This page was translated automatically from Turkish. Read the original

Bu sayfa, 26. Dünya Psikiyatri Kongresi'nde sunulan çalışmanın yapay zekâ desteğiyle hazırlanmış Türkçe özetidir. Çalışmanın sahibi sunan yazarlardır; atıf için orijinal kaynağa başvurun.

EV0678

Inflammatory biomarkers and physical activity as key determinants of depression severity: a machine learning analysis

INFLAMMATORY BIOMARKERS AND PHYSICAL ACTIVITY AS KEY DETERMINANTS OF DEPRESSION SEVERITY: A MACHINE LEARNING ANALYSIS

While physical activity emerged as an independent protective factor against depressive symptoms, vitamin D, D-dimer, hs-CRP, TNF, IL-6, and myoglobin were the biomarkers with the highest relative importance in distinguishing symptom severity in Random Forest models.

The study aimed to examine the multivariate relationships between obesity, systemic inflammation, and mental health using integrative models. In a cross-sectional design, 251 adults were classified into a control group (n=140) and a mental health group (n=111). Psychological symptoms were assessed using the DASS-21; body mass index (BMI) and biomarkers (IL-6, TNF-alpha, hs-CRP, vitamin D, cortisol) were also measured. The relative importance of biomarkers was examined using Random Forest models, while independent associations were examined using logistic and linear regression. Participants who reported regular physical activity had lower symptom scores than those who were sedentary. Despite substantial interindividual variability, the negative association between physical activity and symptom scores was statistically significant, with small to moderate effect sizes. In the Random Forest models, vitamin D and D-dimer ranked highest in relative importance for distinguishing depressive symptom severity, followed by hs-CRP, TNF, IL-6, and myoglobin; these six biomarkers were marked as significant in the poster. cTnI, IL-1beta, CK-MB, NT-proBNP, CRP, and IL-8 had lower relative importance. According to the authors, behavioral factors and inflammatory biomarkers converge in the regulation of psychiatric symptoms, and the inflammatory signature highlights the role of the immune–neuroendocrine axis in psychopathology. Combining biomarker profiling with machine learning may contribute to early stratification strategies in psychiatry.

Tüm WCP 2026 özetleri