Category: Statistics

Bioinformatics. 2026 Jan 22:btag032. doi: 10.1093/bioinformatics/btag032. Online ahead of print.

ABSTRACT

MOTIVATION: Genome-wide association studies (GWAS) at biobank scale are computationally intensive, especially for admixed populations requiring robust statistical models. SAIGE is a widely used method for generalized linear mixed-model GWAS but is limited by its CPU-based implementation, making phenome-wide association studies impractical for many research groups.

RESULTS: We developed SAIGE-GPU, a GPU-accelerated version of SAIGE that replaces CPU-intensive matrix operations with GPU-optimized kernels. The core innovation is distributing genetic relationship matrix calculations across GPUs and communication layers. Applied to 2,068 phenotypes from 635,969 participants in the Million Veteran Program (MVP), including diverse and admixed populations, SAIGE-GPU achieved a 5-fold speedup in mixed model fitting on supercomputing infrastructure and cloud platforms. We further optimized the variant association testing step through multi-core and multi-trait parallelization. Deployed on Google Cloud Platform and Azure, the method provided substantial cost and time savings.

AVAILABILITY: Source code and binaries are available for download at https://github.com/saigegit/SAIGE/tree/SAIGE-GPU-1.3.3. A code snapshot is archived at Zenodo for reproducibility (DOI: [10.5281/zenodo.17642591]). SAIGE-GPU is available in a containerized format for use across HPC and cloud environments and is implemented in R/C ++ and runs on Linux systems.

SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PMID:41572430 | DOI:10.1093/bioinformatics/btag032

Oncologist. 2026 Jan 22:oyag005. doi: 10.1093/oncolo/oyag005. Online ahead of print.

ABSTRACT

BACKGROUND: We assessed the clinical relevance of age and sex as risk factors for health-related quality of life (HRQoL) in patients with adult-type diffuse glioma.

MATERIALS AND METHODS: The CODAGLIO 2.0 database contains 16 randomized trials from 5,369 patients with glioma. Patients’ HRQoL was assessed using EORTC QLQ-C30 and QLQ-BN20 questionnaires. In 8 HRQoL scales, we compared mean HRQoL at baseline with the general population and evaluated factors associated with HRQoL over time using linear mixed models (LMMs). We used anchor-based minimally important difference to interpret clinically relevant changes.

RESULTS: We included 4,301 patients with baseline HRQoL followed until 3 months. Compared to the general population, patients with glioma at baseline had statistically and clinically relevant worse HRQoL, which was still evident after stratifying by age and sex groups. In LMMs, compared to patients aged ≤60 years, those >60 years had statistically significant associations with worse physical functioning: -2.40 (95% confidence interval (CI) -4.14 to -0.71), better social: 4.88 (2.68 to 7.30) and role: 3.79 (1.39 to 6.16) functioning, and less fatigue: -3.43 (-5.44 to -1.33) and pain: -4.56 (-6.18 to -2.93). Compared to men, women had statistically significant associations with worse physical and social functioning and more fatigue and pain. Associations between age, sex, and HRQoL were not clinically relevant. Performance status had clinically relevant associations in 5/8 scales.

CONCLUSION: Patients with glioma have clinically relevant worse HRQoL compared to the general population. There are statistically but not clinically significant associations between age, sex, and certain HRQoL scales.

PMID:41572423 | DOI:10.1093/oncolo/oyag005

J Cardiovasc Imaging. 2026 Jan 22;34(1):2. doi: 10.1186/s44348-025-00063-4.

ABSTRACT

BACKGROUND: Transthoracic echocardiography derived left ventricular ejection fraction (LVEF) is a cornerstone in heart failure risk prevention. However, the lower limits of normal LVEF remains imprecisely defined. We aimed to define normal LVEF ranges by sex, age group, and self-reported race/ethnicity using data from population-based echocardiographic studies.

METHODS: We systematically searched MEDLINE for studies published between January 1, 2000, and January 3, 2025, that reported the mean and standard deviation of LVEF measured by 2D or 3D echocardiography in healthy, community-based adult populations.

RESULTS: In 10 studies (n = 10,427; female sex, 48%), the pooled mean LVEF was 62.8% (95% confidence interval, 61.0%-64.7%), with estimated lower and upper normal limits of 51.8% and 73.2%, respectively. Women had higher mean LVEF (63.7%) than men (61.9%), with corresponding lower normal limits of 52.7% and 51.7%, respectively. LVEF was similar across age groups. Individuals of Asian origin had 2 to 3 percentage points higher LVEF than Black or White individuals, with lower normal limits of 54% for women and 53% for men. Fewer than 1% of women and approximately 1% of men would be expected to have an LVEF below 50%. Across all demographic subgroups, the probability that an LVEF < 50% is within the normal range was < 5%. There was significant heterogeneity of the included studies (e.g., τ² = 8.82, I² = 99.7% for overall analysis) that appeared unexplained by sex, age, or echocardiography modality (2D vs. 3D).

CONCLUSIONS: In healthy adults, the lower limit of normal LVEF is approximately 53% for women and 52% for men, with slightly higher thresholds among individuals of Asian origin. An LVEF < 50% is highly unlikely to reflect normal function, regardless of sex, age, or self-reported race/ethnicity. Given the high statistical heterogeneity, the results should be interpreted with caution.

PMID:41572417 | DOI:10.1186/s44348-025-00063-4

Br J Haematol. 2026 Jan 22. doi: 10.1111/bjh.70342. Online ahead of print.

ABSTRACT

Waldenström’s macroglobulinaemia (WM) is a rare B-cell lymphoproliferative disorder with multiple effective front-line treatment options. However, real-world comparative data on commonly used regimens are limited. We conducted a retrospective cohort study of 348 consecutive, newly diagnosed WM patients treated between 2002 and 2024. Patients received first-line therapy with either bortezomib-dexamethasone-rituximab (BDR, n = 35), Bruton’s tyrosine kinase inhibitors (BTKis, n = 57) or dexamethasone-rituximab-cyclophosphamide (DRC, n = 256). BTKi demonstrated the highest MRR (major response rate, ≥PR) (80.7%), followed by DRC (68.4%) and BDR (40.0%) (p < 0.001). The median PFS and OS did not differ significantly among regimens. TTNT seemed to be longer in the BTKi group (log-rank p = 0.025), with a 74% reduced risk of salvage therapy compared to DRC (aHR = 0.26, p = 0.016). Cumulative WM-related mortality at 5 years was lowest in BTKi-treated patients (4.1% vs. 13.0% DRC vs. 17.1% BDR), though differences were not statistically significant. In this single-centre analysis, both BTKi and DRC led to prolonged disease control in the upfront treatment of patients with WM. Extended follow-up and prospective validation are needed to reveal potential long-term survival differences.

PMID:41572416 | DOI:10.1111/bjh.70342

BMC Psychol. 2026 Jan 22. doi: 10.1186/s40359-026-03992-x. Online ahead of print.

ABSTRACT

BACKGROUND: Improved internet access has exposed rural adolescents in China to a greater risk of internet addiction. However, existing studies seldom examine the relationship between dynamic changes in internet addiction and psychosocial maladjustment. This study aims to explore the transition patterns of internet addiction and its associations with emotional and interpersonal problems over time.

METHODS: A one-year longitudinal survey was conducted among 782 middle school students in rural China. Latent Profile Analysis (LPA) was conducted to identify internet addiction profiles at two time points. Latent Profile Transition Analysis (LPTA) was then used to examine the transition patterns between profiles over time. Subsequently, statistical analyses were conducted to explore how these transitions were associated with emotional and interpersonal problems.

RESULTS: Three profiles of internet addiction were identified: minimal-internet addiction, low-internet addiction, and high-internet addiction. Based on LPTA, most adolescents with higher internet addiction at T1 shifted to lower-severity profiles over time (high → minimal: 35.3%; low → minimal: 39.8%; high → low: 33.3%), while some with initially lower levels transitioned to more severe profiles (minimal → high: 6.9%; low → high: 12.2%; minimal → low: 25.7%). Transition into higher addiction profiles predicted increased depression, anxiety, and poorer relationships with parents, peers, and teachers. Conversely, reductions in addiction were linked to improved depressive symptoms.

CONCLUSIONS: Changes in internet addiction have an impact on adolescent psychosocial maladjustment. Early detection and flexible interventions are essential in rural settings.

PMID:41572413 | DOI:10.1186/s40359-026-03992-x

BMC Psychol. 2026 Jan 22. doi: 10.1186/s40359-026-03994-9. Online ahead of print.

ABSTRACT

BACKGROUND: In Turkey, smartphone use exceeds both European and global averages, posing potential risks to individuals’ physical, social, and psychological well-being. The present research aimed to examine the relationships among smartphone addiction, significance quest, and passion, as well as to identify the mediating role of passion in the relationship between significance quest and smartphone addiction. In this regard, since there was no instrument available to assess passion within the Turkish cultural context, Study 1 adapted the Passion Scale for use in Turkish culture. Study 2 then investigated the associations among smartphone addiction, significance quest, and passion.

METHODS: During the adaptation process of the Passion Scale, SPSS 25.0 was used for the exploratory factor analysis (EFA), and Mplus 7 was employed for the confirmatory factor analysis (CFA). In Study 1, the measurement invariance of the Turkish version of the scale was also examined across groups formed according to gender, age and types of activities performed with passion. The sample sizes for EFA and CFA are 270 and 289, respectively. Majority of participants in both groups are aged between 19 and 25 (EFA: 71.2%; CFA: 66.1%). In Study 2, after testing multivariate statistical assumptions, the SEM-based mediation model was tested using Mplus 7. The study group of Study 2 consisted of 674 individuals from different ages (22.6% of participants were adolescents aged 15-18; 57.1% of participanst aged 19-25; 20.3% of participants aged 26-35).

RESULTS: The findings of Study 1 demonstrated that the two-factor structure of the Passion Scale was confirmed within the Turkish cultural context and that the scale exhibited high reliability. In Study 2, positive and significant relationships were identified between significance quest and smartphone addiction, between significance quest and obsessive passion, and between obsessive passion and smartphone addiction. Moreover, obsessive passion was found to mediate the relationship between significance quest and smartphone addiction. In contrast, no significant associations were observed between significance quest and harmonic passion, nor between harmonic passion and smartphone addiction.

CONCLUSIONS: The results of the study suggest that obsessive passion is one of the factors that strengthens the relationship between the significance quest and smartphone addiction. When an individual with obsessive passion experiences a loss of significance, their usage process may progress toward smartphone addiction, even if they are not yet addicted.

PMID:41572409 | DOI:10.1186/s40359-026-03994-9

Cardiooncology. 2026 Jan 22;12(1):9. doi: 10.1186/s40959-025-00417-3.

ABSTRACT

BACKGROUND: HIV, cancer, and their respective treatments are independently associated with cardiovascular risk, but limited data exist on the intersection of these conditions. The purpose of this study was to gain insights into the cardiovascular risk factor burden and cardiac function in people with HIV (PWH) treated for breast cancer.

METHODS: In a cohort of PWH and breast cancer treated with anthracyclines and/or trastuzumab (2017-2022) in Botswana, we assessed pre-treatment (baseline) left ventricular ejection fraction (LVEF), and prospectively obtained an echocardiogram at least one year after cancer treatment initiation. Wilcoxon signed rank sum test was used to test the differences between baseline and follow-up LVEF.

RESULTS: Thirty-three women were enrolled at a median of 2.1 years (Quartile (Q)1-Q3 1.8-3.1) from their cancer treatment initiation. The median age was 48.0 years (Q1-Q3 44.0-54.0). All but one patient was on antiretroviral therapy (ART); the median ART duration was 11.6 years (Q1-Q3 6.3-15 years) with a median viral load of 30 (Q1-Q3 0-30) and CD4 count of 874 (Q1-Q3 361-1131). At baseline, 70% were obese or overweight, and 24.2% reported hypertension; this increased to 30.3% at follow-up. The median LVEF at baseline was 65% (Q1-Q3 60-68%), and decreased to 62% (Q1-Q3 59-65%) at follow-up; an absolute difference of 2.9%, 95%CI: -5.3 to -0.2% (p = 0.038). There was no report of clinical heart failure.

CONCLUSIONS: Obesity and hypertension are highly prevalent amongst PWH and breast cancer. We also noted a statistically significant, but modest decline in LVEF after cancer therapy initiation. Further studies are needed to prospectively characterize the cardiovascular risk factor burden and changes in cardiac structure and function following cardiotoxic cancer treatment in this population.

PMID:41572399 | DOI:10.1186/s40959-025-00417-3

BMC Nutr. 2026 Jan 22. doi: 10.1186/s40795-026-01258-8. Online ahead of print.

NO ABSTRACT

PMID:41572393 | DOI:10.1186/s40795-026-01258-8

Biomark Res. 2026 Jan 22;14(1):17. doi: 10.1186/s40364-026-00897-0.

NO ABSTRACT

PMID:41572346 | DOI:10.1186/s40364-026-00897-0

Diabetol Metab Syndr. 2026 Jan 22. doi: 10.1186/s13098-025-02039-2. Online ahead of print.

ABSTRACT

BACKGROUND: The impact of discordance between remnant cholesterol (RC) and low-density lipoprotein cholesterol (LDL-c) on diabetes, diabetic kidney disease (DKD), diabetic retinopathy (DR) and cardiovascular disease (CVD) remains unclear. This study aims to explore the association between the discordance and these outcomes, using data from the National Health and Nutrition Examination Survey (NHANES) 1999.1-2020.3 and Clinical Medical College & Affiliated Hospital of Chengdu University.

METHODS: We prespecified a ± 15 percentile-point cutoff for discordance between RC and LDL-c, defined as RC percentile minus LDL-c percentile. Using this rule, 11,826 NHANES participants (cohort 1) and 306 participants from the Clinical Medical College & Affiliated Hospital of Chengdu University (cohort 2) were categorized as low discordance ( ≤ – 15), concordant (- 15 to + 15), or high discordance ( ≥ + 15). Key variables were screened by the Boruta algorithm. Logistic regression analysis models, restricted cubic spline (RCS), and subgroup analyses were used to assess the associations of discordance with outcomes. Receiver operating characteristic (ROC) and three machine learning models were used to assess the predictive value of the discordance for outcomes.

RESULTS: High discordance was significantly associated with increased risks of diabetes (OR -cohort 1: 2.371, 95% CI: 1.848-3.055; OR -cohort 2: 4.064, 95% CI: 1.750-10.020), DKD (OR: 2.593, 95% CI: 1.930-3.521), DR (OR: 2.205, 95% CI: 1.404-3.556), and CVD (OR -cohort 1: 2.299, 95% CI: 1.900-2.791; OR -cohort 2: 3.220, 95% CI: 1.266-8.175). In cohort 1, the RCS analysis showed linear relationships for these outcomes. Hypertension, HOMA-IR ≥ 3.1 and HOMA-β < 100 were identified as significant modifiers in subgroup analyses. In cohort 2, the RCS analysis showed linear relationships for diabetes and non-linear for CVD. All machine learning models demonstrated great predictive value of the discordance for diabetes and CVD.

CONCLUSION: The discordance is a significant predictor of diabetes, diabetic microvascular diseases, and cardiovascular disease.

KEY MESSAGE: (1)The discordance between RC and LDL-c is significantly associated with diabetes, diabetic microvascular diseases, and cardiovascular disease.(2) In NHANES, the RCS analysis showed linear relationships for diabetes, DKD, DR and CVD. In clinical cohort, the RCS analysis showed linear relationships for diabetes and non-linear for CVD.(3)The association of discordance with diabetes, DKD and CVD was more prevalent in individuals with hypertension and HOMA-IR ≥3.1, while individuals with hypertension were more sensitive to DR and individuals with HOMA-β <100 were more sensitive to diabetes.(4)The discordance showed the certain predictive ability for all outcomes.

PMID:41572345 | DOI:10.1186/s13098-025-02039-2