Sleep represents an underutilized data source in medicine. Clinically disturbed sleep is often treated as a symptom of the disorder, but sleep is also a physiological state in which the brain, heart, respiration, and autonomic systems continuously interact. This interaction is measurable and potentially predictive. Consumer sleep tracking is becoming commonplace thanks to wearable devices that capture movement using accelerometry. These devices also use photoplethysmography (PPG), which estimates changes in blood volume from reflected light.