Effects of Telephone-Delivered CBT-I on Sleep: Do Outcomes Differ by - - PowerPoint PPT Presentation

Effects of Telephone-Delivered CBT-I on Sleep: Do Outcomes Differ by Baseline Demographic, VMS, or Mood Symptoms?

MsFLASH-04 study supported by the Fred Hutchinson Cancer Research Center, Seattle, WA (1U01 AG032699)

S.M. McCurry1, K.A. Guthrie2, J.C. Larson2, C.M. Morin3, N.F. Woods1, C.A. Landis1, L.S. Cohen4, K.E. Ensrud5, H. Joffe4, K.M. Newton6, J.L. Otte7, S.D. Reed1, B. Sternfeld8, L. Tinker2, and A.Z. LaCroix9

1University of Washington, Seattle WA 2Fred Hutchinson Cancer Research Center, Seattle WA 3Universite Laval, Quebec, QC, Canada 4Harvard Medical School, Boston MA 5University of Minnesota, Minneapolis, MN 6Group Health Research Institute, Seattle, WA 7Indiana University, Indianapolis, IN 8Northern California Kaiser Permanente 9University of California, San Diego, La Jolla, CA

• All women go through menopause; a majority also experience

insomnia symptoms during this time of life

• Menopause-related sleep disturbance places substantial economic

burden on women and society

• Sleep problems are a leading reason for visits to health care

professionals during menopause

• Effective, cost-efficient, non-pharmacological treatments are needed

that can be integrated into primary care

Significance

Background: MsFLASH 04 Trial

We have shown that telephone-delivered Cognitive Behavior Therapy for Insomnia (CBT-I) in peri- and post-menopausal women:

• Is feasible
• Is efficacious at 8 and 24-weeks for reducing self-reported

insomnia symptoms, improving overall sleep quality, and increasing sleep efficiency compared to a Menopause Education Control (MEC) condition

• Significantly reduced self-reported hot flash interference,

depression symptoms, and stress compared to MEC Little is known about the interaction between sleep outcomes and baseline demographic, VMS, and mood variables

Intervention

Eligibility

Inclusion: Age/menopause Sleep Hot Flashes Accessibility Consent  40-65 years old, in the menopausal transition or postmenopausal  ISI >12 (the observed median in other MsFLASH trials)  > 2 hot flashes per day on average over 2 weeks by a daily diary  Available by telephone during the 8-week intervention  Written informed consent signed Exclusion:  Menstruated in past 60 days or < 2 skipped cycles  Excessive use of alcohol (> 3 drinks/day)  Previous diagnosis of primary sleep disorder (sleep apnea, restless legs syndrome, PLMS, REM behavior disorder, narcolepsy)  Significant current major illness interfering with sleep or intervention participation (e.g., active cancer)  Routine use (4+ days/week) of sedating prescription medications

RCT: Mean ISI Scores with 95% CI by Intervention Status and Time Point

Mean (95% CI) baseline-week 24 change scores

CBTI (n=44) MEC (n=37) Difference ISI -10.7 (-11.9, -9.4) -6.7 (-8.4, -5.0)

• 4.0 (-7.2, -3.5)

P < .001

RCT: Mean PSQI Scores with 95% CI by Intervention Status and Time Point

P < .002

Mean (95% CI) baseline-week 24 change scores

CBTI (n=44) MEC (n=37) Difference PSQI -4.3 (-5.1, -3.5) -2.7 (-3.5, -1.9)

• 1.6 (-2.7, -0.4)

Specific Aims of Current Study

To evaluate whether the efficacy of CBT-I versus MEC for sleep varies by the presence of specific demographic characteristics and co-occurring menopausal symptoms measured at baseline. Sleep outcomes: Insomnia Severity Index (ISI), Pittsburgh Sleep Quality Index (PSQI) Exploratory hypothesis: Response to treatment in general, and to specific interventions will vary according to baseline: a) Demographics (age, education, marital status, menopausal status, duration of sleep disturbances, alcohol and sleep medication use, Berlin sleep apnea score) b) Vasomotor symptoms (# hot flashes/day, VMS bother) c) Mood (depression, PHQ-8; perceived stress, PSS)

Baseline Demographic Subgroups By Treatment Arm*

* No significant differences by treatment arm

CBTI (%) MEC (%)

Age (> 55 years) 55 49 Education: College graduate 77 77 Married / Marriage like relationship 83 74 Alcohol use > 1 drink/day 64 60 Sleep medication use 49 32 Berlin sleep apnea, > 2 pos categories 38 25 Menopause status Postmenopausal 64 64 Perimenopausal / Indeterminate 36 36 Duration of sleep disturbances < 1 year 17 19 1 - < 5 years 53 38 > 5 years 28 42

Baseline Symptom Subgroups by Treatment Arm*

* No significant differences by treatment arm

CBTI (%) MEC (%)

VMS Frequency, HF / day < 4.8 36 32 4.8 - < 8.9 36 30 > 8.9 28 38 VMS Bother, 1 – 4 scale < 2.46 34 32 2.46 - < 2.87 36 30 > 2.87 30 38 PHQ-8 Depression < 10 70 68 > 10 28 30 Perceived Stress Scale (PSS) < 13 32 32 13 – 17 39 26 > 18 36 40

• Between-group baseline characteristics were compared

using chi-square tests.

• Outcome analyses were based on intention to treat (ITT),

including all randomized participants with follow-up data.

• Tests for significant statistical interactions between

treatment groups and each covariate to assess differential treatment responses were done via linear regression models, separately for 8- and 24-week outcomes.

Moderator Subgroup Analysis Plan

8-week Treatment Interactions with Depression and Stress

• Other than depression and stress, no baseline characteristic,

including VMS frequency and bother, was significantly associated with sleep treatment response at either 8- or 24 weeks.

• Observed depression and stress interaction effects were not

maintained at 24 weeks but despite sustained improvements in insomnia symptoms and sleep quality.

• Results support further development and testing of centralized

CBT-I programs for treating menopausal insomnia.

Conclusions