Electroencephalographic (EEG) slow wave activity (SWA) or Delta power (0.5–4 Hz) is a marker of restorative sleep (Dijk, 2009). Delta sleep, also referred to as slow wave sleep (SWS), is widely recognized as the sleep stage most linked to restorative sleep, contributing significantly to sleep quality and continuity (Bonnet, 1987; Dijk and Archer, 2009; Dijk et al., 2006). Consistent with its restorative role, Delta power is highest after prolonged wakefulness and exponentially declines through the night as sleep progresses (Achermann and Borbély, 2003; Achermann et al., 1993; Aeschbach et al., 1996; Akerstedt and Gillberg, 1986; Borbély and Wirz-Justice, 1982; Dijk et al., 1991). The two-process model of sleep regulation describes how Process S—a sleep/wake dependent homeostatic process—predicts the decline of Delta power during sleep while interacting with Process C, which captures the course of circadian factors (Chang et al., 2016; Dijk and Archer, 2009; Franken and Dijk, 2009).

In this context, the synaptic homeostasis hypothesis integrates many molecular events associated with sleep (Tononi and Cirelli, 2003). The firing patterns observed during non-rapid eye movement (NREM) sleep after sustained wakefulness demonstrate an intricate interplay between sleep homeostasis and neuronal activity. During NREM sleep, cortical neurons exhibit bi-stability, with alternating periods of firing and silence (Rodriguez et al., 2016; Vyazovskiy et al., 2009). This increase in synchronous firing is reflected in the EEG as increased Delta power, particularly after prolonged wakefulness, and is an indicator of sleep homeostasis. Broadly, wakefulness is associated with synaptic potentiation, while sleep is associated with synaptic downscaling (pruning) accompanied by a decline in Delta power as sleep progresses (Duncan et al., 2013a; Walker, 2009). Mechanisms underlying sleep homeostatic regulation are believed to have beneficial effects on learning and memory retention and, when dysfunctional, detrimental effects on mood (Tononi and Cirelli, 2014). These mechanisms, in which synapses are alternately pruned and strengthened (Frank et al., 2001; Tononi and Cirelli, 2014; Vyazovskiy et al., 2008), represent an essential component of a day/night cycle of neuroplasticity.

Numerous factors influence Delta power, including age, sex (Armitage, 1995; Armitage et al., 1995; Dijk et al., 1989), or the presence of a mood disorder (Armitage et al., 1995, 2000; Lopez et al., 2012). Age-related declines in Delta power occur in both sexes, although some evidence suggests the decline may be greater in males (Dijk et al., 1989; Ehlers and Kupfer, 1989; Feinberg et al., 1977; Landolt et al., 1996). Conversely, some studies found no age-related difference in Delta power (Nissen et al., 2002). It is also important to note that few studies have explored differences in Delta frequencies. One study of healthy volunteers (HVs) compared Low Delta power (<2 Hz) to the Total Delta band (1–4 Hz) and found that, unlike Total Delta, Low Delta did not decline from the first to the second NREM sleep episode (Achermann and Borbély, 1997). Other studies investigating the effects of sleep deprivation on recovery sleep in HVs obtained mixed results, with one reporting increased Delta power within the broader frequency range of 1–4 Hz (Dijk et al., 1991) and another reporting increased Delta power in the 0–3 Hz frequency range (Feinberg et al., 1988).

Both major depressive disorder (MDD) and diminished mood have been strongly linked to disrupted sleep patterns. These disruptions often include difficulty initiating sleep, frequent awakenings throughout the night (wake after sleep onset (WASO)), and early morning awakening (American Psychiatric Association, 1994). Moreover, they have been associated in various studies with deficiencies in Delta power (Borbély, 1987; Borbély et al., 1984; Campbell and Gillin, 1987; Kupfer et al., 1984, 1986). Alongside deficits in total Delta power, a pattern of decline in Delta power—as indicated by a lower Delta sleep ratio—is present, indicating insufficient dissipation of Delta power from the initial to the subsequent NREM sleep stages (Duncan et al., 2013b; Kupfer et al., 1990). Insomnia can also be associated with less decline in overnight Delta power compared to those without insomnia (Lunsford-Avery et al., 2021).

In depression, insufficient buildup of sleep pressure during wakefulness has been hypothesized to underlie reduced Delta power during NREM sleep (Borbély, 1987; Borbély et al., 1984; Campbell and Gillin, 1987). Although the sleep literature extensively discusses Delta power, the exact definition of the EEG band that characterizes Delta power remains ambiguous. (Campbell et al., 2006). Delta power definitions vary, ranging from 1 to 4 Hz to below 2 Hz (Feinberg et al., 1988), with some studies emphasizing the 0.5–2 Hz frequency range (Borbély et al., 1984). Given the potential for inconsistent results stemming from the use of varying definitions of the Delta EEG band, especially within the context of the literature on MDD, it would be beneficial to explore potential clinical distinctions among different Delta power frequencies.

Broadly, studies have identified an association between reduced Delta power in the lower frequency range (0.5–2Hz) and individuals with MDD (Borbély et al., 1984; Kupfer et al., 1989). In a study of individuals with treatment-resistant depression (TRD), total Delta (0.5–4Hz) was associated with improved mood in response to the rapid-acting antidepressant ketamine, with the highest peak of post-ketamine Delta power observed at 2 Hz (Duncan et al., 2013a), highlighting the importance of low frequency Delta in MDD and TRD. This suggests that differences in the relative expression of low and high frequency Delta power may be clinically important in MDD and TRD. Such findings could help identify treatment strategies tailored to specific deficits associated with Delta power frequencies. Total Delta power (0.5–4 Hz) is most concentrated in the 0.5–2 Hz range (Borbély et al., 1984; Duncan et al., 2013a; Kupfer et al., 1989), raising the question of whether deficits of Low or High Delta are more strongly associated with MDD and TRD. However, few studies have explored whether deficits in Low Delta differ from High Delta in MDD and TRD. One study examined sex and Delta power (1–4.5 Hz) in a group of young, age-matched unmedicated MDD participants and found that, compared to men, women with MDD had increased SWA compared to controls (Plante et al., 2012). To our knowledge, few (if any) human studies have explored the intersection between age, sex, and depression comparing specific Delta frequency bands.

Furthermore, despite the fact that approximately one third of individuals with MDD are resistant to conventional treatments such as selective serotonin reuptake inhibitors (SSRIs), few studies have explored sleep differences in individuals with TRD (Zhdanava et al., 2021). Given the potential importance that sleep impairments and factors associated with non-restorative sleep may play in TRD, this study sought to examine potential distinctions between Low and High Delta power and their associations with age, sex, and disrupted sleep patterns in individuals with TRD. This exploratory study sought to examine the nightly progression of the entire Total Delta power range of 0.5–4 Hz, with an additional focus on the most common frequency associated with MDD—Low Delta (0.5–2 Hz)—which was then compared to the High Delta (2–4 Hz), the upper frequency range within the total band. One hundred unmedicated individuals with TRD and 24 HVs participated in the study. The study's main hypothesis was that there would be distinct differences between Low and High Delta power in individuals with TRD. A second hypothesis was that these differences would be unique to individuals with TRD compared to HVs. The study also explored the extent to which differences related to sex and age were linked to sleep deficits such as reduced TST or increased WASO.