Tag: statistics

Nat Neurosci. 2026 Apr 7. doi: 10.1038/s41593-026-02255-7. Online ahead of print.

ABSTRACT

The brain must infer the state of the external world despite the inherent uncertainty of its sensory inputs and internal processes. Under conditions of heightened uncertainty, it increasingly relies on prior knowledge, derived from accumulated experience with the regularities and statistical structures of the environment. This principle has been formalized by Bayesian inference theories, which are supported by substantial evidence from both behavioral and neuroscience studies. However, direct evidence for the existence of prior knowledge in the brain, and for the encoding of environmental statistics by neural circuits, remains limited. Here we show that cerebellar circuits learn the prior probability distribution of temporal variables during eyeblink conditioning in mice and encode these representations in Purkinje cell simple and complex spike signaling. We further demonstrate that Purkinje cells are involved in eliciting predictive motor behaviors, such as the conditioned eyeblink response, that also reflect the statistics of the experimentally imposed prior distribution of the stimulus. Computational modeling of these results indicates the juxtaposition of counteracting long-term plasticity mechanisms by which cerebellar Purkinje cells could acquire prior knowledge that is shaped by the statistics of different probability distributions. Our results suggest that the cerebellar circuitry may be uniquely poised to learn the probability of events in the world and internalize these as prior knowledge. These findings advance understanding of how neural computations could implement Bayesian inference.

PMID:41946969 | DOI:10.1038/s41593-026-02255-7

NPJ Digit Med. 2026 Apr 7;9(1):289. doi: 10.1038/s41746-026-02470-3.

ABSTRACT

Atrial fibrillation (AF) is the most common sustained arrhythmia and a leading cause of ischemic stroke. Existing risk scores, such as CHA₂DS₂-VASc, offer limited predictive accuracy and fail to capture complex clinical patterns. To improve generalizability and clinical utility, we developed and externally validated clinically interpretable machine learning models using only age, comorbidities, and medication use to predict 1-year stroke risk in patients with newly diagnosed AF. Both logistic regression (LR) and Platt-calibrated extreme gradient boosting (XGB) models achieved high discrimination in internal (AUCs = 0.915 and 0.914) and external validation cohorts (AUCs = 0.877-0.886), significantly outperforming CHA₂DS₂-VASc (AUCs = 0.614-0.621; p < 0.001). Calibration curves and decision curve analysis confirmed strong clinical utility. Long-term follow-up demonstrated superior risk stratification and treatment responsiveness in LR-defined high-risk groups. These models provide accurate, individualized stroke risk estimates to guide direct oral anticoagulant (DOAC) initiation in real-world hospital settings.

PMID:41946928 | DOI:10.1038/s41746-026-02470-3

Sci Rep. 2026 Apr 7. doi: 10.1038/s41598-026-47471-0. Online ahead of print.

ABSTRACT

Almost half of under-five children in health facilities are febrile in sub-Saharan Africa. The non-specific clinical symptoms coupled with; limited diagnostics capacities, overlapping endemicity and the surveillance gaps complicates accurate cause identification of an acute febrile illnesses in resource limited settings. This challenge leads to misdiagnosis, overtreatment, and delays in appropriate management, increasing morbidity and mortality. Health systems are often overloaded, with providers attributing fevers to the most common pathogen, while other emerging infections are the cause. This study unpacked the febrile illness by testing for dengue fever in parallel to malaria. Febrile ill children below 5 years seeking health services public health facilities in Busia and Kisumu Counties were screened using an approved malaria and dengue fever rapid test kits at the outpatient department. Those who screened positive were recruited into the study. A total of 1004 children were screened, 380 met the recruitment criteria. 215 (21.4%) tested positive for P. falciparum alone, 90 (8.9%) tested positive for dengue fever alone while 75 (7.5%) had co-infections. Busia had the highest P. falciparum-only infection (23.4%) while Kisumu had the highest dengue-only infections (12.6%). Dengue fever is a re-emerging neglected tropical, climate change driven disease in malaria endemic regions. Other than creating awareness to build capacity for diagnosis, this study unmasked and confirmed dengue as a major contributor to the non malarial febrile illnesses among children. There is need to revise the screening algorithm for febrile patients to improve arboviral surveillance.

PMID:41946915 | DOI:10.1038/s41598-026-47471-0

Sci Rep. 2026 Apr 7. doi: 10.1038/s41598-026-47584-6. Online ahead of print.

ABSTRACT

To evaluate the effectiveness of combining problem-based learning with the flipped classroom (PBL-FC) teaching method in the “Ocular Trauma” module in ophthalmology education, specifically in enhancing clinical medicine students’ self-directed learning ability and empathy. A quasi-experimental study with a controlled educational intervention was conducted, and third-year clinical medicine students from eight classes were divided into PBL-FC (n = 43) and traditional groups (n = 44). Both groups were assessed before and after class using Self-Directed Learning Ability and Empathy Quotient (EQ)-40 scales. A comparison was also made between theoretical examination scores. The two groups showed no statistically significant differences (P > 0.05) in gender ratio, age, pre-class empathy scores or pre-class self-directed learning ability scores. After class, the PBL-FC group demonstrated significantly higher total self-directed learning ability scores (175.72 ± 16.20 vs. 156.19 ± 12.72) and scores across all sub-dimensions (P < 0.05). In addition, their empathy (46.91 ± 5.56 vs. 41.89 ± 5.61) and theoretical examination scores (77.81 ± 8.63 vs. 72.45 ± 10.55) were also substantially superior to those of the traditional group (P < 0.05). The combined PBL-FC teaching method is effective in enhancing undergraduate medical students’ self-directed learning ability, empathy and theoretical knowledge. This study provides empirical support for the reform of ophthalmology education.

PMID:41946914 | DOI:10.1038/s41598-026-47584-6

Eur J Clin Nutr. 2026 Apr 7. doi: 10.1038/s41430-026-01732-3. Online ahead of print.

ABSTRACT

BACKGROUND: India’s food consumption landscape has transformed over the past decade, with implications for nutrition security. This article quantifies changes in household expenditure, consumption patterns across major food groups, and dietary sources of selected micronutrients.

METHODS: Data from nationally representative Household Consumption Expenditure Surveys (HCES) 2011-12 and 2023-24 were analysed. Food quantities were standardised to adult female equivalents, and micronutrient intake was derived by linking food acquisition data to Indian Food Composition Tables. Models were fitted using a generalised additive mixed model with random effects (bam function, mgcv package in R).

RESULTS: Monthly per capita expenditure has increased across rural and urban areas, while share of food in household expenditure has declined, signalling economic diversification. Within food budgets, share of cereals has fallen sharply, particularly among poorest quintiles, consistent with expansion of food security programmes providing subsidised cereals. Meanwhile, probability and quantity of consumption of nutrient-dense foods- dairy, fruits, and flesh products, has risen across income groups, with larger gains among poorest households. Seasonal and regional disparities in perishable food consumption persist but have narrowed. Despite greater dietary diversity, micronutrient intake remains suboptimal. Estimated daily intake of Iron, Zinc, Calcium, and B Vitamins fell below Estimated Average Requirements for non-lactating adult women, with median inadequacy exceeding 75% for several micronutrients.

CONCLUSION: India’s dietary transition shows progress but persistent gaps. HCES-based estimates provide valuable insights, underscoring need for integrated strategies aligning food policy, social protection, and nutrition-sensitive interventions to improve equitable access to nutrient-dense foods and reduce micronutrient inadequacy.

PMID:41946892 | DOI:10.1038/s41430-026-01732-3

Ann Biomed Eng. 2026 Apr 7. doi: 10.1007/s10439-026-04130-9. Online ahead of print.

ABSTRACT

PURPOSE: Thick-walled vascular models, including block models, are increasingly used for in vitro investigations of fluid dynamics and endovascular device testing, particularly fatigue testing at high frequencies, as valuable alternatives to thin-walled models due to their simpler fabrication. Physiologically compliant models provide more realistic insights into hemodynamics and implant deformation; however, no compact and experimentally validated closed-form equation exists to predict the volumetric compliance of thick-walled vessel models.

METHODS: A closed-form mathematical equation was derived to estimate the volumetric compliance of thick-walled cylindrical vessel models. The equation was experimentally validated through static compliance tests on cylindrical vascular models of varying wall thicknesses fabricated from two silicone elastomers and polyvinyl alcohol hydrogel (PVA-H) with four PVA concentrations (10, 12, 14, and 16 wt/vol.%). Dynamic compliance tests were performed at 1, 5, 10, and 30 Hz to assess the viscoelastic behavior of the materials. Finally, an applicability study was performed by fabricating block-form anatomical aneurysm models targeting physiological volumetric compliance based on the predicted equation.

RESULTS: Predicted and measured values showed strong agreement (R² > 0.95). PVA-H-10 exhibited volumetric compliance within the physiological range (0.4-1.1%/mmHg). Volumetric compliance decreased by over 50% at 30 Hz, confirming viscoelastic behavior with a maximum loss to storage volumetric compliance ratio of 0.14. Measured aneurysm model compliance matched predictions with an error below 7%.

CONCLUSION: This study provided a validated mathematical and experimental framework for fabricating thick-walled vascular models with physiologically relevant volumetric compliance and viscoelasticity for fluid dynamics and endovascular device testing.

PMID:41946863 | DOI:10.1007/s10439-026-04130-9

Cogn Affect Behav Neurosci. 2026 Apr 7. doi: 10.3758/s13415-026-01423-3. Online ahead of print.

ABSTRACT

Pupil dilation responses are reliable physiological markers of arousal in response to unexpected events. We investigated how these responses generalise across different sensory modalities by using transitions between regular and random sequences of visual dots and auditory tones. In Experiment 1, we investigated sequences of visual dots and found that (a) transitions from a regular sequence to a random sequence induced pupil dilations, (b) transitions from one regular sequence to another regular sequence also induced pupil dilations, and (c) transitions from a random sequence to a regular sequence did not reliably induce them. In Experiment 2, we replicated these findings, confirming their reliability and thereby generalizing the literature from the auditory to the visual modality. In Experiment 3, we directly compared pupil dilations in visual and auditory modalities. We observed strong cross-modal similarity in pupil sizes, particularly for transitions between regular and random sequences. We also decomposed the pupil size time series to approximate phasic pupil dilation events. While the patterns of dilation events were quite similar, differences between modalities in dilation size (but not in rates) occurred during transitions from one regular to another regular sequence. Overall, our findings suggest that pupil-linked arousal reflects inference of statistical structure and its violations, exhibiting substantial (albeit not perfect) similarity across modalities.

PMID:41946847 | DOI:10.3758/s13415-026-01423-3

J Clin Sleep Med. 2026 Apr 7;22(1):44. doi: 10.1007/s44470-026-00066-y.

ABSTRACT

PURPOSE: Examine associations between sleep patterns (duration, timing, and regularity) and weight status and explore moderating effects of parenting/family functioning among children with mild sleep-disordered breathing (SDB) from the Pediatric Adenotonsillectomy Trial for Snoring.

METHODS: Two age groups (3-5 and 6-12-year-olds) were examined separately with waist circumference (WCz) and body mass index (BMIz) z-scores as the primary and secondary endpoints, respectively.

RESULTS: Participants were 207 younger and 252 older children in a socio-demographically diverse sample. In adjusted models with older children, shorter sleep was associated with higher WCz, controlling for weekend mid-sleep timing (coefficient estimate beta = – 0.137, p = 0.027) with a similar finding for BMIz (beta = – 0.144, p = 0.041); later timing was related to higher WCz, controlling for sleep duration (beta = 0.122, p = 0.032). Shorter sleep duration was associated with higher WCz among the older children with greater parental distress (interaction p = 0.005), parent-child dysfunctional interactions (interaction p = 0.020), and less authoritative parenting (interaction p = 0.023) compared to those with more optimal parent-child relationships. There were no statistically significant associations or interactions among younger children.

CONCLUSIONS: In school-aged children, shorter sleep duration and later midpoint, but not social jet lag, tended to be associated with weight status. When mutually adjusting for duration and midpoint, a stronger statistically significant association was observed. Children subject to less optimal parenting practices were at the greatest risk of shorter sleep. Accounting for sleep timing and parenting factors that might further increase children’s biological propensity for obesity is recommended in future studies in this area.

CURRENT KNOWLEDGE/STUDY RATIONALE: Initial evidence suggests that shorter sleep duration and sleep-disordered breathing (SDB) increase the risk of becoming overweight in later childhood. This study examined relationships of additional sleep patterns-timing and regularity-and possible moderating effects of parenting/family factors using a well-characterized cohort of children with mild SDB from the Pediatric Adenotonsillectomy Trial for Snoring.

STUDY IMPACT: This study demonstrates that both shorter sleep duration and later timing relate to higher waist circumference and body mass index z-scores among school-aged children. Parenting factors (i.e., stress, parent-child dysfunctional relationships, and less use of an authoritative parenting style) may moderate sleep pattern-weight status relationships and therefore identify children most vulnerable to obesity because of insufficient sleep.

PMID:41946844 | DOI:10.1007/s44470-026-00066-y

J Neurol. 2026 Apr 7;273(5):254. doi: 10.1007/s00415-026-13796-z.

ABSTRACT

OBJECTIVE: Postictal aggression (PIA) is an under-researched phenomenon with potentially harmful consequences for patients and their environment. This study assessed the prevalence of PIA, its associated clinical variables, and its relationship with quality of life (QoL), anxiety, and depression.

METHODS: In this cross-sectional study, consecutive patients with epilepsy (≥ 1 seizure in the past 12 months; age ≥ 16 years) were surveyed using standardized questionnaires in a tertiary epilepsy center. Occurrence of PIA was retrospectively assessed over the preceding 12 months and defined by a score of ≥ 2 on the Overt aggression scale modified (OAS-M). Anxiety and depression were measured with the Hospital anxiety and depression scale (HADS-A, HADS-D) and the Neurological disorders depression inventory in epilepsy (NDDI-E).

RESULTS: A total of 201 patients were included (57% male; median age 47 years; 76% focal epilepsy, 14% idiopathic generalized epilepsy (IGE), 10% unclassified). Twenty-four patients (12%) reported clinically relevant PIA. Patients with PIA had lower QoL (QOLIE-10 median 40 vs. 37, p = 0.031) and more frequent clinically relevant anxiety (HADS-A ≥ 8: 58% vs. 30%, p = 0.019). Depressive symptoms were more common in PIA (NDDI-E ≥ 16: 21% vs. 10%) without statistical significance (p = 0.309). IGE (OR 3.13, 95% CI 1.06-8.98) and anxiety (OR 2.51, 95% CI 1.02-6.26) were independently associated with PIA. There were no associations with antiseizure medications or their dosage. Ten patients reported adverse consequences of PIA, most commonly shame and fear.

CONCLUSION: Almost one in eight patients experienced PIA. It was associated with reduced QoL and increased anxiety, highlighting the psychosocial burden.

PMID:41946837 | DOI:10.1007/s00415-026-13796-z

Mol Psychiatry. 2026 Apr 7. doi: 10.1038/s41380-026-03553-z. Online ahead of print.

ABSTRACT

Biomarkers can potentially improve the diagnosis, monitoring, and treatment of posttraumatic stress disorder (PTSD). However, PTSD biomarkers that are scalable and easily integrated into real-world clinical settings have not been identified. The analysis was conducted between June to November 2024 using genomic samples and laboratory test results recorded in the Mass General Brigham (MGB) Health System. The analysis included 23,743 European ancestry participants from the nested MGB Biobank study. The first exposure was polygenic risk score (PRS) for PTSD, calculated using the largest available European ancestry genome-wide association study (GWAS), employing a Bayesian polygenic scoring method. The second exposure was a clinical diagnosis of PTSD, determined by the presence of two or more instances of PTSD-related diagnostic codes in the longitudinal electronic health records (EHR). The primary outcomes were the inverse normal quantile transformed, median lab values of 241 laboratory traits with non-zero h²_SNP estimates. Sixteen unique laboratory traits across the cardiometabolic, hematologic, hepatic, and immune systems were implicated in both genomic and phenotypic lab-wide association scans (LabWAS). Two-sample Mendelian randomization analyses provided evidence of potential unidirectional causal effects of PTSD liability on hepatic (decreased albumin and total bilirubin), cardiometabolic (decreased HDL cholesterol and increased VLDL cholesterol), and hematologic (decreased mean platelet volume) markers. These findings demonstrate the potential of a triangulation approach to uncover scalable and clinically relevant biomarkers for PTSD.

PMID:41946832 | DOI:10.1038/s41380-026-03553-z