Categories
Nevin Manimala Statistics

Exploring supportive care needs of lung cancer patients in China and predicting with machine learning models

Support Care Cancer. 2025 Jun 13;33(7):573. doi: 10.1007/s00520-025-09619-y.

ABSTRACT

PURPOSE: This study aims to explore the level of supportive care needs among hospitalized lung cancer patients in China, explore the key influencing factors and use machine learning (ML) to develop predictive models for the level of supportive care needs among hospitalized lung cancer patients.

METHODS: This cross-sectional study collected data on the supportive care needs, demographics, and clinical information of 486 hospitalized lung cancer patients. Univariate and multivariate analyses identified factors associated with these needs. Predictive models were developed using six machine learning methods-logistic regression, linear regression, k-nearest neighbors, support vector machine, random forest, and adaptive boosting-to assess their performance, followed by a visualization of feature importance. The code used for model development and analysis is publicly available at https://github.com/zimengcc/predict_cancer_scn.

RESULTS: Among the factors influencing the supportive care needs of hospitalized lung cancer patients, age, education level, occupation, tumor stage, and household per capita monthly income have a significant impact on supportive care needs scores. Multiple linear regression analysis revealed that education level and household per capita monthly income were statistically significant predictors of supportive care needs scores. In the predictive tasks, the random forest model performed the best, with a mean absolute error (MAE) of 4.45 for predicting the total supportive care needs score. Furthermore, to predict the dimension with the highest level of supportive care needs, the model achieved an accuracy of 88. 42%, an F1 score of 87. 49%, and an ROC-AUC of 0.9061.

CONCLUSION: Our study explored the factors influencing the level of supportive care needs among hospitalized lung cancer patients. While the machine learning models demonstrate promising predictive performance, it is important to note that all results were derived solely through cross-validation. Therefore, potential overfitting and overestimation of model performance should be considered when interpreting these findings. Nevertheless, these models may serve as a foundation for developing tools to support personalized care planning in clinical settings.

PMID:40512392 | DOI:10.1007/s00520-025-09619-y

Categories
Nevin Manimala Statistics

Toward a general framework for AI-enabled prediction in crop improvement

Theor Appl Genet. 2025 Jun 12;138(7):151. doi: 10.1007/s00122-025-04928-6.

ABSTRACT

A theoretical framework for AI and ensembled prediction for crop improvement is introduced and demonstrated using the logistic map. Symbolic/sub-symbolic AI-based prediction can increase predictive skill with increase in system complexity. The curse of dimensionality in genomic prediction has been established and hampers genetic gain for complex traits. Artificial intelligence (AI) that fuses symbolic and sub-symbolic approaches to prediction is emerging as an approach that can deal effectively with this problem. By leveraging information across physiological and genetic networks, it is plausible to increase prediction accuracy by harnessing prior knowledge and computation approaches. Ensembles of models de facto implement the diversity prediction theorem, and thus enable breeders identify subnetworks of genetic and physiological networks underpinning crop response to management (M) and environment (E). Here, we introduce a theoretical framework for AI-enabled prediction in crop improvement. This framework brings together elements of dynamical systems modeling, ensembles, Bayesian statistics and optimization. We demonstrate properties of this framework and limits to predictability using a simple logistic map. We show that heritability and level of predictability decrease with increase in system complexity that conforms well with prior empirical evidence. We show that predicting systems states is an inferior strategy to predicting system process rates for complex systems. This holds for both the level of predictability and for the ability to use the data generating functions to produce a view of the system state space that can help breeders develop an intuition for how biological interventions can affect the performance of the crops. By integrating biological knowledge and computational approaches to prediction, it is feasible to increase predictive accuracy in breeding systems and therefore hasten the rate of genetic gain.

PMID:40512386 | DOI:10.1007/s00122-025-04928-6

Categories
Nevin Manimala Statistics

Spontaneous Correction of Facial Torsion in Nonsyndromic Unicoronal Craniosynostosis: Does Time Heal All?

J Craniofac Surg. 2025 Jun 13. doi: 10.1097/SCS.0000000000011367. Online ahead of print.

ABSTRACT

Unicoronal synostosis creates facial asymmetry typified by torsion of the midface to the contralateral side. Whether the facial tort corrects spontaneously after frontal orbital advancement surgery remains debatable. The authors aimed to evaluate the degree of clinically evident spontaneous correction of facial torsion after bifrontal orbital advancement. The authors performed a retrospective review of nonsyndromic unicoronal craniosynostosis patients treated with frontal orbital advancement between 1994 and 2016 by evaluating preoperative and postoperative clinical photographs. Preoperative and postoperative facial torsion was serially measured on AP photographs by angulation of the nasal tip from the mid-sagittal plane. Facial torsion was classified as perceptible or not perceptible for qualitative review and angle measurements were recorded for quantification. Sixty-two patients were included. The mean age at surgery was 11 months with the mean follow-up of 57 months. Two hundred sixty-three photographs were reviewed. Mean nasal tip angulation was 6.2±2.1 degrees at the first review with a correction of 1.6±1.82 degrees observed at the final review. All patients had persistent perceptible facial torsion. A statistically significant correction of facial torsion by 1.6 degrees was present but did not lead to a perceptible improvement in facial torsion. The qualitative evaluation demonstrated that perceptible facial torsion was present in both short-term and long-term postoperative photographs in all cases. Although the severity of perceptible asymmetry was not quantified, a subjective degree of perceptible facial torsion preoperatively was unchanged in long-term follow-up despite statistically significant quantitative correction.

PMID:40512385 | DOI:10.1097/SCS.0000000000011367

Categories
Nevin Manimala Statistics

Effects of loaded deep breathing training combined with PERMA model in lung cancer patients undergoing surgery: a randomized controlled trial

Support Care Cancer. 2025 Jun 13;33(7):574. doi: 10.1007/s00520-025-09607-2.

ABSTRACT

OBJECTIVE: Lung cancer (LC) and surgery cause physical and psychological distress to patients. This study aimed to investigate the impact of loaded deep breathing training combined with the PERMA model (PERMA-LDBT) on physical and psychological symptoms, functional capacity, and quality of life (QoL) in patients undergoing LC surgery.

METHODS: A three-arm randomized controlled trial was conducted. Patients undergoing LC surgery were randomly assigned to the LDBT plus routine care (IG), PERMA-LDBT plus routine care (CIG), and the routine care (CG). Primary variables included exercise capacity (6MWD), dyspnea (mMRC), anxiety and depression (HADS), and quality of life (EORTC-QQL-C30). Secondary variables comprised adherence, postoperative pulmonary complications (PPCs), length of hospitalization, duration of postoperative thoracic drainage tube, and adverse events. Socio-demographic variables comprised age, gender, smoking status, BMI, comorbidity with COPD, pulmonary function, length of hospitalization, cancer stage, and lobectomy location. Data were collected at admission, the day before surgery, two days after surgery, and the day of discharge.

RESULTS: A total of 148 patients were recruited. Forty-six were excluded, and 102 patients were randomized into three groups of 34 each. Significant improvements in dyspnea, exercise capacity, anxiety, depression, and QoL were observed in both the CIG and IG compared to the CG. However, no statistically significant differences were found in PPCs, length of hospitalization, or postoperative chest drain retention time. The CIG demonstrated statistically significant improvements in adherence and anxiety levels compared to the IG.

CONCLUSION: This study provides evidence that PERMA-LDBT intervention can effectively improve perioperative outcomes in LC patients.

TRIAL REGISTRATION NUMBER: (ChiCTR2400080214) and date of registration (2024-01-23) “retrospectively registration”.

PMID:40512375 | DOI:10.1007/s00520-025-09607-2

Categories
Nevin Manimala Statistics

Cross-lagged analysis in nephrology

J Nephrol. 2025 Jun 13. doi: 10.1007/s40620-025-02319-0. Online ahead of print.

ABSTRACT

Cross-lagged analysis is a statistical method employed to examine directional relationships between variables over time, making it especially valuable for addressing causality challenges in clinical research. This method is essential for comprehending complex bidirectional relationships, such as stress and immunity, dietary habits and metabolic conditions, or medication adherence and health outcomes. By analyzing longitudinal data, cross-lagged analysis establishes temporal precedence, tests reciprocal influences, and controls for confounding variables, thereby enhancing causal inferences. In nephrology, this approach can be beneficial for studying the interaction between acute kidney injury (AKI) and chronic kidney disease (CKD), clarifying whether AKI episodes accelerate CKD progression or if pre-existing CKD increases susceptibility to AKI. It also illuminates the relationship between CKD and cardiovascular diseases, investigating whether CKD exacerbates heart failure or vice versa while considering shared risk factors like hypertension and diabetes. Furthermore, cross-lagged analysis can elucidate the kidney-brain connection by examining whether CKD leads to cognitive decline through mechanisms such as uremic toxin accumulation or if neurological dysfunction worsens kidney outcomes through sympathetic overactivation. Cross-lagged analysis accommodates latent variables and measurement errors, allowing researchers to explore how variables interact over time. This method provides a strong framework for understanding dynamic relationships in nephrology, offering critical insights to guide interventions and advance knowledge of disease progression mechanisms.

PMID:40512355 | DOI:10.1007/s40620-025-02319-0

Categories
Nevin Manimala Statistics

Effectiveness and safety of intravenous iron in ulcerative colitis: A real-world study of impact of disease activity and iron preparations

Indian J Gastroenterol. 2025 Jun 13. doi: 10.1007/s12664-025-01775-7. Online ahead of print.

ABSTRACT

BACKGROUND: There is lack of evidence on how disease activity in inflammatory bowel disease influences response to intravenous iron.

METHODS: A single-centre prospective study was conducted to study responses to intravenous iron in ulcerative colitis. Patients with iron deficiency anemia (hemoglobin < 8 g/dL OR < 12 g/dL with intolerance to oral iron or active disease) received protocolized intravenous iron dosing per European Crohn’s and Colitis Organisation (ECCO) guidelines. Ferric carboxymaltose (FCM) or iron isomaltoside was used per patient preference. The outcomes were increase in Hb ≥ 2 g/dL, normalization of Hb at four weeks and normalization of iron indices (serum ferritin >100 ng/mL in active and > 30 ng/mL in inactive disease and transferrin saturation > 20%) assessed at four weeks.

RESULTS: Thirty-two patients (females = 21 [65.6%], mean age = 32.78 ± 12.65 years, active disease = 17 [53.1%]) received IV iron. Twenty-one received FCM, while 11 received iron isomaltoside. Complete normalization of Hb was seen in seven (41%) in active patients and nine (60%) in remission groups (p = 0.47). Normalization of iron profile was seen in 11 (64%) and 12 (80%) patients in two groups, respectively (p = 0.32). Secondary objectives of mean change in Hb and iron indices tended towards better response in inactive group, but were not statistically significant except change in transferrin saturation that was better in inactive group (20.8 ± 9.1% and 14.2 ± 8.1%, p = 0.04). FCM was associated with hypophosphatemia in 6/21 (28.5%) patients vs. none in isomaltoside group. The predictive performance for complete hematological response with the reticulocyte hemoglobin and percentage of hypochromic cells was low (area under the receiver operating characteristic [AUROC] of 0.67 and 0.685, respectively).

CONCLUSION: Although there was a tendency towards better response to intravenous iron in those with remission, the findings were not statistically different. Larger studies are needed to provide conclusive evidence. Iron isomaltoside may be preferable over FCM due to risk of hypophosphatemia with the latter.

PMID:40512340 | DOI:10.1007/s12664-025-01775-7

Categories
Nevin Manimala Statistics

Delineating Life-Course Percentile Curves and Normative Values of Multi-Systemic Ageing Metrics in the United Kingdom, the United States, and China

J Cachexia Sarcopenia Muscle. 2025 Jun;16(3):e13862. doi: 10.1002/jcsm.13862.

ABSTRACT

BACKGROUND: Ageing is a complex and multi-dimensional process that manifests heterogeneities across different organs/systems, individuals and countries. We aimed to delineate the life-course percentile curves and establish the normative values of multi-systemic (e.g., muscle-skeletal, brain, cardiovascular and pulmonary) ageing metrics for people under distinct sociodemographic contexts (i.e., sex, income and education).

METHODS: Three national datasets, the UKB (the United Kingdom), the NHANES (the United States) and the CHARLS (China) were utilized for the analyses. We selected 14 ageing metrics (e.g., body mass index, grip strength, fat-free mass index, bone mineral content [BMC], bone mineral density [BMD], diastolic blood pressure, cognitive function and frailty index_Lab) that represent the functions of different organs/systems and plotted their sex-, educational- and income-specific percentile curves utilizing the GMALSS model. We also estimated the age-specific normative values for each ageing metric in distinct sociodemographic contexts.

RESULTS: The functions of all metrics, except for cognitive function, manifested a progressive decline or maintained stability after adulthood (20s), especially after middle age (40s-50s). The cognitive function showed an evident decline in old age (70s-75s) (e.g., in the CHARLS: the median [IQR] cognitive function scores were 11.6 [9.1, 13.8], 10.3 [7.5, 12.9], 8.3 [5.5, 11.0] at the ages of 60, 70 and 80 for males, respectively). In the stratified analyses, males and females manifested disparities in percentile curves of ageing metrics involving the muscle-skeletal and cardiovascular systems. For instance, BMC and BMD manifested an evident decline after middle age in females, whereas they showed a slow decline after adulthood in males. Notably, we observed substantial income and educational disparities in percentile curves of several ageing metrics within Chinese participants: the ‘low-income’ and ‘low-education’ subgroups manifested an evident decline in ageing metrics (e.g., grip strength and frailty index_Lab) representative of multiple systems. By contrast, these income or educational disparities were not observed in the British and American participants.

CONCLUSIONS: Our investigation delineated the potential heterogeneities and socioeconomic disparities in percentile curves of multi-systemic ageing metrics and provided their age-specific normative values tailored to different sexes and socioeconomic contexts based on three national datasets. This study may serve as a proof-of-concept for understanding the multi-dimensional signature of systemic ageing and calls for policies to promote health equity across nations when facing dramatic global ageing.

PMID:40511588 | DOI:10.1002/jcsm.13862

Categories
Nevin Manimala Statistics

Patients Having Major Abdominal Cancer Surgery Exhibit Significant Acute Muscle Wasting

J Cachexia Sarcopenia Muscle. 2025 Jun;16(3):e13858. doi: 10.1002/jcsm.13858.

ABSTRACT

BACKGROUND: Postoperative myopenia (acute muscle loss) is a significant concern following major cancer resection surgery, contributing to prolonged recovery, increased dependency and impaired quality of life. Despite its clinical relevance, the mechanisms underlying postoperative myopenia and its potential mediators remain poorly understood. This study aims to evaluate the acute changes in muscle size, strength and activity following major cancer surgery and to explore the role of insulin resistance and selenium deficiency as potential mediators of these alterations.

METHODS: A prospective cohort study was conducted involving 52 patients undergoing elective open major abdominal surgery for cancer. Preoperative and postoperative assessments included measurements of rectus femoris cross-sectional area (RFCSA) via ultrasound, handgrip strength (HGS), sniff nasal inspiratory pressure (SNIP) and physical activity using an accelerometer. Blood samples were analysed for markers of muscle metabolism, systemic inflammation, insulin resistance and selenium levels. Statistical analyses were performed to compare preoperative and postoperative values and to explore correlations between these measures and clinical outcomes.

RESULTS: A significant reduction in RFCSA was observed in 50% of patients, with a median decrease of 10.2% within the first week post-surgery, which persisted at the 6-week follow-up. HGS and SNIP also showed significant declines postoperatively, and reduced physical activity was associated with greater muscle loss. Insulin resistance and postoperative selenium depletion were associated with greater reductions in RFCSA.

CONCLUSION: Major cancer surgery is associated with significant acute muscle loss, which is not fully recovered by 6 weeks postoperatively. Insulin resistance and selenium deficiency may contribute to this muscle loss. Further research is needed to investigate potential interventions to prevent or mitigate postoperative myopenia.

PMID:40511585 | DOI:10.1002/jcsm.13858

Categories
Nevin Manimala Statistics

Impact of short-term air pollution exposure on premature rupture of membranes in the North China Plain, 2015-2021: a multicity case-crossover study

J Glob Health. 2025 Jun 13;15:04165. doi: 10.7189/jogh.15.04165.

ABSTRACT

BACKGROUND: The effects of short-term air pollution exposure on hospital admissions for premature rupture of membranes (PROM) were still controversial.

METHODS: Daily data on PROM hospitalisations and air pollutants were collected from 1 January 2015 to 31 December 2021, across five cities in the North China Plain. We investigated the associations between short-term (lag0-7) exposure to air pollutants, including fine particulate matter (PM2.5), inhalable particulate matter (PM10), nitrogen dioxide (NO2), sulfur dioxide (SO2), carbon monoxide (CO), and ozone (O3), as well as composite air pollution indicators, including air quality index (AQI) and air quality composite index (AQCI). We further assessed the modifying effects of age (<35 and ≥35 years), season (cold and warm seasons), and the ‘Three-Year Action Plan’ (before and after implementation) on the above associations.

RESULTS: A total of 16 503 mothers with PROM were included. For each interquartile range (IQR) increase, the strongest relative risks (RRs) were 1.024 (95% confidence interval (CI) = 1.001-1.047, P < 0.05) in PM2.5 in lag3 (IQR = 48 μg/m3), RR = 1.036 (95% CI = 1.009-1.063, P < 0.05) in CO in lag4 (IQR = 0.49 mg/m3), RR = 1.024 (95% CI = 1.003-1.044, P < 0.05) in AQI in lag3 (IQR = 53), and RR = 1.026 (95% CI = 1.001-1.050, P < 0.05) in AQCI in lag3 (IQR = 2.9). The effects of exposure to PM2.5, NO2, SO2, CO, AQI, and AQCI on PROM were stronger in mothers aged <35 years, and the effects of exposure to PM10, NO2, and SO2 on PROM were stronger during the warm season (all P modification<0.05).

CONCLUSIONS: Short-term air pollution exposure was related to elevated hospital risk of PROM. The effects were more pronounced among mothers aged <35 years and during the warm season.

PMID:40511528 | DOI:10.7189/jogh.15.04165

Categories
Nevin Manimala Statistics

Rheumatoid arthritis and risk of thyroid cancer: A nationwide cohort study

Arthritis Care Res (Hoboken). 2025 Jun 13. doi: 10.1002/acr.25582. Online ahead of print.

ABSTRACT

OBJECTIVE: Despite extensive cancer-related research in patients with rheumatoid arthritis (RA), there remains a paucity of studies on thyroid cancer in RA patients. We investigated risk of thyroid cancer in RA patients using updated definitions to identify RA patients and incident cases of thyroid cancer with consideration of RA serologic status.

METHODS: Using a nationwide database, we identified 40,895 patients with newly diagnosed RA between 2010 and 2017, and matched them by sex and age at a 1:5 ratio to a control population of 204,475 non-RA individuals. The association of thyroid cancer and RA with consideration of serostatus was investigated using Cox regression analyses. Stratified analysis by sex and age were conducted using the same Cox modeling.

RESULTS: During 5.5 years of follow-up, compared to the matched control group, the adjusted hazard ratio (aHR) of thyroid cancer in overall RA patients was 0.98 (95% confidence interval [CI] 0.85-1.13). Compared to the seronegative RA patients, seropositive RA patients did not show a significantly different risk of thyroid cancer (aHR 1.03, 95% CI 0.78-1.36). Stratified analyses by age or sex showed no statistical interaction.

CONCLUSION: Patients with newly diagnosed RA did not show a different risk of thyroid cancer compared to a matched control group. The risk of thyroid cancer incidence was not affected by serologic status of RA, sex, or age.

PMID:40511527 | DOI:10.1002/acr.25582