Abstract B028: Co-occurring social determinants of endometrial cancer disparities in All of Us
Bibliographic record
Abstract
Abstract Background: Endometrial cancer (EC) incidence is rising, especially among Black women and those under 50. Social determinants of health (SDoH), non-medical factors shaped by lived environments, contribute to EC disparities, yet are often studied in isolation, limiting our understanding of how real-world SDoH patterns influence EC risk. Methods: Using bipartite network modeling, a machine learning algorithm that autonomously identifies clusters of co-occurring exposures, we analyzed co-occurring SDoH indicators in a case-control EC cohort from the All of Us database (vs. 8). We hypothesized that Black women would be overrepresented in clusters marked by socioeconomic disadvantage and that these clusters would be associated with increased prevalence odds of EC, particularly early-onset EC (EC diagnosis at age < 50 years). Age-matched cases and controls (1:2 ratio) with complete SDoH data were weighted to address survey bias. Logistic regression assessed associations between cluster membership and EC prevalence. Results: Our study cohort (ncases = 530; ncontrols = 1,060) was composed of 1,394 (90.9% of cases; 86.0% of controls) White participants, 95 (3.8% of cases; 7.1% of controls) Black participants, and 101 (5.3% of cases, 6.9% of controls) participants of other races. Of 129 participants aged < 50 years, 43 (33%) had early onset EC based on diagnosis at age < 50 years. Five distinct SDoH clusters emerged with a biclustered modularity Q of 0.216 (p = 0.047), indicating modest but statistically significant community structure. Detected clusters ranged from Cluster 1—characterized by poor English language proficiency, Cluster 2— defined by high discrimination, delayed care, and limited affordability of care, Cluster 3— defined by inadequate healthcare coverage and poor neighborhood-level infrastructure, Cluster 4—characterized by socioeconomic hardship, and Cluster 5—reflecting low social support and poor neighborhood cohesion. Cluster 4 (socioeconomic hardship) and Cluster 5 (low social support) had the highest proportions of EC cases (40.7% and 35.6%, respectively; 28.2%-31.6% for clusters 1-3). Clusters 2 and 4 had the highest within-cluster proportions of Black patients (8.4% and 9.1%, respectively; 3.8%-5.6% for all other clusters). In race-adjusted models, Cluster 4 membership was significantly associated with increased odds of EC (OR = 1.54, 95% CI: 1.18–2.02), while Black race showed a protective association after adjusting for cluster membership (OR = 0.48, 95% CI: 0.29–0.79). Race-adjusted logistic regression revealed that Cluster 4 membership conferred over a fourfold increased odds of early-onset EC (OR = 4.11, 95% CI: 1.65–10.23). Conclusion: Co-occurring SDoH exposures, particularly socioeconomic disadvantage, are strongly associated with EC risk and early onset. After adjusting for SDoH cluster membership, Black individuals were less likely to be diagnosed with early-onset EC. These findings underscore the need for intersectional approaches to address EC disparities. Citation Format: Oyomoare Osazuwa-Peters, Drew Neish, Jesus Gonzalez. Bosquet, Rebecca Previs, Tomi Akinyemiju. Co-occurring social determinants of endometrial cancer disparities in All of Us [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: The Rise in Early-Onset Cancers—Knowledge Gaps and Research Opportunities; 2025 Dec 10-13; Montreal, QC, Canada. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(23_Suppl):Abstract nr B028.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".