Tag: statistics

Kidney Blood Press Res. 2026 Mar 18:1-21. doi: 10.1159/000551589. Online ahead of print.

ABSTRACT

INTRODUCTION: The identification of non-diabetic kidney disease (NDKD) in diabetic patients is critically important. Unlike diabetic nephropathy, NDKD often requires additional therapeutic interventions beyond standard diabetes care. There is a need to develop computational methods using electronic medical record data to identify NDKD in diabetic patients for whom kidney biopsy is not an option.

METHODS: The study included 1136 diabetic patients who underwent kidney biopsy at a tertiary teaching hospital. We collected 103 parameters from electronic medical records, including demographic characteristics, physical examination results, laboratory tests, and the status of diabetic retinopathy. We developed seven models to detect NDKD, including k-nearest neighbors, random forest, extreme gradient boosting (XGB), lasso Logistic regression, support vector machine, naïve bayes, and multilayer perceptron (MLP), in the training set (n=908), and compared their performances in the testing set (n=228). The SHapley Additive exPlanations (SHAP) approach was used to analyze the importance of features.

RESULTS: Biopsy-confirmed NDKD was present in 53% of the 1136 participants. In the testing set, the area under the receiver operating characteristic curve (AUC) for NDKD detection using XGB, Lasso regression, and MLP reached 0.8, with performances that were stable regardless of whether variable normalization was performed. Among them, XGB revealed the highest AUC (0.833; 95% CI: 0.800 to 0.864) without feature normalization, which was statistically superior to the other models according to DeLong’s tests. After feature normalization, SVM achieved the highest AUC of 0.841 (95% CI: 0.817to 0.861) among all models. In addition to established predictive factors for NDKD (e.g., hematuria and absence of diabetic retinopathy), SHAP analysis identified several features, such as low IgG levels, that contributed significantly to the differentiation models.

CONCLUSION: Despite performance variations in different modeling techniques, machine learning models may have the potential to facilitate the detection of NDKD for patients with contraindications for kidney biopsy. Further efforts are warranted to improve accuracy and facilitate their translation into clinical practice.

PMID:41849636 | DOI:10.1159/000551589

JMIR Form Res. 2026 Mar 18;10:e83090. doi: 10.2196/83090.

ABSTRACT

BACKGROUND: Smartwatches have gained popularity for their potential to provide accurate measurements of various physiological parameters. However, the validity of energy expenditure (EE) across different smartwatch models remains a topic of ongoing investigation. Discrepancies between results obtained from different models and gold standard methods are particularly critical across varying exercise intensities and types, as validation studies have demonstrated overestimation when wearable activity monitors are compared with indirect calorimetry.

OBJECTIVE: This study investigated the accuracy of 2 versions of the Samsung smartwatch (Galaxy Watch [GW] 6 and 7) in measuring EE during intermittent moderate-intensity running exercises, using indirect calorimetry as the gold standard method.

METHODS: This study included 148 healthy adults, comprising 80 men and 68 women. Participants performed intermittent treadmill running, consisting of walking at 5 km·h⁻¹ for 1 minute and running between 8 and 16 km·h⁻¹ for 2 minutes, based on participant preference, for a total duration of 27 minutes. The GW6 and GW7 models were used and EE was measured by indirect calorimetry using a wearable portable metabolic gas analysis system (K5; Cosmed), which is considered a gold standard method.

RESULTS: No statistically significant differences were found between the GW models and the K5. The K5 showed a mean EE of 213.60 (SD 43.04) kilocalories, compared with 219.53 (SD 35.70) kilocalories for the GW6 and 202.67 (SD 47.42) kilocalories for the GW7 (all P>.05). Good Spearman correlations (0.63-0.70) and moderate intraclass correlation coefficients (0.65-0.74) were found. Mean absolute percentage error values ranged from 10.10% to 12.55%. Bland-Altman analysis revealed limits of agreement for all comparisons (K5 vs GW6 and GW7, -61.93 to 65.80 kcal).

CONCLUSIONS: The GW6 and GW7 devices showed moderate validity for estimating EE during intermittent running exercises, demonstrating the suitability of the GW as a low-cost and practical wearable option for daily physical activities.

PMID:41849635 | DOI:10.2196/83090

Cardiology. 2026 Mar 18:1-24. doi: 10.1159/000551373. Online ahead of print.

ABSTRACT

BACKGROUND: Heart failure (HF) and various arrhythmias frequently co-occur in clinical practice, suggesting shared pathophysiological mechanisms. However, the extent and nature of their common genetic architecture remains incompletely understood. This study aimed to systematically investigate the genetic correlations and shared causal loci between HF-related traits and multiple arrhythmia phenotypes.

METHODS: We utilized GWAS summary statistics from European cohorts to analyze HF-related traits and ten common arrhythmias. Global genetic correlations were assessed using LDSC and HDL. Local genetic correlations were further investigated using LAVA, HESS, and SUPERGNOVA to identify regional overlaps. Pleiotropic loci were identified using PLACO, with Bayesian colocalization analysis (stringent threshold PP.H4 ≥ 0.75) to assess shared causality. Bidirectional Mendelian randomization (MR) was conducted to explore causal relationships, utilizing a discovery threshold (P < 5×10⁻⁶) and a validation threshold (P < 5×10⁻⁸) with independent FinnGen data.

RESULTS: Significant genome-wide genetic correlations were identified between HF and seven arrhythmia traits, with the strongest association for atrial fibrillation (LDSC rg = 0.42, P = 5.1×10⁻¹⁸; HDL rg = 0.63, P = 5.9×10⁻³⁷). Local genetic correlation analyses identified multiple genomic regions of significant overlap, particularly converging on a major hotspot at the 4q25/PITX2/ENPEP locus across all three methods. Pleiotropic analysis identified several high-confidence shared loci, including regions harboring BAG3 (PP.H4 = 0.990) and ZFHX3 (PP.H4 = 0.938). Bidirectional MR revealed significant causal effects of AF on HF development (IVW OR = 1.22, P = 4.83×10⁻¹⁸) and HF on reduced heart rate variability (P = 1.86×10⁻⁴), both validated in independent cohorts.

CONCLUSIONS: Our findings demonstrate substantial and complex shared genetic architecture between HF and multiple arrhythmia phenotypes. These insights identify specific pleiotropic genes, regional correlation hotspots, and causal pathways, potentially informing future precision medicine approaches for cardiovascular disease prevention and treatment.

PMID:41849624 | DOI:10.1159/000551373

Health Rep. 2026 Mar 18;37(3):15-27. doi: 10.25318/82-003-x202600300002-eng.

ABSTRACT

BACKGROUND: The accurate monitoring of population mental health requires repeated assessments using valid and reliable measures. The Warwick-Edinburgh Mental Well-being Scale (WEMWBS) and its short form (SWEMWBS) are widely used positive mental health (PMH) measures ([S]WEMWBS is used hereafter to refer to both). This study tested their validity among Canadian adults using representative health survey data.

DATA AND METHODS: Cross-sectional data from the 2024 Canadian Community Health Survey – Rapid Response on Sleep Quality and Positive Mental Health of adults (18 years and older) living in the provinces were used. The distributions of (S)WEMWBS responses and scores were examined. Confirmatory factor analysis (CFA) and bifactor exploratory structural equation modelling (ESEM) were conducted to assess factorial validity. Measurement invariance was tested across gender and age. Differences in (S)WEMWBS scores by gender, age, and other mental health indicators were examined. Cronbach’s alphas were used to investigate internal consistency.

RESULTS: (S)WEMWBS scores had relatively normal distributions, with no floor and minimal ceiling effects. A bifactor ESEM and bifactor CFA model for the WEMWBS and SWEMWBS, respectively, fit the data best, with indices suggesting that they were essentially unidimensional. Evidence was found for measurement invariance across gender and age. Older adults had higher (S)WEMWBS scores on average, as did men on the WEMWBS. The (S)WEMWBS had acceptable internal consistency and were associated with other mental health indicators.

INTERPRETATION: The (S)WEMWBS appear to be valid and reliable PMH measures for Canadian adults. The (S)WEMWBS could be regularly included in health surveys to support the surveillance of population-level changes in PMH.

PMID:41849617 | DOI:10.25318/82-003-x202600300002-eng

JAMA Netw Open. 2026 Mar 2;9(3):e260762. doi: 10.1001/jamanetworkopen.2026.0762.

ABSTRACT

IMPORTANCE: Current guidance from the American Diabetes Association recommends liver stiffness measurement (LSM) only when the Fibrosis-4 (FIB-4) index is elevated. However, LSM may provide additional valuable information compared with FIB-4, which is known to underperform in certain populations, such as individuals with type 2 diabetes.

OBJECTIVE: To assess whether liver fibrosis evaluated by LSM is associated with increased mortality in individuals with and without diabetes.

DESIGN, SETTING, AND PARTICIPANTS: This cohort study included adult patients with complete vibration-controlled transient elastography (VCTE) and controlled attenuation parameter (CAP) data. Baseline data, including demographics and routine laboratory tests, were obtained from the 2017 to 2018 National Health and Nutrition Examination Survey and linked to data from the National Center for Health Statistics and National Death Index up to December 31, 2019. Patients with a history of liver disease other than metabolic dysfunction-associated steatotic liver disease were excluded. Data were analyzed from December 2024 to December 2025, accounting for the complex survey design using examination weights.

EXPOSURES: Liver disease based on CAP results and LSM by VCTE. CAP results 274 dB/m or higher indicate a diagnosis of metabolic dysfunction-associated steatotic liver disease (MASLD) and LSM results 9.7 kPa or higher indicate advanced liver fibrosis.

MAIN OUTCOMES AND MEASURES: All-cause mortality. Mortality risk was expressed as hazard ratios (HRs) with 95% CIs calculated using Cox proportional hazards regression models.

RESULTS: A total of 4102 adult patients (mean [SEM] age, 47 [1] years; 50.7% female), were included in the study. The mean (SEM) body mass index (BMI), calculated as weight in kilograms divided by height in meters squared, was 29.5 (0.3). Diabetes was present in 14.5% of participants. After a mean (SEM) follow-up of 24 (2) months, 59 patients (1.4%) had died. Patients who died during follow-up vs those who did not were older (mean [SEM] age, 62 [3] years vs 47 [1] years; P < .001), had a higher prevalence of diabetes (35.7% vs 14.2%; P = .01), and were more likely to be of non-Hispanic White race (85.1% vs 62.7%, P = .002). Increased risk of all-cause mortality was associated with the coexistence of diabetes with MASLD (adjusted hazard ratio [AHR], 2.77; 95% CI, 1.16-6.65; P = .03) and diabetes with advanced liver fibrosis (AHR 6.41; 95% CI, 1.03-39.85; P = .047). In patients with diabetes, LSM (AHR, 1.06; 95% CI, 1.04-1.09; P < .001) but not FIB-4 index remained associated with all-cause mortality even after adjusting for other clinical variables, such as age, sex, BMI, and hemoglobin A1c.

CONCLUSIONS AND RELEVANCE: In this cohort study, LSM was an independent risk factor for all-cause mortality in individuals with diabetes, even after a relatively short follow-up. Implementing LSM to screen for liver fibrosis as part of routine diabetes management could aid in early identification of patients with high mortality risk.

PMID:41848734 | DOI:10.1001/jamanetworkopen.2026.0762

JAMA Netw Open. 2026 Mar 2;9(3):e260908. doi: 10.1001/jamanetworkopen.2026.0908.

ABSTRACT

IMPORTANCE: Distinguishing primary lung squamous cell carcinoma (SCC) from squamous metastases to the lung is a clinical challenge due to histopathologic similarities. Accurate diagnosis is essential to guide treatment decisions.

OBJECTIVE: To assess the utility of an artificial intelligence (AI) approach that includes evaluation of key orthogonal evidence in distinguishing primary lung SCCs from metastatic tumors of other tissue origins.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used GPSai, a tissue-of-origin AI model run automatically on each sample submitted for molecular profiling, to flag potential misdiagnoses among research-eligible cases submitted as lung SCC. Molecularly profiled cases within the Caris Life Sciences clinicogenomic database from January 1, 2024, to January 31, 2025, were queried. All cases were reviewed by board-certified pathologists.

MAIN OUTCOMES AND MEASURES: The primary outcome was the rate of misdiagnosis among presumed lung SCCs confirmed by pathologist review and orthogonal evidence, which included clinical history and clinical findings, GATA3 and uroplakin II immunohistochemistry for urothelial carcinoma, UV variant signature for cutaneous SCC, CD5 and CD117 (c-KIT) immunohistochemistry for thymic carcinoma, and human papillomavirus positivity for orogenital SCC (eg, head and neck, cervical).

RESULTS: Through a combination of AI and orthogonal evidence, 123 (3.1%) misdiagnoses were confirmed among 3958 cases submitted as presumed lung SCC (patients misdiagnosed: median [range] age, 71 [39 to >89]; 76.4% male). The cohort included 50 cutaneous SCCs (40.7%), 33 orogenital SCCs (26.8%) (including 25 head and neck [75.8%]), 20 urothelial carcinomas (16.3%), 15 thymic carcinomas (12.2%), 4 NUT carcinomas (3.3%), and 1 prostate SCC (0.8%). Ninety-two of the 123 patients (74.8%) had clinical history or findings consistent with the new diagnosis. Eighty-eight cases (71.5%) had differences in guideline-preferred first-line systemic therapies following the diagnosis change.

CONCLUSIONS AND RELEVANCE: In this cross-sectional study of patients diagnosed with lung SCC, a meaningful number of patients experienced misdiagnosis, which was identified using a multipronged AI-assisted approach. Diagnosis changes prompted by AI and orthogonal evidence may assist clinicians in prognostication and therapy selection.

PMID:41848733 | DOI:10.1001/jamanetworkopen.2026.0908

JAMA Netw Open. 2026 Mar 2;9(3):e262201. doi: 10.1001/jamanetworkopen.2026.2201.

ABSTRACT

IMPORTANCE: Biologic manufacturer copay coupon programs reduce patient out-of-pocket spending but may undermine biosimilar competition. Whether biologic manufacturers deploy coupons strategically to facilitate shifting patients to reformulated drugs to extend market exclusivity (ie, product hopping), remains unclear.

OBJECTIVE: To examine coupon use and coupon value for biologics line extensions whose originators are at risk of biosimilar competition.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used data from national prescription drug claims from the IQVIA Formulary Impact Analyzer, October 1, 2017, through September 30, 2019, including 43 088 commercially insured patients with 382 550 claims across 47 biologic products. Analyses were conducted February to September 2025.

EXPOSURES: Candidate product hop status interacted with the time since originator approval (≥12 vs <12 years). Candidate product hops were defined as biologic line extensions approved between 2017 and 2019; comparators were biologics with earlier approval.

MAIN OUTCOME AND MEASURES: The primary outcome was manufacturer coupon use. Secondary outcomes included coupon amount, share of patient costs covered by coupons, and postbuydown out-of-pocket spending. Linear probability models with product line random effects estimated associations among candidate hop status, time since originator approval (≥12 vs <12 years), and coupon outcomes.

RESULTS: Among 43 088 patients (22 113 [51.3%] female; mean [SD] age, 47.6 [15.2] years), coupon use for candidate hops increased relative to comparators as originators approached 12 years since approval (change, 19.0 [95% CI, 1.7-36.4] percentage points; P = .03). For products with older originators, coupons for candidate hops offset 38.8 (95% CI, 28.0 to 49.7) percentage points more of patient cost-sharing than comparators (P < .001), and postbuydown costs were $124 (95% CI, -$183 to -$65) lower (P < .001). Candidate hops accounted for 11 122 of 382 550 claims (2.9%).

CONCLUSIONS AND RELEVANCE: In this cross-sectional study, manufacturer coupon use for candidate hop biologics diverged from comparators as originators neared exclusivity loss, suggesting coupons may facilitate product hopping. Regulators should consider monitoring coupon programs during biologic exclusivity transitions.

PMID:41848732 | DOI:10.1001/jamanetworkopen.2026.2201

JAMA Netw Open. 2026 Mar 2;9(3):e262279. doi: 10.1001/jamanetworkopen.2026.2279.

ABSTRACT

IMPORTANCE: A greater understanding of the factors associated with occupational distress for female physicians will allow organizations to more effectively realize the benefits of improving physician occupational well-being.

OBJECTIVE: To identify factors associated with occupational distress among female physicians.

DESIGN, SETTING, AND PARTICIPANTS: This survey study was conducted at 15 US academic medical institutions who participate in the Healthcare Professional Well-Being Academic Consortium. A cross-sectional, voluntary survey was administered to attending physicians from October 2019 to July 2021 to assess occupational well-being. Data were analyzed from April to September 2023.

EXPOSURE: Physician gender was used as the exposure variable.

MAIN OUTCOMES AND MEASURES: Burnout and professional fulfillment were measured using the Professional Fulfillment Index, which included subscales to assess leadership support, personal-organizational values alignment, control over schedule, electronic health record helpfulness, and self-valuation. A secondary data analysis was performed to describe markers of occupational well-being by gender. Mediation models were used to assess the degree to which contributing factors mediated the difference in occupational distress between female and male physicians.

RESULTS: The survey was completed by 19 088 of 37 511 attending physicians (response rate of 50.8%), among whom 16 731 self-identified as male or female and were included in the analysis (8197 female [43%] and 8534 male [45%]), and 2357 (12%) either did not report their gender or reported a gender other than male or female. Respondents’ ages ranged from younger than 29 to older than 60 years, with 40 to 49 years being the most common age bracket for both female (2788 participants [34%]) and male (2448 participants [29%]) physicians. Compared with male physicians, female physicians experienced more burnout (3338 of 8052 participants [42%] vs 2769 of 8404 participants [33%]; P < .001) and less professional fulfillment (2786 of 8133 participants [34%] vs 3852 of 8463 participants [46%]; P < .001). In mediation models including 5 factors associated with occupational well-being, the direct association between burnout and gender was no longer significant (B = 0.04; 95% CI, -0.01 to 0.08; P = .14), while the gender difference in professional fulfillment remained significant (B = -0.09; 95% CI, -0.14 to -0.04; P = .001). The indirect association of gender with burnout through self-valuation (B = 0.27; 95% CI, 0.24 to 0.29; P < .001) accounted for the majority (63%) of the total difference in burnout between female and male physicians.

CONCLUSIONS AND RELEVANCE: In this survey study of attending physicians, the gender difference in burnout was fully mediated by 5 factors associated with occupational well-being, with self-valuation being the largest mediator. These same factors did not fully explain the difference in professional fulfillment, suggesting that further research is needed.

PMID:41848730 | DOI:10.1001/jamanetworkopen.2026.2279

JAMA Netw Open. 2026 Mar 2;9(3):e262284. doi: 10.1001/jamanetworkopen.2026.2284.

ABSTRACT

IMPORTANCE: Social media is a dominant source of health information, including vaccine information, but little is known about which message characteristics audiences prefer when deciding what social media content to engage with and about how to measure these preferences.

OBJECTIVE: To quantify the attributes of existing social media vaccination posts that are associated with the preference for and confidence in vaccination-related content.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used a combination of the discrete choice experiment (DCE) and the Swiss tournament approaches to evaluate preferences, which were collected in an online survey between March and May 2024. Participants (aged ≥18 years) from a previous California-based COVID-19 study were recontacted through social media between January and August 2024. Participants completed the DCE and Swiss tournament test.

EXPOSURES: Participants viewed pairs of vaccination-related social media posts that varied systematically by attributes, such as messenger, source, tone, artwork type, age group of messenger, and topic.

MAIN OUTCOMES AND MEASURES: Participants selected which post they were most likely to engage with (like, share, or comment). Preferences were analyzed using a conditional logit model to estimate relative importance of post attributes.

RESULTS: Among 243 participants (mean [SD] age, 36.4 [12.4] years; 127 males [52.5%]), 153 (63.0%) reported favorable attitudes toward vaccines in general, 167 (68.7%) toward influenza vaccines, and 117 (48.2%) toward COVID-19 vaccines. Daily social media use was common (228 [93.8%]), and most participants preferred visual content, such as pictures (206 [84.8%]) or short videos (184 [75.7%]). In the DCE, social media vaccination posts from public health agencies such as California Department of Public Health (adjusted odds ratio [AOR], 1.80; 95% CI, 1.57-2.07) and University of California San Francisco (AOR, 1.67; 95% CI, 1.21-2.29) and posts depicting health care workers (AOR, 1.28; 95% CI, 1.12-1.47) or older adults (AOR, 1.48; 95% CI, 1.22-1.80) were more likely to be preferred. Humorous tone was associated with reduced preference (AOR, 0.45; 95% CI, 0.36-0.57), whereas COVID-19 (AOR, 2.35; 95% CI, 1.95-2.87) and influenza (AOR, 1.78; 95% CI, 1.54-2.06) topics were associated with increased likelihood for preference. Artwork type showed no significant association with preference.

CONCLUSIONS AND RELEVANCE: This cross-sectional study found that participants preferred social media-based vaccination posts that were factual, featured health care professionals, and were sourced from public health organizations. Understanding communication preferences through innovative experimental methods can inform the design of digital public health communications.

PMID:41848729 | DOI:10.1001/jamanetworkopen.2026.2284

JAMA Cardiol. 2026 Mar 18. doi: 10.1001/jamacardio.2026.0212. Online ahead of print.

ABSTRACT

IMPORTANCE: Premature onset of menopause is associated with increased short-term risk of coronary heart disease (CHD), but the associated long-term CHD risk and whether this differs by self-identified race are not known.

OBJECTIVE: To calculate lifetime risk estimates of incident CHD and to estimate years lived free of and with CHD by premature menopause status stratified by self-identified race.

DESIGN, SETTING, AND PARTICIPANTS: This prospective population-based cohort study was conducted with 163 600 person-years of follow-up, from 1964 to 2018. Individual-level data from postmenopausal women (aged 55-69 years) who self-identified their race as Black or White across 6 US cohorts were included. All participants were free of CHD at baseline and had data on menopausal status and CHD outcomes. Individuals who self-reported surgically induced menopause were excluded.

EXPOSURE: Premature onset of natural menopause (age <40 years).

MAIN OUTCOME AND MEASURES: The primary outcome was CHD (fatal and nonfatal myocardial infarction). The following analyses were performed: (1) modified Kaplan-Meier analysis to estimate lifetime risks, (2) adjusted competing Cox models to estimate joint cumulative risks for CHD or non-CHD death, and (3) Irwin restricted mean survival time to estimate mean years lived free of CHD and with CHD.

RESULTS: Of the 3522 Black women and 6514 White women included, mean (SD) age at baseline was 61.2 (4.3) years and 60.0 (4.4) years, respectively. Premature natural menopause occurred more frequently in Black women (545 [15.5%]) compared with White women (313 [4.8%]). Premature menopause was associated with a higher lifetime risk of incident CHD, with hazard ratios of 1.41 (95% CI, 1.04-1.90) for Black women and 1.39 (95% CI, 1.03-1.87) for White women. Mean years lived free of CHD were 18.2 years (95% CI, 17.5-18.9) for Black women with premature menopause compared to 19.1 years (95% CI, 18.8-19.4) for Black women without premature menopause; a similar pattern was seen in White women with and without premature menopause, but neither met statistical significance.

CONCLUSIONS AND RELEVANCE: In this cohort study, premature menopause was associated with 40% higher lifetime risk of CHD in Black and White women. This suggests that premature onset of menopause is an important risk-enhancing factor for lifetime risk and should be routinely assessed in clinical practice to consider intensification of preventive efforts.

PMID:41848694 | DOI:10.1001/jamacardio.2026.0212