2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Saturday, December 12, 2026
05:30 p.m. - 07:30 p.m.
MCCNO - Hall F
Access, Representation, Population Health, Technology and Procedures, Treatment Considerations, Combination Therapy, Drug Development, Supportive Care, Biological Therapies, Transfusion, Non-Biological Therapies, Devices, Pharmacology, Women’s Health, Adverse Events, Artificial Intelligence (AI), Computer Vision, Machine Learning, Biological Processes, Genomics, Immune Mechanism, Clinical Practice (Health Services and Quality), Patient Safety, Quality of Care, Systems-Based Hematology, Value, Diseases, Genetic Disorders, Hemoglobinopathies, Sickle Cell Disease, Sickle Cell Trait, Thalassemia, Infectious Diseases, Viral, Bacterial, Research, Translational Research, Clinical Research, Clinical Trials, Epidemiology, Health Outcomes Research, Real-World Evidence, Study Population, Human, Adult, Embryo/Fetus, Pediatric, Pregnant, Young Adult
PAPER

1175 Comparison of depression and anxiety in code-based versus genotype-defined sickle cell disease in the all of us research program

Yong Eun, MD, Kenneth Rivlin, MD, PhD2
1 Columbia University Vagelos College of Physicians and Surgeons, New York, New York, United States, NYC Health + Hospitals / Harlem, New York, New York, United States
2 Albert Einstein College of Medicine, Bronx, New York, United States, NYC Health + Hospitals / Jacobi, Bronx, New York, United States

Background: Sickle cell disease (SCD) is frequently complicated by depression and anxiety, and the National Alliance of Sickle Cell Centers sets annual mental health screening as a standard of care. Reported prevalence estimates vary widely. A systematic review of 36 studies linked that variation to the assessment tool, recruitment setting, and timing of assessment (Oudin Doglioni 2021). A prior All of Us analysis found that a code-based SCD definition had a positive predictive value of 0.31 against genotype (Linton 2025). Gaps remain in how the case definition used to identify SCD and the frequency of healthcare visits relate to psychiatric estimates. Using whole-genome HBB genotype, we compared coded depression and anxiety between genotype-defined SCD and genotype-defined sickle cell trait, then examined how a code-based definition and visit frequency related to the estimates.

Methods: We conducted a cross-sectional analysis of All of Us controlled-tier, version 8, adults with electronic health record (EHR) linkage and whole-genome sequencing. Genotype-defined SCD (n=269) was compared with genotype-defined sickle cell trait (n=5,274). Code-defined SCD was identified separately among sequenced participants (n=422) by 3 or more distinct SCD diagnosis dates, hydroxyurea, transfusion, or an inpatient SCD diagnosis. The 3-date criterion was adapted from a validated claims definition (Reeves 2014), with treatment and inpatient criteria added. The code- and genotype-defined SCD groups could overlap. Outcomes were depression, anxiety, and their composite, each requiring 2 or more dated occurrences. After 1:4 propensity-score matching on genetic ancestry, age, and sex, logistic models adjusted for the matching factors, socioeconomic measures, and area deprivation. An overlap-preserving person-level bootstrap estimated the difference between the code- and genotype-defined log odds. To evaluate the relation with care frequency, we repeated the code-based model with adjustment for the number of distinct visit-days. Counts below 20 were suppressed under All of Us policy.

Results: Coded depression or anxiety was more common in genotype-defined SCD than in sickle cell trait (odds ratio [OR] 1.97, 95% confidence interval [CI] 1.46 to 2.66). The OR was 2.01 (95% CI 1.47 to 2.74) for depression and 2.38 (95% CI 1.73 to 3.28) for anxiety. Code-defined SCD produced a larger composite estimate (OR 2.83, 95% CI 2.26 to 3.55), exceeding the genotype-defined estimate by 0.37 on the log-odds scale (95% CI 0.19 to 0.58). After adjustment for visit-days, the code-based composite fell to OR 1.31 (95% CI 0.996 to 1.72). Depression fell to OR 1.17 (95% CI 0.89 to 1.54), while anxiety remained elevated at OR 1.45 (95% CI 1.08 to 1.94).

Conclusion: Coded depression and anxiety were more common in genotype-defined SCD, showing that the signal did not depend on diagnosis codes to define SCD. The larger estimate in code-defined SCD fell substantially after visit-day adjustment, consistent with greater detection during more frequent care. Anxiety retained a smaller elevation, whereas depression did not. Visit-days is a coarse measure that may also capture disease severity. Because only the code-based model was adjusted for visit-days, the adjusted estimates are a sensitivity analysis rather than fully harmonized comparisons. Psychiatric estimates from diagnosis-code-defined SCD cohorts should be interpreted with frequent healthcare contact in mind.

Yong Eun, MD: No relevant conflicts of interest to declare.