Chao Li
· Associate Professor of MathematicsVerifiedNew York University · Mathematics
Active 1998–2026
About
Chao Li is an Associate Professor at the Courant Institute of Mathematical Sciences at New York University. He holds a Ph.D. from Stanford University, where his dissertation advisors were Rick Schoen and Brian White. His academic journey includes an undergraduate degree in Mathematics from Peking University, mentored by Huijun Fan, and an instructional position at Princeton University prior to his current appointment. His research interests encompass differential geometry, partial differential equations, and geometric measure theory, with recent work focusing on minimal surfaces, scalar curvature, and mathematical general relativity. His scholarly contributions include numerous publications and collaborations on topics such as stable minimal hypersurfaces, scalar curvature deformation, and geometric analysis, supported in part by the National Science Foundation and a Sloan research fellowship.
Research topics
- Medicine
- Internal medicine
- Surgery
- Emergency medicine
- Environmental health
- Immunology
Selected publications
Metrika · 2026-04-02
articleSenior authorCorrespondingStatistical Methods in Medical Research · 2026-03-31
articleWe investigate efficient estimation strategies for partially linear transformation models with time-dependent covariates under interval censoring. The unknown monotone function is approximated using a monotone <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" overflow="scroll"> <mml:mi>B</mml:mi> </mml:math> -spline basis to enable flexible semiparametric modeling, and we develop a computationally efficient nested hybrid EM algorithm that integrates Newton’s method with isotonic regression. To support large-sample inference, we propose a straightforward variance–covariance estimation procedure for the regression parameters and introduce a score test to assess the adequacy of the proportional hazards (PH) specification within the broader class of transformation models. The numerical performance of the penalized estimators is examined extensively and compared with both the time-invariant covariate model by Lu et al. and the semiparametric transformation model by Zeng et al. Finally, the proposed methodology is applied to data from the National Alzheimer’s Coordinating Center (NACC) to demonstrate its practical utility in a real-world clinical setting.
2025-04-29
preprintOpen access<p dir="ltr">Objective: To determine whether cellular matrix products result in better healing rates than acellular matrix products for non-healing diabetic foot ulcers.</p><p dir="ltr">Research Design and Methods: Dermagraft® and Oasis® Longitudinal Comparative Efficacy Study (DOLCE) was a randomized, single-blinded, three-arm, controlled trial. Patients (aged ≥18 years) with a full thickness, nonhealing diabetic foot ulcer who met inclusion/exclusion criteria were enrolled.</p><p dir="ltr">Results: Of 169 eligible patients, 138 were enrolled and 117 randomized. For 12 weeks, patients received standard care, cellular matrix, or acellular matrix. The primary outcome was the percentage of wounds healed by 12 weeks. Of the 117 subjects, 41 were in the cellular matrix group, 48 in the acellular matrix group, and 28 in the standard care group. There were 21 withdrawals, but 7 had reached the first primary endpoint. Complete re-epithelialization of the ulcer by 12 weeks occurred in 59% of the 117 total subjects: 49% in the cellular matrix group, 69% in the acellular matrix group, and 57% in the standard of care group (Chi-square test, p = 0.16). At 28 weeks, the percentages were 25 (61%) in the CM group, 27 (56%) in the ACM group, and 18 (64%) in the SOC group healed (p = 0.78). No differences were found in wound recidivism or adverse event occurrence between groups.</p><p dir="ltr">Conclusions: No difference in efficacy was found between standard of care, acellular matrix, and cellular matrix, suggesting that standard of care can reduce the economic burden of diabetic foot ulcer treatment.</p>
Communication in Statistics- Theory and Methods · 2025-06-14
articleSenior authorCorrespondingJournal of Clinical Oncology · 2025-05-28
article12131 Background: Psychological distress is highly prevalent among cancer survivors, and it interferes with their ability to recover after treatment and resume normal life activities. Yoga is a promising therapy that may reduce psychological distress and facilitate optimal recovery for survivors. However, accessibility can be limited for survivors who have higher risks of infection due to immunosuppression or high travel burden. Virtual delivery of yoga may increase accessibility for survivors. Methods: We conducted a decentralized, digital, phase II randomized controlled trial (RCT) examining the efficacy of virtual yoga compared to usual care for improving psychological distress among survivors. Participants were cancer survivors who completed primary treatment (e.g., surgery, chemotherapy, radiation therapy) within the last 2-60 months. Participants were randomized to receive virtual yoga or usual care. The Zoom platform was used to virtually deliver the Yoga for Cancer Survivors (vYOCAS) intervention. vYOCAS is a 4-week intervention based on gentle Hatha and restorative yoga. Each yoga session was delivered by a certified yoga instructor in small groups (2-4 survivors/group) for 75 minutes, twice a week. Psychological distress was assessed via the Profile of Mood States (POMS) at baseline and post-intervention. POMS evaluated tension-anxiety, depression, anger-hostility, fatigue, confusion, and overall mood. T-tests and ANCOVAs with baseline as a covariate were used to evaluate within- and between-group changes, respectively. Results: 42 survivors (93% female; mean age 58.5±11.6 years; 60% breast cancer; 17% residing in small town/underserved areas) were randomized and completed the study. On average, participants attended 6.2 of 8 prescribed yoga sessions. 44% of vYOCAS participants reported additional home practice of 62.8 minutes over 4 weeks. vYOCAS participants reported significant decreases in psychological distress (tension-anxiety: -1.5±0.6; depression: -1.2±0.4; fatigue: -2.4±0.7; overall mood: -7.3±2.4; all p < 0.05) at post-intervention. Usual care participants did not demonstrate similar improvements. ANCOVA results also revealed that vYOCAS participants experienced significantly greater improvements in fatigue (-2.1±0.8, p = 0.02) and overall mood (-6.4±3.1, p = 0.04) compared to usual care participants. No intervention-related adverse events were reported and the majority of survivors would recommend virtual yoga to others. Conclusions: vYOCAS is safe, feasible, and amenable for cancer survivors. vYOCAS may also significantly improve psychological distress.Clinicians should consider recommending virtual yoga therapy for survivors with psychological distress to overcome barriers related to accessibility. Future phase III decentralized digital RCTs are needed to confirm these findings. Clinical trial information: NCT04458194 .
Penalized estimation of general frailty Poisson models for recurrent count events
Statistical Methods in Medical Research · 2025-12-02
articleSenior authorWe study spline-based efficient estimation of frailty models for panel count data using a penalization technique. An easy-to-implement and computationally efficient two-stage iterative expectation-maximization algorithm is proposed for the analysis. A general quasi-likelihood estimation that does not specify the stochastic model of the underlying counting process is developed to provide flexibility for model fitting. A powerful score test is discussed to detect the presence of overdispersion in count data. The proposed methods are assessed via an extensive simulation and further illustrated by analyzing data from a non-melanoma skin cancer chemoprevention study.
JNCI Cancer Spectrum · 2025-10-29 · 2 citations
articleOpen accessBACKGROUND: Cancer-related cognitive impairment is a significant concern, yet data assessing long-term changes from pretreatment baselines are limited. METHODS: In this large nationwide study, patients with breast cancer and lymphoma and controls were assessed at pre-chemotherapy, post-chemotherapy, 6-month, 1-year, and 2-year follow-ups. Cognitive function was measured using self-reported and performance-based assessments. The analysis included 225 participants: breast cancer (41 patients/36 controls) and lymphoma (73 patients/75 controls). Cognitive trajectories were estimated using longitudinal linear mixed models, adjusting for demographic, clinical, and psychosocial factors. A minimal clinically important difference cutoff identified changes over time in perceived cognitive impairment. RESULTS: Patients with breast cancer reported greater cognitive complaints than controls, with declines in FACT-Cog Total (β = -17.17, P < .001), Perceived Cognitive Impairment (β = -11.09, P < .001), and Perceived Cognitive Abilities (β = -3.60, P = .03) from pre-chemotherapy to 2-year follow-up. Declines were observed in memory (Rey Auditory Verbal Learning Test [RAVLT] Immediate Recall: β = -1.53, P = .04) and executive function (phone fluency: β = -1.53, P = .02) at 1 year. Patients with lymphoma also showed more complaints, with declines in Perceived Cognitive Impairment (β = -6.19, P = .01), poorer RAVLT Immediate (β = -1.62, P < .001), Delayed Recall (β = -1.72, P < .001), and phone fluency (β = -2.46, P < .001) from pre-chemotherapy to 1-year follow-up. CONCLUSION: Compared with controls, patients with breast cancer and lymphoma showed persistent cognitive worsening up to 1-year posttreatment, and patients with breast cancer reported perceived impairment at 2 years compared with controls.
Journal of Clinical Oncology · 2025-05-28
article12132 Background: Immune checkpoint inhibitors (ICIs) have transformed cancer treatment, extending patient survival. However, side effects such as endocrinopathies are common and have severe, sometimes irreversible outcomes if not managed promptly. Predictive factors for endocrinopathies are poorly understood. We examined whether demographic and clinical characteristics are linked with the development of ICI-induced endocrinopathies. Methods: The DiRECT Cohort (URCC21038, NCT05364086) is an ongoing observational trial of cancer patients scheduled to receive anti-PD-(L)1 ICI therapy and enrolled through the URCC NCORP Research Base nationwide network. This analysis was based on 1,525 patients with toxicity data assessed 12/31/2024. Endocrinopathies were graded using the Common Terminology Criteria for Adverse Events (CTCAE) criteria version 5.0. Toxicity data was collected after each infusion of ICI. Demographics (age, sex, BMI, race) and clinical factors (cancer type, cancer stage, treatment agent, autoimmune disease, and significant comorbidities) were collected at baseline. We tested bivariate associations with chi-square tests and multivariable associations with logistic regression; statistical significance was set at a p-value of 0.05. Results: Of the 1.525 participants, 533 (35%) had lung cancer, 263 (17.3%) had breast cancer, 812 (53.3%) were aged ≥ 65, 1142 (75.4%) White, 828 (54.4%) women, 1009 (66.3%) had BMI < 30, 802 (52.7%) had at least one comorbidity, and 128 (8.4%) had an autoimmune disease, 831 (55.2%) had stage IV cancer, and 930 (61.1%) were on pembrolizumab. 252 (16.5%) developed endocrinopathies of any grade: Hyperthyroidism, 61 (24.2%); Hypothyroidism, 156 (61.9%); Thyrotoxicosis, 7 (2.75%); and Adrenal insufficiency, 11 (4.37%). The most common of these were hypo/hyperthyroidism, of which 9% were grade ≥2. In bivariate analyses, grade >2 endocrinopathies were associated with younger age (11% age < 65 vs. 7% age ≥65, p = 0.012), female sex (10% in women vs. 4% in men, p = 0.001), and obesity (12% of those with BMI >30 vs 7% with BMI < 30, p = 0.001). We found no significant associations with other factors evaluated. In multivariable logistic regression analyses, younger age, female sex, and obesity were significant predictors, with higher odds of endocrinopathy for those of age < 65 (OR: 1.58, 95% CI: 1.35-1.85), female sex (OR: 1.75, 95% CI: 1.48-2.08), with BMI ≥30 (OR: 1.97, 95% CI: 1.68-2.31). Conclusions: In this nationwide observational trial, younger age, female sex, and obesity were associated with a higher likelihood of developing grade ≥2 ICI-induced endocrinopathies. Future work will focus on identifying biomarkers predictive of endocrinopathy and developing predictive models for risk stratification.
Diabetes Care · 2025-04-29 · 4 citations
articleOBJECTIVE: To determine whether cellular matrix (CM) products result in better healing rates than acellular matrix (ACM) products for nonhealing diabetic foot ulcers. RESEARCH DESIGN AND METHODS: The Dermagraft and Oasis Longitudinal Comparative Efficacy Study (DOLCE) was a randomized, single-blinded, three-arm controlled trial. Patients (aged ≥18 years) with a full-thickness nonhealing diabetic foot ulcer who met inclusion/exclusion criteria were enrolled. RESULTS: Of 169 eligible patients, 138 were enrolled and 117 randomly assigned. For 12 weeks, patients received standard of care (SOC), CM, or ACM. The primary outcome was the percentage of wounds healed by 12 weeks. Of the 117 participants, 41 were in the CM group, 48 in the ACM group, and 28 in the SOC group. There were 21 withdrawals, but seven had reached the first primary end point. Complete re-epithelialization of the ulcer by 12 weeks occurred in 59% of the 117 total participants: 49% in the CM group, 69% in the ACM group, and 57% in the SOC group (P = 0.16 by χ2 test). At 28 weeks, 25 participants (61%) in the CM group, 27 (56%) in the ACM group, and 18 (64%) in the SOC group had healed (P = 0.78). No differences were found in wound recidivism or adverse event occurrence between groups. CONCLUSIONS: No difference in efficacy was found between SOC, ACM, and CM, suggesting that SOC can reduce the economic burden of diabetic foot ulcer treatment.
Journal of Clinical Oncology · 2025-05-28 · 1 citations
article556 Background: Cancer-related fatigue (CRF) is reported by ~75% of patients receiving chemotherapy for breast cancer. CRF has been linked to inflammation. Chronic, low grade systemic inflammation is a polygenic trait, and a polygenic risk score for inflammation (iPRS) might be associated with risk of CRF. Methods: Using data from the UK Biobank,we developed an iPRS using the INFLA-score, a composite measure of serum C-reactive protein, white-cell count, platelet count, and neutrophil-lymphocyte ratio. The iPRS was evaluated for association with CRF among women with non-metastatic breast cancer enrolled in one of two completed multi-site clinical trials of the University of Rochester Cancer Center NCI Community Oncology Research Program (NCORP) Research Base. CRF was measured before and after standard-of-care chemotherapy using the Multidimensional Fatigue Symptom Inventory-Short Form (MFSI-SF). Linear regression evaluated the change in MFSI-SF score from pre- to post-chemotherapy; logistic regression evaluated a binary outcome of any vs no worsening of scores. Analyses were adjusted for patient and treatment factors. Results: The NCORP cohort included 802 women who received chemotherapy (anthracycline-based = 51.8%; previous surgery = 85.0%) at a median age of 55 years (range = 22 to 81). There was an increase in MFSI in 55% of the women, with a mean increase of 8 (range=-64 to 71) from prechemotherapy (mean=8, range = -24 to 83) to post-chemotherapy (mean=15.3, range = -24 - 88), indicating an overall increase in CRF. The iPRS was associated with a significant decrease in MFSI-SF (β=-3.29; 95%CI=-6.25 to -0.34; P =0.029; covariate-adjusted β=-2.71; 95%CI=-5.50 to 0.08; P =0.057) and lower odds of worsening CRF (OR=0.66; 95%CI=0.47-0.93; P =0.016; covariate-adjusted OR=0.67; 95%CI=0.47 to 0.96; P =0.029). The negative relationship between the iPRS and change in CRF was partially explained by the finding that women with an iPRS in the highest quartile have worse pre-chemotherapy MFSI-SF scores (β=4.33; 95%CI=0.23 to 8.43; P =0.038). Conclusions: Women with genetic predisposition to low-grade systemic inflammation, indicated by a higher iPRS, have worse CRF pre-chemotherapy that does not worsen, and may improve, over the course of treatment while women with a lower iPRS have less CRF pre-chemotherapy and are at greatest risk of developing new or worsening CRF during treatment. If validated, the iPRS could identify patients in need of supportive care interventions to reduce CRF. This work was supported by the National Institutes of Health National Cancer Institute Contract No. HHSN261201500003I, Task Order No. HHSN261000039 and by UG1CA189961, URCC NCORP Research Base.
Frequent coauthors
- 39 shared
Melissa N. Loja
- 36 shared
Todd E. Rasmussen
Mayo Clinic in Arizona
- 36 shared
Thomas M. Scalea
University of Maryland, Baltimore
- 36 shared
Joseph J. DuBose
- 36 shared
John B. Holcomb
University of Alabama at Birmingham
- 36 shared
M. Margaret Knudson
University of California, San Francisco
- 36 shared
Yu Liu
Jinan University
- 36 shared
Amanda Sammann
University of California, San Francisco
Labs
Education
B.S., Mathematics
Peking University
Ph.D.
Stanford University
Awards & honors
- Sloan research fellowship
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Chao Li
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