Background: Depression is a major cause of morbidity and death globally, and its incidence is increasing. Increases have been linked to modern dietary changes, stress, loneliness, sleep deprivation and endocrine dysfunction. Its diagnostic criteria span psychological and physical symptoms, including low mood, negative thinking and changes in sleep, appetite and weight. Additionally, there are strong relationships between depression and physical health, particularly heart disease. Key biomarkers identified to operate at the nexus of physical and mental health include cortisol, oxytocin, omega-3 fatty acids and leptin. While these show potential for novel therapeutic strategies, studies show inconsistent results, likely due to complex interactions between biopsychosocial variables and heterogeneous symptom subtypes. Structural equation modelling (SEM) is an analytical technique allowing examination of complex interactions between variables that predict important outcomes, such as illness. SEM has been extensively applied to coronary heart disease to identify points for interventions among numerous lifestyle and physical factors. As yet, similar, comprehensive models of depression are lacking. Developing and testing complex models including key biological and psychosocial variables may lead to improved interventions.