catalase, glutathione peroxidase, glutathione, and hydroxyl radical levels all rose significantly, while total antioxidant capacity fell. Notably, hydrogen peroxide levels, although improved, remained significantly lower in the patient group than in healthy controls even after four weeks of treatment, hinting that some aspects of redox balance recover more slowly than mood symptoms resolve.

The most clinically consequential result concerns prediction. Correlation analysis showed that the reduction in YMRS scores over the four weeks was significantly negatively associated with changes in TAC levels, with a correlation coefficient of −0.254 and a P value below 0.05. In other words, patients whose total antioxidant capacity declined the most tended to show the greatest symptomatic improvement. When the researchers turned to multivariate linear regression, three factors emerged as significant predictors of symptom improvement: the baseline YMRS score, the four-week change in TAC, and sex, all with P values below 0.05. The direction of the TAC finding fits the compensatory interpretation, since a falling TAC would parallel the normalization of an initially overactivated antioxidant response as mania subsides.

Several caveats deserve emphasis. The correlation coefficient, while statistically significant, describes a modest association, meaning that changes in TAC explain only a small fraction of the variability in treatment response. The study involved 70 patients at a single hospital, and the four-week window captures early response rather than long-term outcome. The researchers measured peripheral blood, not brain tissue, so the findings speak to systemic redox biology rather than directly to oxidative processes within neural circuits. Sex emerged as a predictor, but the study was not designed to dissect why, and the authors did not report mediation analyses that would clarify how demographic and biological variables interact. As with any observational follow-up study, association does not establish that antioxidant dynamics drive recovery; they may be a downstream echo of it.

Even with those limitations, the work carries real significance for a field hungry for biomarkers. Psychiatry has few laboratory tests to guide treatment decisions, and patients with bipolar mania often endure a trial-and-error process before an effective regimen is found. A marker that tracks, and potentially anticipates, clinical improvement could eventually help clinicians identify early non-responders, adjust treatment sooner, and monitor recovery objectively. The study also adds nuance to the oxidative stress hypothesis by showing that the relationship between antioxidant measures and illness is dynamic: TAC is elevated in acute mania and falls as symptoms lift, a pattern that static cross-sectional comparisons alone would have obscured.

The research, published open access on 6 October 2026 under DOI 10.1186/s12888-026-08734-2, was supported by funding from the Suzhou Key Technologies Program, the Suzhou Clinical Medical Center for Mood Disorders, and related regional programs. The authors emphasize that oxidative stress profiles in patients experiencing bipolar mania differ significantly from those of healthy individuals, with TAC emerging as a potential predictor of clinical improvement. Larger, multi-center studies with longer follow-up will be needed to confirm whether measuring total antioxidant capacity at the bedside can genuinely forecast who will respond to treatment. But the study marks a step toward a future in which a routine blood draw, read alongside clinical scales, becomes part of the standard assessment of manic illness, turning an abstract biochemical theory into a practical instrument of psychiatric care.

Subject of Research: Oxidative stress markers as predictors of treatment response in bipolar mania

Article Title: Total antioxidant capacity as a predictor of treatment response in bipolar mania: evidence from a 4-week follow-up study

Article References: Sun, Y., Wang, R., Wang, X., Yuan, S., Zhang, X., Zhang, X., & Sun, W. (2026). Total antioxidant capacity as a predictor of treatment response in bipolar mania: evidence from a 4-week follow-up study. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08734-2

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08734-2

Keywords: bipolar disorder, mania, oxidative stress, total antioxidant capacity, biomarkers, treatment response, catalase, glutathione peroxidase, glutathione, hydrogen peroxide, psychiatry, follow-up study