Tag: pubmed

JAMA Netw Open. 2026 Jul 1;9(7):e2623265. doi: 10.1001/jamanetworkopen.2026.23265.

ABSTRACT

IMPORTANCE: Relapse limits survival after allogeneic hematopoietic cell transplant. Although human leukocyte antigen-matched related donors (MRDs) are traditionally preferred, mismatched unrelated donors (MMUDs) using posttransplant cyclophosphamide may provide stronger graft-vs-leukemia effects, potentially altering donor selection strategies.

OBJECTIVE: To compare outcomes between MRD and MMUD transplant in the posttransplant cyclophosphamide era and assess whether MMUD grafts are associated with lower relapse rates.

DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study (January 1, 2017, to August 31, 2024) assessed 1038 patients with acute leukemia or myelodysplastic syndromes and/or myeloproliferative neoplasms undergoing their first peripheral blood hematopoietic cell transplant using posttransplant cyclophosphamide. The MRD cohort was from a single tertiary center, and the MMUD cohort was from the Center for International Blood and Marrow Transplant Research. Data were analyzed between January 19, 2026, and April 2, 2026.

EXPOSURE: Transplant from MRD (n = 306) vs MMUD (n = 732).

MAIN OUTCOMES AND MEASURES: Disease-free survival (DFS), overall survival (OS), relapse, nonrelapse mortality, and graft-vs-host disease (GVHD). The study used inverse probability of treatment weighting with overlap weights and bootstrapping to address structural confounding, particularly donor age. Cox proportional hazard models were used to compare outcomes between groups and confirm that phrasing with the author.

RESULTS: Among 1038 patients (526 [50.7%] female), donors in the MRD group were older (median [IQR] age, 57 [45-64] years) than in the MMUD group (median [IQR] age, 28 [24-35] years) (P < .001). Recipient age was similar (median [IQR] age, 60 [47-66] vs 58 [47-65] years; P = .40). In patients with a high or very high Disease Risk Index (DRI), MMUD transplant was associated with significantly lower relapse hazard vs MRD (weighted hazard ratio [HR], 0.56; 95% CI, 0.33-0.94; P = .03), although DFS (HR, 0.71; 95% CI, 0.46-1.11; P = .14) and OS (HR, 0.85; 95% CI, 0.53-1.37; P = .51) were similar. Conversely, in patients with low to intermediate DRIs, hazard estimates for DFS (HR, 1.24; 95% CI, 0.92-1.66; P = .16) and OS (HR, 1.36; 95% CI, 0.99-1.88; P = .06) lacked precision. MMUD was associated with lower risk of grade III to IV acute GVHD (HR, 0.56; 95% CI, 0.32-0.98; P = .04) but higher chronic GVHD (HR, 2.82; 95% CI, 2.02-3.95; P < .001).

CONCLUSIONS AND RELEVANCE: In this cohort study, MMUD transplant was associated with lower relapse rates compared with MRD in patients with high-risk malignant tumors, potentially reflecting enhanced alloreactivity, although accompanied by increased chronic GVHD. These findings indicated that the immunologic role of human leukocyte antigen mismatch varies by disease risk, suggesting that MMUD grafts offered a distinct risk-benefit profile that warrants further investigation.

PMID:42475099 | DOI:10.1001/jamanetworkopen.2026.23265

JAMA Netw Open. 2026 Jul 1;9(7):e2623622. doi: 10.1001/jamanetworkopen.2026.23622.

ABSTRACT

IMPORTANCE: Radical cystectomy for bladder cancer is associated with high complication rates. Evidence supporting perioperative specialized immunonutrition for reducing complications is limited.

OBJECTIVE: To determine whether perioperative specialized immunonutrition reduces 30-day complications compared with standard oral nutrition support among adults undergoing radical cystectomy.

DESIGN, SETTING, AND PARTICIPANTS: This multicenter, randomized, double-blind, phase 3 clinical trial (SWOG S1600) conducted at 13 US academic and community sites within the National Cancer Institute National Clinical Trials Network and National Community Oncology Research Program enrolled patients from March 5, 2019, through October 19, 2023, with 30-day complications assessed through day 90 and survival outcomes through 2 years. Participants included adults with bladder cancer scheduled for radical cystectomy who could swallow oral liquids and did not have severe malnutrition. Statistical analysis was performed from January 2024 to December 2026.

INTERVENTIONS: Participants consumed drinks 3 times daily for 5 days before and after surgery. Both drinks used the same base formula. The immunonutrition product contained added l-arginine, ω-3 fatty acids, and dietary nucleotides. Adherence was 94.5% (52 of 55), measured using the plasma arachidonic acid to eicosapentaenoic acid biomarker.

MAIN OUTCOMES AND MEASURES: Analysis was performed on a modified intent-to-treat basis. The primary outcome was any 30-day complication (Clavien-Dindo grade ≥1). A 65% complication rate and a 35% relative reduction were prespecified. Secondary outcomes included high-grade complications at 30 and 90 days, adverse events, and 2-year disease-free and overall survival.

RESULTS: Of the 203 participants enrolled, the median age was 68.8 years (range 32.0-95.2 years) and 162 (79.8%) were male. Of 203 randomized (99 immunonutrition, 104 standard), 17 withdrew, and 8 lacked 30-day data, yielding 178 evaluable participants (90 immunonutrition, 88 standard). Any 30-day complication occurred in 62.2% of patients (56 of 90) receiving immunonutrition and 58.0% of patients (51 of 88) receiving standard nutrition (odds ratio [OR], 1.18; 95% CI, 0.64-2.18). High-grade 30-day complications occurred in 11.1% of patients (10 of 90) receiving immunonutrition vs 12.5% of patients (11 of 88) receiving standard nutrition (OR, 0.86; 95% CI, 0.34-2.17). Two-year overall survival was 87.4% among patients receiving immunonutrition vs 78.2% among patients receiving standard nutrition, and disease-free survival was 77.0% among patients receiving immunonutrition vs 67.5% among patients receiving standard nutrition. Adverse events occurred for 34.1% of patients (31 of 91) receiving immunonutrition and 20.5% of patients (18 of 88) receiving standard nutrition, with nausea most common.

CONCLUSIONS AND RELEVANCE: In this randomized clinical trial of adults undergoing radical cystectomy, perioperative specialized immunonutrition did not reduce 30-day complications compared with standard oral nutrition. Differences in 2-year survival favored immunonutrition but were not statistically significant. Further mechanistic evaluation and subgroup analyses are warranted.

TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03757949.

PMID:42475098 | DOI:10.1001/jamanetworkopen.2026.23622

JAMA Netw Open. 2026 Jul 1;9(7):e2623952. doi: 10.1001/jamanetworkopen.2026.23952.

ABSTRACT

IMPORTANCE: The Patient Health Questionnaire-9 (PHQ-9) is widely used to assess depression, and the General Anxiety Disorder-7 (GAD-7) is widely used to assess anxiety, yet the extent to which subtle differences in instruction delivery influence symptom scores is poorly understood.

OBJECTIVE: To determine whether emphasizing instructional wording alters PHQ-9 and GAD-7 responses among US adults with depression or anxiety.

DESIGN, SETTING, AND PARTICIPANTS: This randomized clinical trial was conducted from April 17 to May 30, 2025. Participants included US adults with self-reported clinician-diagnosed or treated major depressive disorder or generalized anxiety disorder, recruited via online advertisements and assessed by telephone.

INTERVENTIONS: Participants completed the PHQ-9 and the GAD-7 twice within the same session. Control participants received standard instructions for both administrations. Intervention participants received standard instructions initially and, for the second administration, were reminded to attend to the wording of the questionnaire instructions, with emphasis placed on the word bothered.

MAIN OUTCOMES AND MEASURES: The primary outcome was changes in PHQ-9 and GAD-7 total scores. Secondary outcomes were a score reduction greater than 20% and categorical change in score severity (eg, moderate to mild).

RESULTS: Among 200 total participants (n = 100 in the intervention group and n = 100 in the control group; mean [SD] age, 41.5 [15.4] years; 146 female [73%]), those in the intervention group demonstrated greater score reductions than those in the control group on the PHQ-9 (difference in changes: -2.60 [95% CI, -3.30 to -1.91]; P < .001) and the GAD-7 (difference in changes: -2.63 [95% CI, -3.35 to -1.99]; P < .001). Intervention participants had higher odds of achieving a score reduction greater than 20% (PHQ-9: adjusted odds ratio [AOR], 14.29 [95% CI, 5.95-34.48], P < .001; GAD-7: AOR, 10.20 [95% CI, 5.08-20.41], P < .001) and categorical improvement (PHQ-9: AOR, 9.17 [95% CI, 4.35-19.61], P < .001; GAD-7: AOR, 7.14 [95% CI, 3.57-14.29], P < .001) compared with control participants.

CONCLUSIONS AND RELEVANCE: In this randomized clinical trial of instructional emphasis on PHQ-9 and GAD-7 scores among participants with anxiety or depression, findings suggest that variation in how questionnaire instructions are emphasized can influence PHQ-9 and GAD-7 scores, underscoring the need for standardized administration in clinical, research, and digital settings.

TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT06956378.

PMID:42475096 | DOI:10.1001/jamanetworkopen.2026.23952

JAMA Netw Open. 2026 Jul 1;9(7):e2624047. doi: 10.1001/jamanetworkopen.2026.24047.

ABSTRACT

IMPORTANCE: Burnout is highly prevalent among emergency medicine (EM) residents and is associated with negative outcomes for physician well-being and patient care. Emerging evidence suggests sexual minority trainees may be at increased risk, but differences within EM residency are not well characterized.

OBJECTIVE: To determine if burnout prevalence among EM residents differs by sexual orientation after adjustment for key confounders.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study included US resident physicians who took the 2024 American Board of Emergency Medicine In-Training Exam (ITE) postexamination survey and answered the question on sexual orientation.

EXPOSURE: Sexual minority status (heterosexual or straight and sexual minority [lesbian, gay, bisexual, queer or questioning, asexual, pansexual, or other sexual identity]), adjusted for age, gender, race and ethnicity, postgraduate training year, program length, and region.

MAIN OUTCOMES AND MEASURES: The primary outcome was burnout, assessed with an abbreviated 6-item Copenhagen Burnout Inventory: internal (personal or work-related) and external (patient-related) burnout. Prevalence ratios (PRs) were estimated for burnout stratified by sexual orientation, utilizing Poisson regression with robust standard errors.

RESULTS: Of 9478 residents who took the ITE, 7852 respondents (median [IQR] age, 30 [28-32] years; 4416 male [56%]; 3364 female [43%]; 56 nonbinary [1%]) were included, with 928 (12%) identifying as a sexual minority and 6924 (88%) identifying as heterosexual. Compared with heterosexual residents, sexual minority residents had higher prevalence of any burnout prevalence (431 of 803 residents [54%] vs 2786 of 5967 residents [47%]; P < .001) and internal burnout (311 of 802 residents [39%] vs 1902 of 5979 residents [32%]; P < .001). After adjustment, sexual minority status was associated with higher prevalence of internal burnout (PR, 1.12; 95% CI, 1.02-1.24) and any burnout (PR, 1.09; 95% CI, 1.01-1.17). Among sexual minority subgroups, bisexual residents had the highest adjusted prevalence of any burnout (PR, 1.17; 95% CI, 1.05-1.30), and queer residents had higher prevalence of internal burnout (PR, 1.28; 95% CI, 1.02-1.62).

CONCLUSIONS AND RELEVANCE: In this cross-sectional study of 7852 EM residents, those who identified as a sexual minority reported higher burnout, particularly internal burnout, compared with heterosexual peers. These findings indicate that targeted support for sexual minority trainees is needed to improve physician well-being, training experience, and patient care.

PMID:42475095 | DOI:10.1001/jamanetworkopen.2026.24047

JAMA Netw Open. 2026 Jul 1;9(7):e2624095. doi: 10.1001/jamanetworkopen.2026.24095.

ABSTRACT

IMPORTANCE: Food insecurity is associated with worse cardiovascular health. There is variation in state Supplemental Nutrition Assistance Program (SNAP) eligibility and administration policies that influence participation rates, but how such policies impact cardiovascular mortality is not known.

OBJECTIVE: To evaluate the association of SNAP-related policies with county-level cardiovascular mortality rates.

DESIGN, SETTING, AND PARTICIPANTS: In this cross-sectional study, using state-level SNAP policy data from the US Department of Agriculture, a SNAP policy index was created (range, 0-10), with higher values denoting adoption of policies associated with higher SNAP participation. Mortality data from the National Center for Health Statistics and population data from the Census Bureau were used to calculate county-level, age-adjusted, cardiovascular mortality rates for adults aged 20 years and older.

EXPOSURES: Changes in an index of state SNAP policies from 2006 to 2019.

MAIN OUTCOMES AND MEASURES: The primary outcome was county-level cardiovascular mortality rates. Using the g-computation procedure, a robust causal inference method, and a marginal structural model, the association between changes in the SNAP policy index and county-level cardiovascular mortality rates was estimated. Mortality rates were also estimated under different potential trajectories for the SNAP index.

RESULTS: Between 2006 and 2019, there were a total of 11 409 251 cardiovascular deaths among adults in 3127 US counties (5 713 305 deaths [50.1%] occurred among women; 9 266 123 deaths [81.2%] among adults aged ≥65 years). The median (IQR) SNAP policy index value increased from 3.7 (2.7-4.2) to 7.1 (5.6-7.5). After accounting for differences in demographic and economic factors, a 1-point absolute increase in the SNAP policy index was associated with a 1.02% (95% CI, 0.47%-1.60%; P < .001) lower county cardiovascular mortality rate. If all states had followed a high-generosity SNAP policy trajectory, compared with if they had followed a low-generosity trajectory, an estimated 46 123 (95% CI, 28 587-65 096) fewer adult cardiovascular deaths are estimated to have occurred in 2019.

CONCLUSIONS AND RELEVANCE: In this cross-sectional study of SNAP policies and cardiovascular mortality, an index of policies that can increase SNAP participation was associated with statistically significantly lower cardiovascular mortality rates. Adoption of such policies may play a role in improving cardiovascular health among lower-income adults.

PMID:42475094 | DOI:10.1001/jamanetworkopen.2026.24095

JAMA Netw Open. 2026 Jul 1;9(7):e2624114. doi: 10.1001/jamanetworkopen.2026.24114.

ABSTRACT

IMPORTANCE: Shifts in the demographic composition of US military veterans underscore the need for updated population-based estimates of suicidal thoughts and behaviors (STBs) to accurately characterize suicide risk in this population.

OBJECTIVE: To update prevalence estimates of and factors associated with STBs among US military veterans and to examine mental health treatment utilization among veterans with a history of STBs, stratified by use of the Veterans Health Administration as a primary source of health care.

DESIGN, SETTING, AND PARTICIPANTS: Population-based cross-sectional study of US military veterans drawn from the 2025 to 2026 National Health and Resilience in Veterans Study.

MAIN OUTCOMES AND MEASURES: Past 2-week and past-year suicidal ideation (SI), lifetime planning, and lifetime suicide attempt(s).

RESULTS: Among 2636 veterans (mean [SD] age, 62.4 [16.2] years; 2349 [88.5%] male), the prevalence of past 2-week SI was 13.4% and past-year SI was 14.0%, with past 2-week SI significantly higher in 2025 to 2026 than in 2019 to 2020 (13.4% vs 9.0%; z = 5.69; P < .001). In contrast, prevalence estimates of lifetime suicide planning (8.0% vs 7.3%; z = 1.06; P = .29) and suicide attempts (4.9% vs 3.9%; z = 1.97; P = .048) were similar to those observed in 2019 to 2020. After adjustment for sociodemographic and military characteristics, factors most strongly associated with past 2-week SI based on relative importance analyses included functional disability (odds ratio [OR], 1.10; 95% CI, 1.04-1.16), lifetime major depressive disorder (OR, 2.22; 95% CI, 1.63-3.02), and lifetime posttraumatic stress disorder (OR, 1.53; 95% CI, 1.09-2.13). Factors most strongly associated with past-year SI included functional disability (OR, 1.09; 95% CI, 1.07-1.11), lifetime major depressive disorder (OR, 2.33; 95% CI, 1.72-3.16), and lifetime posttraumatic stress disorder (OR, 1.62; 95% CI, 1.17-2.25). Factors most strongly associated with lifetime suicide planning included lifetime posttraumatic stress disorder (OR, 2.61; 95% CI, 1.82-3.76), lifetime major depressive disorder (OR, 2.61; 95% CI, 1.84-3.71), and younger age (OR, 2.11; 95% CI, 1.36-3.28). Factors most strongly associated with lifetime suicide attempts included greater adverse childhood experiences (OR, 1.39; 95% CI, 1.29-1.51), lifetime major depressive disorder (OR, 3.76; 95% CI, 2.31-6.10), and functional disability (OR, 1.04; 95% CI, 1.02-1.07).

CONCLUSIONS AND RELEVANCE: In this cross-sectional study of US military veterans, results suggest that SI was more prevalent in 2025 to 2026 relative to estimates from 2019 to 2020, whereas the prevalence of suicide planning and attempts remained stable. Younger and female veterans appear to carry disproportionate risk for STBs. Targeting key risk factors in vulnerable subpopulations may help mitigate suicide risk.

PMID:42475092 | DOI:10.1001/jamanetworkopen.2026.24114

JAMA Netw Open. 2026 Jul 1;9(7):e2624175. doi: 10.1001/jamanetworkopen.2026.24175.

ABSTRACT

IMPORTANCE: Previous studies suggest that spouses of individuals with dementia may be at increased risk of developing dementia, but evidence regarding subgroup variation and absolute risk remains limited.

OBJECTIVE: To examine whether spouses of individuals with dementia have an increased risk of developing dementia and to assess whether this association is modified by demographic, socioeconomic, and family context factors.

DESIGN, SETTING, AND PARTICIPANTS: This nationwide population-based cohort study used Taiwan’s National Health Insurance Research Database from 1998 to 2022, with follow-up through 2022. To address prior limitations of small sample size and limited subgroup evaluation, spouses of individuals with incident dementia were identified and matched 1 to 4 to unexposed controls by birth year, sex, income, and urbanicity. The analysis was conducted between July 15, 2025, and May 15, 2026.

EXPOSURE: Spousal exposure to incident dementia.

MAIN OUTCOMES AND MEASURES: Incident all-cause dementia, Alzheimer disease, and vascular dementia. Hazard ratios (HRs) and 5-year absolute risk differences were estimated using Cox regression and g-computation.

RESULTS: The matched cohort included 955 105 individuals, with 118 904 exposed women and 72 117 exposed men as well as 475 616 unexposed women and 288 468 unexposed men. The largest proportion were aged 65 to 74 years at cohort entry, for women (47 084 [39.6%]) and men (28 477 [39.5%]). Spouses of individuals with dementia had a higher risk of all-cause dementia than unexposed spouses. For women, the HR was 1.74 (95% CI, 1.70-1.78), with a 5-year adjusted risk difference of 2.75 percentage points (pp) (95% CI, 2.63-2.87 pp). For men, the HR was 1.69 (95% CI, 1.64-1.73), with a 5-year adjusted risk difference of 3.49 pp (95% CI, 3.30-3.69 pp). Similar patterns were observed in analyses of Alzheimer disease and vascular dementia. Relative risks were more prevalent at younger ages, whereas absolute risk differences increased with age. For all-cause dementia, the association was attenuated among individuals with higher income or more children. Associations were otherwise broadly similar across sex and urbanicity.

CONCLUSIONS AND RELEVANCE: In this cohort study of married individuals in Taiwan, spouses of individuals with dementia had a higher risk of dementia than unexposed spouses. The association varied across family and socioeconomic contexts, highlighting the relevance of household context in dementia risk assessment.

PMID:42475091 | DOI:10.1001/jamanetworkopen.2026.24175

J Magn Reson Imaging. 2026 Jul 20. doi: 10.1002/jmri.70459. Online ahead of print.

ABSTRACT

BACKGROUND: Progressive supranuclear palsy (PSP) is both an underdiagnosed and a frequently misdiagnosed disorder. To remedy these diagnostic limitations, MRI has been used to investigate brain morphology to differentiate PSP from Parkinson’s disease (PD) and multiple system atrophy (MSA). However, while nigrostriatal degeneration and tau aggregation are prominent in PSP, and result in iron accumulation and atrophy in the region, substantia nigra (SN) atrophy remains an underexplored diagnostic marker.

PURPOSE: To investigate the diagnostic utility of QSM-derived SN parameters for PSP.

STUDY TYPE: Retrospective.

POPULATION: 123 (59 Males/64 Females) PD patients, 48 (26 M/22 F) MSA patients, and 22 (11 M/11 F) PSP patients were included in the main dataset. MSA patients include 20 parkinsonian type (MSA-P) (11 M/9 F) and 18 cerebellar type (MSA-C) (9 M/9 F) of MSA with 10 undetermined subtype. The external validation set included 12 (6 M/6 F) healthy controls, 13 PD (7 M/6 F), and 10 PSP (6 M/4 F) patients.

FIELD STRENGTH AND SEQUENCE: 3 T, MPRAGE T₁-weighted imaging and multi-echo gradient echo imaging (mGRE) for QSM.

ASSESSMENT: Group level differences of SN volume, magnetic susceptibility, and susceptibility-to-volume ratio (SVR) among PSP, PD, and MSA groups, and these metrics’ differentiating power are measured. Bivariate logistic regression with T₁-based morphological markers alongside SN metrics is performed.

STATISTICAL TESTS: Mann-Whitney U test for group comparisons. Receiver operating characteristic analysis with bootstrapping. p < 0.05 after Bonferroni correction is defined as statistically significant result.

RESULTS: The SN SVR was significantly higher in the PSP group compared to the MSA group and the MSA-P group. Moreover, using pons measurements with SN SVR in a bivariate model resulted in the highest differentiation between the PSP and MSA groups (AUC = 0.88, 95% CI: [0.78-0.95]), particularly driven by the increased differentiation between the PSP and MSA-P groups (AUC = 0.88, 95% CI: [0.75-0.98]). External validation supported the generalizability of SN SVR, yielding 100% sensitivity and 80% specificity for differentiating PSP from HC and PD.

DATA CONCLUSION: QSM-based SN morphometry can complement T₁-weighted imaging for Parkinsonism assessment.

EVIDENCE LEVEL: 3.

TECHNICAL EFFICACY: Stage 2.

PMID:42475080 | DOI:10.1002/jmri.70459

Pain. 2026 Jul 20. doi: 10.1097/j.pain.0000000000004064. Online ahead of print.

ABSTRACT

Complex regional pain syndrome (CRPS) is a chronic, treatment-refractory neuropathic disorder associated with substantial functional impairment and socioeconomic burden. Although socioeconomic conditions have been implicated in vulnerability to chronic pain, evidence on the association between income dynamics and the risk of developing CRPS remains limited. Using the Korean National Health Insurance Service database, we conducted a nationwide cohort study including 3,437,021 adults in South Korea who underwent national health screening in 2012 and were followed through to 2022. Income-related indicators-including baseline income level, cumulative income exposure, income change, and income volatility during the 5 years preceding baseline-were assessed. Incident CRPS was identified using Korean Standard Classification of Diseases codes combined with registration in the Rare Intractable Disease program. Compared with the highest income quartile, individuals in the lowest income quartile had a higher CRPS risk (adjusted hazard ratio [aHR], 3.74; 95% confidence interval [CI], 3.07-4.56). Risk increased with longer duration of low-income status (aHR, 1.91; 95% CI, 1.48-2.45 for 5 vs 0 years) and decreased with longer duration of high-income status (aHR, 0.55; 95% CI, 0.44-0.70 for 5 vs 0 years). In addition, high-income volatility (highest vs lowest quartile aHR, 1.48; 95% CI, 1.23-1.78) and downward income changes (peak aHR, 1.84; 95% CI, 1.24-2.73) were independently associated with increased CRPS risk. These findings suggest that income dynamics represent an important socioeconomic risk factor for CRPS and highlight the relevance of longitudinal socioeconomic vulnerability in the development of disabling chronic pain conditions.

PMID:42475049 | DOI:10.1097/j.pain.0000000000004064

Psychol Methods. 2026 Jul 20. doi: 10.1037/met0000860. Online ahead of print.

ABSTRACT

The ability of a student can be conceptualized as either a continuously varying entity (e.g., conventional analysis using dichotomous item response theory [IRT] models; Lord & Novick, 1968) or a bundle of latent classes (e.g., cognitive diagnostic models [CDMs]; Rupp et al., 2010; von Davier & Lee, 2019). This article builds on recent efforts to focus on predictive differences between measurement models in an attempt to examine the degree to which such approaches, which utilize quite distinctive notions regarding the nature of ability, produce different predictions of response behavior in the real world. We first present two simulation studies in which data are generated from variants of CDMs that differ in sample size, attribute hierarchical structures, and attribute estimation methods. We illustrate that, as we would expect given that they were used to generate data, CDM-based predictions uniformly outperform those of IRT models. We then compare the performance of CDM- and IRT-based approaches across 11 empirical data sets previously analyzed using CDMs. Our findings indicate that overfitting is a pervasive issue across CDM-based predictions, particularly with the generalized deterministic inputs, noisy “and” gate model. Furthermore, in no case does the CDM show superior performance when using the maximum a posteriori estimator, and only six out of 11 data sets show improved model fit for CDMs over the two-parameter logistic model when using the marginal mastery probabilities estimator. Researchers and practitioners may need to balance the diagnostic appeal of CDMs with the fact that their complexity can come at the cost of predictive accuracy. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

PMID:42475042 | DOI:10.1037/met0000860