Objective: Excess cardiovascular mortality has been demonstrated in patients with major depression and bipolar disorder. Cardiovascular risk factors, such as the components of the metabolic syndrome, also appear more prevalent in populations with mood disorders. Depressive symptoms have been suggested to elevate risk of cardiovascular disease in several studies. On prior analyses from the Collaborative Depression Study (CDS), manic but not depressive symptoms showed a dose-response relationship with cardiovascular mortality. We sought to assess the relationship between treatment exposure or symptom burden and measures of vascular function.
Methods: Through prior participation in the National Institute of Mental Health CDS, participants’ phenomenological and treatment histories were rigorously assessed more than 25 years. At a separate visit, participants were assessed for body mass index, body fat percentage, fasting lipids, C-reactive protein, Interleukin-6 (IL-6), insulin levels, and glucose. Vascular function and stiffness were assessed using pulse wave analysis, pulse wave velocity, and conduit vessel function. Endothelium-dependent and endothelium-independent vascular function of conduit vessels were assessed non-invasively via ultrasound measurement of brachial artery diameter following flow-mediated and nitroglycerin-mediated vasodilation. Linear regression models determined the relationship between symptom burden or treatment exposure and vascular function, controlling for age and gender.
Results: To date, 29 participants from the CDS have completed vascular assessments. The mean (SD) age of participants is 62 (8) with 90% of participants being female and 52% having a diagnosis of bipolar disorder. The mean (SD) body mass index is 30 (7) and 55% have smoked tobacco. An interim analysis shows the proportion of time spent with clinically significant depressive symptoms was associated with a higher augmentation index corrected for heart rate of 75 (AIX@75) on pulse wave analysis, controlling for age and gender with manic symptoms additionally modeled (F1,24=4.2, partial R2=0.15, p=0.05). In addition, manic symptom burden was associated with insulin resistance (F1,24=3.69, partial R2=0.19, p=0.03). Cumulative exposure to second generation antipsychotics, divalproex, and/or lithium were strongly associated with elevated body mass index (F1,24=2.29, partial R2=0.18, p=0.03) and triglycerides (F1,24=2.82, partial R2=0.25, p=0.01) as well as insulin resistance (F1,24=2.83, partial R2=0.25, p=0.01), but not vascular function.
Conclusions: These results provide evidence that mood disorders may contribute to vascular damage in a dose-dependent fashion. Patients with more depressive symptoms had greater augmentation of estimated aortic systolic pressure. We also confirm prior findings that newer antipsychotic medications may worsen components of the metabolic syndrome. Assessment of vascular function may represent a useful method for the assessment of cardiovascular risk associated with mental illness. These methods may be further useful in the assessment of mediators, moderators, and correlates of cardiovascular risk in mood disorders.