Michael P Lavalley
· PhD ProfessorVerifiedBoston University · Biostatistics
Active 1995–2025
About
Michael LaValley, PhD, is a Professor of Biostatistics at the Boston University School of Public Health. He teaches courses on meta-analysis, logistic regression, and survival analysis. His research efforts are focused on meta-analysis of study results, analysis of ordered categorical data, survival analysis, clinical trials, and outcome assessment in arthritis studies. He serves as the Research Director of the Boston University Core Center for Clinical Research and is the co-lead of the Analysis Core for the Multicenter Osteoarthritis Study (MOST). Dr. LaValley's academic background includes a BA in Mathematics from American International College, an MA in Mathematics from Ohio State University, and a PhD in Statistics from Penn State University. He joined Boston University after completing a postdoctoral fellowship in Biostatistics at the Harvard School of Public Health.
Research topics
- Internal medicine
- Medicine
- Surgery
- Pathology
- Physical therapy
- Anatomy
- Physical medicine and rehabilitation
Selected publications
Clinical Biomechanics · 2025-11-22
articleOpen accessOsteoarthritis and Cartilage · 2025-04-01
articleChiropractic & Manual Therapies · 2025-02-20 · 5 citations
articleOpen accessINTRODUCTION: Limited adoption of first line treatments for low back pain (LBP) in primary care settings may contribute to an overreliance on pain medications by primary care providers (PCPs). While chiropractic care typically includes recommended nonpharmacologic approaches (e.g., manual therapy, exercise instruction, advice on self-care), implementation strategies to increase adoption of chiropractic care for LBP in primary care clinics are understudied, particularly in underserved communities. METHODS: We will use a stepped-wedge cluster randomized controlled pilot trial design to evaluate the feasibility of a multi-level implementation strategy to increase adoption of chiropractic care for LBP in primary care clinics at community health centers. Key barriers and facilitators identified by site champions and other key stakeholders will help us to develop and tailor implementation strategies including educational materials and meetings, developing a network of local chiropractors, and modifying the electronic health record to facilitate referrals. Three primary care clinics will be randomized to receive the implementation strategy first, second, or third over a fourteen-month study period. At our first clinic, we will have a four-month pre-implementation period, a two-month implementation deployment period, and a subsequent eight-month follow-up period. We will stagger the start of our implementation strategy, beginning in a new clinic every two months. We will evaluate the proportion of patients with LBP who receive a referral to chiropractic care in the first 21 days after their index visit with PCP. We will also evaluate adoption of other guideline concordant care (e.g., other nonpharmacologic treatments) and non-guideline concordant care (e.g., opioids, imaging) over the study period. DISCUSSION: LBP is currently the leading cause of disability worldwide. While there are several treatment options available for individuals with LBP, patients in underserved populations do not often access recommended nonpharmacologic treatment options such as chiropractic care. The results from this study will inform the development of practical implementation strategies that may improve access to chiropractic care for LBP in the primary care context. Furthermore, results may also inform policy changes needed to expand access to chiropractic care in underserved communities. CLINTRIALS.GOV NCT#: NCT06104605.
Arthritis Care & Research · 2025-08-12 · 1 citations
articleOpen accessOBJECTIVE: Pain sensitization is common in knee osteoarthritis (OA) and is associated with pain severity and functional limitations. Whether pain sensitization is induced by OA or it may be an inherent trait remains unclear. We evaluated pain sensitization trajectories and their relations to symptoms in people with or at risk of knee OA. METHODS: We used data from four study visits of the Multicenter Osteoarthritis Study over a nine-year period. We agnostically identified pain sensitization trajectory groups defined by wrist and knee pressure pain thresholds (PPTs) over nine years using group-based trajectory methods stratified by sex. We evaluated the relations of the wrist and knee PPT trajectory groups to Western Ontario and McMaster Universities Arthritis Index (WOMAC) knee pain and function at the final visit using separate linear regression models. RESULTS: ) and identified three distinct trajectory groups in both women and men: low, moderate, and high PPTs. Overall, the PPT trajectories changed minimally over time in each group. Compared to the low PPT trajectory groups (the most sensitized groups), the moderate and high PPT groups had better WOMAC pain and function at the final assessment nine years later. CONCLUSION: We identified stable but distinct pain sensitization trajectories over nine years despite likely changes in disease course and treatments over time, suggesting that individuals with knee OA may be predisposed to having different degrees of sensitization as an inherent trait, which in turn likely influences their pain experience and functional status.
Open Forum Infectious Diseases · 2025-06-19 · 7 citations
articleOpen accessAbstract Background Infections following colonization of multidrug-resistant gram-negative bacteria (MDR-GNB), particularly Enterobacterales with extended-spectrum beta-lactamases (ESBL-E) or carbapenem-resistant Enterobacterales (CRE), represent a major global health threat. Our aim was to assess quality of evidence and provide estimates on rate of infection following colonization with multidrug-resistant gram-negative bacteria. Methods We performed an umbrella review of systematic reviews and meta-analyses. Quality was assessed using the AMSTAR 2 tool, and a meta-analysis was performed to estimate rate of infection. Results An initial search for systematic reviews and meta-analyses yielded 847 results, with 17 articles ultimately included. After exclusion of 2 studies for overlapping results and very low quality, the pooled incidence of infection following colonization across the studies was 22% for ESBL-E and 22% for CRE. Few reviews included high-quality findings on mortality or transmission following colonization. Additionally, only a limited number of reviews included findings related to MDR Pseudomonas aeruginosa or carbapenem-resistant Acinetobacter baumannii. Conclusions Our results suggest a substantial rate of infection following colonization of multidrug-resistant gram-negative bacteria. These findings can inform individual patient counseling, future decolonization innovation, clinical trial design, and regulatory approval of new decolonization agents. However, the heterogeneity of the included populations may limit the generalizability of these findings.
Addiction Science & Clinical Practice · 2025-02-05
articleOpen accessBACKGROUND: Unhealthy alcohol use, a spectrum of use inclusive of risky consumption and alcohol use disorder (AUD), is a leading cause of preventable death in the United States. Most people with unhealthy alcohol use do not receive evidence-based treatment. This four-arm factorial design randomized trial will assess whether population health management (PHM) and clinical care management (CCM) support for primary care providers (PCPs) are associated with improved AUD treatment engagement among their patients, beyond electronic health record (EHR) prompting and decision support alone. METHODS: PCPs from an urban safety-net hospital-based primary care clinic are randomized to one of four groups (1) EHR best practice advisory (BPA) and clinical decision support tools for unhealthy alcohol use (BPA), (2) BPA plus population health manager support, (3) BPA plus clinical care manager support, and (4) all three. All PCPs will have access to the EHR BPA and decision support tools which provide chart-based advisories and order set navigation. PCPs assigned to receive PHM support will receive quarterly panel-level feedback on AUD treatment metrics for their patients. PCPs assigned to receive CCM support will receive CCM facilitation of AUD treatment processes including medication counseling, referrals, and support through direct patient interactions. The primary outcome will be the percent of patients engaged in AUD treatment among those with a new AUD diagnosis on a PCP's panel. Secondary outcomes include the percent of patients with a new diagnosis of AUD who (1) initiated AUD treatment, (2) were prescribed AUD medications within 90 days, and (3) numerical counts of a range of AUD health services (outpatient encounters, specialty AUD care encounters, referrals, and acute healthcare utilization) in this sample. We will assess the primary outcome and the acute healthcare utilization secondary outcomes using Medicaid claims; the remaining secondary outcomes will be assessed using EHR data. DISCUSSION: The study will evaluate how a targeted EHR innovation alone, compared with population health and care management enhancements alone or in combination, impact engagement in AUD treatment, a national quality of care measure. Findings will advance understanding of supports needed to improve systems of care for AUD in general settings. TRIAL REGISTRATION: ClinicalTrials.gov identifier/registration number (NCT number): NCT05492942.
Secular Changes in Widespread Pain Across Two Generations: The Framingham Heart Study
Innovation in Aging · 2025-12-01
articleOpen accessAbstract Ten to eleven percent of adults have widespread musculoskeletal pain (WSP) with a higher prevalence in older persons. It is unknown whether WSP prevalence has changed nor whether increased obesity or longevity especially of women could explain this change. We studied these questions in 2 generations of the community-based Framingham Heart Study cohorts. We studied the original cohort of the Heart Study in 1992-1995 and Generation2, the offspring of the original cohort and their spouses in 2019-2021, at a similar age to the original cohort. A minoritized community sample (OMNI cohort) was also assessed in 2019-2021. Participants were asked whether they had joint pain on most days and, those saying yes used a homunculus to identify sites of pain. We then determined whether a participant met the WSP definition: pain on the left and right sides of the body; pain above and below the waist, and spinal or low back pain. We tallied the prevalence of WSP in each cohort, adjusting for confounders using logistic regression to test for cohort differences. The mean age of participants, mostly women, was in the 70s (table 1). Just over 14% of both original and OMNI cohorts reported WSP but 19-20% of Generation2 had WSP. Comparing Generation2 to the Original Cohort, WSP increased in prevalence (adjOR = 1.97 (95% CI: 1.49 - 2.61, p <.001). The prevalence of WSP has risen in older persons, an increase not explained by increasing pain with age, increased longevity of women nor increasing BMI.
ACR Open Rheumatology · 2025-07-29 · 1 citations
articleOpen accessOBJECTIVE: To assess the feasibility of a randomized controlled trial evaluating a gait retraining program to reduce peak tibial acceleration on knee pain and impact loading in adults with knee osteoarthritis. METHODS: Participants (n = 44) were randomized to a gait retraining or standard walking program. Walking duration increased from 10 to 30 minutes as feedback faded over eight sessions. Gait retraining participants received real-time biofeedback to reduce peak tibial acceleration by 20%. Feasibility criteria included rates of recruitment, enrollment, and retention and number of adverse events. Knee pain and overground impact loading were assessed at baseline and one week after the last treadmill session. Analysis of covariance models compared group differences in peak tibial acceleration, pain, and impact loading after the walking program, controlling for baseline values. RESULTS: Most feasibility criteria were met. From 2019 to 2023, 867 individuals were screened (~22 individuals per month), and 46 were enrolled and randomized (n = 23 per group). No adverse events were identified. Peak tibial acceleration reduced by 0.13g and 0.09g (gravitational equivalents) for the gait retraining and standard walking groups, respectively. Greater reductions in pain were observed for the standard walking group compared to the gait retraining group. Changes in impact loading were not significant in either group. No between-group differences were observed for peak tibial acceleration, knee pain, or impact loading. CONCLUSION: A full-scale randomized clinical trial is feasible with modification. However, gait retraining to reduce peak tibial acceleration was no more effective than a standard walking program for reducing knee pain and impact loading.
Osteoarthritis Imaging · 2025-11-01 · 1 citations
articleOpen accessTo evaluate whether a composite outcome measure could better detect the effects of adverse patellofemoral morphology on knee structure over two-years than single outcome measures. We used data from the Multicenter Osteoarthritis Study (MOST) to analyze the association between measures of patellofemoral morphology and composite outcomes that combined data on cartilage damage and bone marrow lesion enlargement in two subregions of the patellofemoral joint using proportional odds regression to account for ordinality. We used Z scores to assess sensitivity to change. In the cohort of 240 MOST participants with a mean age of 51 years, BMI of 29 kg/m 2 we found that for tibial tubercle to trochlear groove distance (TT-TG), entry point to trochlear groove angle (EPTG), and entry point to transition point angle (EPTP) cartilage worsening as a single measurement outperformed composite outcomes with higher Z scores. Alternatively, we found that patella tilt angle (PTA) performed best when predicting bone marrow lesion worsening as a single measurement than compared to the composite outcomes. For patellofemoral morphological measures, composite outcomes that combined data on cartilage damage and bone marrow lesion enlargement failed to increase sensitivity to change over outcomes for single structures alone. Other composite outcomes may be more successful in increasing the sensivitiy to change, and this warrants further investigation.
Journal of Racial and Ethnic Health Disparities · 2025-10-03
article
Recent grants
NIH · $29.4M · 2018
NIH · $10.5M · 2019–2029
Core B-Clinical Data Collection and Management Core
NIH · $79.4M · 2023–2028
NIH · $451k · 2020
Frequent coauthors
- 128 shared
David T. Felson
Boston University
- 38 shared
Alfred Mahr
- 36 shared
Tuhina Neogi
Boston University
- 34 shared
Michael C. Nevitt
University of California, San Francisco
- 30 shared
Yuqing Zhang
- 30 shared
Carla Maldini
- 29 shared
Jingbo Niu
Baylor College of Medicine
- 27 shared
Cora E. Lewis
University of Alabama at Birmingham
Labs
Awards & honors
- Professor Receives Distinguished Scholar Award for Rheumatol…
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Michael P Lavalley
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup