Published online by Cambridge University Press: 01 December 2014
Although poor sleep accompanies depression, it is unknown which specific sleep abnormalities precede depression. This is similarly the case for depression developing during interferon-α (IFN-α) therapy. Because vulnerability becomes evident in those who slept poorly before IFN-α, we prospectively determined which specific aspect of sleep could predict subsequent depression.
Two nights of polysomnography with quantitative electroencephalogram (EEG) were obtained in 24 adult, euthymic subjects – all subsequently treated with IFN-α for hepatitis C. Every 2 weeks, a Beck Depression Inventory-II (BDI-II) score was obtained, and the maximal increase in BDI-II from pre-treatment baseline – excluding the sleep question – was determined.
The delta sleep ratio (DSR; an index of early-night restorative delta power) was inversely associated with BDI-II increases (p<0.01), as was elevated alpha power (8–12 Hz; p<0.001). Both delta (0.5–4 Hz) and alpha power exhibited high between-night correlations (r=0.83 and 0.92, respectively). In mixed-effect repeated-measure analyses, there was an interaction between alpha power and DSR (p<0.001) – subjects with low alpha power and elevated DSR were resilient to developing depression. Most other sleep parameters – including total sleep time and percentage of time in slow wave sleep – were not associated with subsequent changes in depression.
Both high DSR and low alpha power may be specific indices of resilience. As most other aspects of sleep were not associated with resilience or vulnerability, sleep interventions to prevent depression may need to specifically target these specific sleep parameters.