Retinal photography: A window into the cardiovascular-brain link in adolescent bipolar disorder

- Bipolar disorder is associated with excessive and premature cardiovascular disease.
- Retinal vessels correlate with mood and cognition in adolescent bipolar disorder.
- Retinal vessels correlate with vascular risk factors in adolescent bipolar disorder.
- Retinal vessels correlate with endothelial function in adolescent bipolar disorder.
- These correlations were not observed in healthy control adolescents.
- Retinal vascular photography may yield insights on the cardiovascular-bipolar link.

ObjectiveThe burden of cardiovascular disease in bipolar disorder (BD) exceeds what can be explained by traditional cardiovascular risk factors (CVRFs), lifestyle, and/or medications. Moreover, neurocognitive deficits are a core feature of BD, and are also related to CVRFs. We examined retinal vascular photography, a proxy for cerebral microvasculature, in relation to CVRFs, peripheral microvascular function, and neurocognition among BD adolescents.MethodsSubjects were 30 adolescents with BD and 32 healthy controls (HC). Retinal photography was conducted using a Topcon TRC 50 DX, Type IA camera, following pupil dilation. Retinal arteriolar and venular caliber was measured, from which the arterio-venular ratio (AVR) was computed. All measures were conducted masked to participant diagnosis. Peripheral arterial tonometry measured endothelial function. Neurocognition was assessed using the Cambridge Neuropsychological Tests Automated Battery.ResultsAVR was not significantly different between groups (Cohen's d=0.18, p=0.103). Higher diastolic blood pressure (BP) was associated with lower (worse) AVR in BD (r=−0.441, p=0.015) but not HC (r=−0.192, p=0.293). Similarly, in the BD group only, higher (better) endothelial function was associated with higher AVR (r=0.375, p=0.041). Hierarchical regression models confirmed that, independent of covariates, retinal vascular caliber was significantly associated with diastolic BP and endothelial function in BD. Within the BD group, mood scores were significantly negatively correlated with AVR (β=−0.451, p=0.044).LimitationsThis study's limitations include a small sample size, a cross-sectional study design, and a heterogeneous sample.ConclusionRetinal photography may offer unique insights regarding the cardiovascular and neurocognitive burden of BD. Larger longitudinal studies are warranted.