Laura Pinheiro
Assistant Professor
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Articles (6)
Ultra-processed foods (UPFs) are associated with adverse health outcomes, but measurement of UPF intake in epidemiological studies remains challenging. Dietary assessments typically employ recall questionnaires, which may lead to measurement errors and misclassifications. Here, we develop a plasma metabolite index of UPF consumption to objectively measure UPF intake. Using two years of authentic food purchasing records from the ChooseWell 365 cohort, we identify metabolites associated with long-term UPF intake. We then apply the UPF-metabolite index to the REasons for Geographic and Racial Differences in Stroke (REGARDS) cohort, a population-based cohort of U.S. adults followed for incident health outcomes. The UPF-metabolite index was associated with leading causes of morbidity and mortality, including incident stroke, all-cause mortality, and cancer mortality. Mediation analysis revealed that the UPF-metabolite index accounts for 62% of the association between UPF intake and stroke risk. These findings suggest that the UPF-metabolite index offers an objective method for assessing UPF intake and its contribution to diet-related disease risk.
Year:
2026
There has perhaps never been more urgency for health services researchers, community partners, and policymakers to work together to address inequitable cancer outcomes in the United States. In this commentary, we first sought to provide a brief overview of how health disparities research has changed over the past 50 years from descriptive studies of downstream drivers of disparities to upstream factors that have downstream impacts on disparities. Then, we used health disparities across the breast cancer care continuum as a base case to describe the interaction of individual, community, and structural factors that drive cancer outcomes. The accompanying article led by Dr. Fowler and colleagues describes the critical role that neighborhood factors, which often occur upstream of observed health outcomes, play in breast cancer mortality disparities between Black and White women in Alabama. We agree with the authors’ suggestions for the health disparities research field to pursue multilevel strategies that leverage community-engaged research. By partnering with community-based organizations and policymakers, we propose that health services researchers, who are well positioned to provide rigorous and objective evaluations of social policies and programs and their impact on cancer outcomes, can play a critical role in achieving health equity. See related article by Fowler et al., p. 1740
Year:
2025
BACKGROUND As TikTok (ByteDance) grows as a major platform for health information, the quality and accuracy of Arabic-language cancer prevention content remain unknown. Limited access to culturally relevant and evidence-based information may exacerbate disparities in cancer knowledge and prevention behaviors. Although large language models offer scalable approaches for analyzing online health content, their utility for short-form video data, especially in underrepresented languages, has not been well established. OBJECTIVE We aimed to characterize and evaluate the quality of Arabic-language TikTok videos on cancer prevention and explore the use of large language models for scalable content analysis. METHODS We used the TikTok research application programming interface and a GPT-assisted keyword strategy to collect Arabic-language TikTok videos (2021-2024). From an initial collection of 1800 TikTok videos, 320 were eligible after preprocessing. Of these, the top 25% (N=30) most-viewed were analyzed and manually coded for content type, cancer type, uploader identity, tone and register, scientific citation, and disclaimers. Video quality was assessed using the Patient Education Materials Assessment Tool for Audiovisual Materials for understandability and actionability, and the Global Quality Scale (GQS). GPT-4 was used to generate artificial intelligence annotations, which were compared to human coding for select variables. RESULTS The top 25% (N=30) most-viewed videos amassed a total of 21.6 million views. Diet and alternative therapies were most common (15/30, 50%), which included recommendations to reduce hydrogenated oils, increase fruit and vegetable intake, and the use of traditional remedies such as garlic and black seed. Only 6.6% (2/30) of videos cited scientific literature. General cancer (15/30, 53%), breast (5/30, 17%), and cervical (4/30, 13%) cancers were most frequently mentioned. Doctors led 30% (9/30) of videos and were more likely to produce higher quality content, including significantly higher global quality scores (GQS=4, median 4, IQR 4-4 vs 3, median 3, IQR 2-3, P=.06). Over half of the videos had low understandability (16/30, 53%) and actionability (18/30, 60%). Emotionally framed content had the highest engagement across likes and shares, although this did not reach statistical significance (P=.08 and P=.05, respectively). However, emotional tone was significantly associated with lower GQS scores (P=.01). GPT-4 showed high agreement with human coders for cancer type (Cohen κ=1.0), strong agreement for GQS (κ=0.94), but low agreement for tone classification (κ=0.15), due to misclassification of emotional delivery from text-only input. CONCLUSIONS Arabic-language TikTok cancer prevention content is highly engaging but variable in quality, with emotionally framed videos attracting substantial attention despite lower informational value. Artificial intelligence-assisted tools show strong potential for scalable, multilingual health content analysis, but multimodal approaches are needed to accurately interpret tonal and audiovisual features.
Year:
2025
Introduction Black Americans have greater coronary heart disease (CHD) burden than White Americans, disparities that are largely socially determined. Discriminatory societal practices that systematically disadvantage Black Americans are forms of structural racism but few studies have examined structural racism and incident CHD. We sought to determine associations between three validated measures of structural racism and incident CHD, hypothesizing that greater state-level structural racism is associated with incident CHD for Black but not White individuals. Methods We used data from the national REasons for Geographic And Racial Differences in Stroke (REGARDS) cohort, which enrolled 30,239 Black and White community dwelling adults between 2003-7 who were contacted every 6 months with retrieval of medical records and expert adjudication of myocardial infarction and cause of death. Incident CHD was defined as myocardial infarction or death due to CHD. Structural racism variables included Black:White % living below the Federal poverty line, Black:White % uninsured, and the Dissimilarity Index (DI), a measure of residential racial segregation. Structural racism variables were dichotomized at the median. Separate race-stratified Cox proportional hazards models examined associations between each measure of structural racism and incident CHD. Results The 24,099 participants free of CHD at baseline included 10,286 Black and 13,813 White participants. Mean age at baseline was 64 years, 58% were women, and 47% had annual household income <$35,000. Greater structural racism was significantly associated with incident CHD for Black but not White participants. For high Black:White poverty, Black HR=1.17 (95% CI 1.01-1.35), White HR=0.93 (0.83-1.06); for high Black:White uninsurance, Black=HR 1.34 (1.06-1.70), White HR=1.20 (0.98-1.47); for high DI, Black HR=1.17 (1.01-1.35), White HR=0.99 (0.88-1.12). Findings suggest that structural racism variables indirectly influence CHD via individual-level income and education. Results were similar for men and women and for older and younger individuals. Significant associations were observed for fatal but not nonfatal CHD events. Conclusions Structural racism was associated with higher incidence of CHD for Black but not White individuals. If these associations are causal, changing state level laws to combat poverty in Black communities, expand Medicaid, and reduce segregation could potentially lessen Black:White disparities in CHD.
Year:
2024
Background Multimorbidity is increasingly prevalent in lower-and middle-income countries (LMICs). Health-related quality of life (HRQOL) has been inversely associated with multimorbidity but is understudied in LMICs. We report cardiovascular disease (CVD) multimorbidity in Haiti and its association with HRQOL. Methods We used data from the Haiti CVD Cohort, a population-based longitudinal cohort of adults. CVD multimorbidity was 2+ CVD risk factors/diseases at enrollment. HRQOL was measured using the Short Form-12, yielding physical (PCS)/mental (MCS) component summary scores between 0-100, with higher scores indicating better HRQOL. We used linear regressions to assess the association between CVD multimorbidity and HRQOL and between individual CVD comorbidities and HRQOL. Additionally, we examined how gender and education modified the main effect. Results Among 2,996 participants, the median age was 40 (IQR: 27-55), 58.0% were female, and 70.3% earned <1 USD/day. CVD multimorbidity prevalence was 24.1%; compared to those without CVD multimorbidity, those with CVD multimorbidity were older (median 56.0 [IQR: 47.0, 53.0]) and female (70.5%). Adjusted models revealed CVD multimorbidity was inversely related to PCS (−2.7 [95% CI: −3.8, −1.6]) and MCS (−1.0 [95% CI: −1.8, −0.2]). Heart failure and hypertension showed the strongest CVD morbidities associated with poor HRQOL. In the interaction analysis, among men, CVD multimorbidity was associated with a 4.3-point lower PCS score, and among those with less education, CVD multimorbidity was associated with a 4.6-point lower PCS score than no CVD multimorbidity. Conclusions Our data are among the first to describe HRQOL data with high CVD multimorbidity in a young population in urban Haiti, and CVD multimorbidity was associated with decreased HRQOL. ( https://www.clinicaltrials.gov/ct2/show/NCT03892265 , # NCT03892265 ) Research Perspective – This research raises the question of whether people with CVD multimorbidity in LMICs need specific interventions that are tailored to their high levels of comorbidity and if women and those with low socioeconomic status are at extra risk for experiencing the adverse effects of multimorbidity. – Future research should investigate cut-off points for describing a population average relevant to an LMIC.
Year:
2024
PURPOSE A metastatic breast cancer (mBC) diagnosis can affect physical and emotional well-being. However, racial and ethnic differences in receipt of outpatient psychosocial care and supportive care medications in adults with mBC are not well described. METHODS Adults with mBC were identified in the INSIGHT-Clinical Research Network, a database inclusive of >12 million patients receiving care across six New York City health systems. Outpatient psychosocial care was operationalized using Common Procedure Terminology codes for outpatient psychotherapy or counseling. Psychosocial/supportive care medications were defined using Rx Concept Unique Identifier codes. Associations between race/ethnicity and outpatient care and medication use were evaluated using logistic regression. RESULTS Among 5,429 adults in the analytic cohort, mean age was 61 years and <1% were male; 53.6% were non-Hispanic White (NHW), 21.4% non-Hispanic Black (NHB), 15.9% Hispanic, 6.1% Asian/Native Hawaiian/Pacific Islander (A/NH/PI), and 3% other or unknown. Overall, 4.1% had ≥one outpatient psychosocial care visit and 63.4% were prescribed ≥one medication. Adjusted for age, compared with NHW, Hispanic patients were more likely (odds ratio [OR], 2.14 [95% CI, 1.55 to 2.92]) and A/NH/PI patients less likely (OR, 0.35 [95% CI, 0.12 to 0.78]) to have an outpatient visit. NHB (OR, 0.59 [95% CI, 0.51 to 0.68]) and Asian (OR, 0.36 [95% CI, 0.29 to 0.46]) patients were less likely to be prescribed medications. CONCLUSION Despite the prevalence of depression, anxiety, and distress among patients with mBC, we observed low utilization of psychosocial outpatient care. Supportive medication use was more prevalent, although differences observed by race/ethnicity suggest that unmet needs exist.
Year:
2024
Collaborators (13)
Erica Phillips
Professor of Clinical Medicine
Weill Cornell Medicine
Safa Elkefi
Project Director
Stevens Institute of Technology
Katherine Reeder-Hayes
Professor with Tenure
University of North Carolina at Chapel Hill
Dawn L. Hershman
Columbia University Irving Medical Center
Monika M. Safford
Weill Cornell Medicine
W. Taylor Kimberly
-
Lily Yan
Assistant Professor
Weill Cornell Medicine
Anne Thorndike
Massachusetts General Hospital
Ana-Lucia Garcia-Guarniz
Assistant Professor in Neurology
Rhode Island Hospital
Perla Chebli
Assistant Professor
NYU School of Medicine
Suzanne Judd
University of Alabama, Tuscaloosa
Madeline Sterling
Weill Cornell Medical College
Rehana Rasul
City University of New York School of Public Health
Make your next move.
Prepare your application
Review drafts before you send them.