Inflammation-Induced Temperature Shifts and Mitochondrial Function: Rethinking Biomarker Interpretation in Depressive Disorders Authors: Alexander Karabatsiakis, Rafael Pfeiffer, Annalena Steinlechner, Karin de Punder Published: 2026 Depression is a severe mental disorder that has developed into a major threat to the healthcare systems around the globe, significantly impairing the affective, cognitive, and somatic functioning of those suffering. Even worse, depression not only affects mental health but also increases the risk of physical diseases in the long term [1–3]. For example, depression has been associated with various immune- and age-associated complications, including autoimmune diseases and inflammatory conditions, adverse cardiovascular and cerebrovascular health, diabetes mellitus and other metabolic conditions, neurodegenerative disorders, and even cancer [1, 3, 4]. Despite decades of intense and multidisciplinary research efforts, depression remains a psychiatric condition with an incompletely mapped and poorly understood pathophysiology. As a result, markers (biological or digital) for direct applications in diagnosing and disease management are still lacking, including those that would enable predictive, preventive, and more personalized (3P) measures, rather than reactive treatment. On a positive note, converging evidence from preclinical and clinical studies consistently identified several functional systems and cross-talk axes of biological relevance in depression, including the neuroendocrine and central nervous systems, as well as related communication pathways, e.g., the brain-immune axis [2, 3, 5, 6] that serve as promising starting points for intensified research efforts [7, 8]. Among the most recently studied biomarker candidates for stress-associated depressive conditions is mitochondrial bioenergetics [7, 8], which describes how and to what extent mitochondria generate biochemical energy in the form of adenosine triphosphate (ATP), and can be assessed in vitro using various biological samples, including cells and tissues, and even whole organisms like C. elegans, as demonstrated recently [9–12]. In clinical research, peripheral blood mononuclear cells are predominantly used, and reduced mitochondrial function in immune cells, often classified as “mitochondrial dysfunction,” together with higher levels of pro-inflammatory signaling and low-grade inflammation [5, 6, 13, 14], processes known to induce sickness behavior [15], have been implicated in the pathophysiology of depression. Despite the robustness of these observations, it remains unclear how these factors interrelate within disease pathophysiology, particularly with respect to the immune system. Clarifying their interconnections, rather than considering each factor in isolation, is essential for a more accurate interpretation of the observations reported in the literature regarding “mitochondrial dysfunction.” As a starting point for a more reflective and critical perspective on the terminology of mitochondrial “dysfunction,” basic physiology teaches us that a biological system’s function depends on two core factors: pH and temperature [16–18]. Both physical and mental challenges can cause an accumulation of acidic metabolites, primarily lactate and pyruvate, as well as ketone bodies (beta-hydroxybutyrate) and Krebs cycle intermediates (succinate, fumarate, citrate). Together with an increase in protons (H+) during physiological adaptation to cellular challenges, these processes increase the likelihood of disrupting intracellular pH homeostasis (acidosis), thereby negatively affecting the physiological function of pH-sensitive proteins, as schematically shown in Figure 1a. As a result, these pH drifts can impair the physiological function of enzymes, depending on the severity and chronicity of stress exposure [16–18]. Another lesson from physiology is that enzyme activity also depends on temperature, bringing the cardinal signs of inflammation into the picture: rubor (redness), tumor (swelling), dolor (pain), and functio laesa (loss of function), all local, and heat (calor), whose systemic counterpart is fever, directly relevant to our perspective [19, 20]. Stress, pH, and temperature. a Stress causes (i) an accumulation of acetic metabolites, including lactate, together with protons (H+), that lower the pH of the cell and endanger homeostasis and function. As a result, the physiological activity of pH-sensitive enzymes is lowered, contributing to a performance loss in depression that might show equivalence to the severity of clinical symptoms (ii). b To compensate for the limiting influences of lower pH on physiology, the body can react with an inflammation-mediated increase in temperature to compensate or reconstitute performance loss (iii). Therefore, temperature should also be considered when comparing physiological performance in pH-drift-associated pathophysiological conditions, such as depression. Taking reported changes of pH homeostasis [21], increased pro-inflammatory signaling [2, 5, 22], and increases in body temperature (the “good site” of inflammation) with the severity of depressive symptoms into account [23], it is now important to return to the literature, reflecting that almost all results from in vitro experiments with human material are reported for the standardized temperature of 37 °C only. Consequently, as depicted in Figure 1b, it is also necessary to include temperature (and pH) in the qualitative and quantitative discrimination between function and dysfunction, as elevated body temperature may exert a compensatory effect on mitochondrial function. In line with this expectation, environmentally induced increases in body temperature (whole-body hyperthermia treatment) show robust antidepressant effects with as-yet-unknown underlying mechanisms [24], thereby further highlighting the need to pay closer attention to temperature, enzymes, and physiological performance. As a result, it seems reasonable to expect that assessing mitochondrial “function” and classifying it as “function” or “dysfunction” requires greater caution. In terms of more personalized measures, we propose a broader investigation of mitochondrial function in the context of (mental) health and disease, accounting for important variables such as pH and body (core) temperature when analyzing mitochondrial function in vitro (cellular respirometry) and in vivo (e.g., by functional near-infrared spectroscopy in the CNS). As NIRS can also detect temperature at the skull surface by thermometry, we propose considering body temperature and skull surface temperature as markers of systemic inflammation and as important covariates in these analyses [25]. First results from empirical work addressing the presented concept, through the comparative analysis of mitochondrial function in peripheral blood mononuclear cell at 37 °C and 39 °C, not only support our assumptions about temperature-dependent mitochondrial function [26] but also further underscore the need for a multistep characterization of mitochondrial oxygen consumption rates across a broader range of physiological temperature windows, as proposed in Figure 2. Conclusion and recommendation. More caution is needed when interpreting and communicating results from experiments measuring mitochondrial bioenergetics at 37 °C, as they might provide only a single insight into the regulation of mitochondrial function. In conclusion, the distinction between function and dysfunction in qualitative and quantitative assessments of mitochondrial function in immune cells requires clarification of whether the observations in depression reflect adaptive or maladaptive processes. Because the regulation of mitochondrial function across cells, tissues, and organs at the systemic level is insufficiently addressed and therefore inadequately understood, cautious interpretation is warranted until additional empirical insights from research on brain-immune-endocrine cross-communication and other relevant initiatives are available. Future research will provide the needed answers. Conflict of Interest Statement The authors have no conflicts of interest to declare. Alexander Karabatsiakis was a member of the journal’s Editorial Board at the time of submission. Funding Sources This study was not supported by any sponsor or funder. Author Contributions Alexander Karabatsiakis elaborated on the concept and addressed the research question empirically. Rafael Pfeiffer, Annalena Steinlechner, and Karin de Punder contributed to the manuscript by providing relevant input and literature to its realization. All authors approved the final version of the manuscript.