by Nattapong Jaimchariyatam, MD, MSc, FCCP; Carlos L. Rodriguez, MD; and Kumar Budur, MD, MS
Dr. Jaimchariyatam is from the Division of Pulmonary and Critical Care Medicine, Department of Medicine, Faculty of Medicine, Chulalongkorn University in Bangkok, Thailand, Excellence Center for Sleep Disorders, King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand, and Sleep Disorders Center, Neurological Institute, Cleveland Clinic, Cleveland, Ohio; Dr. Rodriguez is from the Sleep Disorders Center, Neurological Institute, Cleveland Clinic in Cleveland, Ohio; and Dr. Budur is from the Sleep Disorders Center, Neurological Institute, Cleveland Clinic in Cleveland, Ohio, and Takeda Global Research and Development, Inc., Deerfield, Illinois.

Innov Clin Neurosci. 2011;8(7):35–49

Funding: No funding was received for the preparation of this article.

Financial disclosures: The authors have no conflicts of interest relevant to the content of this article.

Key words: Alpha-delta sleep, major depressive disorder, daytime sleepiness

Abstract: Objective. Major depressive disorder is associated with sleep disturbances. An electroencephalographic pattern of alpha wave intrusion in delta wave sleep (alpha-delta sleep) is observed in some subjects with major depressive disorder. The treatment-resistant symptoms in major depressive disorder, nonrestorative sleep and fatigue, are associated with alpha-delta sleep. The objective of this study is to identify the prevalence and clinical correlates of alpha-delta sleep in major depressive disorder.

Design. Retrospective study

Setting. Sleep Disorders Center, Cleveland Clinic, Cleveland, Ohio
Participants. Polysomnograms were conducted on 150 subjects 18 years of age or older (75 with and 75 without major depressive disorder) were reviewed.

Measurements. The percent of delta waves with alpha intrusion was collected and analyzed.

Results. Subjects with major depressive disorder compared to nondepressed subjects had a higher sleep efficiency (83.0±9.6; 78.1±8.2%), shorter rapid eye movement latency (85.0±44.5; 189.9±25.6 min), less slow wave sleep (8.3±3.0; 13.5±6.2%), and greater rapid eye movement (24.7±7.0; 19.2±8.2%), and all of these findings were statistically significant. Patients with major depressive disorder had higher alpha-delta sleep (23.4±14.2%; 2.3±6.7%, p<0.01). Patients with major depressive disorder were categorized into high and low alpha-delta sleep based on percentage of alpha-delta sleep present in slow wave sleep (alpha-delta sleep was present ?15% or ?15% of slow wave sleep, respectively). Patients with major depressive disorder with high alpha-delta sleep were at 3.15 greater odds (1.22–8.14; p=0.018) to have excessive daytime sleepiness.

Conclusion. Patients with major depressive disorder have a higher prevalence of alpha-delta sleep. Alpha-delta sleep is associated with daytime sleepiness in patients with major depressive disorder. Study limitations include the retrospective nature of the project and the fact that the principle investigator, who scored and interpreted alpha intrusion, was not blind to group membership.

Background

Major depressive disorder (MDD) is a common condition with a reported lifetime prevalence of 5 to 12 percent in men and 10 to 25 percent in women.[1] A variety of factors have been associated with an increased risk for mood disorders, including psychosocial stress, chronic illness and pain, alcohol/substance abuse, and physical or psychological trauma.[2–9] Most patients with MDD complain of insomnia, mainly difficulty falling/staying asleep, early morning awakenings and nonrestorative sleep.[10] Although a significant number of depressed patients frequently report increased daytime sleepiness/fatigue, there is little evidence to show that they have increased daytime sleepiness when measured objectively.[11–13] Mood disorders and sleep disturbances have a bidirectional relationship. Not only are sleep disturbances common in patients with mood disorders, they are predictive of individuals at higher risk for the development of depression.10 The lifetime prevalence of MDD in subjects with sleep disturbances demonstrated significantly higher rates of MDD in individuals with sleep complaints: 25.3 pecent of those who had hypersomnia and 54.3 percent of those with both insomnia and hypersomnia had MDD, compared to 2.7 percent in individuals with no sleep complaints.[11,14] Nevertheless, symptoms of depression tend to persist even when sleep abnormalities show some signs of improvement.[15]

MDD has been studied polysomnographically more than any other psychiatric disorder, and the polysomnographic findings are characterized by a shift of rapid eye movement (REM) sleep earlier in the night, resulting in reduction of REM latency, increase in total REM time/REM density, disrupted sleep continuity, and diminished slow wave sleep (SWS).[11,16,17]

An electroencephalogram (EEG) pattern of alpha intrusion into nonrapid REM (NREM) sleep was first noted in 1973 in patients with psychiatric disorders.[18] The EEG appearance was that of intrusion of prominent alpha activity (frequency of 8–13 cycles per second) on delta waves (frequency of 0.5–2.0 per second with amplitude greater than 75 microvolts) (Figure 1, Figure 2). A similar pattern, termed alpha-delta sleep, has been found to be related to a complaint of nonrestorative sleep in patients with musculoskeletal pain or fibrositis and nondepressed patients with chronic fatigue.[19–21] Other disorders in which alpha-delta sleep, nonrestorative sleep, and pain or fatigue symptoms are evident include rheumatoid arthritis and systemic lupus erythematosus.[22–24]

Alpha-delta sleep is also reported in patients with psychophysiological insomnia and other primary sleep disorders (e.g., periodic limb movement disorder, circadian rhythm sleep disorders, sleep apnea, and narcolepsy).[24–29] Among these conditions, however, daytime symptoms, such as daytime fatigue and excessive daytime sleepiness, are thought to be multifactorial in origin, and nonrestorative sleep is known to contribute to daytime sleepiness and fatigue.[30]

Abnormalities of sleep, as observed on a polysomnogram, offer unique insight into some of the physiological processes in the brain and the pathophysiology of the disease state. Although a lot of excitement was generated regarding REM sleep abnormalities in MDD (short REM latency and increased REM density), these findings were by no means specific to MDD and do not have any diagnostic or therapeutic value.[31,32] Alpha-delta sleep is observed in patients with MDD, and patients with fibromyalgia often suffer from MDD as well.[33] Much is known about the characteristic features of alpha-delta sleep in patients with rheumatological/pain disorders, but to date no studies have systematically assessed alpha-delta sleep in patients with MDD using objectively defined criteria that can be considered reliable. Ware et al[34] were one of the first to describe alpha intrusions in sleep of patients with depression. They found that alpha intrusion during NREM sleep in depressed patients was “common.” They did not specifically look at alpha intrusion in slow wave or delta sleep, but rather they studied any alpha intrusion in NREM sleep. In addition, they graded the severity of alpha intrusion in NREM sleep on a four point scale: 1=0 to 25 percent of NREM sleep has alpha intrusion; 2=26 to 50 percent alpha intrusion; 3=51 to 75 percent; and 4=76 to 100 percent. Five out of 12 depressed patients had a rating of 3 or 4 (i.e., alpha intrusion occurred in more than 50% of the NREM sleep). Although commendable given this work was done in 1986, some of the major drawbacks include criteria for alpha intrusion that are not considered appropriate today, small sample size, and lack of a control group.

Manu et al[21] studied alpha-delta sleep in patients with a chief complaint of chronic fatigue. Eight of the 30 patients had MDD and three of these eight patients (37%) had alpha-delta sleep pattern. In this study, the criterion to detect the presence of alpha-delta sleep was “prominent alpha frequency activity occurring tonically during nonrapid eye movement sleep.” This criterion does not have any objective measures and was not validated. This study did account for the “first-night effect” in the majority of the patients by doing sleep studies on two or three successive nights. However, this study again lacked any control subjects.

Alpha-delta sleep, as measured by objectively defined and reliable criteria, if consistently found in patients with MDD, could emerge as an important feature and perhaps a biological marker. Fatigue and nonrestorative sleep are often residual symptoms in patients with MDD, and these are considered to be the most resistant to the medical treatment of MDD. Since alpha-delta sleep is also associated with similar symptoms (i.e., non-restorative sleep and fatigue), treatment aimed at normalizing the alpha intrusion in slow-wave sleep could potentially benefit MDD patients. Sodium oxybate has demonstrated normalization of alpha-delta sleep in patients with fibromyalgia, and this is associated with a significant improvement of pain, fatigue, and subjective sleep. Therefore, elucidating the nature, severity, and correlates of alpha-delta sleep in patients with MDD can have significant diagnostic and therapeutic implications.

We hypothesized that patients with MDD have a higher prevalence of alpha-delta sleep compared to age- matched non-MDD subjects. If so, we decided to identify various correlates of alpha-delta sleep in patients with MDD who had subjective complaint of excessive daytime sleepiness.

Methods

A retrospective review of 150 polysomnogram (PSG) studies, consisting of 75 subjects with MDD and 75 nondepressed subjects, who had polysomnograms done at Cleveland Clinic Sleep Disorders Center from July 2005 to October 2008 was conducted.

The Cleveland Clinic Sleep Disorders Center database had information (both raw data and interpretations) on a total of 14,549 patients for the period from July 2005 to October 2008. A total of 960 patients out of 14,549 patients (6.6%) matched the criteria for a current active medical history of MDD. Out of these 960 patients, the first 410 patient medical charts were reviewed, including the clinical interview, sleep questionnaire, and polysomnographic raw data, to identify 75 patients who met all the inclusion criteria for the study.

Similarly, for the nondepressed patient selection, the first 341 charts (out of 13,589 patients) were reviewed to identify 75 subjects who met all the inclusion criteria for the study. The most common reason for exclusion of subjects in both MDD and non-MDD group was the presence of significant sleep apnea.

All the subjects included in this study had a single-night polysomnogram conducted in an accredited sleep lab attended by a polysomnogram technician. All polysomnograms were performed according to the standard American Academy of Sleep Medicine (AASM) procedure, including video recording, left and right electrooculogram (EOG), central and occipital electroencephalogram (EEG), mental and submental electromyogram (EMG), intercostal EMG, left and right anterior tibialis EMG, electrocardiogram (ECG), snoring sensor, continuous airflow measurement with thermistor, nasal pressure transducer, chest and abdominal effort, and oxygen saturation with pulse oximetry.

Sleep stages were re-scored according to standard AASM 2007 criteria on 30-second epoch, which were scored as wakefulness, sleep stage NREM 1, 2, and 3 (stage N1, N2, N3, respectively) and REM sleep.[35]

Given that the sleep studies done prior to 2007 were scored according to the “old” PSG scoring criteria, all of the identified polysomnograms of patients who met the inclusion criteria were then re-scored in accordance with the 2007 AASM scoring criteria. Re-scoring of the polysomnograms was accomplished by the principal investigator. In addition, the standard sleep definitions, as per the International Classification of Sleep Disorders, Second Edition (ICSD-2) and standard mental disorders definitions as described in Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV), were applied for the purpose of this study.33,36 Specifically, insomnia was defined as either difficulty falling asleep, difficulty staying asleep, difficulty maintaining sleep, and experiencing nonrefreshing sleep despite adequate opportunity to sleep, as well as associated daytime impairments such as fatigue, headaches, tension, anxiety, and lack of concentration.[38]

Alpha-delta sleep, the most important parameter for our study, has unclear criteria. As discussed previously, most researchers have defined alpha-delta sleep as “prominent” alpha frequency occurring during slow wave sleep. None of the published studies until now have used objective criteria to define alpha-delta sleep. Manu et al[31] evaluated the correlation of alpha-delta sleep and various other medical conditions that can cause fatigue (including depression), but the criteria used to identify alpha-delta sleep was “prominent alpha sleep during slow wave sleep.” Similarly, Ware et al[34] did not specifically look at alpha intrusion in slow wave or delta sleep but rather considered any alpha intrusion in NREM sleep.

In view of the lack of objective criteria for alpha-delta sleep and a lack of guidance from any of the major sleep or psychiatry professional bodies, including American Academy of Sleep Medicine, American Sleep Research Society, and American Psychiatry Association, it was decided to define objective, straight forward criteria that are reproducible. Accordingly, we defined alpha-delta sleep as the duration of delta waves with superimposed alpha rhythm divided by the total duration of slow wave sleep multiplied by 100 (i.e., percent of slow wave sleep duration). Alpha rhythm was defined as trains of sinusoidal 8 to 13Hz activity recorded predominantly over the occipital region, and delta (slow) wave activity was defined as waves of frequency 0.5 to 2Hz and peak to trough amplitude greater than 75 microvolts recorded predominantly over the frontal regions, consistent with the AASM guidelines.[35] This criterion was used to determine the presence and extent of alpha-delta sleep in all of the polysomnograms reviewed in this study. Of note, the principal investigator who scored and interpreted the studies was also involved with the selection or exclusion of subjects who fulfilled the inclusion/exclusion criteria (i.e., the principal investigator was not blinded).

To ensure reproducibility of the criteria used in this study, a sample of 15 polysomnograms were randomly selected from the 150 studies identified for our study. A board-certified sleep specialist who was not involved in this study scored alpha-delta sleep using the above criteria.

Inclusion criteria: Subjects included in the study met the following criteria: 1) were 18 years of age or older; 2) had an apnea-hypopnea index (AHI) less than 5/hour (apnea was defined as airflow cessation for at least 10 seconds, and hypopnea was defined as the nasal pressure signal excursions [or those of the alternative hypopnea sensor] drop by ?50% of baseline for at least 10 seconds and was associated with an oxygen desaturation of at least 3% and/or followed by a cortical arousal); 3) had adequate polysomnogram (PSG) data, defined as total sleep time (TST) of more than 6.5 hours and a sleep efficiency (SE) of 65 percent or more.

Exclusion criteria: Subjects excluded from the study met the following criteria: 1) had comorbid disorders known to be associated with alpha-delta sleep, including fibromyalgia, systemic lupus erythematous, rheumatoid arthritis, polymyalgia rheumatica, fibrositis, myositis, chronic pain syndrome, and chronic fatigue syndrome; 2) had significant sleep disorders (e.g., sleep apnea syndrome (both obstructive and central), narcolepsy, circadian rhythm sleep disorders, restless legs syndrome, and periodic limb movement disorders; and 3) were undergoing positive airway pressure (PAP) titration studies or split-night studies.

Statistical analyses. Data on the following demographic and disease characteristic variables were collected: age, gender, body mass index (BMI), neck circumference, past medical history, and Epworth Sleepiness Scale (ESS) scores.

Data on polysomnographic variables of interest were also collected including sleep efficiency, sleep and REM latencies, sleep stage distribution, arousal index, apnea-hypopnea index (AHI), periodic limbs movements index (PLMI), periodic limbs movements arousal index (PLMAI), nadir oxygen desaturation, and alpha-delta sleep (as percentage of delta wave sleep).

Variables of interest with normal distribution were summarized and expressed as mean but otherwise as median or ratio (%). To ascertain the normal distribution of all continuous variables, a Kolmogorov-Smironov test was performed. Independent sample t-tests were performed so as to compare basic demographics, sleep parameters, and alpha-delta sleep between subjects with MDD and subjects without MDD. The effect of comorbid disorders, medication, gender, and age groups on the prevalence of alpha-delta sleep was determined using univariate analysis of variance (ANOVA). The relationship between alpha-delta sleep and ESS was qualified using Pearson correlation.

Upon data review, we observed a tendency for subjects with a relatively lower percentage of alpha-delta sleep to be associated lower ESS scores. In the post-hoc analysis, the subjects with MDD were divided into two groups (high and low alpha-delta sleep) by means of a median split method (> or <15% of alpha delta sleep) and the basic demographic data, sleep parameters, and ESS in both groups were compared using independent sample t-test.

Multiple linear regression analyses and scatter plot were performed to evaluate the contribution of each basic demographic and sleep variable to models predicting the percentage of alpha-delta sleep. In addition, the impact of collinearity among the variables in the regression model was measured by means of tolerance and the variance inflation factor (VIF). The multiple logistic regression analysis was subsequently performed to determine whether alpha-delta sleep correlates with excessive daytime sleepiness (defined as ESS?10) after adjusting for other factors. The statistical significance was defined as p?0.05 for all statistical tests.

The sample size was estimated to detect a 10-percent difference in the alpha delta sleep in patients with and without MDD. A total sample size (n) of 138 subjects (69 in each group) gives an alpha of 0.5 and a power of 90 percent. The above statistical computations were performed using the SPSS version 11.5 and JMP version 6.1.

Results

The scoring criterion for alpha-delta sleep was validated in the study. A sample of 15 polysomnograms were randomly selected from the 150 studies identified and a board certified sleep specialist who was not involved in this study scored alpha-delta sleep using the criteria defined in the study. The random sample had six studies with a history of MDD and nine studies with no history of MDD. The alpha-delta sleep scoring was similar and mean alpha-delta score was slightly higher in the studies scored by the principle investigator. The variation was 2.66±1.82 percent (mean±SD).

One hundred and fifty subjects were included in this study, of which 75 (50%) were subjects with MDD and the other 75 (50%) were subjects without MDD (Figure 3). When stratified by age groups (18–29, 30–39, 40–49, 50–59 and ?60 years), the distribution of subjects was again similar in the two groups (Figure 4). The baseline demographic and disease characteristics of subjects in the MDD and non-MDD groups were not significantly different except for gender and ESS score. The majority of subjects with MDD were female, somewhat consistent with the higher prevalence of depression in women (78.67% vs. 54.67%, p<0.05); in addition, compared to nonMDD subjects, subjects with MDD reported more sleepiness, although the ESS score remained within normal limits (9.09±4.09 vs. 6.0±4.55, p<0.001, Table 1).

Compared to non-MDD subjects, subjects with MDD had significantly higher sleep efficiency (p=0.01), shorter REM sleep latency (p<0.001), higher REM sleep time (p<0.001) and less slow wave sleep distribution (p<0.001, Table 2). Furthermore, leg movements were more predominant in subjects with MDD compared to non-MDD subjects (p=0.02). Between group difference in percent alpha delta sleep is illustrated in Figure 5. Alpha-delta sleep was significantly higher in MDD subjects compared to non-MDD subjects (p<0.01). This difference persisted even after stratifying each of the groups into five different age groups (Table 3). Gender did not significantly affect the prevalence of alpha-delta sleep, but female subjects with MDD had relatively more alpha-delta sleep compared to male subject with MDD (p=0.057); even within the non-MDD group, female subjects had relatively more alpha-delta sleep than male subjects (Figure 6).

Within the MDD group, increasing age had an inverse correlation with alpha-delta sleep, particularly at age 60 and higher (p<0.001, Figure 7). On the other hand, history of comorbid insomnia or comorbid psychiatric disorders (other than MDD) did not significantly affect the prevalence of alpha-delta sleep in subjects with MDD (Figure 8, Figure 9).

Of the 75 MDD subjects in this study, 30 patients had never been prescribed any medication, 26 were on selective serotonin reuptake inhibitors (SSRI), nine on tricyclic antidepressants (TCA), four on benzodiazepines (BZD), three on BZD and SSRI, and three on BZD plus SSRI and TCA. Use of SSRI in patients with MDD was associated with a significant decrease in alpha delta-sleep compared to subjects with MDD who were not on any medicine (p<0.001, Figure 10)

Of all the demographic and disease characteristic variables of interest, alpha-delta sleep showed a statistically significant but weak association with ESS scores (r=0.246, p=0.034). A scatter plot generated according to the multiple linear regression model also demonstrated a somewhat linear correlation between ESS and the percentage of alpha-delta sleep
(r2=0.06, p<0.05, Figure 11).

In post-hoc analysis within the MDD group, a median split was performed on the percentage of alpha-delta sleep in delta wave sleep to examine whether a high percentage of alpha-delta sleep (defined as ?15% of alpha-delta sleep) had any unique characteristics compared to subjects with low percentage of alpha-delta sleep (defined as <15% of alpha-delta sleep) (Figure 12). The demographic and disease characteristics (Table 4) and polysomnographic characteristics (Table 5) were examined among patients with high and low alpha-delta sleep within the MDD group. The mean age of MDD patients with high alpha-delta sleep was significantly lower than that of MDD patients with low alpha-delta sleep (36.92±10.65 vs. 47.36±15.77, p=0.001). Interestingly, although the mean ESS score of all MDD patients remained within normal limits (9.09±4.09, Table 1), the MDD patients with high alpha-delta sleep had a significantly higher ESS score compared to the MDD patients with low alpha-delta sleep (10.36±4.09 vs. 7.92±3.79, p=0.009). Furthermore, multiple logistic regression analysis also demonstrated that the MDD patients with high alpha-delta sleep had a 3.15 greater odds of having excessive daytime sleepiness, defined as ESS of 10 or greater, compared to the MDD patients with low alpha-delta sleep (Table 6).

Discussion

This polysomnogram-based study was conducted to evaluate the prevalence of alpha-delta sleep in patients with MDD, as well as to determine various demographic and clinical correlates of alpha-delta sleep in patients with MDD. In this study, we found that MDD subjects (compared to subjects with no MDD) had a higher prevalence of alpha-delta sleep. In addition, MDD subjects, compared to subjects with no MDD, had a higher sleep efficiency, shorter REM sleep latency, less slow wave sleep, higher REM sleep (as a percentage of total sleep), and more frequent leg movements.

We found it interesting that subjects with MDD have a significantly higher sleep efficiency compared to non-MDD subjects in this study. In general, subjects with MDD complain of decreased amount and poor quality of sleep. However, a minority of patients with MDD, perhaps 10 to 15 percent, demonstrate high sleep efficiencies and report spending more time in bed. This finding is often associated with complaints of anergia and psychomotor slowing. The nature of depression in these patients is sometimes consistent with seasonal affective disorder, which is more prevalent in higher latitudes where winters are longer and more severe.[34] Given that the majority of patients seen at the Cleveland Clinic come from northeast Ohio, which has cold climatic conditions and a relatively higher prevalence of seasonal affective disorder, this finding of higher sleep efficiency, although unexpected, is not completely surprising.[38]

We found that subjects with MDD had significantly higher alpha-delta sleep compared to non-MDD subjects, according to our criteria for alpha-delta sleep.

As discussed earlier in the background section, previous studies have demonstrated inconsistent findings regarding the association between MDD and alpha-delta sleep. However, most studies defined alpha-delta sleep as “prominent” alpha frequencies occurring during slow wave sleep. Since the previous studies had not used objective criteria (as was done in the current study), this may have contributed to the inconsistent findings. When alpha-delta sleep in patients with MDD was first described, it was thought to be associated with a heterogeneous group of psychiatric patients with somatic malaise and fatigue.[26] Subsequently, alpha-delta sleep was noted to occur in a majority (70%) of patients with fibromyalgia and among healthy subjects deprived of non-REM sleep.[19] Interestingly, in the later group, alpha-delta sleep also correlated with the presence of depression and irritability. Ware et al[34] found that alpha-delta sleep was present in 42 percent of patients with major depression. In addition, Hudson et al[39] showed that mood disorders are often diagnosed in fibromyalgia patients who are known to have alpha-delta sleep. Although Manu et al[21] found no correlation between alpha-delta sleep (criteria for alpha-delta sleep defined as “prominent alpha frequencies in slow wave sleep”) and chronic fatigue syndrome, fibromyalgia, major depression, primary sleep disorders, or Lyme disease, they did find that alpha-delta sleep was more evident among subjects with chronic fatigue but without major depressive disorders. However, that study had several limitations, including a small sample size of 30 patients and lack of a control group. The current study with its larger sample size revealed for the first time that subjects with MDD have a higher prevalence of alpha-delta sleep compared to non-MDD subjects at least according to the newly proposed criteria for alpha-delta sleep (duration of delta wave sleep with superimposed alpha frequency divided by the total duration of slow wave sleep multiplied by 100).

Compared to non-MDD subjects, subjects with MDD had significantly more alpha-delta sleep in every age group. Within the MDD group, increasing age, particularly 60 years and higher, and the use of SSRIs was associated with decreased alpha-delta sleep. On the other hand, among different age groups of non-MDD subjects, there was no significant difference in the prevalence of alpha-delta sleep. Therefore, it is likely that differences between the sleep of patients with MDD and normal controls is not just influenced by age but also by other factors. Age strongly affects the relationship between sleep parameters and mood disorders, and the interaction of depression and age on sleep parameters suggests that depression may accelerate the effects of aging on sleep.[16,40] The relationship between a generalized decrease in alpha-delta sleep and aging in subjects with MDD but not in non-MDD subjects is not clearly understood. One possible explanation for this finding was noted in a previous study that showed that sleep parameters that change with age do not change in patients with depression. For example, elderly patients with depression do not tend to have a significant reduction in slow wave sleep when compared to control subjects.[2,6]

Although there were limited numbers of MDD subjects using tricyclic antidepressants (n=9), and benzodiazepines (n=4), the number of patients using SSRIs (n=26) was comparable to the number patients not using any medication (n=30; p=0.085). A history of SSRI medication use was associated with decreased alpha-delta sleep. This relationship is probably attributable to the effects of SSRIs on MDD disease activity or on the sleep architecture itself. However, duration of treatment may also contribute to these findings, and several studies have identified various changes in sleep architecture associated with SSRIs and selective serotonin and norepinephrine reuptake inhibitors (SNRIs). Knott et al[41] demonstrated that acute paroxetine (an SSRI) did not alter the EEG in MDD patients but chronic treatment was associated with significant alterations as shown by diffuse decrease in alpha power and increases in slow (delta and theta) and anterior fast (beta) wave power. Kluge et al[42] found that duloxetine (an SNRI) increases N3 sleep and suppresses REM sleep in patients with major depression. McClelland et al[43,44] studied the acute effects of paroxetine (6 hours after taking it) on EEG in normal subjects, which revealed a reduction in alpha waves and an increase in delta waves. However, data on the chronic effects of SSRIs/SNRIs on EEG in otherwise healthy subjects is lacking. Therefore, in addition to an affect on disease severity, SSRIs/SNRIs may have an affect on EEG power itself, and thus this may play an important role in decreasing alpha-delta sleep in the subjects with MDD found in our study.

Alpha-delta sleep was not significantly affected by gender or comorbid insomnia. This finding is somewhat surprising in that women with depression, in general, have a relatively higher prevalence of insomnia, fatigue, and pain-related symptoms. These symptoms, especially fatigue and pain, were found to be associated with alpha-delta sleep. Based on this association, we had expected a higher prevalence of alpha-delta sleep in women with depression.

MDD is often associated with other comorbid psychiatric disorders, especially anxiety disorders. We found that patients with MDD and no history of other comorbid psychiatric disorders had a relatively higher alpha-delta sleep compared to MDD subjects with a history of other comorbid psychiatric disorder, such as bipolar disorder and anxiety disorders (Figure 9). However, none of these differences were statistically significant (p=0.776). The difference in the amount of alpha-delta sleep observed between these groups is hard to explain given that in general more than 80 percent of subjects with MDD have comorbid anxiety. In fact, what was surprising was the relatively large number of MDD subjects with no comorbid anxiety.

Alpha-delta sleep is thought to be associated with a state of hypervigilance and hence difficulties with initiating and maintaining sleep or experiencing nonrestorative sleep associated with daytime fatigue (but not necessarily excessive daytime sleepiness), similar to patients with psychophysiological insomnia. Similarly, most patients with MDD complain of insomnia, increased daytime fatigue, and sometimes sleepiness. However, These patients do not consistently show evidence of increased daytime sleepiness when tested objectively.11–13 Although the present study revealed significant differences between ESS scores of MDD and non-MDD subjects (9.09±4.09 vs. 6.01±4.55, p<0.001) the total ESS score remained within the normal range (<10). This is consistent with what we had expected based on studies done in patients with MDD and other studies that evaluated the effects of alpha-delta sleep. The reason why some MDD subjects experience daytime sleepiness remains unanswered. As mentioned above, alpha-delta sleep is thought to be associated with nonrestorative sleep, which, in turn, may result in daytime sleepiness or fatigue. The post-hoc analysis offers some insight if the severity of alpha-delta sleep has any influence on the symptoms of excessive daytime sleepiness. Subjects with MDD were divided into two groups: high alpha-delta sleep (defined as ?15% of alpha-delta sleep) and low alpha-delta sleep (defined as <15% of alpha-delta sleep) by a median split method. We found that MDD subjects with high alpha-delta sleep had higher ESS scores and they were at 3.15 greater odds to have excessive daytime sleepiness (ESS?10) compared to MDD subjects with low alpha-delta sleep. The influence of alpha-delta sleep on daytime sleepiness in MDD subjects persisted even after controlling for age, neck circumference, BMI, sleep efficiency, AHI, and arousal index. The cause-effect relationship between alpha-delta sleep and excessive sleepiness is not well understood. One interesting hypothesis proposed by Englebienne and De Meirleir44 is that an acquired channelopathy with loss of intracellular potassium could lead to metabolic and intracellular abnormalities, resulting in central fatigue and sleep disturbances, such as alpha intrusion seen in fibromyalgia patients. However, data from Hoof et al[45] did not support the inclusion of sleep disturbances, including alpha intrusion, in their list of potential consequences of the suggested channelopathy. Since the etiology of alpha-delta sleep itself is not understood, any possible explanation on the association/ causation of alpha-delta sleep and excessive daytime sleepiness is likely to be only speculative.

In the post-hoc analysis, one interesting finding was the slightly higher AHI in subjects with MDD with low alpha-delta sleep compared to subjects with MDD with high alpha delta sleep. The AHIs in low and high alpha delta sleep groups were 2.46±1.53 and 1.62±1.17, respectively. This difference was statistically significant at p=0.009 and 95 percent CI 0.22 to 1.47. This difference may be mediated by the age difference in the two groups of subjects. The mean age in low and high alpha delta sleep groups were 47.36 and 36.92 years, respectively. It is well known that age is a risk factor for sleep apnea and increasing age in general is associated with an increase in the AHI.[46]

There were several unique features that enhanced the utility of this study. First,we studied patients with MDD and had non-MDD subjects serving as controls. None of the subjects in our study had any disorders other than MDD that are known to be associated with alpha-delta sleep. This is in contrast to the previous studies that had a significant number of patients with MDD with comorbid conditions that are known to have prominent alpha-delta sleep, such as fibromyalgia and chronic fatigue syndrome. We rescored all polysomnograms with AASM 2007 criteria, which is thought to reduce inter-reader variability. Finally, we proposed a new objective method for characterizing alpha-delta sleep (and validated it) instead of using highly subjective and nonreproducible “prominent” alpha-delta sleep criteria that was used in almost all of the studies thus far reviewing this topic.

However, our study does have several limitations. This is a retrospective study with its attendent deficiencies with an inability to influence data collection (such as any sleep deprivation prior to the sleep study). Only 960 out of 14,549 subjects (6.6%) in the sleep center database had a current active diagnosis of MDD, which is less than what is observed in the general population. It is possible that only those patients who had significant depression were diagnosed and hence studied. The other possibility is that a number of subjects in the non-MDD group may have had mild or moderate depression that was not clinically diagnosed. In addition, it is important to realize that the “first-night” effect (the altered quality and duration of sleep observed on the first night of a sleep study) may have affected the result of the polysomnograms. Ideally, each of these patients should have had two nights of polysomnography, with the data obtained from the second night used for the analysis. While this would minimize first night effect, it would have been prohibitively expensive. However, it is probably safe to assume that both MDD and non-MDD subjects were equally affected by first night effect.[48] There is some literature that has shown that depressed patients often do not show evidence of the first night effect.[49,50] This apparent absence of first night effect may be secondary to the abnormal sleep patterns of depressed patients being superimposed upon and effectively concealing the characteristic features of first night effect. Another limitation is that the principal investigator who scored and interpreted the alpha-delta sleep was not blind to group membership. However, when the alpha-delta criteria were validated by using a sample of studies by another scorer (board certified in sleep medicine and blinded to the disease status of the subjects), scores/interpretations were similar to those obtained by the prinicipal investigator. We had not planned (and hence did not have enough sample size) to analyze the association of alpha-delta sleep with various comorbidities, medications, and daytime sleepiness, and therefore a relatively small sample size was available for analyses in these subgroups. Using a median split method to classify subjects with MDD into two groups is not ideal as a great deal of information is lost when a continuous variable is converted into binary categorical variables.

Conclusion

This study shows that subjects with MDD have a higher prevalence of alpha-delta sleep. Compared to non-MDD subjects, subjects with MDD have a shorter REM latency, more total REM sleep, less slow wave sleep, and tend to have higher ESS scores (but within the normal range). Within the MDD group, SSRI/SNRIs and higher age decrease alpha-delta sleep. In contrast, a history of comorbid insomnia or other comorbid psychiatric disorders did not significantly affect the prevalence of alpha-delta sleep. Alpha-delta sleep seems to be associated with daytime sleepiness in subjects with MDD. Finally, subjects with MDD with high alpha-delta sleep (as defined as ?15% of delta waves with superimposed alpha) were at 3.15 greater odds to have excessive daytime sleepiness (ESS?10) compared to those with low alpha-delta sleep.

In terms of future directions, it may be interesting to control for gender since MDD has a higher prevalence in women and women tend to complain more about fatigue and nonrestorative sleep, both of which are associated with alpha-delta sleep. Further research to find an association between the severity of depression and alpha-delta sleep and between fatigue and alpha-delta sleep may be helpful in further delineating the relationship between MDD and alpha-delta sleep. Finally, a longitudinal follow-up study of patients with depression with or without SSRI treatment may help with understanding the nature and course of alpha-delta sleep in patients with MDD.

References
1. Boyd JH, Weissman MM. Epidemiology of affective disorders: a reexamination and future directions [review]. Arch Gen Psychiatry. 1981;38:1039–1046.
2. Schmitz B. Depression and mania in patients with epilepsy. Epilepsia. 2005;46(Suppl 4):45–49.
3. Gainotti G, Antonucci G, Marra C, Paolucci S. Relation between depression after stroke, antidepressant therapy, and functional recovery. J Neurol Neurosurg Psychiatry. 2001;71:258–261.
4. Lemke MR, Fuchs G, Gemende I,
et al. Depression and Parkinson’s disease. J Neurol. 2004;251
(Suppl 6):24–27.
5. Davis L, Uezatob A, Newella JM, Frazierd E. Major depression and comorbid substance use disorders. Curr Opin Psychiatry. 2008;21:14–18.
6. Davis LL, Frazier E, Husain MM,
et al. Substance use disorder comorbidity in major depressive disorder: a confirmatory analysis of the STAR_D cohort. Am J Addict. 2006;15:278–285.
7. Katona C, Peveler R, Dowrick C. Pain symptoms in depression: definition and clinical significance. Clin Med. 2005;5:390–395.
8. Kleim B, Ehlers A. Evidence for a curvilinear relationship between posttraumatic growth and posttrauma depression and PTSD in assault survivors. J Trauma Stress. 2009;22(1):45–52.
9. Rugulies R. Depression as a predictor for coronary heart disease a review and meta-analysis. Am J Prev Med. 2002;23(1):51–61.
10. Ford DE, Kamerow DB. Epidemiologic study of sleep disturbance and psychiatric disorders: an opportunity for prevention? JAMA. 1989;262: 1479–1484.
11. Benca RM. Mood disorders. In: Kryger MH, Roth T, Dement WC (eds). Principles and Practice of Sleep Medicine, Fourth Edition. Philadelphia: Elsevier Saunders;2005:1311–1326.
12. Reynolds CF III, Coble PA, Kupfer DJ, et al. Application of the multiple sleep latency test in disorders of excessive sleepiness. Electroencephalogr Clin Neurophysiol. 1982;53:443–452.
13. Billiard M, Dolenc L, Aldaz C, et al. Hypersomnia associated with mood disorders: a review perepective.
J Psychosom Res. 1994;38(Suppl 1): 41–47.
14. Breslau N, Roth T, Rosenthal L,
et al. Sleep disturbance and psychiatric disorders: a longitudinal epidemiological study of young adults. Biol Psychiatry. 1996;39:411–418.
15. Jarrett RB, Rush AJ, Khatami M,
et al. Does the pretreatment polysomnogram predict response to cognitive therapy in depressed outpatients? A preliminary report. Psychiatry Res. 1990;33:285–299.
16. Benca RM, Obermeyer WH, Thisted RA, et al. Sleep and psychiatric disorders: a meta-analysis. Arch Gen Psychiatry. 1992;49:651–668.
17. Reynolds CF III, Kupfer DJ. Sleep research in affective illness: state of the art circa 1987. Sleep. 1987;10:199–215.
18. Hauri P, Hawkins DR. Alpha-delta sleep. Electroencephalogr Clin Neurophysiol. 1973;34:233–237.
19. Moldofsky H, Scarisbrick P, England R, et al. Musculoskeletal symptoms and nonREM sleep disturbance in patients with “fibrositis syndrome” and healthy subjects. Psychosom Med. 1975;37:341–351.
20. Bronco J, Atalaia A, Paiva T. Sleep cycles and alpha-delta sleep in fibromyalgia syndrome. J Rheumatol. 1994;21:1113–1117.
21. Manu P, Lane TJ, Matthews DA,
et al. Alpha-delta sleep in patients with chief complaint of chronic fatigue. South Med J. 1994;87: 465–470.
22. Mahowald MW, Mahowald ML, Bundlie SR, et al. Sleep fragmentation in rheumatoid arthritis. Arthritis Rheum. 1989;32: 974–978.
23. Drewes AM, Svendsen L, Taagholt SJ, et al. Sleep in rheumatoid arthritis: a comparison with healthy subjects and studies of sleep/wake interactions. Br J Rheumatol. 1998;37:71–81.
24. Anch AM, Lue FA, MacLean AW,
et al. Sleep physiology and psychological aspects of the fibrosis (fibromyalgia) syndrome. Can J Psychol. 1991;45:178–184.
25. Schneider-Helmert D, Kumar A. Sleep, its subjective perception, and daytime performance in insomniacs with a pattern of alpha sleep. Biol Psychiatry. 1995;37: 99–105.
26. Pivik RT, Harman KA. Reconceptualization of EEG alpha activity during sleep: all alpha activity is not equal. J Sleep Res. 1995;4:131–137.
27. Connemann BJ, Mann K, Pascual-Marki RD, Roschke J. Limbic activity in slow wave sleep in a healthy subject with alpha-delta sleep. Psychiatry Res. 2001;107: 165–171.
28. Honda M, Koga E, Ishikawa T, et al. Alpha-delta sleep in a case of non-24-h sleep-wake syndrome: quantitative electroencephalogram analysis of alpha and delta band waves. Psychiatry Clin Neurosci. 1997;51:387–392.
29. Saskin P, Moldofsky H, Lue FA. Periodic movements in sleep and sleep-wake complaint. Sleep. 1985;8:318–324.
30. Sarzi-Puttini P, Rizzi M, Andreoli A, et al. Hypersomnolence in fibromyalgia syndrome. Clin Exp Rheumatol. 2002;20:69–72.
31. Kupfer DJ. REM latency: a psychobiologic marker for primary depressive disease. Biol Psychiatry. 1976;11:159–174.
32. Akiskal HS, Lemmi H, Yeravanian B, et al. The utility of the REM latency test in psychiatric diagnosis: a study of 81 depressed outpatients. Psychiatry Res. 1982;7:101–110.
33. International Classification of
Sleep Disorders, Second Edition. Diagnostic and Coding Manual. Westchester, Ill: American Academy of Sleep Medicine; 2005.
34. Ware JC, Russell JJ, Campos E. Alpha intrusions into the sleep of depressed and fibromyalgia syndrome (fibrosis) patients. Sleep Res. 1986;15:210.
35. Iber C, Ancoli-Israel S, Chesson A, Quan SF (eds). The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology, and Technical Specification, First Edition. Westchester, IL: American Academy of Sleep Medicine;
2007.
36. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorder, Fourth Edition, Text Revision. Washington, DC: American Psychiatric Press Inc.; 2000.
37. Detre TP, Himmelhoch J, Swartzburg M, Kupfer DJ. Hypersomnia and manic depressive disease. Am J Psychiatry. 1972;128:1303–1305.
38. Lam RW, Levitt AJ, Levitan RD, et al. The Can-SAD study: a randomized controlled trial of the effectiveness of light therapy and fluoxetine in patients with winter seasonal affective disorder. Am J Psychiatry. 2006;163(5):805–812.
39. Hudson JI, Pope HG Jr. Fibromyalgia and pshychopathology: is fibromyalgia a form of “affective spectrum disorder?” J Rheumatol Suppl. 1989;19:15–22.
40. Reynold CF III, Kupfer DJ,
Thase ME, et al. Sleep, gender
and depression: an analysis of gender effects on the electroencephalographic sleep of 302 depressed outpaients. Biol Psychiatry. 1990;28:673–684.
41. Knott V, Mahoney C, Kennedy S, Evans K. EEG correlates of acute and chronic paroxetine treatment in depression. J Affect Disord. 2002;69:241–249.
42. Kluge M, Schüssler P, Steiger A. Duloxetine increases stage 3 sleep and suppresses rapid eye movement (REM) sleep in patients with major depression. Eur Neuropsychopharmacol. 2007; 17(8):527–531.
43. McClelland G, Raptopoulos P, Jackson D. The effect of paroxetine on the quantitative EEG. Acta Psychiatr Scand Suppl. 1989;80: 50–52.
44. De Meirleir K, De Becker P, Niji J, et al. CFS etiology, the immune system, and infection. In: Englebienne P, De Meirleir K (eds). Chronic Fatigue Syndrome: A Biological Approach. New York: CRC Press;2002:201–208.
45. Van Hoof E, De Becker P,
Lapp C, et al. Defining the occurrence and influence of
alpha-delta sleep in chronic
fatigue syndrome. Am J Med Sci. 2007;333(2):78–84.
46. Bixler EO, Kales A, Cadieux RJ,
et al. Sleep apneic activity in older healthy subjects. J Appl Physiol. 1985;58:1597–1601.
47. McClelland G, Raptopoulos P. EEG and blood level of the potential antidepressant paroxetine after
a single oral dose to normal volunteers. Pharmacology. 1984;83: 327–329.
48. Carskadon MA, Dement WC. Normal human sleep: an overview. In: Kryger MH, Roth T, Dement WC (eds). Principles and Practice of Sleep Medicine, Fourth Edition. Philadelphia: Elsevier Saunders; 2005:13–23.
49. Akiskal HS, Lemmi H, Yerevanian Â, et al. The utility of the REM latency test in psychiatric diagnosis: a study of 81 depressed outpatients. Psychiatry Res. 1982;7:101–110.
50. Annsseatt M, Kupfer DJ, Reynolds CF III. Internight variability of REM latency in major depression. Implications for the use of REM latency as a biological correlate. Biol Psychiatry. 1985;20:489–505.
?