Pergamon

0005-7916(94)00073-5

J. Behav. Ther. &Exp. Psychiat. Vol. 26. No. I. pp. 17-23, 1995. Elsevier Science Ltd

Printed in Great Britain

P R E D I C T I O N O F O U T C O M E IN B E H A V I O R A L L Y B A S E D I N S O M N I A

T R E A T M E N T S

DONALD L. BLIWISE

Sleep Disorders Center, Stanford University Medical School

LEAH FRIEDMAN

Department of Psychiatry and Behavioral Sciences, Stanford University Medical School

JAMIE C. NEKICH

Sleep Disorders Center, Stanford University Medical School

JEROME A. YESAVAGE

Department of Psychiatry and Behavioral Sciences, Stanford University Medical School

Summary - - Factors related to successful behavioral intervention for individuals with insomnia are not well understood. In this study we employed the Neuroticism, Extraversion and Openness (NEO) Personality Inventory, to predict successful treatment outcome. Two behavioral treatments for insomnia, sleep restriction therapy (SRT) and relaxation training (RT) were employed in 32 elderly insomniacs. Following two baseline weeks, subjects underwent four weeks of individual treatment. Daily telephone call-ins generated data on sleep times and sleep latency. Follow-up occurred three months after the end of treatment. Results indicated that subjects showing the greatest improvement in total sleep time with both treatments were more traditional, conventional and rigid. Improvement in sleep onset latency was unrelated to NEO Scores. SRT appeared to be more effective for increasing total sleep time in these older subjects.

Poorly understood in behavioral treatment for insomnia are factors re la ted to successful treatment. Lacks and Powlishta (1989) reported that insomniacs with less overall psychopathology as assessed by the MMPI showed greater benefits from behavioral treatments. Since the MMPI validity scales were designed as measures of test- taking attitudes and not personality characteristics (Graham, 1977) their use in the latter context may be suspect. In this s tudy we emp loyed the Neuroticism, Extraversion and Openness (NEO)

Personality Inventory (Costa & McCrae, 1985) to determine whether personality characteristics may be associa ted with improvemen t in several behav iora l ly based insomnia t rea tments . A preliminary report of the study has appeared elsewhere and was based on the initial 22 cases (Friedman, Bliwise, Yesavage & Salom, 1991). The report expands upon this prior work by presenting NEO data on all five domains and analyses prediction of outcome separately by treatment group.

Requests for reprints should be addressed to Donald L. Bliwise, Sleep Disorders Center, Department of Neurology, Emory University School of Medicine, P.O. Drawer V, Suite 6000-WMB, Atlanta, GA30322, U.S.A.

17

18 DONALD L. BLIWISE et al.

Methods

Subjects

Subjects were 32 elderly (M age 68.7, SD =

9.7) men (N = 10) and women (N = 22) who received either sleep restriction therapy (SRT; N =16) or relaxation therapy (RT; N = 16). The 32 subjects were derived from a multi-stage screening process consisting of initial recruitment from advertisements in senior citizen organizations and f rom presenta t ions in independen t l iving residences for seniors. These recruitment efforts yielded 192 prospective subjects who completed a sleep ques t ionnai re . Results of this survey, particularly as they relate to the issue of defining oneself as having "insomnia" or merely obtaining "insufficient sleep" have been reported previously (Friedman, Bliwise, Tanke, Salom & Yesavage, 1992). From the 192 individuals, we successfully recruited 32 individuals for participation in the treatment protocol. Admittance into the study was cont ingent upon absence of serious medical diseases including cancer or pr ior s troke, excessive snoring or other symptoms suggestive of sleep apnea, regular use of medications possibly interfering with sleep such as theophylline, anti- histamines, or beta-blockers, or heavy use of a lcohol (greater than two drinks per day). Additionally, we eliminated individuals likely to be clinically depressed as evidenced by history of psychiatric hospitalization, history of suicidal behavior or a Geriatric Depression Scale score (Yesavage et al., 1983) of higher than 9. All subjects had Mini-Mental State Exam (Folstein,

Folstein, & McHugh, 1975) scores of 27 or higher except for one subject in SRT with a score of 23 which was thought to reflect a mild depression. A comparison of the 32 subjects with the remaining 160 individuals not included in the study is shown in Table 1 and suggests that the admitted subjects had more severely disturbed sleep.

Procedures

All subjects completed the NEO Personality Inventory (Costa & McCrae, 1985) prior to entry into the protocol . There were no significant differences between SRT and RT groups on any of the f ive domains of the NEO (neurot ic ism, ex t ravers ion , openness , agreeableness , conscientiousness). Age and gender composition were also comparable and did not significantly distinguish the groups.

Both SRT and RT involved once a week therapy sessions over four weeks and a final wrap- up session at the end of treatment. SRT consisted of the customary package of limiting time in bed, avoiding naps and maintaining set wake-up times as described by Spielman, Saskin, and Thorpy (1987). Time in bed was extended whenever the floating mean sleep efficiency over the preceding five days equalled or exceeded 85%. To maintain the cooperation of these elderly subjects some flexibil i ty in SRT procedures was employed, predominantly in initial assignment of allowable time in bed. For example, if initial time in bed assignment based on the 85% requirement resulted in an assigned duration which the subject found unacceptable (e.g., in some cases less than five

Table 1

Comparison of Study Subjects with Prospective Subject Pool

Variable

Prospective subjects Study subjects not in study

M (SD) M (SD) (N = 32) (N = 160) t p

Nights per week with sleep latency problem Nights per week with sleep maintenance problem Total sleep obtained at night (mins) Time to fall asleep (mins) Number of awakenings

3.9 (2.7) 3.0 (2.6) 5.5 (2.0) 4.7 (2.4)

315.0 (61.4) 350.6 (90.6) 55.1 (62.9) 44.4 (49.4)

2.5 (1.3) 2.5 (1.3)

1.80 1.68 2.12 1.07 .49

.07

.10

.03

.28

.62

Prediction of Outcome in Insomnia 19

hours), we selected the lowest possible time which subjects would tolerate. Such modifications of traditional SRT appeared necessary to maintain subjects' compliance. RT consisted of a standard approach of alternating tension and relaxation of various muscle groups except that muscle groups in the neck and back were excluded, as in our exper ience these frequently lead to spasm in elderly subjects (Yesavage & Jacob, 1984).

Daily record keeping was accomplished in both groups by calls placed twice daily (morning and evening) to a telephone answering machine to report bed time, wake-up times, sleep latencies, and time awake during the night Baseline data were derived from 14 days of daily call-ins prior to start of t rea tment . T rea tmen t per iod data consisted of the final 14 days of treatment (weeks 3 and 4). Missing data were infrequent but, when they occurred, Baseline and Treatment period data were computed on the basis of the number of complete days of data (at least 7). Three months fo l lowing the last week of act ive t reatment , subjects once again called in to the machine for two weeks. Two-week means for sleep onset latencies (SOLs) and total sleep times (TSTs) over these three periods (baseline, treatment, follow-up) constituted the dependent variables in this study. Unlike our previous report, we elected to focus only on these two variables in these analyses. Previously we had also presented data on variables such as sleep efficiency, wake after sleep onset, and time in bed after awakening. Our rationale for not presently including these variables was that they all are subject to artifactual manipulation by

the treatment instructions, particularly SRT. For example, wake after sleep onset can be reduced and sleep efficiency increased merely by having subjects spend less time in bed. Treatment-related changes in sleep onset latency and total sleep time, however, might be expected to be less subject to this type of instructional bias.

Data were analyzed in repeated measures ana lyses of va r iance (ANOVAs) p e r f o r m e d sepa ra te ly for SOL and TST. Predic t ion of improvement employed partial correlations in which baseline TST (or SOL) was controlled and then each NEO scale was correlated with follow- up TST (or SOL).

Results

Treatment Effects

Table 2 shows means and standard deviations of TSTs and SOLs (in minutes) for the two t rea tment groups for basel ine, end of act ive treatment, and 3-month follow-up. There were no group differences in TST or SOL at Baseline. A N O V A s indica ted that TST at the end of treatment did not differ significantly from baseline IF(l,30) = .79, NS] however at 3-month follow-up there were both significant effects of time [F( 1,30) = 53.74, p < .001] and an interaction of treatment group by time [F(1,30) = 6.93, p < .02] indicating that the SRT group showed differentially greater improvement relative to the RT group. ANOVAs for SOL showed significant time effects, both at

Table 2

Summary of Treatment Effects

M(SD) M(SD) M(SD) Baseline Treatment Follow-up

Total sleep time (minutes) SRT (N = 16) 339.8 (63.8) 352.5 (43.9) 400.4 (45.9) RT (N = 16) 335.4 (52.1) 337.0 (79.4) 364.0 (59.6)

Sleep onset latency (minutes) SRT (N= 16) 44.3 (38.8) 21.4 (14.5) 17.5 (10.9) RT (N= 16) 47.4 (34.4) 31.5 (29.5) 26.2 (21.2)

Note: F-values are shown in text.

20 DONALD L. BLIWISE et al.

the end of treatment [F(1,30) = 11.33, p < .005] and at the end of follow-up IF(l,30) = 20.63, p < .001], but without any treatment group by time interaction.

In order to allow some appreciation of the differential extent of individuals' improvement under treatment, Baseline to 3-month Follow-up changes in TST were categorised by 30 minute increments. For individuals in SRT, the mean and median gains in TST were 58.2 (SD = 39.7) and 47.0 minutes; 13 individuals showed TST improvement of over 30 minutes and six showed TST improvement in excess of 60 minutes. In contrast, RT mean and median gains in TST were 28.6 minutes (SD = 29.2) and 24.5 minutes; six individuals showed TST improvement of over 30 minutes and three showed TST improvement in excess of 60 minutes. Under SRT all subjects reported some improvement in TST at 3-months fol low-up, however , under RT two subjects actually slept less than at baseline.

Prediction of Outcome at Three-Month Follow-Up

To investigate improvement more fully under these two behavioral treatments, we performed partial correlations (Table 3) predicting TST and SOL at 3-month fol low-up. Sleep latency at

Table 3

Partial Correlations Between NEO Scales and Reported Sleep at Follow-up

NEO scale N E O A C

Totalsleeptime at follow-up SRT -.13 -.48* -.52* .06 .02 RT -.27 -.28 -.70t -.31 -.37

Sleep onset latency at follow-up SRT .06 .15 .30 -.23 -.33 RT -.42 .35 -.10 .15 .10

*p < .05; tp < .005. Correlations partialed for baseline total sleep time and baseline sleep onset latency, respectively. Abbrevia t ions are as fol lows: N = neurot ic ism; E = extravers ion; O = openness ; A = agreeableness ; C = conscientiousness.

follow-up was unrelated to any NEO scale for SRT or RT, however a different situation emerged for total sleep time. For both SRT and RT higher initial scores on the openness scale of the NEO were associated with relatively less improvement in total sleep time. Additionally, for SRT subjects, tendencies towards introversion were associated with greater reported improvement in TST.

Discussion

These results imply that, although reported treatment efficacy appeared somewhat different across the two treatments employed here (slightly favor ing SRT), persona l i ty charac ter i s t ics predicting improvement were similar for these two behavioral interventions. Specifically, elderly poor sleepers who could be described as conventional, traditional, rigid and unimaginative, reported greater success with the directive and structured treatments used here, at least insofar as total sleep time was concerned. It remains unclear why sleep onset latency improvement was unpredicted by any personality dimensions. The extent to which this total sleep time result reflects the age of some other demographic characteristic of this sample is unclear at this point, as is whether any successful behavioral treatment for insomnia might be related to such personality characteristics. Few papers have appeared which report variables predicting successful outcomes in insomnia t reatment , although in recent years a shift towards "tailoring" insomnia treatments for particular patients has emerged (Espie, Brooks & Lindsay, 1989; Hauri, 1991; Lacks & Powlishta, 1989; Morin, 1993; Morin & Kwentus, 1988; Thoresen, Coates, Kirmil-Gray & Rosekind, 1981), as has the use of treatment packages involving combined elements of sleep restriction, stimulus control, relaxation, and lifestyle manipulation (Edinger, Hoelscher, Marsh, L ipper & Ionescu-Pioggia , 1992; Hoelscher & Edinger, 1988; Lacks, 1987; Morin, Kowatch, Barry & Walton, 1993). Previous studies with the MMPI suggesting that lower F and higher L and K validity scale configurations predicted better response to behavioral treatments

Prediction of Outcome in Insomnia

for insomnia (Edinger, Stout & Hoelscher, 1988; Shealy, Lowe & Ritzler, 1980) have probably presaged the present results, though the inference of such personal i ty character is t ics is, in all likelihood, better grounded in the NEO.

Several other aspects of these results deserve further note. The evidence for the differential treatment efficacy of SRT emerged at the 3-month follow-up rather than at the end of the active t reatment . Al though this may appear counterintuit ive, evidence suggesting greater ef f icacy subsequent to the end of behavioral treatments for insomnia is not unusual (Lacks, Bertelson, Sugerman & Kunkel, 1983; McCluskey, Milby, Switzer , Wil l iams & Wooten, 1991; Sanavio, 1988; Thoresen et al., 1981; Woolfolk & McNulty, 1983) even, in some cases, on follow- ups of one year or more (Espie, Lindsay, Brooks, Hood & Turvey, 1989; Lacks & Powlishta, 1989). It may well be that the four weeks of treatment essentially served as a supervised training exercise after which subjects then elected to implement SRT or RT on their own.

An acknowledged limitation of this study is our reliance on self-report as an outcome. Although data suggest that self-reports of an individual's sleep are somewhat related to polysomnography (Coates et al., 1982; Hoch et al., 1987a) many studies have suggested a far less than perfect co r re spondence be tween the two cr i ter ia (Carskadon et al., 1976; Freedman & Papsdorf, 1976; Monroe, 1967). Some investigators have adopted behaviora l rat ings made by other observers (Morin & Azrin, 1987) when possible to supplement self-reports. We did not employ such measures in our study.

The present results suggest that SRT is probably a more successful treatment for geriatric insomnia, as measured phenomenologically, than RT. Why this should be the case is not altogether clear. One possible explanation is that since SRT borrows heavily from stimulus control treatment tor insomnia (avoidance of napping, fixed wake- up times) and because stimulus control has long been considered the most effective behavioral treatment for insomnia (Borkovec, 1982; Lichstein & Fischer, 1985), SRT works for these reasons. In

21

fact, not all authors concur that stimulus control p rocedures work best. Thus, Lacks and Powlishta's (1989) incisive re-analysis of the data f rom insomniacs in their program suggested equivalent benefits of all treatments at three- month follow-up. Furthermore, there is good reason to believe that the mechanisms involved in the success of SRT in this study may bear only rough similarity to the mechanisms involved in stimulus control. First, stimulus control typically involves some procedures (e.g., leave bed if unable to sleep) (Bootzin & Nicassio, 1978) which are not a component of SRT (Spielman et al., 1987). In our study this instruction was not a componen t of e i ther SRT or RT, and some researchers have even demonstrated equivalent success in both young adults (Zwart & Lisman, 1979) and the elderly (Davies, Lacks, Storandt & Bertelson, 1986) for countercontrol procedures in which subjects actually stay in bed if unable to sleep but merely avoid behaviors incompatible with sleep. Second, it is conceivable that the sleep deprivation developed during SRT (the predictable consequence of a constantly late bedtime, an early wake-up time, and avoidance of naps) may be the powerful driving force towards longer and better sleep in the elderly. Several experimental studies of sleep deprivation in elderly persons recorded in the sleep lab (Bonnet & Rosa, 1987; Carskadon & Dement, 1985; Reynolds et al., 1987) suggest improved sleep length and quality following deprivation. Similarly, more ecologically based studies of: (a) elderly nuns leading a highly rout in ized life style with f ixed s l eep -wake schedules (Hoch et al., 1987b), and (b) inpatient wards including geriatric patients who are placed on a moderately restrictive time in bed protocol (Edinger, Lipper & Wheeler, 1989) show parallel results. Hence SRT may work simply because it efficiently induces sleep loss, although sleepiness induced by such an intervention (e.g., lapses while driving) must also be considered (Mitler et al., 1988).

A c k n o w l e d g e m e n t s - - Support by AG 06066, AG 10643, MH 45143, MH 35182 and the Medical Research Services of the Department of Veterans Affairs.

22 DONALD L. BLIWlSE et al.

References

Bonnet, M. H., & Rosa, R. R. (1987). Sleep and performance in young adults and older normals and insomniacs during acute sleep loss and recovery. Biological Psychology, 25, 153-172.

Bootzin, R. R., & Nicassio, P. M. (1978). Behavioral treatments for insomnia. In M. Hersen, R. M. Eisler, & P. M. Miller (Eds.), Progress in Behavior Modification (Vol. 6, pp. 1-45). New York: Academic Press.

Borkovec, T. D. (1982). Insomnia. Journal of Consulting and Clinical Psychology, 50, 880-895.

Carskadon, M. A., & Dement, W. C. (1985). Sleep loss in elderly volunteers. Sleep, 8, 207-221.

Carskadon, M. A., Dement, W. C., Mitler, M. M., Guilleminault, C., Zarcone, V. P., & Spiegel, R. (1976). Self-reports versus sleep laboratory findings in 122 drug- free subjects with complaints of insomnia. American Journal of Psychiatry, 133, 1382-1388.

Coates, T. J., Killen, J. D., George, J., Marchini, E., Silverman, S., & Thoresen, C. (1982). Estimating sleep parameters: a multi-trait-multimethod analysis. Journal of Consulting and Clinical Psychology, 50, 345-352.

Costa, P. T., & McCrae, R, R. (1985). The NEO Personality Inventory Manual. Odessa, Florida: Psychological Assessment Resources.

Davies, R., Lacks, P., Storandt, M., & Bertelson, A. D. (1986). Countercontrol treatment of sleep-maintenance insomnia in relation to age. Psychology andAging, 1, 233-238.

Edinger, J. D., Hoelscher, T. J., Marsh, G. R., Ionescou- Pioggia, M., & Lipper, S. (1992). A cognitive-behavioral therapy for sleep-maintenance insomnia in older adults. Psychology and Aging, 7, 282-289.

Edinger, J. D., Lipper, S., & Wheeler, B. (1989). Hospital ward policy and patient's sleep patterns: a multiple baseline study. Rehabilitation Psychology, 34, 43-50.

Edinger, J. D., Stout, A. L., & Hoelscher, T. J. (1988). Cluster analysis of insomniacs' MMPI profiles: relation of subtypes to sleep history and treatment outcome. Psychosomatic Medicine, 50, 77-87.

Espie, C. A., Brooks, D. N., & Lindsay, W. R. (1989). An evaluation of tailored psychological treatment of insomnia. Journal of Behavior Therapy and Experimental Psychiatry, 20, 143-153.

Espie, C. A., Lindsay, W. R., Brooks, D. N., Hood, E. M., & Turvey, T. (1989). A controlled comparative investigation of psychological treatments fqr chronic-sleep onset insomnia. Behavior Research and Therapy, 27, 79-88.

Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). "Mini-mental state": a practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189-198.

Freedman, R., & Papsdorf, J. D. (1976). Biofeedback and progressive relaxation treatment of sleep-onset insomnia: a controlled, all-night investigation. Biofeedback and Self- Regulation, 1, 253-271.

Friedman, L., Bliwise, D. L., Yesavage, J. A., & Salom, S. R. (1991). A preliminary study comparing sleep restriction and relaxation treatments for insomnia in older adults. Journal of Gerontology: Psychological Sciences, 46, P1-8.

Friedman, L. F., Bliwise, D. L., Tanke, E. D., Salom, S. R., & Yesavage, J. A. (1992). A survey of self-reported poor sleep

and associated factors in older individuals. Behavior, Health and Aging, 2, 13-20.

Graham, J. R. (1977). The MMPI: a practical guide. New York: Oxford University Press.

Hauri, P. J. (1991). Case studies in insomnia. New York: Plenum Medical Book Company.

Hoch, C. C., Reynolds III, C. F., Kupfer, D. J., Berman, S. R., Houck, P. R., & Stack, J. A. (1987a). Empirical note: self- report versus recorded sleep in healthy seniors. Psychophysiology, 24, 293-299.

Hoch, C. C., Reynolds III, C. F., Kupfer, D. J., Houck, P. R., Berman, S. R., & Stack, J. A. (1987b). The superior sleep of healthy elderly nuns. International Journal of Aging and Human Development, 25, 1-9.

Hoelscher, T. J., & Edinger, J. D. (1988). Treatment of sleep- maintenance insomnia in older adults: sleep period reduction, sleep education, and modified stimulus control. Psychology and Aging, 3, 258-263.

Lacks, P. (1987). Behavioral treatment for persistent insomnia. New York: Pergamon Press.

Lacks, P., & Powlishta, K. (1989). Improvement following behavioral treatment of insomnia: Clinical significance, long-term maintenance, and predictors of outcome. Behavior Therapy, 20, 117-134.

Lacks, P., Bertelson, A. D., Sugerman, J., & Kunkel, J. (1983). The treatment of sleep-maintenance insomnia with stimulus control techniques. Behavior Research and Therapy, 21, 291-295.

Lichstein, K. L., & Fischer, S. M. (1985). Insomnia. In M. Hersen & A. S. Bellack (Eds.), Handbook of clinical behavior therapy with adults (pp. 319-352). New York: Plenum Press.

McCluskey, H. Y., Milby, J. B., Switzer, P. K., Williams, V., & Wooten, V. (1991). Efficacy of behavioral versus triazolam treatment in persistent sleep-onset insomnia. American Journal of Psychiatry, 148, 121-126.

Mitler, M. M., Carskadon, M. A., Czeisler, C. A., Dement, W. C., Dinges, D. F., & Graeber, R. C. (1988). Catastrophes, sleep and public policy: consensus report. Sleep, 11, 100-109.

Monroe, L. J. (1967). Psychological and physiological differences between good and poor sleepers. Journal of Abnormal Psychology, 72, 255-264.

Morin, C. M. (1993). Insomnia: psychological assessment and management. New York: Guilford Press.

Morin, C. M., & Azrin, N. H. (1987). Behavioral and cognitive treatments of geriatric insomnia. Journal of Consulting and Clinical Psychology, 56, 748-753.

Morin, C. M., Kowatch, R. A., Barry, T., & Walton, E. (1993). Cognitive-behavior therapy for late-life insomnia. Journal of Consulting and Clinical Psychology, 61,137-146.

Morin, C. M., Kwentus, J. A. (1988). Behavioral and pharmacological treatment for insomnia. Annals of Behavioral Medicine, 10, 91-100.

Reynolds III, C. F., Kupfer, D. J., Hoch, C. C., Houck, P. R., Stack, J. A., Berman, S. R., Campbell, P. I., & Zimmer, B. (1987). Sleep deprivation as a probe in the elderly. Archives of General Psychiatry, 44, 982-990.

Sanavio, E. (1988). Pre-sleep cognitive intrusions and treatments of onset-insomnia. Behavior Research and Therapy, 26, 451-459.

Shealy, R. C., Lowe, J. D., & Ritzler, B. A. (1980). Sleep onset

Prediction of Outcome inlnsomnia 23

insomnia: personality characteristics and treatment outcome. Journal of Consulting and Clinical Psychology, 48, 659-661.

Spielman, A. J., Saskin, P., & Thorpy, M. J. (1987). Treatment of chronic insomnia by restriction of time in bed. Sleep, IO, 45-56.

Thoresen, C. E., Coates, T. J., Kirmil-Gray, & Rosekind, M. R. (1981). Behavioral self-management in treating sleep- maintenance insomnia. Journal of Behavioral Medicine, 4, 41-52.

Woolfolk, R. L., & McNulty, T. F. (1983). Relaxation treatment for insomnia: a component analysis. Journal of Consulting and Clinical Psychology, 51,495-503.

Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, O., Huang, V., Adey, M., & Leirer, V. O. (1983). Development and validation of a geriatric depression screening scale: a preliminary report. Journal of Psychiatric Research, 17, 3749.

Yesavage, J., & Jacobs, R. (1984). Effects of relaxation and mnemonics on memory, attention, and anxiety in the elderly. Experimental Aging Research, 10, 211-214.

Zwart, C. A., & Lisman, S. A. (1979). Analysis of stimulus control treatment of sleep-onset insomnia. Journal of Consulting and Clinical Psychology, 47, 113-118.