Insomnia and depression: How much is the overlap?

Abstract

Background:

Previous literature considers insomnia as one of the features, predictor, and also as a residual symptom of depression. However, chronic insomnia and major depressive disorder (MDD) have overlapping features making differentiation between two difficult.

Materials and Methods:

Forty subjects in each of the three categories-MDD, insomnia (I) and combined diagnoses (MDD-I) were recruited in this study after excluding potential confounders. Diagnosis of MDD was made following Diagnostic and Statistical Manual 5 edition (DSM-5), while the International Classification of sleep disorders 3 edition criteria of insomnia were used for diagnosing insomnia. The severity of insomnia and depression was assessed using the Insomnia Severity Index (ISI) and Patient Health Questionnaire-9 (PHQ-9), respectively. Fatigue was assessed using the Fatigue Severity Scale (FSS), which was translated in Hindi for this study. All subjects were also asked regarding effect of good sleep at night on daytime symptoms, especially on mood.

Results:

Subjects in MDD group were younger than the other two. Insomnia group was significantly different from the other two groups on most of the measures according to the DSM-5 criteria for MDD. MDD group had lesser frequencies of initial insomnia, middle insomnia, dissatisfaction with sleep and overall distress during the day. MDD-I group had a higher prevalence of daytime sleepiness and hyperactivity/impulsivity. PHQ-9 score was the lowest in the insomnia group. Despite statistically significantly different, ISI score was clinically comparable. The severity of fatigue was comparable across three groups. Contrary to the MDD group, subjects in insomnia and MDD-I group reported significant improvement in daytime symptoms after a good sleep for even one night.

Conclusion:

There is considerable overlap of symptoms between insomnia and MDD. Subjects having insomnia report significant improvement in daytime and mood symptoms after good sleep, contrary to subjects with MDD.

INTRODUCTION

Relationship between insomnia and depression goes in more than one way. A body of literature suggests that persistent insomnia paves way for depression; some evidence describe insomnia as a residual symptom of depression, and still, other studies suggest that both have bidirectional relationship.[1,2,3]

Insomnia has been found to herald the onset of depression across many studies.[1,4,5] Buysse et al.[5] in a 20 years long study found that subjects with “insomnia” did not develop “depression” and vice versa at any point of time. In addition, while subjects with “insomnia” developed “insomnia with depression” at later point of time, subjects with “depression” did not develop insomnia, challenging the bi-directional relationship.

Another line of evidence suggests that a significant number of subjects with depression report insomnia as a residual symptom.[2,6] However, evidence suggests that these two have more intricate relationship. Interestingly, the type of therapy offered for the management of depression-pharmacotherapy or behavior therapy does not influence the persistence of depression, even though some antidepressants have hypnotic potential.[6] Further, depressive symptoms in these patients do not improve till sleep disturbances are resolved.[2] These evidence raise the possibility that these subjects might be suffering from two independent disorders-depression as well as insomnia, bringing theories of “bidirectional relationship” as well as “residual symptom” in question.

Insomnia has been found to be associated with depressive symptoms and suicidal ideation across different populations.[7,8] Epidemiological data suggest that a significant number of subjects complain of daytime symptoms after a night with poor sleep, characterized by poor concentration and mood changes varying from depressed, to anxious or irritable.[9] A recent study has suggested that in patients with insomnia, improvement in depression was primarily related to improvement in these daytime symptoms.[10] Carney et al.[11] suggested that depressive symptoms among patients with insomnia may actually reflect the effect of insomnia rather than true depression. Moreover, many of the daytime symptoms of insomnia overlap with symptoms of depression.[12,13]

Finally, cognitive behavior therapy for insomnia (CBT-I) has been found to improve not only sleep quality but also the depressive symptoms.[14] Prescription of CBT-I among subjects with insomnia has differential effects on depressive symptoms, and suicidal ideations e.g., improvement in suicidal ideations is short-lasting, while improvement in depressive symptoms appears to be long-lasting, as along as 18 months.[15] Moreover, improvement in depressive symptomatology appears to be related to be a function of sleep efficiency and symptoms of insomnia, rather than improvement in overall sleep quality.[16,17] These evidences also explain why subjects with insomnia developed depression, while subjects with depression did not develop insomnia at a later point.[5]

Together these evidence suggest that depression occurring consequent to insomnia may not be true clinical depression. Instead, it could be “distress,” related to the poor sleep at night and consequent daytime symptoms of insomnia that have significant overlap with symptoms of depression. Previous studies have examined this issue using items in Beck's depression inventory and indicated that daytime symptoms were consequent to poor night time sleep.[11,18] However, to the best of our knowledge, this has never been systematically examined using standard criteria for insomnia and depression.[12,13]

Thus, the present study was planned to compare the prevalence of symptoms of insomnia according to the International Classification of Sleep Disorders 3^rd Edition (ICSD-3) and that of depression as per the Diagnostic and Statistical Manual 5 edition (DSM-5) criteria among subjects with “insomnia,” “depression” and “insomnia with depression.” In addition, the severity of insomnia, depression, and fatigue was compared across three groups.

MATERIALS AND METHODS

This study was conducted after obtaining approval from institutional ethics committee of Himalayan Institute of Medical Sciences, Dehradun. Subjects fulfilling criteria for major depressive disorder (MDD) according to DSM-5 and for insomnia according to ICSD-3 criteria were screened from the outpatients department of department of Psychiatry and Sleep Clinic.[12,13] Those fulfilling inclusion criteria were screened for the exclusion criteria. Exclusion criteria included subjects having any other medical disorder including neurocognitive disorder, stroke, neurodegenerative disorders, hypothyroidism, diabetes mellitus, coronary artery disease, congestive heart failure, chronic obstructive lung disease, bipolar disorder, anxiety disorder, psychotic disorder; those taking psychotropic medications; those suffering from other sleep disorders, for example, obstructive sleep apnea, restless legs syndrome. Those found eligible to participate in the study were explained the aims of the study and were requested to participate. Written informed consent was obtained from those who were willing to participate.

After this, all included subjects were screened for the presence of symptoms that characterize MDD according to DSM-5 and for symptoms of insomnia according to ICSD-3.[12,13] Subjects were divided into three groups – those with MDD, insomnia (I), and insomnia with MDD-I.[12,13] In each group, consecutive 40 subjects were included. Their demographic data and information regarding their illness were gathered. Physical examination was done.

The severity of insomnia was assessed using the Hindi version of insomnia severity index (ISI).[19,20] The severity of MDD was measured using the Patient Health Questionnaire-9 (PHQ-9).[21] Each subject in all the groups was asked, “Does good sleep at night significantly improves your symptoms the next day?”

Insomnia severity index

It contains seven items that assess insomnia over the past 1 month.[19,20] First three items ask for trouble in initiating, maintaining sleep, and early morning awakening. Other items address dissatisfaction with sleep, daytime functions, recognition of insomnia by others, and finally, distress caused by insomnia. These are scored on a five-point Likert's scale (0 = no problem to 4 = very severe problem). Score of 0–7 depicts the absence of insomnia, 8–14 shows subthreshold insomnia, 15–21 represents moderate, and 22–28 shows severe insomnia.

Patient Health Questionnaire

PHQ-9 is one of the modules of PRIME-MD that is useful for the diagnosis of certain psychiatric disorder and is self-reported. PHQ-9 contains nine DSM-5 criteria for depression, which are to be scored on a Likert scale of 0–3 (not at all to nearly every day). Scores of 5, 10, 15, and 20 on PHQ-9 represent the severity of depression-mild, moderate, moderately severe, and severe, respectively.[21]

Fatigue Severity Scale

It is a 9-item self-report measure that measures fatigue on a visual analog scale of 0–7. These items represent the occurrence of fatigue, interference with daily functions due to fatigue, subjective perception of fatigue, and effect of fatigue on motivation.[22] This has been validated across various diagnostic groups.[22,23]

Translation and validation of Fatigue Severity Scale

The English version of Fatigue Severity Scale (FSS) was translated to Hindi by two bilingual experts separately, and version 1 (V1) and version 2 (V2) were made. After independent translation, both versions were compared. Both translators compared each item for language and meaning of important words until a consensus was reached, and a third version (V3) was developed. V3 was back-translated to English by two different bilingual experts, resulting in version V4 and version V5. Discrepancies, if any, between these two versions were discussed and a final back-translated version, version V6, was developed and compared with the original scale. The back-translated version (V6) was brought as close as possible to the original scale and appropriate changes in V3 were made, wherever required in the presence of all translators. Thus, Hindi version ready for testing in a small sample was obtained (V7).

A pilot study was conducted using version V7 of FSS on 40 subjects suffering from MDD and were asked to point out the difficulties they faced in comprehending the items on the scale. The suggestions made during the pilot study were taken into account and another version, version V8, was developed. This version [Annexure 1] was used in the current study.

Statistical analysis

Statistical analysis was performed using Statistical Package for Social Sciences (SPSS v. 21.0; IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp.). Descriptive analysis was performed. The Chi-square analysis was used to compare proportions among three groups. Post hoc analysis to compare cell values of proportions across the table was done using moment-corrected standardized residual technique.[24] One-way ANOVA with post hoc Tukey was used to compare numerical values across groups. Reliability analysis of the FSS was performed using Cronbach's alpha.

RESULTS

This study included 40 subjects in each of the three categories-MDD, insomnia and depression with insomnia. Subjects in MDD group were younger than other two (33.80 ± 13.34 years in MDD; 38.90 ± 10.55 years in I and 40.65 ± 11.35 in MDD-I group; P = 0.03). All groups were comparable with regard to gender distribution (P = 0.15) and women constituted nearly half of each group.

In insomnia group, 62.5% of subjects had chronic insomnia while remaining had short-term insomnia. In MDD-I group, chronic insomnia (57.5%) dominated over short-term insomnia. In the MDD group 10% subjects and in MDD-I group 12.5% had depressive symptoms <1 month, while rest had a longer illness.

Post hoc analysis suggested that insomnia group was significantly different from other two groups on most of the measures according to DSM-5 criteria for MDD-sad mood, loss of interest, weight change, change in appetite, fatigue, feeling of worthlessness and guilt, concentration difficulties and suicidal ideas [Figure 1a]. Similarly, the MDD group had lesser frequencies of initial insomnia, middle insomnia, and dissatisfaction with sleep and overall distress during the day compared to the other two groups. MDD-I group had a higher prevalence of daytime sleepiness, and hyperactivity/impulsivity compared to the other two groups [Figure 1b]. Comparison of scores on various parameters-PHQ-9, ISI and FSS is depicted in Table 1.

Figure 1.

(a) Comparison of symptoms of depression (Diagnostic and Statistical Manual 5 edition) across various diagnostic categories (b) Prevelence of symptoms of insomnia (International Classification of Sleep Disorders 3^rd Edition) across various diagnostic categories

Table 1.

Comparison of groups on “relationship between sleep and mood and mean scores of scales”

	Major depressive disorder (n=40)	Insomnia (n=40)	MDD + insomnia (n=40)	P
Affirmative response to “Do you find a significant change in your mood after a good night sleep?,” n (%)	1 (2.5)	37 (92.5)	38 (95)	<0.001
ISI*	12.43±3.320	16.60±2.629	15.33±2.645	<0.01
PHQ-9*	15.48±2.970	6.600±2.753	13.05±3.419	<0.01
FSS*	38.65±6.447	40.63±7.217	39.90±6.155	0.41

*Scores represented as mean±SD. ISI – Insomnia Severity Index; FSS – Fatigue Severity Scale; PHQ-9 – Patient Health Questionnaire-9; SD – Standard deviation; MDD – Major depressive disorder

Translation of Fatigue Severity Scale in Hindi

This was a cross-cultural translation; hence, certain difficulties were faced during the process which have been discussed below. Certain modifications in the words were made in the final translated version (V8) so that the contextual meaning of the original version was retained. For example, first item was originally translated as “jab main thaka hua hota hun mere manobal mein kami aa jati hai” but in final version “manobal mein kami aa jati hai” was changed to “manobal kam ho jata hai” so that the final translated question was close to the original question “My motivation is lower when I am fatigued.” Third item was originally translated as “main asani se thak jata hun” but in the final version “thak jata hun” which was changed to “thakan hoti hai” so that the final question matched with the original question of “I am easily fatigued” Fourth item was originally translated as “thakan se main sharirik kaam nahi kar pata hun” but in the final version “thakan se main” was changed to “thakan mujhe” and “kaam nahi kar pata hun” was replaced with “kaam karne mein bahdha dalti hai”. Sixth item was originally translated as “thakan mujhe lagatar sharirik karya karne se rokti hai” but as it was “my fatigue” in the original FSS hence “thakan” was replaced by “meri thakan” and “karya” was replaced by “kaam” which is more frequently used by the subjects in the catchment area. Item number 7 was translated as “thakan kuch karyon evam mukhya zimmedariyon ko nibhane mein badha dalti hain” but in the final version “thakan” was replaced by “thakavat” and “karyon” was replaced by “kartavyon” as it more related to the original word “duties.” The eighth question was originally translated as “thakavat mere teen pramukh lakshano mein se ek hai” but in the final version “pramukh” was replaced by “mukhya” as it's more commonly used by the patients [Annexure 1].

Reliability analysis suggested Cronbach's alpha of 0.93 for the nine items of FSS. Table 2 shows the Item-Total Statistics for Hindi version of FSS. Most of the items had moderate-to-good correlation with each other [Table 3].

Table 2.

Item-total statistics for Hindi version of Fatigue Severity Scale

	Scale mean if item deleted	Scale variance if item deleted	Corrected item-total correlation	Squared multiple correlation	Cronbach’s alpha if item deleted
Q1	34.10	33.07	0.66	0.56	0.93
Q2	34.58	33.48	0.76	0.76	0.92
Q3	34.60	33.12	0.78	0.81	0.92
Q4	34.30	32.83	0.73	0.67	0.92
Q5	34.08	32.94	0.70	0.71	0.92
Q6	34.10	32.66	0.70	0.54	0.92
Q7	34.28	33.23	0.74	0.69	0.92
Q8	34.55	31.28	0.90	0.83	0.91
Q9	34.43	33.79	0.72	0.64	0.92

Table 3.

Inter-item correlation matrix for Hindi version of Fatigue Severity Scale

	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9
Q1	1.00	0.69	0.63	0.41	0.56	0.50	0.48	0.65	0.42
Q2	0.69	1.00	0.79	0.49	0.57	0.52	0.52	0.72	0.60
Q3	0.63	0.79	1.00	0.56	0.44	0.63	0.63	0.77	0.59
Q4	0.41	0.49	0.56	1.00	0.60	0.65	0.57	0.76	0.66
Q5	0.56	0.57	0.44	0.60	1.00	0.50	0.68	0.70	0.49
Q6	0.50	0.52	0.63	0.65	0.50	1.00	0.55	0.68	0.55
Q7	0.48	0.52	0.63	0.57	0.68	0.55	1.00	0.68	0.66
Q8	0.65	0.72	0.77	0.76	0.70	0.68	0.68	1.00	0.69
Q9	0.42	0.60	0.59	0.66	0.49	0.55	0.66	0.69	1.00

DISCUSSION

This study suggested that there is overlap of symptoms among subjects with depression, insomnia and depression with insomnia. It was seen for not only cognitive but also somatic symptoms according to both ICSD-3 and DSM-5 criteria. Second, PHQ-9 score suggested that subjects with insomnia had at least mild-to-moderate symptoms of depression. Thus, there is a possibility that “distress” reported by subjects with insomnia may be mistaken for clinical depression. However, certain features, namely sadness, guilt, suicidal ideas, change in weight, altered appetite, and hypersomnia can discriminate both groups reliably. In addition, good sleep at night did not improve symptoms in MDD group, contrary to groups having insomnia [Table 1]. Finally, fatigue was comparable across groups.

Insomnia and depression have overlapping symptoms.[25] Earlier study has reported that insomnia subjects with and without depression were comparable with regard to sleep complaints, irritability, fatigue, poor concentration as well as suicidal ideation.[11] Subjects across different groups were comparable with reference to these variables in the present study as well, however, with the exception of suicidal ideations. This discrepancy could be related to the exclusion of subjects with severe illness in depressed group in the earlier study.[11] Similar to the findings of this study sadness, guilt, hypersomnia, and change in appetite have been reported to differentiate depression from insomnia in the past, too.[11] Thus, results of the present study reiterate previous findings among Indian subjects. In addition, the present study found that a significant improvement in daytime symptoms after good sleep for even one night indicates insomnia rather than depression. This differentiation is important considering the fact that sleep disturbance is frequently mistaken for depression in general practice, as reported in the past as well![26]

Due to overlap of symptoms between the two, symptoms of insomnia may be mistaken for depression. Two different lines of evidences favor this theory- first, results of sleep deprivation studies and second, studies done in patients suffering from insomnia.

Results of sleep deprivation studies report short-term effect of sleep deprivation on mood, cognition, and somatic symptoms. Experimental studies using induced multiple nocturnal awakenings protocol (i.e., partial acute sleep deprivation) were found to reduced positive affect; induced anhedonia as well as catastrophic thinking and lastly, increased arousal while good sleep was found to have the opposite effect.[27,28] Total sleep deprivation of even one night i.e., acute total sleep deprivation has been found to have a similar effect on mood, with greater propensity among women.[29] In addition, acute total sleep deprivation has been found to worsen concentration.[30] Thus, even acute sleep deprivation, whether partial or complete can induce mood, somatic, and cognitive symptoms; symptoms that overlap with that of depression.

However, due to difference in pathophysiology, finding among subjects with clinical insomnia might be different from experimental sleep deprivation studies. While sleep-deprivation is associated with the loss of sleep with persistent ability to sleep, insomnia is characterized by persistent hyperarousal.[31] Despite this pathophysiological difference, daytime symptoms in subjects with “pure insomnia” have shown findings similar to that of experimental studies. Among subjects suffering from insomnia, night-time symptoms of insomnia have been found to be associated with cross-sectional as well as incidental depression.[10,18,32,33] In addition, these patients have daytime consequences of night-time insomnia that include both cognitive and somatic symptoms and these have been reported in clinical as well as in general population.[9,18,34] Longitudinal studies have suggested that incidental depression among insomnia patients has been found to be a function of the frequency of insomnia.[32] However, one study spanning over 20 years reported that subjects with “pure insomnia” never developed “pure depression,” rather they had higher odds to develop “depression with insomnia.”[5] There is a possibility that persisting insomnia increased the magnitude and burden of symptoms, which might be perceived as depression. This could be one reason why subjects suffering from insomnia have been found to qualify for mild depression in previous as well as the present study.[11] Studies directed at nonpharmacological management of insomnia provide further support to our argument. These studies have reported improvement in depression after improvement of night-time sleep. CBT-I has been found to improve both cognitive as well as somatic symptoms in addition to mood, consequently improving depression.[18,10]

Since, stress-diathesis theory has been implicated in both depression and insomnia, effect of stress must be discussed here.[35,36] Insomnia itself may be a stressful situation that could activate stress circuitry or vice versa.[37] Second, time spent awake is often mis-utilized in thinking about past failures and problems in life, which could further increase the stress and induce depressogenic thoughts.[33,38,39] In addition, subjects with insomnia have been found to ruminate even during daytime compounding the problem and adding to the existing stress.[40] Activation of stress circuitry can precipitate sadness, cognitive difficulties, as well as somatic symptoms that overlap with symptoms of depression.[36] This “distress” arising out of transient or persistent insomnia can be mistaken for symptoms of “clinical depression.”[11] A number of evidences favor this idea- first, number of subjects belonging to insomnia group in the present study reported significant improvement in mood after a good night sleep; second, as already discussed, the onset of depressive symptoms follows the onset of insomnia; third, depression has been found to be a function of the severity of insomnia, especially the frequency of awakenings at night and frequency of symptoms; and finally, significant improvement in depressive symptoms has been noted after cognitive behavior therapy directed at insomnia.[16,18,32,33,41] Improvement in depressive symptoms with CBT-I has been reported even in subjects with multiple sclerosis, suggesting that sleep loss is a stressful situation, which may be directly implicated for the presence of depressive symptoms and fatigue.[37,36,42]

Another important finding in this study was comparable fatigue among three groups. Similar findings have been reported earlier as well.[11] Fatigue in insomnia subjects has been found to be related to subjective measures of sleep rather than objective measures.[43] Subjects in all three groups in the present study also had comparable scores on ISI, a subjective measure of sleep quality. Second, fatigue has been found to correlate with hyperarousal as well as health-related quality of life among patients with insomnia.[43,44] As already discussed, hyperarousal is the key factor in the initiation and perpetuation of insomnia that could, in turn, lead to the development of depressive symptoms. Together, all these factors assert that insomnia may be a problem as important as MDD.

Like any other scientific investigation, this study also had some methodological limitations. First, the study sample was small. Second, this was a cross-sectional study. Hence, causation between insomnia and depressive symptoms could not be established. It would have been better to assess the effect of improvement in sleep on depressive symptoms across groups. Third, subjects included in this study were not having any other comorbidity and were free of medications; hence, care should be taken while extrapolating results of this study to patients seen in clinics. Fourth, the insomnia group had subjects with short term as well as chronic insomnia. This could have resulted in reduced scores on PHQ-9 in the present study.

CONCLUSION

The present study suggests that insomnia may be mistaken for depression owing to overlapping features. However, certain features such as sadness, guilt, suicidal ideations, and hypersomnia may differentiate the two conditions. In addition, another differentiating feature is a significant improvement in symptoms after a night of good sleep. Considering the daytime burden of symptoms and effect on the quality of life, insomnia is a major public health burden.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Annexure 1: Fatigue Severity Scale (Hindi version)