SEARCH

SEARCH BY CITATION

Keywords:

  • Beck Depression Inventory;
  • depression;
  • Hamilton Rating Scale for Depression;
  • insomnia;
  • quality of life

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENT
  8. References

Insomnia and major depressive episodes (MDE) have each been associated with quality of life (QOL) deficits. In this study we examined insomnia as an independent predictor of QOL deficits during MDE, and used a secondary analysis of cross-sectional data. The study was based at the inpatient psychiatric ward and included 88 adults (mean age 53; 78% women). We assessed insomnia severity with the 21-item Hamilton Rating Scale for Depression (HRSD) and the Beck Depression Inventory (BDI). Measurements of QOL in the week prior to admission included activities of daily living (ADLs), instrumental ADLs (IADLs), daily living and role functioning, and relation to self and colleagues (the last two both subscales of the Basis 32). Linear regression models used the insomnia items as independent variables and the QOL measures as the dependent variables, after adjusting for age and nonsleep related depression severity. The results showed that 93% of patients endorsed insomnia on the observer-rated HRSD, and 97% endorsed sleep disturbance in the self-rated BDI. However, the insomnia items on the HRSD and BDI showed poor concurrent validity. Increasing severity of insomnia on the HDRS was associated with better QOL, while increasing severity of insomnia on the BDI was associated with worse QOL. We conclude that the BDI and HRSD do not produce equivalent measures of insomnia severity in depressed inpatients, and each insomnia measure has a unique relationship with QOL.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENT
  8. References

Earlier cross-sectional studies have shown an association between insomnia and work-related problems. Civilian insomniacs are more likely than good sleepers to have absenteeism and diminished productivity in the workplace ( Leigh 1991; Schweitzer et al. 1992 ), and military insomniacs are less likely to receive promotions ( Johnson and Spinweber 1983). These economic consequences are but one facet of the more complex concept of quality of life (QOL). Lawton has suggested that QOL has many dimensions, including psychological well-being, quality of the living environment, behavioral competence, spiritual satisfaction, and others ( Lawton 1991). A recent report by Zammit et al. (1999 ) demonstrated that insomnia is associated with QOL deficits in many different spheres of life activity.

A major depressive episode (MDE) is typically accompanied by a complaint of insomnia ( Nelson and Charney 1980), and MDE has been shown in cross-sectional studies to be associated with clinically significant QOL deficits ( Wells et al. 1989 ; Ormel et al. 1994 ; Alexopoulos et al. 1996 ; Dunham and Sager 1994; Iliffe et al. 1993 ; Pyne et al. 1997 ; Ormel et al. 1998 ; McCall et al. 1999 ).

Many methods are available to measure insomnia in depressed patients, but perhaps the most commonly used are the insomnia items imbedded within standard depression rating scales. We compared the performance of the sleep items imbedded within the Hamilton Rating Scale for Depression and the Beck Depression Inventory, and then we examined whether these sleep items were associated with QOL deficits in depressed patients, independent of these patients’ noninsomnia symptoms of depression.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENT
  8. References

Patients

Eighty-eight patients on an inpatient psychiatric unit gave written, informed consent to participate in this project. The project was approved by the investigational review board. This study is a secondary analysis of a data set that is primarily intended to evaluate the longitudinal changes in QOL in depressed patients after psychiatric hospitalization. The mean age of the patients ± SD was 53.4 ± 15.4 (69 women and 19 men). A diagnosis of MDE was confirmed with the Structured Clinical Interview for DSM III-R administered by a Master’s or doctoral level mental health provider ( American Psychiatric Association 1987; Spitzer et al. 1992 ). All patients were free of substance abuse, had no prior or current history of schizophrenia or schizo-affective disorder, and had no more than minimal cognitive deficits. Further description of the demographic and clinical characteristics of this sample can be found elsewhere ( McCall et al. 1999 ).

Psychometric measures

All psychometric assessments of depression severity (including insomnia symptoms) and QOL were completed within three business days of hospital admission. All measurements were intended to reflect the patient’s condition during the week prior to admission. Simultaneous independent measurements were made on a random subset of patients for the purpose of inter-rater reliability.

Depression severity

Depression severity was assessed with the 21-item Hamilton Rating Scale for Depression (HRSD) and the 21-item Beck Depression Inventory (BDI) ( Beck et al. 1998 ; Hamilton 1960). The HRSD is observer-rated and was administered following a semistructured interview ( Williams 1988). The BDI is self-rated by the patient.

Sleep disturbance

Sleep disturbance is recorded on both the HRSD and the BDI. The HRSD separately records the intensity of initial, middle and late insomnia, and each of these items is scored 0, 1, or 2, with a score of 0 indicating no difficulty (Table 1). The maximum score for all the HRSD sleep items was 6, representing 9% of the maximum total score for the HRSD. Sleep disturbance on the BDI is measured in a single item scored 0, 1, 2, or 3 with a score of 0 indicating no difficulty with sleep. The sleep disturbance item represents 5% of the maximum total score. The BDI sleep item is inconsistent in its scaling, allowing a score of ‘1’ for any reported newly developed sleep problem, but allows scores of ‘2’ or ‘3’ only for late insomnia (Table 1).

Table 1.  Hamilton Rating Scale for Depression (HRSD) and Beck Depression Inventory (BDI) sleep items Thumbnail image of
QOL: Competence in self-care

The integrity of activities of daily living (ADLs) is captured in two rating scales: the Physical Self-Maintenance Scale (PSMS) and the Instrumental Activities of Daily Living Scale (IADL). The reliability and validity of these scales has been established by Lawton and Brody (1969). Subjects were asked whether they had performed each of the behaviors during the week prior to admission. The PSMS scale assesses basic skills such as toileting, feeding, bathing, dressing, grooming and ambulation. A perfect score on the PSMS scale is ‘30’, with lower scores indicating a progressive deficit. The IADL scale measures higher ADL functions such as using the telephone, shopping, cooking, housekeeping, laundry, transportation, and responsibility for medications and finances. A perfect score for the IADL is ‘31’, with lower scores indicating a progressive deficit. The PSMS and IADL scales were used by four different raters in this study. One rater recorded verbatim what the patients said about their self-care abilities and then provided these statements to the other raters who then independently scored the scales. Inter-rater reliability for the IADL ranged from 0.97 to 0.99 (P < 0.005) among the four raters. Inter-rater reliability for the PSMS ranged from 0.89 to 0.98 (P < 0.001).

QOL: Role functioning and relationships

We conceptualized QOL as occurring in strata of increasing levels of sophistication. The more primitive behaviors are measured by the PSMS and IADL scales as outlined above. We measured the more complex strata of behavior (i.e. work, recreation, and relationships) with two QOL subscales from the Basis-32: ‘daily living and role functioning’ and ‘relation to self and others’ ( Eisen et al. 1994 ). The Basis-32 has been shown to have good internal consistency among its subscales, to have good test–retest reliability, to have concurrent validity, and to be sensitive to change ( Eisen 1995). The ‘relation to self and others’ subscale and the ‘daily living and role functioning’ subscale are each comprised of 7 items. We used the average of these 7 items as the score. The ‘daily living and role functioning’ subscale measures self-rated difficulties in the subject’s primary role (homemaker, breadwinner, or student), as well as difficulties in leisure activities and in achieving independence/autonomy. The ‘relation to self and others’ subscale measures difficulties in feeling close to others, difficulties in family relationships, and difficulties in relationships outside of the family. A score of ‘4’ indicates extreme difficulty and a ‘0’ indicates no difficulty for each of the Basis-32 subscales.

Statistical methods

Inter-rater reliability of the total HRSD score is evaluated using the intraclass correlation coefficient while reliability of the individual HRSD sleep items is established using Kappa coefficients of agreement for categorical data ( Cohen 1960). Internal consistency of the items comprising the HRSD and BDI is measured using coefficient alpha ( Cronbach 1951). Concurrent validity of these scales is established using the Pearson product moment correlation coefficient. The associations between the individual HRSD and BDI sleep items is assessed using Fisher’s exact test of independence.

Unadjusted means and standard deviations for each of the QOL measures within each category of insomnia are presented. Prior cross-sectional analyses demonstrated that overall depression severity and age are the most robust predictors of QOL in this depressed sample ( McCall et al., 1999 ); therefore we assess the independent contribution of the sleep symptoms to QOL by performing linear regression after adjusting for age and for the corresponding nonsleep items of the HRSD (‘HRSD-nonsleep’) or the BDI (‘BDI-nonsleep’). Initial comparisons between each of the insomnia categories and the reference category (‘no difficulty’) are based on two-sided Z-tests with a type I error rate of 0.05.

A final set of comparisons were made after applying a Bonferroni adjustment for multiple comparisons. We also present the unadjusted relationships between QOL and depression severity from a linear regression analysis of depression severity on each of the QOL measures.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENT
  8. References

Frequency of sleep complaints

The frequency of sleep complaints measured by the HRSD and BDI are presented in Table 2. Ninety-three percent of the patients were rated by the HRSD as having some difficulty (score of 1 or 2) with initial, middle, or late insomnia. Eighty percent were rated as having some difficulty with initial insomnia, 80% with middle insomnia, and 67% with late insomnia. Ninety-seven percent of patients reported some type of sleep disturbance on the BDI.

Table 2.   Cross tabulations of the Hamilton Rating Scale for Depression (HRSD) sleep items with the Beck Depression Inventory (BDI) sleep item. Entries in the table are the number of patients reporting symptoms Thumbnail image of

Inter-rater reliability, internal consistency, and concurrent validity of sleep measurements

Inter-rater reliability for the HRSD was assessed on a subsample of 27 subjects. Our research assistants demonstrated high inter-rater reliability for the total HRSD score (ICC=0.98, n=27, P < 0.001). The initial, middle and late insomnia items were also reliability scored (kappas equal 0.78, 0.88, 0.94, n=27, all P < 0.001, respectively). The mean total HRSD score for the entire sample (n=88) was 28.2 (SD=6.5) while the mean total BDI score was 33.5 (SD=10.6). The internal consistency of the total HRSD and BDI scores was established by Cronbach alphas of 0.59 and 0.82, respectively.

Concurrent validity for the BDI was established by its moderately high correlation with the HRSD (r=0.60, n=88, P < 0.001). Similarly, the BDI-nonsleep score was moderately correlated with the HRSD-nonsleep score (r=0.62, n=88, P < 0.001).

The cross tabulations of the HRSD sleep items with the BDI sleep item are presented in Table 2. For analyses, categories 0 and 1 for the BDI sleep item were combined into a single category because of the small number of individuals (n=3) reporting no sleep problems. Therefore, individuals reporting no sleep problems or reporting not sleeping as well as usual are combined into a single category. Fisher’s exact tests of independence indicate no statistically significant associations between the BDI and HRSD sleep items (P=0.727, 0.490, 0.268 for initial, middle and late insomnia, respectively).

Prediction of QOL from depression severity

Greater depression severity as measured by the BDI and HRSD was associated with poorer functioning on the PSMS (P=0.08 and 0.21, respectively) and IADL (P=0.16 and 0.22, respectively), but these results were not statistically significant. However, higher levels of depression were significantly associated with worse daily living and role functioning (P < 0.001) and relationships with self and colleagues (P < 0.001).

Prediction of QOL from the sleep items

After adjusting for the corresponding nonsleep items and age, subjects reporting waking during the night (middle insomnia category 2) and subjects unable to fall asleep again if they get out of bed (late insomnia category 2) had better IADL functioning compared to those with no difficulty (P=0.02 and 0.04, respectively) (Tables 3 and 4. Initial insomnia (categories 1 and 2) and waking in the early hours of the morning (late insomnia category 1) were both significantly associated with better daily living and role functioning (P=0.03, 0.05, 0.01, respectively). However, sleep disturbance as measured by the BDI was significantly associated with poorer daily living and role functioning (P=0.02 and 0.0002 for categories 2 and 3 compared to the reference, respectively). After a Bonferroni adjustment for multiple comparisons the only result retaining statistical significance was the difference in daily living and role functioning between those reporting waking several hours earlier than usual and unable to get back to sleep on the BDI (category 3) and those with little or no sleep difficulty (category 0,1).

Table 3.  Unadjusted Mean (Standard Deviation) QOL Scores at Each Level of Insomnia Severity Thumbnail image of
Table 4.  Linear regression beta coefficients (standard error) adjusted for nonsleep items and age Thumbnail image of

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENT
  8. References

This sample of severely depressed psychiatric inpatients had high frequencies of insomnia complaints. Our finding of a 93% frequency of insomnia complaints in this sample of depressed inpatients is similar to Nelson and Charney’s report (1980) that 85% of their sample of depressed inpatients had insomnia. The level of depressive symptoms were reliably rated for total symptoms, sleep symptoms, and nonsleep symptoms in the HRSD. The overall HRSD and BDI scales were moderately correlated at a magnitude consistent with other reports. However, the HRSD sleep items and BDI sleep item were not significantly associated.

Poor correlation between various measures of sleep disturbance is a frequent finding in sleep research. Polysomnographic (PSG) measures may markedly differ from subjective reports of sleep in insomniacs ( McCall and Edinger 1992; Carskadon et al. 1976 ; Frankel et al. 1976 ). Subjective improvement on the HRSD sleep items during treatment of depressed insomniacs may be accompanied by deterioration of PSG measures of sleep continuity ( Gillin et al. 1997 ). Still, we had expected to find good correlation between the subjective sleep assessment in the BDI and the observer-rated sleep items of the HRSD. Our failure to find the expected correlation may reflect the inconsistent scaling of the BDI sleep item, or the inherent differences between self-ratings and observer ratings.

Our findings of good inter-rater reliability, internal consistency, and concurrent validity make it unlikely that failure to find an association between the HRSD and BDI sleep items is related to problems with our rating methods or the internal psychometrics of the scales. The poor association between the HRSD and BDI sleep items suggests they provide different sorts of information about sleep and justify separate analyses for predicting QOL.

The finding of better IADL and ‘daily living and role functioning’ scores with increasing levels of HRSD insomnia defies intuition and any easy explanation. However, these associations were no longer significant when allowing for multiple comparisons with the Bonferroni adjustment. In contrast, the finding of poorer functioning on ‘daily living and role functioning’ with increasing severity on the BDI sleep item is in the expected direction and is the only association that remained statistically significant when allowing for multiple comparisons. Our finding of worsening QOL in depressed patients with increasing levels of insomnia is consistent with other studies showing that insomnia is a predictor of other poor outcomes in MDE such as suicide ( Fawcett et al. 1990 ). Our findings provide support for the practice of specifically targeting insomnia symptoms during the treatment of MDE ( Nierenberg et al. 1994 ).

The most important finding of this investigation is that non-PSG measures of insomnia severity are not interchangeable in depressed insomniacs, yielding poor concurrent validity and unpredictable associations with QOL measures. Our findings are limited by our observation of only inpatients. Future investigations of QOL and insomnia may require multiple measurements of sleep complaints such as sleep logs and global reports, perhaps in addition to an objective measure such as PSG or actigraphy. Also, similar studies in depressed outpatients would be desirable.

ACKNOWLEDGEMENT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENT
  8. References

Supported by NIMH awards MH51552 and MH01090.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENT
  8. References
  • 1
    Alexopoulos, G. S., Vrontou, C., Kakuma, T., Meyers, B. S., Young, R. C., Klausner, E., Clarkin, J. Disability in Geriatric Depression. Am. J. Psychiatry, 1996, 153: 877 885.
  • 2
    American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. American Psychiatric Association, Washington, DC, 1987 (third edition, revised).
  • 3
    Beck, A. T., Steer, R. A., Garbin, M. G. Psychometric properties of the Beck Depression Inventory: twenty-five years of evaluation. Clin. Psychol Review, 1998, 8: 77 100.
  • 4
    Carskadon, M. A., Dement, W. C., Mitler, M. M., Guilleminault, C., Zarcone, V. P., Spiegel, R. Self-report versus sleep laboratory findings in 122 drug-free subjects with complaints of chronic insomnia. Am. J. Psychiatry, 1976, 133: 1382 1388.
  • 5
    Cohen, J. A coefficient of agreement for nominal scales. Educ Psychol Meas., 1960, 20: 37 46.
  • 6
    Cronbach, L. J. Coefficient alpha and the internal structure of tests. Psychometrika, 1951, 16: 297 334.
  • 7
    Dunham, N. C. & Sager, M. A. Functional status, symptoms of depression, and the outcomes of hospitalization in community-dwelling elderly patients. Arch. Family Med., 1994, 3: 676 681.
  • 8
    Eisen, S. V., Dill, D. L., Grob, M. C. Reliability and validity of a brief patient-report instrument for psychiatric outcome evaluation. Hosp Comm Psychiatry, 1994, 45: 242 247.
  • 9
    Eisen, S. V. Assessment of subjective distress by patients’ self-report versus structured interview. Psychol Reports, 1995, 76: 35 39.
  • 10
    Fawcett, J., Scheftner, W., Fogg, L. et al. Time Related Predictors of Suicide. In Major Affective Disorder. Am. J. Psychiatry, 1990, 147: 1189 1194.
  • 11
    Frankel, B. L., Coursey, R. D., Buchbinder, R., Snyder, F. Recorded and reported sleep in chronic primary insomnia. Arch. Gen Psychiatry, 1976, 33: 615 623.
  • 12
    Gillin, J. C., Rapaport, M., Erman, M. K., Winokur, A., Albala, B. J. A comparison of nefazodone and fluoxetine on mood and on objective, subjective, and clinician-rated measures of sleep in depressed patients: a double-blind, 8-week clinical trial. J. Clin. Psychiatry, 1997, 58: 185 192.
  • 13
    Hamilton, M. A rating scale for depression. J. Neurol. Neurosurg Psychiatry, 1960, 23: 56 62.
  • 14
    Iliffe, S., Tai, S. S., Booroff, A., Goldenberg, E., Morgan, P., Gallivan, S. Assessment of elderly people in general practice. 4. depression, functional ability and contact with services. Br. J. Gen Prac., 1993, 43: 371 374.
  • 15
    Johnson, L. C. & Spinweber, C. L. Quality of sleep and performance in the Navy: a longitudinal study of good and poor sleepers. In: C. Guilleminault, E. Lugaresi (Eds) Sleep/Wake Disorders: Natural History, Epidemiology, and Long-Term Evolution. Raven Press, New York 1983: 13–28.
  • 16
    Lawton, M. P. A multidimensional view of quality of life in frail elders. In: J. E. Birren, J. E. Lubben, J. C. Rowe, D. E. Deutchman (Eds) The Concept and Measurement of Quality of Life in the Frail Elderly. Academic Press, New York, 1991.
  • 17
    Lawton, M. P. & Brody, E. M. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist, 1969, 9: 179 186.
  • 18
    Leigh, P. Employee and job attributes as predictors of absenteeism in a national sample of workers: the importance of health and dangerous working conditions. Soc. Sci. Med., 1991, 33: 127 137.
  • 19
    McCall, W. V. & Edinger, J. D. Subjective total insomnia: an example of sleep state misperception. Sleep, 1992, 15: 71 73.
  • 20
    McCall, W. V., Cohen, W., Reboussin, B., Lawton, P. Effects of mood and age on quality of life in depressed inpatients. J. Affective Disorders, 1999, 55: 107 114.
  • 21
    Nelson, J. C. & Charney, D. S. Primary affective disorder criteria and the endogenous-reactive distinction. Arch. Gen Psychiatry, 1980, 37: 787 793.
  • 22
    Nierenberg, A. A., Adler, L. A., Peselow, E. et al. Trazodone for antidepressant-associated insomnia. Am. J. Psychiatry, 1994, 151: 1069 1072.
  • 23
    Ormel, J., Von Korff, M., Ustun, T. B., Pini, S., Korten, A., Oldehinkel, T. Common mental disorders and disability across cultures. JAMA, 1994, 272: 1741 1748.
  • 24
    Ormel, J., Kempen, G. I., Brilman, E. I., Van Sonderen, E., Relyveld, J. Functioning, well-being, and health perception in late middle-aged and older people: comparing the effects of depressive symptoms and chronic medical conditions. J. Am. Geriatrics Soc., 1998, 46: 39 48.
  • 25
    Pyne, J. M., Patterson, T. L., Kaplan, R. M., Ho, S., Gillin, J. C., Golshan, S., Grant, I. Preliminary longitudinal assessment of quality of life in patients with major depression. Psychopharmacol. Bull., 1997, 33: 23 29.
  • 26
    Schweitzer, P. K., Engelhardt, C. L., Hillaker, N. A. et al. Consequences of poor reported sleep. Sleep Res., 1992, 21: 22 (Abstract).
  • 27
    Spitzer, R. L., Williams, J. B. W., Gibbon, M., First, M. B. The Structured Clinical Interview for DSM-III-R, I. History, Rationale, Description. Arch. Gen Psychiatry, 1992, 49: 624 629.
  • 28
    Wells, K. B., Stewart, A., Hays, R. D., Burnam, M. A., Rogers, W., Daniels, M., Berry, S., Greenfield, S., Ware, J. The functioning and well-being of depressed patients. JAMA, 1989, 262: 914 919.
  • 29
    Williams, J. B. A structured interview guide for the Hamilton Depression Rating Scale. Arch. Gen Psychiatry, 1988, 45: 742 747.
  • 30
    Zammit, G. K., Weiner, J., Damato, N., Sillup, G. P., McMillan, C. A. Quality of life in people with insomnia. Sleep, 1999, 22 (Suppl.): S379 S385.