About

Margaret (Molly) Baldwin Conroy, MD, MPH, FACSM, FAHA, is a Primary Care Provider and General Internist at the Madsen Internal Medicine Clinic at the University of Utah Health. Her clinical interests include physical activity and lifestyle counseling, weight management, cardiovascular disease prevention, hypertension management, and women’s health. She received her undergraduate degree from Stanford University and her MD from the University of Pennsylvania. Dr. Conroy completed an Internal Medicine residency in the Primary Care Track at the University of California, San Francisco, and a General Internal Medicine Fellowship at Harvard Medical School/Massachusetts General Hospital. She is Board Certified by the American Board of Internal Medicine and recognized as a Fellow of both the American College of Sports Medicine and the American Heart Association.

Research topics

• Medicine
• Internal medicine
• Nursing
• Political Science
• Physical therapy
• Psychology
• Economic growth
• Environmental health
• Intensive care medicine
• Gerontology
• Family medicine
• Social psychology

Selected publications

• Actigraphic estimates of sleep duration in those reporting sleeping less than 7 h

  Journal of Clinical Sleep Medicine · 2026-05-22

  articleOpen access

  OBJECTIVES: Sleep duration < 7 h increases risk for chronic disease, which makes identifying short sleep duration critical to public health. The goal of this study is to evaluate objective short sleep duration among individuals who self-report insufficient sleep to test and evaluate predictors of the subjective-objective sleep duration difference. METHODS: This study presents baseline data from a sleep extension study involving adults aged 18-65, fluency in English, and self-reported sleep duration ≤ 7 h and elevated blood pressure. Objective sleep duration was measured with actigraphy, and subjective-objective sleep difference was calculated as the difference between self-reported habitual sleep duration and actigraphically measured sleep duration. Data were analyzed using regression models, Bland-Altman plots and exploratory spline-based logistic regression models. RESULTS: Among 195 adults (age m = 42 ± 11 years), 54% had objective sleep duration < 7 h, and on average participants underestimated their sleep by 29 m. Under-reporting of self-reported sleep compared to actigraphy was associated with poorer objective sleep, including longer sleep onset latency (p < .001) and higher wake after sleep onset (p < .001), but also higher sleep efficiency (p < .001). In addition, perceived stress (p < .01) and self-reported sleep disturbance (p < .01) were associated with underestimation. A Bland-Altman analysis showed larger negative differences at longer objective sleep durations, consistent with both sleep perception patterns and statistical regression effects. Exploratory spline-based logistic modeling indicated a U-shaped relation with the lowest predicted probability of a large subjective-objective difference occurring at approximately 6.3 h of objective sleep duration. CONCLUSION: These findings highlight the importance of objective assessments to determine short sleep duration. Poorer subjective and objective sleep and higher stress may intensify perceptions of inadequate sleep, contributing to under-reporting. Sleep duration < 7 h increases risk for chronic disease, which makes identifying short sleep duration critical to public health. The goal of this study is to evaluate objective short sleep duration among individuals who self-report insufficient sleep and test and evaluate predictors of the subjective-objective sleep duration difference. Results demonstrate that half of individuals with self-reported sleep duration < 7 h did not have objective short sleep duration; most participants tended to underestimate their sleep duration, in particular those with poorer objective sleep, sleep disturbance, and higher stress. This difference between self-reported and objective sleep highlights the complexity of identifying individuals and populations with objectively short sleep duration for research and public health interventions.

  PublisherDOI

• Sleep Duration and Quality as Predictors of Weight Loss and Adherence During a Behavioral Weight Loss Intervention

  Behavioral Sleep Medicine · 2026-03-26

  article

  OBJECTIVES: To investigate the independent and combined associations between sleep duration and quality with weight loss and adherence in a behavioral weight loss intervention. METHOD: Adults with overweight/obesity (N = 296) participated in a 12-month intervention. Weight and sleep (Pittsburgh Sleep Quality Index) were assessed at baseline, 6 and 12 months. Analyses examined associations between baseline sleep and sleep changes with 6- and 12-month weight loss and intervention adherence (attendance, physical activity, caloric intake). RESULTS: < .02). CONCLUSION: Baseline short sleep duration and poor sleep quality independently and jointly predicted poorer weight loss and lower adherence in a behavioral weight loss intervention. Screening for sleep disturbance may identify individuals who could benefit from additional support to improve adherence and outcomes.

  PublisherDOI

• Social needs screening in hepatology clinic: a qualitative study of patient experience

  BMC Gastroenterology · 2026-04-07

  articleOpen access
  PublisherDOI

• Effects of reducing sedentary behaviour on heart rate variability and cardio‐metabolic biomarkers in desk workers with untreated high blood pressure

  Experimental Physiology · 2026-04-02

  articleOpen access

  Lower sedentary behaviour (SB) has been associated with reduced cardiovascular disease and mortality. Yet, few trials have evaluated cardiovascular mechanisms of benefit. We examined the effects of a SB reduction intervention on heart rate variability (HRV) and cardio-metabolic biomarkers. We evaluated secondary outcomes in RESET-BP, a randomized clinical trial testing the effects of a 3-month SB reduction intervention on blood pressure (BP) in desk workers with untreated high BP. The intervention included health coaching, sit-stand desks, wearable activity prompters and text messages that reduced SB by increasing standing (∼1 h/day) and stepping (∼5 min/day). Resting HRV was assessed in a subset of participants. Fasting blood plasma and serum were drawn to evaluate several cardio-metabolic biomarkers, including glucose, insulin, lipids, aldosterone and plasma renin activity. Analysis of covariance models examined the effect of the intervention on outcomes. Participants with baseline and follow-up measures of HRV (n = 146) and cardio-metabolic biomarkers (n = 188) were included. No significant intervention effects were observed for any measures of HRV, glucose, insulin or lipids (all P > 0.05). Three-month changes in aldosterone (difference = -1.5 ng/dL; P = 0.0370) and plasma renin activity (difference = -0.2 ng/mL/h; P = 0.0462) were significantly lower in the intervention compared to the control condition at the 3-month follow-up. Reducing SB with mostly standing did not improve HRV or cardio-metabolic biomarkers. However, the observed intervention effects on plasma renin and aldosterone encourage future research to confirm that SB reduction suppresses the renin-angiotensin-aldosterone system activity.

  PublisherDOI

• Effectiveness of Nonpharmacologic Treatments for Chronic Low Back Pain

  Annals of Internal Medicine · 2026-04-20

  article

  BACKGROUND: Many treatments are recommended for chronic low back pain (cLBP), but comparative effectiveness and adaptive interventions have not been adequately studied. OBJECTIVE: To compare the effectiveness of physical therapy (PT) and cognitive behavioral therapy (CBT) as first-stage treatment and switching treatments versus mindfulness as second-stage treatment. DESIGN: Multisite sequential, multiple-assignment, randomized trial with 52-week follow-up. (ClinicalTrials.gov: NCT03859713). SETTING: Three health care systems. PARTICIPANTS: Adults with cLBP. INTERVENTION: Eight weeks of PT or CBT in stage I. Nonresponders were randomly assigned again to 8 weeks of stage II treatment. MEASUREMENTS: Co-primary outcomes were function measured with the Oswestry Disability Index (ODI; range, 0 to 100) and pain intensity (range, 0 to 10) at 10 (stage I), 26, and 52 (stage II) weeks. RESULTS: The sample comprised 749 participants. After 10 weeks, there was greater improvement in function in the PT group (adjusted mean ODI difference, 2.8 [96% CI, 0.38 to 5.1]) and no difference in pain intensity (adjusted mean difference, 0.32 [99% CI, -0.07 to 0.71]). The mean difference in ODI was below the minimum important difference of 6. After 52 weeks, there were no differences in stage II treatments for nonresponders for either function (adjusted mean ODI difference, 0.43 [96% CI, -0.29 to 2.4]) or pain intensity (adjusted mean difference, -0.05 [96% CI, -0.58 to 0.48]). LIMITATIONS: Treatment initiation was lower than expected, particularly for CBT and for nonresponders. Participants were not blinded. Sample size was reduced due to the COVID-19 pandemic. CONCLUSION: Patients with cLBP may benefit from PT as first-line treatment. Among nonresponders, there were no differences in second-stage treatment with mindfulness or switching. PRIMARY FUNDING SOURCE: Patient-Centered Outcomes Research Institute (PCORI).

  PublisherDOI

• 0277 Sleep Duration, Efficiency, and Nocturnal Blood Pressure Among a Sample of Participants with Short Sleep Duration and Elevated Blood Pressure

  SLEEP · 2025-05-01

  articleOpen access

  Abstract Introduction Multiple aspects of sleep, including duration and efficiency, are important for blood pressure (BP) regulation and cardiovascular health. While prior research links sleep efficiency (SE) and total sleep time (TST) to BP outcomes, less is known about these associations among populations with chronic short sleep duration. This study examined associations between SE, TST, and 24-hour ambulatory BP in individuals with elevated BP and sleep duration &amp;lt; 7 hours per night. Methods Participants were screened for a randomized sleep extension intervention among individuals with elevated BP (□120 mmHg systolic and/or 80 mmHg diastolic) and habitual short sleep. SE and TST were measured via 7-day wrist actigraphy. Participants completed 24-hour ambulatory BP monitoring, with variables including average awake and asleep SBP/DBP and dipping ratios (asleep BP/awake BP, classified as non-dipping if ≥0.9). Associations between sleep and BP outcomes were analyzed using linear regression for continuous outcomes (e.g., SBP, DBP, dipping ratios) and Poisson regression with a log-link for binary outcomes (e.g., Asleep Hypertension, Non-Dipping SBP, Non-Dipping DBP). Two models were tested: Model 1 (unadjusted) and Model 2 (adjusted for age, sex, and race). Results The study included 114 participants (mean age: 42 years, SD: 11; 40% female). Longer TST was significantly associated with more favorable nocturnal BP measures. For every 10-minute increase in TST, the odds of Asleep Hypertension decreased by 6.3% (OR = 0.937, 95% CI [0.896, 0.980], p = 0.009), and the odds of Non-Dipping DBP decreased by 8.7% (OR = 0.913, 95% CI [0.844, 0.990], p = 0.031). SE was significantly associated with DBP Dipping Ratio in unadjusted models (b = 0.0035, 95% CI [0.0001, 0.0068], p = 0.040) but was marginal after adjustment (b = 0.0033, p = 0.059). Neither TST nor SE were associated with awake BP measures. Conclusion Longer TST was consistently associated with better nocturnal BP, including lower odds of asleep hypertension and non-dipping DBP. The relationship with SE and BP values diminished significance after adjustment, suggesting confounding. Results align with prior literature, highlighting the importance of promoting adequate sleep duration to mitigate cardiovascular risks in populations with short sleep and elevated BP. Support (if any) Supported by grant R01NR018891.

  PublisherOA PDFDOI

• Feasible and acceptable social drivers of health screening among patients with chronic liver disease

  Hepatology Communications · 2025-06-27 · 2 citations

  articleOpen access

  BACKGROUND: Social drivers of health (SDoH) contribute to health disparities among patients with chronic liver disease (CLD). Little is known about the feasibility and acceptability of SDoH screening in hepatology clinics. This study aimed to define SDoH prevalence among CLD patients, identify a feasible and acceptable screening approach, and assess the convergent validity of a locally developed screener. METHODS: Among adult patients with CLD receiving care in hepatology clinics, 2 SDoH screeners were administered to eligible participants: (1) default electronic medical record (EMR) questions and (2) Screener for Intensifying Community Referrals for Health (SINCERE). The primary outcomes were (1) prevalence of SDoH, (2) SDoH screening feasibility and acceptability, and (3) factors associated with screening acceptability. As a secondary outcome, the convergent validity of SINCERE to EMR was assessed. RESULTS: Among 250 participants, the mean age was 56 years, 56% were women, 22% were Hispanic, 7% were American Indian/Alaska Native, 58% had cirrhosis, 29% completed high school or less, 22% were unemployed/disabled, and 29% had an annual income <$35,000. Based on SINCERE, 26% had food insecurity, 8% transportation needs, 43% financial strain, 5% lack of social support, and 24% housing instability. Most respondents (69%) were comfortable or very comfortable completing SDoH screening. Using the McNemar test, there were statistically significant differences between screeners for financial strain and housing instability. CONCLUSIONS: Among CLD patients at our center, SDoH were prevalent, and screening within the hepatology clinic was feasible and acceptable. To detect social needs, SINCERE, a locally developed screener, had overall acceptable convergent validity. These data support SDoH screening in hepatology clinics. Future multicenter studies evaluating the effective implementation of SDoH screening for CLD patients, including contextualized care plans and connection to available resources, should be conducted.

  PublisherDOI

• Abstract 4369002: Blood Pressure Variability and Implications for Trial Screening

  Circulation · 2025-11-03

  article

  Background: Blood pressure (BP) is a dynamic measure influenced by contextual and behavioral factors, with well-established evidence showing notable differences between office (OBP) and home (HBP) measurements . While OBP is often used to diagnose and manage hypertension and can be also used to identify eligible patients for clinical trials to enhance trial pragmatism through efficient use of electronic health records, caution is warranted against relying on a single OBP measurement for participant screening. In this study, we explored explanatory factors that associated with BP variability. Hypothesis: We hypothesize that OBP and HBP will vary, and that variability cannot be explained by change in anti-hypertensives alone. Methods: We compared OBP and HBP for 1000 patients above age 65 from two health systems participating in an NIH funded pragmatic, hybrid effectiveness implementation study. Patients were identified using a pragmatic recruitment strategy based on OBP. BP variability was defined as the SBP difference (i.e. SBP between qualifying OBP and baseline HBP) and further grouped into low (SBP less than 10 mmHg), moderate (SBP between 10 - 20 mmHg), and high (SBP more than 20 mmHg). We also examined changes in anti-hypertensives or their doses that could contribute to the BP variability. Results: The study cohort had a mean (SD) age of 73.3 (5.6), with 61% female, 88% White, 10% Black and 2% Hispanic participants. The mean (sd) baseline DBP and SBP was 135 (17) mmHg, and 70 (16) bpm, respectively. The median (Q1, Q3) run-in period was 13 (8, 20) days between eligibility OBP and baseline visit with an HBP. 285, 292, 423 participants had low (0 – 10), moderate (10 – 20), and high (&gt; 20) BP variability. Among those with moderate to high BP variability, 54 (8%) increased the use of antihypertensives (AHT) during the run-in period. In addition, female sex [1.62 (1.23 – 2.12)], low OBP [1.07 (1.05 – 1.08)], vision deficit [1.35 (1.01 – 1.81)], valvular disease [1.51 (1.03 – 2.21)] and smoking [0.50 (0.28 – 0.87)] are shown to be significantly associated with high BP variability. Conclusion: These analyses highlight the substantial variability in BP readings in the same person and offers critical insights to inform the design of robust pragmatic trials in hypertension research. For screening for enrollment in clinical trials, multiple HBP readings, may be better than a single OBP reading.

  PublisherDOI

• Response by Barone Gibbs et al to Letters Regarding Article, “Effects of Sedentary Behavior Reduction on Blood Pressure in Desk Workers: Results From the RESET-BP Randomized Clinical Trial”

  Circulation · 2025-07-07

  letter
  PublisherDOI

• Impact of sedentary behaviour reduction on desk-worker workplace satisfaction, productivity, mood and health-related quality of life: a randomised trial

  Occupational and Environmental Medicine · 2025-02-01 · 5 citations

  article

  OBJECTIVES: Sedentary behaviour (SB) is related to lower worker health-related quality of life (HRQOL), mental health and productivity. However, it is unknown whether reducing SB improves these outcomes. This study assessed whether a 3-month SB reduction intervention improved or was associated with dose-response changes in workplace satisfaction, productivity, mood and HRQOL. METHODS: Inactive desk workers with elevated blood pressure were randomised to a 3-month SB reduction intervention (n=135) or control (n=136). The intervention used a sit-stand desk, wrist-worn activity prompter and bi-monthly individual coaching to primarily replace work SB with standing and stepping. SB measured via a thigh-mounted activPAL3 micro, workplace satisfaction, productivity, mood and HRQOL were assessed at baseline and 3 months. Analyses of covariance compared changes in outcomes between groups with adjustment for baseline values. In both groups, associations between changes in work and non-work SB and outcomes were examined using linear regression. RESULTS: Compared with control, reducing SB did not significantly change workplace satisfaction, productivity or HRQOL. Among mood measures, only vigour improved (1.01 vs 0.1 points, p=0.0302). Among all participants, reductions in non-work SB were associated with improvements in workplace satisfaction (+0.15 and +0.27 points/hour of non-work standing and stepping, respectively) and productivity (+0.47 points/hour of non-work stepping), while changes in SB during work were not related. CONCLUSIONS: Interventions reducing non-work rather than work SB may be more impactful for improving workplace satisfaction and HRQOL outcomes. Future studies should explore workplace-based SB reduction interventions that are longer and target non-work time to improve similar outcomes. TRIAL REGISTRATION NUMBER: NCT03307343.

  PublisherDOI

Recent grants

• Maintaining Activity and Nutrition through Technology-Assisted Innovation in Prim

  NIH · $2.4M · 2012–2018

• NIH Grant K23HL085405

  NIH · $602k · 2011

• Promoting Lifestyle Change via Tailored mHealth to Improve Health

  NIH · $3.1M · 2017–2024

• Utah Stimulating Access to Research in Residency Transition Scholar (StARRTS)

  NIH · $337k · 2018–2022

Frequent coauthors

• Bret H. Goodpaster

  Translational Research Institute for Metabolism and Diabetes

  59 shared

• Robert M. Boudreau

  University of Pittsburgh

  53 shared

• Anne B. Newman

  University of Pittsburgh

  53 shared

• Tamara B. Harris

  National Institute on Aging

  50 shared

• Pedro Velasquez‐Mieyer

  LifeDoc Health

  49 shared

• Michael Nevitt

  University of California, San Francisco

  49 shared

• Todd M. Manini

  University of Florida

  49 shared

• Marjolein Visser

  Amsterdam Neuroscience

  49 shared

Education

• B.A.

  Stanford University

• M.D.

  University of Pennsylvania

• Other, General Internal Medicine

  Harvard Medical School/Massachusetts General Hospital

• Other, Internal Medicine

  University of California, San Francisco