About

Constance H. Fung is a HS Clinical Professor of Medicine at UCLA. Her research focuses on sleep disorders, particularly nocturia, insomnia, and obstructive sleep apnea in older adults. She has been involved in multiple clinical trials aimed at improving sleep quality and treatment adherence among older populations, including the use of consumer wearable devices to augment sleep therapy and strategies for discontinuing hypnotic medications. Her work also explores behavioral treatments for sleep disturbances, the management of comorbid sleep conditions, and the development of decision aids for veterans with sleep disorders. Dr. Fung's contributions include advancing understanding of sleep-related issues in geriatric populations and developing integrated behavioral interventions to improve health outcomes.

Research topics

• Medicine
• Psychiatry
• Internal medicine
• Physical therapy
• Family medicine
• Psychology
• Clinical psychology

Selected publications

• Delivering Cognitive Behavioral Therapy for Nightmares (CBT-N) in primary care within the Veterans Health Administration: A preliminary report on clinician-perceived barriers and benefits

  Journal of Clinical Sleep Medicine · 2026-05-22

  article
  PublisherDOI

• Directly engaging patients with glucose management: integrated visualization of personal device data for use between traditional visits

  Human Factors in Healthcare · 2026-03-21

  articleOpen access

  Keeping glucose below a threshold, as well as minimizing variation, is critical for individuals living with type 2 diabetes to avoid harm to organs and increase length and quality of life. Wearable devices provide patients with the opportunity to gain insight into how their physical activity and sleep quality relate to glucose levels monitored on a continuous basis. With integrated displays combining data from multiple wearable devices as well as self-reported data and assessment, there is an opportunity to see how related choices interact together to affect glucose levels on a continuous basis and averaged over the day, including documenting more information about diet, sleep, mood, and other activity choices. We conducted interviews with patients to assess their understanding of and anticipated strategies to use a personalized integrated display in-between traditional visits without the synchronous support of a clinician. We aimed to: 1) Reveal what information and uses for an integrated report are valuable for patients and 2) Elicit formative suggestions for improvement of the design of the report. An interview protocol was employed to investigate the reactions of patients to an integrated report ( N = 11). The results underscore that integrated visualizations primarily support patients in making sense of their own data obtained from multiple, independent wearable devices. With an integrated report, they can link their health behaviors to glucose variation, motivating self-reflection and behavioral change, and enhancing self-efficacy. Further, participants valued access to their own data and made recommendations to improve the usability of diet and mood information display. We discuss how integrated visualizations can enhance supervisory control and shared decision-making by empowering patients to act as informed partners in shared decision making with clinicians and diabetes self-management between visits.

  PublisherDOI

• Developing a human factors-informed onboarding workflow for digital health programs: lessons from a virtual type 2 diabetes self-management program

  Human Factors in Healthcare · 2026-05-05

  articleOpen access

  Participation in digital health programs often depends on successful setup and early use of digital tools, yet individuals with lower digital familiarity may encounter barriers during virtual onboarding. To support engagement in Closing The Loop (CTL), a virtual self-management program for type 2 diabetes (T2D), we developed and refined a structured approach to remote setup of devices and mobile applications. This paper describes the onboarding process, the barriers encountered during virtual setup, and how these observations informed a more systematic workflow for guided setup, troubleshooting, and verification of data-sharing during early program use. Individuals with T2D receiving care at the VA Connecticut Healthcare System were mailed a smartwatch and user guide before a scheduled virtual onboarding session. During onboarding, participants configured devices and mobile applications required for program participation. Participants completed a survey assessing demographic characteristics and self-reported digital health literacy using the eHealth Literacy Scale (eHEALS). Field notes from onboarding sessions were synthesized into case summaries to characterize onboarding experiences, identify common barriers, and inform iterative refinement of a more structured onboarding workflow. All 12 participants completed onboarding. Participants were 25% women, and 58% were aged 60 years or older. Common onboarding barriers included device syncing (42%), application permissions (42%), difficulty locating downloaded applications (25%), and inaccessible accounts (25%). Session duration ranged from under 20 minutes to over 60 minutes, with two-thirds of participants completing onboarding in 20 to 60 minutes. Case summaries suggested that participants with lower self-reported digital health literacy generally required more stepwise guidance and longer onboarding sessions. Observations from these sessions informed refinement of the onboarding process into the Structured Onboarding Session (SOS) and TechList, a more systematic workflow for guided setup, troubleshooting, and verification of data-sharing during early program use. This quality improvement project identified recurring barriers encountered during remote onboarding and informed development of a more systematic workflow for early setup of digital health tools. Findings underscore the importance of anticipatory troubleshooting, structured branching, and verification of actual data flow or task completion rather than installation alone. The SOS and TechList formalize onboarding as a structured implementation component within a virtual T2D self-management program. The resulting workflow may offer transferable design principles for digital health interventions that require device setup, account creation, data-sharing, or coordination across multiple apps and platforms.

  PublisherDOI

• Perceived Pain Following Hypnotic Deprescribing in Older Adults

  Journal of the American Geriatrics Society · 2026-01-10

  articleOpen accessSenior authorCorresponding

  BACKGROUND: Older adults with chronic insomnia often use benzodiazepine receptor agonists (BZRAs) despite known associated risks and non-pharmacological alternatives such as cognitive behavioral therapy for insomnia (CBTI). CBTI reduces insomnia severity and could potentially improve other outcomes such as the impact of pain on daily activities, even when BZRAs are deprescribed. Yet concerns that deprescribing may worsen pain (which is often comorbid with insomnia) can be a barrier to engagement in BZRA deprescribing. This study examined changes in pain outcomes associated with deprescribing BZRAs in the context of concurrent CBTI. METHODS: Secondary data analysis was conducted using data from a randomized clinical trial that successfully decreased BZRA use in older adults. Participants (n = 188), who were largely older (68% ≥ 65 years, 55 ≤ range ≤ 91) and male (65%), completed CBTI concurrently with a deprescribing intervention (blinded encapsulated BZRA taper or open pill cutter taper). Participants completed the Brief Pain Inventory (BPI) at baseline, one week posttreatment (1 WK), and at a six-month (6 M) follow-up. Analyses included mixed effects models among all participants and a subset aged 65+ as well as comparison of model results to minimal clinically important difference (MCID) thresholds. RESULTS: Mixed effects models demonstrated that pain severity did not change significantly over time, broadly or in participants aged ≥ 65 years. Significant reductions in pain interference in day-to-day living at 1 WK were observed broadly, although these reductions did not meet the MCID threshold and were no longer significant at 6 M follow-up. CONCLUSIONS: Combined BZRA deprescribing and CBTI did not meaningfully worsen pain in older adults. These results highlight the opportunity for using a combination of CBTI and deprescribing methods in patients with insomnia and comorbid pain, as well as a need for additional interventions to specifically address pain in older adults with chronic insomnia.

  PublisherOA PDFDOI

• Clin‐ <scp>STAR</scp> Corner: Practice Changing Advances at the Interface of Artificial Intelligence/Machine Learning and Geriatrics

  Journal of the American Geriatrics Society · 2026-05-23

  articleSenior author

  ABSTRACT Artificial intelligence (AI) methods, including machine learning (ML), are transforming healthcare by enabling personalized interventions that integrate multimodal data to support rehabilitation, preventive care, and remote monitoring. Despite their broad potential, older adults remain underrepresented in model development, raising concerns about bias and limited generalizability. As AI/ML adoption expands, it is essential to critically appraise emerging tools to ensure ethical, equitable, person‐centered implementation in aging populations and long‐term randomized evaluations. A structured MEDLINE search identified randomized controlled trials (RCTs) published between January 2023 and December 2025 that evaluated AI/ML‐based interventions in adults ≥ 65 years. Of 31 records identified, 19 underwent abstract screening, seven underwent full‐text assessment, and four articles were identified as meeting all criteria for inclusion, which focused on scalability, clinical impact, and methodological rigor. Across postoperative rehabilitation and preventive care, AI/ML interventions demonstrated meaningful clinical benefits. Transformative applications of AI/ML described robotic systems quantifying weight‐bearing, algorithm‐guided paired rehabilitation interventions, paired exercises in geriatric hip fracture rehabilitation, smartphone platforms delivering continuously personalized exercise programs, and conversational chatbots providing tailored vaccine counseling. The included RCTs show that AI‐driven interventions can enhance physical recovery, support psychosocial well‐being, and improve uptake of preventive measures compared with conventional approaches. Key considerations for future implementation include digital health literacy, long‐term follow‐up, and potential bias arising from healthier or more motivated study populations. As evidence grows, geriatric innovation must prioritize safety, ethics, and equitable access to ensure that technological precision enhances, rather than replaces, person‐centered care.

  PublisherDOI

• An integrated behavioral treatment for improving nocturia and insomnia symptoms in older adults (MINT): study protocol for a multi-site randomized clinical trial

  Trials · 2026-02-16

  articleOpen access1st authorCorresponding

  BACKGROUND: Nocturia (i.e., waking to void during the primary sleep period) of two or more times per night affects nearly one-third of older adults and can have a severe impact on sleep, contributing to insomnia symptoms. Current treatment approaches for nocturia often overlook non-lower urinary tract factors that may contribute to nighttime awakenings. Nocturia management, for example, may benefit from more effective integration of cognitive behavioral therapy for insomnia (CBT-I) principles that address other factors underlying insomnia symptoms, and early evidence suggests it also reduces nocturia and the bother it causes. Because nocturia treatment crosses specialties, coordinated delivery of urological and sleep therapies is a treatment barrier. The overall purpose of this trial is to determine whether a promising coordinated, integrated behavioral, non-pharmacological, non-surgical treatment that simultaneously addresses both the urological and insomnia factors contributing to nocturia is efficacious for improving nocturia, sleep, and daytime function. METHODS: This multicenter parallel-group randomized, efficacy trial compares a 5-week integrated behavioral treatment program delivered by a single interventionist (psychologist, nurse practitioner, or physician assistant) to a health education control program in adults aged 60 years or older (proposed n = 192) recruited from sites in Atlanta and Los Angeles, who report typically getting up to urinate two or more times per night (International Consultation on Incontinence Questionnaire-Overactive bladder [ICIQ-OAB] nocturia item) and insomnia symptoms (Insomnia Severity Index > 7). The integrated program includes components of CBT-I and pelvic floor muscle exercise-based behavioral therapy for nocturia. The primary outcome is ICIQ-OAB-measured nocturia frequency 4 months after randomization. Secondary outcomes are sleep diary-measured wake after sleep onset (mean minutes) and Insomnia Severity Index total score. DISCUSSION: The interdisciplinary trial team has developed a program aimed at improving nocturia symptoms and overall sleep of older adults in an efficient and safe manner. The integrated behavioral program has the potential to address nocturia, which is a challenging symptom because it has many etiologies that cross multiple specialties. Findings will provide rigorous evidence of the efficacy of the integrated behavioral treatment program to reduce nocturia frequency as well as sleep disturbance in older adults. TRIAL REGISTRATION: Clinicaltrials.gov NCT06110091, registered 10/25/2023.

  PublisherDOI

• Experiential Avoidance is Associated with Insomnia Symptoms and Related Consequences Among Veterans with Posttraumatic Stress Disorder

  Behavioral Sleep Medicine · 2026-04-15

  articleOpen access

  OBJECTIVES: Experiential avoidance is the unwillingness to come into contact with aversive internal experiences. Trauma exposure is associated with greater experiential avoidance and insomnia symptoms. Experiential avoidance may perpetuate insomnia symptoms in patients with posttraumatic stress disorder (PTSD). We examined the relationship between experiential avoidance and insomnia symptoms among veterans with PTSD (based on the Clinician-Administered PTSD Scale for DSM-5). METHOD: = 54.7 years; 86.0% male) who attributed their sleep disturbance onset to experiences of trauma on the CAPS-5. Experiential avoidance, insomnia, sleep disturbance, daytime sleepiness, and daytime consequences were measured with the Brief Experiential Avoidance Questionnaire (BEAQ), Insomnia Severity Index (ISI), Pittsburgh Sleep Quality Index (PSQI), Epworth Sleepiness Scale (ESS), and International Classification of Sleep Disorders (ICSD) items. We conducted multiple linear regressions with age, sex, and BEAQ as the independent variables and sleep variables as the dependent variables. RESULTS: There were significant positive associations between the BEAQ and the ISI, PSQI daily disturbance factor, ESS, and ICSD daytime consequences. CONCLUSIONS: Greater experiential avoidance was associated with worse insomnia symptoms and consequences, particularly daytime dysfunction. Experiential avoidance may be an overlooked, but relevant treatment target for patients with comorbid insomnia and PTSD.

  PublisherOA PDFDOI

• Sleep aid usage following benzodiazepine receptor agonist tapering and cognitive behavioral therapy for insomnia in middle-aged and older adults

  Journal of Clinical Sleep Medicine · 2026-05-21

  articleSenior author
  PublisherDOI

• 0527 Differences in Subjective and Objective Measures of Sleep by Marital Status in a Sample of US Veterans

  SLEEP · 2025-05-01

  articleOpen access

  Abstract Introduction Marriage confers a degree of benefit for a range of health factors, including sleep behaviors and outcomes. Veterans are more likely to report sleep difficulties (e.g., insomnia) and have less social support than the general population. As such, it is important to examine associations between marital status and the occurrence of sleep issues among veterans. Methods This study utilizes baseline visit data from veterans in Los Angeles reporting sleep concerns who were enrolled across four structured cognitive behavioral therapy for insomnia intervention trials. Measures included the Pittsburgh Sleep Quality Index (PSQI) total and factor scores, Insomnia Severity Index (ISI), sleep efficiency (measured with actigraphy), and the Patient Health Questionnaire (PHQ-9). Overall, 1,411 veterans had available data (74.1% male; mean age: 63.4 years; 43.3% married/living as married, 21.9% separated, 17.0% single/never married, 10.5% divorced, 7.2% widowed). Multilevel models were constructed to evaluate associations between marital status and sleep measures after controlling for sex, age, and depressive symptoms. Individuals were nested within the four different sleep trials and “married” served as the reference group. Results Multilevel modeling showed that separated individuals had scores indicative of worse sleep compared to married individuals (PSQI total score: β = 0.18, p &amp;lt;.01, N = 1,181; PSQI Sleep Efficiency factor score: β = 0.18, p =.017, N = 1,188; sleep efficiency per actigraphy: β = -0.23, p = 0.014, N = 845). Differences by marital status were not found for the PSQI Perceived Sleep Quality factor, PSQI Daily Disturbances factor, or ISI. Divorced, widowed, and single/never married compared to married did not show differences for any of the sleep measure outcomes. Conclusion Using cross-sectional data, we found that veterans who are separated have more severe sleep concerns compared to veterans who are married/living as married. Given this difference, separated veterans may represent a group that are particularly at-risk for sleep difficulties and they may benefit from evaluation of sleep concerns and management of sleep disorders. Findings align with past work showing that marital status can impact measures of health. Support (if any) VA OAA (Gold); VA SWI 24-001, VA OAA (Erickson), RCS-20-191, NIH K24 HL143055 (Martin); VAGLAHS CSHIIP and GRECC

  PublisherOA PDFDOI

• Insomnia Symptom Improvement as a Mediator for Mental Health Symptom Reduction Following Behavioral Insomnia Treatment Among Women Veterans

  Behavior Therapy · 2025-02-20

  articleOpen access
  PublisherOA PDFDOI

Recent grants

• Improving Older Adults' Decision Making For Obstructive Sleep Apnea Treatment

  NIH · $786k · 2013–2019

• A novel mechanism for helping older adults discontinue use of sleeping pills

  NIH · $2.7M · 2018–2024

Frequent coauthors

• Cathy A. Alessi

  VA NY Harbor Healthcare System

  461 shared

• Jennifer L. Martin

  Captain James A. Lovell Federal Health Care Center

  438 shared

• Michael N. Mitchell

  Geriatric Research Education and Clinical Center

  232 shared

• Yeonsu Song

  Geriatric Research Education and Clinical Center

  228 shared

• Karen Josephson

  VA Greater Los Angeles Healthcare System

  218 shared

• Joseph M. Dzierzewski

  National Sleep Foundation

  192 shared

• Stella Jouldjian

  VA Greater Los Angeles Healthcare System

  157 shared

• Michelle Zeidler

  Geriatric Research Education and Clinical Center

  125 shared

Education

• M.D.

  University of California, Los Angeles

  1990

• B.S.

  University of California, Los Angeles

  1986