ISI indicates Insomnia Severity Index.
aParticipants could be ineligible for multiple reasons.
Insomnia Severity Index categories represent the following score ranges: no insomnia, 0 to 7; subthreshold insomnia, 8 to 14; moderate insomnia, 15 to 21; and severe insomnia, 22 to 28. CBT-I indicates cognitive behavioral therapy for insomnia; MEC, menopause education control.
eTable 1. Sensitivity Analyses for Insomnia Severity Index Outcome Results by Treatment Condition
eTable 2. Sensitivity Analyses for Pittsburgh Sleep Quality Index Outcome Results by Treatment Condition
Customize your JAMA Network experience by selecting one or more topics from the list below.
McCurry SM, Guthrie KA, Morin CM, et al. Telephone-Based Cognitive Behavioral Therapy for Insomnia in Perimenopausal and Postmenopausal Women With Vasomotor SymptomsA MsFLASH Randomized Clinical Trial. JAMA Intern Med. 2016;176(7):913–920. doi:10.1001/jamainternmed.2016.1795
Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Effective, practical, nonpharmacologic therapies are needed to treat menopause-related insomnia symptoms in primary and women’s specialty care settings.
To evaluate the efficacy of telephone-based cognitive behavioral therapy for insomnia (CBT-I) vs menopause education control (MEC).
Design, Setting, and Participants
A single-site, randomized clinical trial was conducted from September 1, 2013, to August 31, 2015, in western Washington State among 106 perimenopausal or postmenopausal women aged 40 to 65 years with moderate insomnia symptoms (Insomnia Severity Index [ISI] score, ≥12) and 2 or more daily hot flashes. Blinded assessments were conducted at baseline, 8, and 24 weeks postrandomization. An intent-to-treat analysis was conducted.
Six CBT-I or MEC telephone sessions in 8 weeks. Participants submitted weekly electronic sleep diaries and received group-specific written educational materials. The CBT-I sessions included sleep restriction, stimulus control, sleep hygiene education, cognitive restructuring, and behavioral homework; MEC sessions provided information about menopause and women’s health.
Main Outcomes and Measures
Primary outcome was scores on the ISI (score range, 0-28; scores ≥15 indicate moderate to severe insomnia). Secondary outcome was scores on the Pittsburgh Sleep Quality Index (score range, 0-21; higher scores indicate worse sleep quality). Additional outcomes included sleep and hot flash diary variables and hot flash interference.
At 8 weeks, ISI scores had decreased 9.9 points among 53 women receiving CBT-I (mean [SD] age, 55.0 [3.5] years) and 4.7 points among 53 women receiving MEC (age, 54.7 [4.7] years), a mean between-group difference of 5.2 points (95% CI, –6.1 to –3.3; P < .001). Pittsburgh Sleep Quality Index scores decreased 4.0 points in women receiving CBT-I and 1.4 points in women receiving MEC, a mean between-group difference of 2.7 points (95% CI, –3.9 to –1.5; P < .001). Significant group differences were sustained at 24 weeks. At 8 and 24 weeks, 33 of 47 women (70%) and 37 of 44 (84%) in the CBT-I group, respectively, had ISI scores in the no-insomnia range compared with 10 of 41 (24%) and 16 of 37 (43%) in the MEC group, respectively. The CBT-I group also had greater improvements in diary-reported sleep latency, wake time, and sleep efficiency. There were no between-group differences in frequency of daily hot flashes, but hot flash interference was significantly decreased at 8 weeks for the CBT-I group (–15.7; 95% CI, –20.4 to –11.0) compared with the MEC group (–7.1; 95% CI, –14.6 to 0.4) (P = .03), differences that were maintained at 24 weeks for the CBT-I group (–22.8; 95% CI, –28.6 to –16.9) and MEC group (–11.6; 95% CI, –19.4 to –3.8) (P = .003).
Conclusions and Relevance
Telephone-based CBT-I improved sleep in perimenopausal and postmenopausal women with insomnia and hot flashes. Results support further development and testing of centralized CBT-I programs for treating menopausal insomnia.
clinicaltrials.gov Identifier: NCT01936441
Sleep disturbances are a common and often bothersome menopausal symptom1,2 that increase throughout the menopausal transition and early postmenopause.3,4 Insomnia is associated with increased depression, impaired daytime function, reduced libido, and increased use of health care, creating a substantial burden for women and society.5-10 Women with combined vasomotor and insomnia symptoms have more emergency department visits and lower physical and mental quality of life than women without sleep disturbances.11 Insomnia is also associated with increased risk for obesity, diabetes, stroke, and coronary artery disease,12,13 conditions that increase long-term disease and economic burdens in menopausal women.14
Evidence-based behavioral treatments for insomnia symptoms in perimenopausal and postmenopausal women are lacking. In routine practice, insomnia is most often treated with medications.15-17 However, because of adverse effects,18 not all women desire or benefit from medications to treat insomnia.19 Cognitive behavioral therapy for insomnia (CBT-I) is a well-established, evidence-based approach.20-25 However, in-person CBT-I is rarely available in the settings in which most women receive care. Practical considerations such as cost, transportation, time required for most in-person therapies, and scheduling challenges further affect the accessibility of in-person CBT-I.
We present results from a single-site, randomized clinical trial of a telephone CBT-I intervention vs telephone-based menopause education control (MEC). We hypothesized that CBT-I would be more efficacious than MEC for improving sleep 8 and 24 weeks after randomization.
Question What is the efficacy of brief telephone-based cognitive behavioral therapy compared with menopause education control for insomnia symptoms in perimenopausal and postmenopausal women?
Findings In this randomized clinical trial of 106 women, Insomnia Severity Index scores decreased 9.9 points in women receiving cognitive behavioral therapy and 4.7 points in women receiving menopause education control, a significant difference. Significant group differences were sustained at 24 weeks.
Meaning Results support further development and testing of centralized cognitive behavioral therapy programs for treatment of menopausal insomnia in women.
The study was conducted from September 1, 2013, to August 31, 2015, within the Menopause Strategies Finding Lasting Answers for Symptoms and Health (MsFLASH) research network.26 Women in western Washington State were mailed recruitment postcards from November 4, 2013, to June 1, 2014, that included a telephone screening contact number. Respondents aged 40 to 65 years reporting significant insomnia symptoms and 2 or more hot flashes daily during the previous 2 weeks were mailed a consent form and questionnaires, including 2-week sleep and hot flash diaries. Menopausal status was defined as being postmenopausal, having had no menstrual periods within the past 12 months, having had a bilateral oophorectomy, or being age 55 years or older with hysterectomy or endometrial ablation. Perimenopausal was defined as having had at least 1 menses in the past 12 months or being younger than age 55 years with hysterectomy or endometrial ablation without bilateral oophorectomy.
Eligible women scored 12 or higher on the Insomnia Severity Index (ISI)27 at both telephone screening and on mailed questionnaires. Women were excluded if they had a primary sleep disorder diagnosis, consumed more than 3 alcoholic drinks daily, had a current major illness interfering with sleep, had a job involving shift work (>3 times per week), or routinely (>3 times per week) used prescription sleeping medications. Women reporting use of over-the-counter sleep aids, melatonin, or herbal sleep remedies were not excluded. Screening, eligibility, and participation are shown in Figure 1. The study was approved by the institutional review boards of Fred Hutchinson Cancer Research Center and University of Washington, both in Seattle.
Eligible women were block-randomized to receive CBT-I or MEC. Participants were told that the study compared 2 educational treatments for sleep problems in women with hot flashes, but treatments differed in their approach. Participants were not informed how their group differed from the other.
The CBT-I and MEC interventions both consisted of six 20- to 30-minute telephone sessions conducted over 8 weeks (weeks 1-4, 6, and 8). Participants were invited to have their first session in person at a research office, but women were also permitted to have the first session by telephone. Treatment materials, including the “Menopause: Time for a Change” booklet,28 and additional group-specific reading materials were distributed at the first session or mailed before the first telephone session. In the first session, women were taught how to complete an online daily sleep diary, which was submitted to study interventionists (called coaches) the day before each telephone session.
The CBT-I protocol provided information about age-related sleep changes, sleep hygiene, sleep restriction, and stimulus control procedures (Table 1).27 Participants were instructed to keep a compressed schedule of bed and rising times. Initial sleep restriction windows were set to match the average sleep time reported in baseline screening logs but no less than 5.5 hours in bed. The sleep window was extended by 15 minutes per week when the electronic sleep diary indicated an average 85% sleep efficiency (time asleep divided by time in bed) or greater during the previous week. Stimulus control instructions strengthened the association between bed and sleep by reducing time spent in bed on nonsleep activities. Sleep hygiene education included information about improving bedtime routines and identifying behavioral and environmental factors that negatively affected sleep. Cognitive techniques were taught to reduce physiological arousal at bedtime and to change unrealistic beliefs about sleep loss.29
The MEC protocol included educational content and readings relevant to women’s health and quality of life. Sessions were designed to reduce uncertainty about changes occurring during menopause and to help women identify strategies for symptom self-management. Sessions explicitly excluded active interventions hypothesized to mediate the effect of CBT-I treatment on sleep.30 Individual sessions were conducted in an informative, supportive format in which the coaches remained neutral but did not make recommendations. Weekly sleep logs were submitted. There was no practice or instruction in CBT-I principles (eg, no recommendations to restrict time in bed).
Telephone sessions were led by 2 female coaches with a master’s degree (1 social worker, 1 psychologist) without prior experience in menopause education or CBT-I. Coaches received 1-day training for each intervention, led by experts in CBT-I (C.M.M.) and menopause education (N.F.W.).
Both coaches delivered both interventions; all telephone sessions were recorded. Training included review by the primary investigator (S.M.M.) of all 6 recordings for 2 pilot cases (1 CBT-I, 1 MEC) for each coach. Thereafter, 2 sessions for each participant (1 randomly selected, 1 chosen by either the coach or primary investigator) were reviewed to maintain treatment fidelity and ensure there was no contamination between treatment conditions. Coaches completed weekly content checklists to ensure adherence to key session components. Feedback on reviewed audio recordings was discussed in weekly team meetings.
Blinded assessments were conducted at baseline and 8 and 24 weeks after the intervention. They included primary and secondary sleep outcomes as well as additional sleep and hot flash outcomes described below. Treatment satisfaction was measured at 8 weeks. Assessment packets and diaries were mailed to women with a prepaid return envelope. Women who failed to return packets within 4 weeks of the scheduled collection date were contacted by telephone to gather data on primary and secondary sleep outcomes. Research staff involved in data collection and analysis had no knowledge of treatment group assignment.
Variables included age, race, educational level, marital status, menopausal stage, depression symptoms,31 use of sleep medications, and duration of sleep disturbances.
The primary outcome was score on the ISI,27,32 a 7-item questionnaire assessing global insomnia severity. Items are rated 0 to 4 (total score range, 0-28); a score of 15 or higher is considered moderate to severe insomnia in clinical populations.27 A score higher than 10 is considered optimal for detecting cases of insomnia in community samples,32 and a 6-point within-group reduction is a clinically meaningful change.33
Scores on the 19-item Pittsburgh Sleep Quality Index (PSQI)34 was a secondary sleep outcome. Total scores range from 0 to 21; higher scores indicate worse sleep quality. A decrease to a PSQI score less than 5 or 3-point reduction in score is considered clinically meaningful.35,36 Both the ISI and PSQI have been used in previous MsFLASH network trials.16,37
Daily sleep diaries included bed and rise time, sleep latency (time to fall asleep), and number and duration of nighttime awakenings.38 In a separate diary, participants recorded the frequency, severity, and bother of nighttime and daytime hot flashes. Sleep and hot flash diary results were calculated from 2 weeks of baseline data and 1 week of data at 8 and 24 weeks. The Hot Flash Related Daily Interference Scale39 includes 10 areas of daily functioning that may be affected by hot flashes. Items are rated on a 10-point scale; higher scores indicate worse interference.
At 8 weeks, participants rated the credibility, acceptability, and perceived effectiveness of their intervention by answering the following questions27: Did this treatment and its rationale make sense to you? How acceptable did you consider this treatment? How suitable was this treatment for improving your quality of life despite having menopausal symptoms? How effective did you expect this treatment to be? How well were you able to adhere to this treatment program? How would you rate the quality of your working relationship with your menopause counselor? All items were rated on a 7-point scale; higher scores indicate greater satisfaction.
The intent-to-treat analysis included all participants who provided follow-up data, regardless of adherence to treatment assignment. Baseline characteristics were compared between arms using t tests or χ2 tests. Treatment contrasts for the ISI, PSQI, sleep diary, and hot flashes outcomes were computed as Wald statistics from repeated-measures linear regression models of each outcome by intervention arm, time, and baseline value of the outcome. Repeated-measures logistic regression models were performed to compare incidence of good sleep quality (PSQI score <5) by arm. Participants who contributed baseline and either 8-week or 24-week data were included in these analyses. Robust SEs were calculated via generalized estimating equations to account for correlations between repeated measures from each participant. Treatment effect sizes (difference in mean outcome between groups, divided by the pooled SD) were computed for the ISI and PSQI scores. Treatment satisfaction ratings were compared by arm using t tests.
Two sensitivity analyses of the ISI and PSQI were conducted. First, outcomes data submitted more than 4 weeks past due were excluded from analysis. Second, missing outcomes data for both groups were imputed based on the observed MEC data, using multiple imputation under the assumption that data for participants in the CBT-I group who discontinued follow-up early would mirror data from participants from the MEC group who discontinued.40 All analyses were conducted using SAS, version 9.4 (SAS Institute Inc).
A sample size of 45 participants per group was chosen to provide 90% power to detect a 4-point difference in change in ISI scores between the randomized groups, assuming an SD of 5.6 based on observed scores in an earlier MsFLASH study16 and a t test with 2-sided significance set at P < .05. We planned to enroll 50 women per group to compensate for up to 10% loss to follow-up.
A total of 106 participants (mean [SD] age, 54.8 [4.2] years) were randomly assigned to the 2 intervention arms. The 2 arms did not differ significantly by age, race, educational level, marital status, menopausal stage, use of sleep medications, duration of sleep disturbances (Table 2), or any baseline sleep or hot flash outcome measure.
Participants in both the CBT-I and MEC groups attended an average of 5.7 sessions (range, 1-6). Sessions averaged 22.8 minutes (range, 16.4-32.6). There were no significant differences in the number of telephone sessions or session length by intervention arm or coach.
There were no between-group differences in the number of withdrawals or reasons for treatment discontinuation (Figure 1). Follow-up ISI data were collected on 88 participants (83.0%) at 8 weeks and 81 (76.4%) at 24 weeks; 10 women (3 CBT-I, 7 MEC) completed telephone ISI and PSQI forms at week 8, and 7 women (3 CBT-I, 4 MEC) completed telephone forms at week 24.
At baseline, 31 women (58%) in the CBT-I group and 33 (62%) in the MEC group had ISI scores in the range of moderate (score, 15-21) to severe (score, 22-28) insomnia (Figure 2). From baseline to 8 weeks, ISI scores decreased 9.9 points in women receiving CBT-I and 4.7 points in women receiving MEC, a mean between-group difference of 5.2 points (95% CI, –6.1 to –3.3; P < .001). Significant differences between the groups were sustained at 24 weeks (Table 3).
At baseline, 92.5% of women (49 in each arm) had PSQI scores of 5 or higher, indicating poor sleep quality. On posttreatment follow-up, PSQI scores decreased 4.0 points in women receiving CBT-I and 1.4 points in women receiving MEC, a mean between-group difference of 2.7 points (95% CI, –3.9 to –1.5; P < .001), approaching a 3-point clinically significant difference.35 Significant differences between the groups were sustained at 24 weeks (Table 3). Women in the CBT-I group were significantly more likely than those in the MEC group to have good sleep quality (PSQI score, ≤5) at week 8 (odds ratio, 5.6; 95% CI, 2.3-14.8; P < .001) and week 24 (odds ratio, 3.7; 95% CI, 1.4-9.5; P = .006).
Women in the CBT-I group also had significantly greater 8- and 24-week improvements in diary-reported sleep latency, wake time, and sleep efficiency compared with those in the MEC group, although relative differences between treatment groups were attenuated at 24 weeks.
Standardized mean differences (ie, effect sizes) for ISI and PSQI scores at 8 weeks were 1.04 and 0.84 SDs, respectively, indicating large treatment effects for CBT-I. At 8 weeks, 33 of 47 women (70%) in the CBT-I group had total scores in a range indicating no clinically significant insomnia (score, 0-7) compared with only 10 of 41 women (24%) randomized to receive MEC; at 24 weeks, 37 of 44 women (84%) in the CBT-I group vs 16 of 37 (43%) in the MEC group were in the no insomnia range (Figure 2).
Study results for ISI and PSQI scores were not significantly different from the primary analyses when protocol violators were excluded, and were also robust to sensitivity analyses for missing data (eTable 1 and eTable 2 in the Supplement).
There were no significant differences between treatment group ratings of hot flash frequency (daily or nighttime), severity, or bother at either 8 or 24 weeks. The Hot Flash Related Daily Interference Scale score was significantly decreased at 8 weeks for those in the CBT-I group (–15.7; 95% CI, –20.4 to –11.0) compared with those in the MEC group (–7.1, 95% CI, –14.6 to 0.4) (P = .03). Significant between-group differences were maintained at 24 weeks (CBT-I group, –22.8; 95% CI, –28.6 to –16.9; MEC group, –11.6; 95% CI, –19.4 to –3.8; P = .003). When the Hot Flash Related Daily Interference Scale was analyzed excluding the single sleep item, results were comparable between groups.
Average ratings of perceived suitability, acceptability, effectiveness, and trainer quality for both intervention arms at the posttreatment assessment were high (mean range, 4.2-6.7 on a 1-7 scale). There were no differences between CBT-I and MEC in acceptability, treatment adherence, or relationship quality with the menopause coach. Ratings for CBT-I were significantly higher than for MEC regarding whether the treatment made sense (P = .005), whether it was suitable for improving quality of life despite having menopausal symptoms (P = .009), and perceived treatment effectiveness (P < .001).
Behavioral interventions for women with moderate menopause-associated insomnia and vasomotor symptoms are lacking. In this randomized clinical trial, brief telephone-based CBT-I resulted in significant 8- and 24-week improvements in self-reported insomnia symptoms, overall sleep quality, sleep latency, wake time after sleep onset, and sleep efficiency compared with MEC. Although CBT-I has been found to be efficacious for improving sleep in populations with other comorbid conditions,41 this is one of the first studies, to our knowledge, to show that CBT-I helps healthy women with hot flashes sleep better. A recent small trial found that 6 sessions of CBT-I significantly improved sleep outcomes compared with placebo in middle-aged breast cancer survivors with chronic insomnia.42 A few other small studies reported on psychologist-led groups and self-help cognitive behavioral strategies for improving hot flashes and night sweats but did not target sleep.43,44
Our study found no between-group differences in self-reported hot flash frequency, severity, or bother but found that CBT-I reduced self-reported hot flash interference at 8 and 24 weeks relative to MEC. This finding may indicate that for women receiving CBT-I, the cognitive strategies taught to reduce daytime dysfunction associated with sleep loss generalized to how they responded to vasomotor symptoms. Alternatively, improved sleep could have improved tolerance of hot flashes.
A strength of the study was the telephone-based MEC, which controlled for nonspecific treatment effects, including therapist attention and treatment duration but explicitly excluded active interventions hypothesized to mediate the effect of treatment on sleep.30 Menopause education control had high ratings of acceptability and adherence as well as low dropout treatment rates equivalent to the CBT-I group, suggesting that it was a well-received attention control intervention.
Our study does not provide a comparison with placebo or active medication treatments for insomnia. Cognitive behavioral therapy and pharmacotherapy are considered effective for treating chronic insomnia,15,45 with medications offering an advantage owing to immediate treatment effects, but CBT-I produces superior long-term outcomes.46,47 There have been no head-to-head trials comparing CBT-I vs medication for perimenopausal and early postmenopausal women with insomnia symptoms. Reductions in ISI score with CBT-I in our trial approached those observed in previous studies examining the effect of eszopiclone on insomnia symptoms in menopausal women48 and were larger than have been reported in placebo-controlled trials of the effects of escitalopram, venlafaxine, or low-dose estradiol on sleep in this population.16,49 Future direct comparison of outcomes and cost-effectiveness with pharmacotherapies for insomnia and hot flashes are warranted.
Our findings support the potential for training nonsleep specialists to deliver telephone-based CBT-I to women with insomnia and vasomotor symptoms in a variety of primary and women’s health care settings. Telephone-based CBT-I allows upscaling to reach large populations of menopausal women seeking treatment for sleep problems. Centralized telephone CBT-I should be tested as a dissemination model, similar to effective telephone-based counseling programs for smoking cessation.
This study does have some limitations. The program was delivered in the Seattle, Washington, area, and women responding to recruitment mailings were predominantly college educated and white, limiting generalizablity to other populations. Participants did not undergo formal evaluation for primary sleep disorders; therefore, we were unable to examine whether the effect of CBT-I was consistent across women with and without these conditions. In studies of this type, it is not possible to mask interventionists to treatment assignment. However, all outcomes were collected by research staff blinded to treatment assignment. As expected, lower posttreatment ratings of treatment effectiveness with MEC indicated some nonequivalence between treatment groups in perceived effect on insomnia symptoms.
Sleep and vasomotor outcomes were based on self-report, the most salient and relevant efficacy indicators for clinical practice and women themselves. However, future studies incorporating polysomnography as a screening and outcome measure would have value. Changes in ISI and PSQI scores among women receiving CBT-I were significant and clinically robust after treatment, although the differences relative to MEC were somewhat attenuated at 24 weeks. Trials of the efficacy of CBT-I used in conjunction with other treatments to manage hot flashes are warranted to identify the optimal strategy for achieving long-term improvement in sleep-related symptoms among menopausal women.
Telephone-based CBT-I effectively improved sleep in perimenopausal and postmenopausal women with insomnia and vasomotor symptoms, both immediately after treatment and at 24 weeks of follow-up. These results support further development and testing of centralized CBT-I programs for treatment of menopausal insomnia in women.
Accepted for Publication: March 17. 2016.
Corresponding Author: Susan M. McCurry, PhD, Department of Psychosocial and Community Health, University of Washington, 6200 NE 74th St, Ste 42, Seattle, WA 98115 (email@example.com).
Published Online: May 23, 2016. doi:10.1001/jamainternmed.2016.1795.
Author Contributions: Dr Guthrie had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: McCurry, Guthrie, Morin, Woods, Landis, Ensrud, Joffe, Cohen, Hunt, Newton, Reed, LaCroix.
Acquisition, analysis, or interpretation of data: McCurry, Guthrie, Morin, Woods, Landis, Ensrud, Larson, Joffe, Cohen, Hunt, Otte, Reed, Sternfeld, Tinker, LaCroix.
Drafting of the manuscript: McCurry, Guthrie, Reed, LaCroix.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Guthrie, Larson.
Obtained funding: Joffe, Cohen, Newton, Reed, Sternfeld, LaCroix.
Administrative, technical, or material support: McCurry, Hunt.
Study supervision: McCurry, Morin, LaCroix.
Conflict of Interest Disclosures: Dr Morin reported receiving personal fees for consultancy from Merck and grants from Novartis. Dr Joffe reported receiving grant support from Merck and consulting fees from NeRRe Therapeutics, Merck, SAGE Therapeutics, and Mitsubishi Tanabe. Dr Cohen reported receiving grant support from AstraZeneca Pharmaceuticals, Alkermes, Bristol-Myers Squibb/Otsuka, Sunovion Pharmaceuticals Inc, Bayer HealthCare Pharmaceuticals, Ortho-McNeil Janssen Pharmaceuticals Inc, Pfizer Inc, Forest Laboratories Inc, Cephalon Inc, GlaxoSmithKline, Takeda/Lundbeck, the National Institute on Aging, the National Institutes of Health, and the National Institute of Mental Health, and personal fees for consultancy from JDS Therapeutics LLC, Noven Pharmaceuticals, and PamLab LLC. No other disclosures were reported.
Funding/Support: The Menopause Strategies Finding Lasting Answers for Symptoms and Health research network was established under a National Institutes of Health cooperative agreement to conduct studies of the efficacy of treatments for the management of menopausal hot flashes. The studies were sponsored by the National Institute on Aging in collaboration with the Eunice Kennedy Shriver National Institute of Child Health and Development, the National Center for Complementary and Alternative Medicine, and the Office of Research on Women’s Health. This study was supported by grant U01AG032699 from the National Institute on Aging, National Institutes of Health.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Martha Cagley, MS, and Amy Cunningham, MS, the University of Washington Northwest Research Group on Aging, assisted in conducting this study as intervention coaches. Janet Carpenter, PhD, RN, Center for Research and Scholarship, Indiana University, provided early support and intellectual contributions to the development of this research project. Mss Cagley and Cunningham were paid on the study grant as research interventionists. Dr Carpenter was not compensated for her contribution.