Biopsychosocial Indicators of Depression and Suicidality in Medical Students
- Author(s): Carlos, Kathleen Marie
- Advisor(s): Uban, Kristina
- Riis, Jenna
- et al.
Background: Physicians die by suicide at a rate that is higher than the general population. This increased risk for suicide may begin during medical school. Medical students experience depression and suicidal thoughts and behaviors at a higher rate than their age-matched peers. This dissertation examines biopsychosocial risk factors for depression and suicidality in a sample of medical students. This dissertation also explores the use of salivary biomarkers as screening tools for mental distress by examining associations between indicators of depression and suicidality, salivary cortisol, and salivary C-reactive protein (CRP). Methods: An in-person, self-report survey was completed by 134 medical students. The survey contained questions about sociodemographic characteristics, family history of depression and suicide, and validated measures of health behaviors and psychosocial well-being. Participants also completed two consecutive days of saliva collection at home. Saliva samples were returned by 109 (81%) of the 134 medical students enrolled in the study. Cortisol and CRP were assayed from whole saliva in duplicate. Chapter 2 characterizes the prevalence of indicators of mental distress and potential risk factors, and examines sex differences using chi-square and Mann Whitney U tests. Additionally, associations between indicators of mental distress and potential risk factors including perceived stress, sleep quality, impostor syndrome, alcohol abuse, and financial distress (e.g. difficulty paying bills, financial need, and food insecurity) are examined using logistic regression. Chapter 3 examines associations between daily cortisol activity (i.e., daily output and diurnal slope) and indicators of mental distress using a linear multilevel model to asses variation in cortisol activity across people, sampling days, and individual salivary assessments. Chapter 4 explores associations between mental distress indicators and: 1) inflammatory activity (i.e., waking and evening CRP and diurnal CRP slope) using logistic regression and 2) the coordination between hypothalamic-pituitary-adrenal axis (HPA) and inflammatory function (i.e., waking and evening cortisol/CRP, the interaction between area under the curve with respect to ground (AUCg) for cortisol and CRP, and the interaction between diurnal slopes for cortisol and CRP). Sex-specific differences in waking and evening cortisol/CRP between mental distress groups (i.e., positive versus negative DS) are examined using Mann-Whitney U test. The interaction of cortisol AUCg and CRP AUCg and the interaction between cortisol diurnal slope and CRP diurnal slope are explored through logistic regression. Results: I found a high prevalence of DS, SI, history of suicidality, and factors shown to increase risk for suicide (i.e., poor sleep quality, high perceived stress and impostor syndrome, and high alcohol use). Positive DS was associated with high perceived stress, poor sleep quality, impostor syndrome, and bill payment difficulty after adjustment for demographic covariates and family history of depression. SI and suicide risk were associated with impostor feelings even after adjustment, and history of suicidality was associated with impostor feelings and poor sleep quality even after adjustment. I also found that suicide risk was associated with higher diurnal cortisol output. Additionally, I found that waking CRP was positively associated with SI in males and females and DS in males, and negatively associated with suicide risk in females. Higher waking cortisol/CRP was also associated with a positive DS and suicide risk in females. The interaction between cortisol and CRP activity (i.e., AUCg and diurnal slope) was not associated with mental distress. Conclusion: These findings suggest a high prevalence of factors that are associated with increased risk for suicide in this sample, and that sleep quality, perceived stress, feelings of impostor syndrome, and bill payment difficulty may be important targets for future intervention studies in medical students. Additionally, this dissertation provides preliminary evidence that HPA and inflammatory activity are disrupted in a sample of young adults with mental distress. Future longitudinal research is needed to assess temporality of these findings, confirm these associations in a larger study sample, and provide additional evidence to inform mental health intervention and prevention efforts in this high-risk population.