Category: Statistics

Netw Neurosci. 2023 Jan 1;7(1):299-321. doi: 10.1162/netn_a_00284. eCollection 2023.

ABSTRACT

Executive functioning (EF) is a higher order cognitive process that is thought to depend on a network organization facilitating integration across subnetworks, in the context of which the central role of the fronto-parietal network (FPN) has been described across imaging and neurophysiological modalities. However, the potentially complementary unimodal information on the relevance of the FPN for EF has not yet been integrated. We employ a multilayer framework to allow for integration of different modalities into one ‘network of networks.’ We used diffusion MRI, resting-state functional MRI, MEG, and neuropsychological data obtained from 33 healthy adults to construct modality-specific single-layer networks as well as a single multilayer network per participant. We computed single-layer and multilayer eigenvector centrality of the FPN as a measure of integration in this network and examined their associations with EF. We found that higher multilayer FPN centrality, but not single-layer FPN centrality, was related to better EF. We did not find a statistically significant change in explained variance in EF when using the multilayer approach as compared to the single-layer measures. Overall, our results show the importance of FPN integration for EF and underline the promise of the multilayer framework toward better understanding cognitive functioning.

PMID:37339322 | PMC:PMC10275212 | DOI:10.1162/netn_a_00284

Dis Colon Rectum. 2023 Jun 20. doi: 10.1097/DCR.0000000000002686. Online ahead of print.

ABSTRACT

BACKGROUND: The association of micropapillary pattern with oncologic outcomes has not been fully studied in patients with colon cancer.

OBJECTIVE: We evaluated the prognostic value of micropapillary pattern, especially for patients with stage II colon cancer.

DESIGN: A retrospective comparative cohort study using propensity score matching.

SETTING: This study was conducted at a single tertiary center.

PATIENTS: The patients with primary colon cancer undergoing curative resection from October 2013 to December 2017 were enrolled. The patients were grouped into micropapillary pattern (+) or micropapillary pattern (-).

MAIN OUTCOME MEASUREMENTS: Disease-free survival and overall survival.

RESULTS: Of the eligible 2,192 patients, 334 (15.2%) were micropapillary pattern (+). After 1:2 propensity score matching, 668 patients with micropapillary pattern (-) were selected. Micropapillary pattern (+) group showed significantly worse 3-year disease-free survival (77.6% vs. 85.1%, p = 0.007). Three-year overall survival of micropapillary pattern-positive and micropapillary pattern-negative did not show a statistically significant difference (88.9% vs. 90.4%, p = 0.480). In multivariable analysis, micropapillary pattern -positive was an independent risk factor for poor disease-free survival (hazard ratio 1.547, p = 0.008). In the subgroup analysis for 828 patients with stage II disease, 3-year disease-free survival deteriorated significantly in micropapillary pattern (+) patients (82.6% vs. 93.0, p < 0.001). Three-year overall survival was 90.1% and 93.9% in micropapillary pattern (+) and micropapillary pattern (-), respectively (p = 0.082). In the multivariable analysis for patients with stage II disease, micropapillary pattern (+) was an independent risk factor for poor disease-free survival (hazard ratio 2.003, p = 0.031).

LIMITATIONS: Selection bias due to the retrospective nature of the study.

CONCLUSIONS: Micropapillary pattern (+) may serve as an independent prognostic factor for colon cancer, especially for patients with stage II disease.

PMID:37339285 | DOI:10.1097/DCR.0000000000002686

J Clin Endocrinol Metab. 2023 Jun 20:dgad371. doi: 10.1210/clinem/dgad371. Online ahead of print.

ABSTRACT

BACKGROUND: Thyroid function has been associated with metabolic syndrome (MetS) in a number of observational studies. In spite of that, the direction of effects and the exact causal mechanism of this relationship is still unknown.

METHODS: We performed a two-sample bidirectional Mendelian randomization (MR) study using summary statistics from the most comprehensive genome-wide association studies (GWAS) of thyroid-stimulating hormone (TSH, n = 119,715), free thyroxine (fT4, n = 49,269), MetS (n = 291,107), as well as components of MetS: waist circumference (n = 462,166), fasting blood glucose (n = 281,416), hypertension (n = 463,010), triglycerides (TG, n = 441,016) and high-density lipoprotein cholesterol (HDL-C, n = 403,943). We chose the multiplicative random-effects inverse variance weighted (IVW) method as the main analysis. Sensitivity analysis included weighted median and mode analysis, as well as MR-Egger and Causal Analysis Using Summary Effect estimates (CAUSE).

RESULTS: Our results suggest that higher fT4 levels lower the risk of developing MetS (OR = 0.96, P = 0.037). Genetically predicted fT4 was also positively associated with HDL-C (β=0.02, P = 0.008), while genetically predicted TSH was positively associated with TG (β=0.01, P = 0.044). These effects were consistent across different MR analyses and confirmed with the CAUSE analysis. In the reverse direction MR analysis, genetically predicted HDL-C was negatively associated with TSH (β=-0.03, P = 0.046) in the main IVW analysis.

CONCLUSIONS: Our study suggests that variations in normal-range thyroid function are causally associated with the diagnosis of MetS and with lipid profile, while in the reverse direction, HDL-C has a plausible causal effect on reference-range TSH levels.

PMID:37339283 | DOI:10.1210/clinem/dgad371

Proc Natl Acad Sci U S A. 2023 Jul 4;120(27):e2303168120. doi: 10.1073/pnas.2303168120. Epub 2023 Jun 20.

ABSTRACT

We briefly review the majorization-minimization (MM) principle and elaborate on the closely related notion of proximal distance algorithms, a generic approach for solving constrained optimization problems via quadratic penalties. We illustrate how the MM and proximal distance principles apply to a variety of problems from statistics, finance, and nonlinear optimization. Drawing from our selected examples, we also sketch a few ideas pertinent to the acceleration of MM algorithms: a) structuring updates around efficient matrix decompositions, b) path following in proximal distance iteration, and c) cubic majorization and its connections to trust region methods. These ideas are put to the test on several numerical examples, but for the sake of brevity, we omit detailed comparisons to competing methods. The current article, which is a mix of review and current contributions, celebrates the MM principle as a powerful framework for designing optimization algorithms and reinterpreting existing ones.

PMID:37339185 | DOI:10.1073/pnas.2303168120

Can Assoc Radiol J. 2023 Jun 20:8465371231180815. doi: 10.1177/08465371231180815. Online ahead of print.

ABSTRACT

Background and Purpose: Evidence has emerged for an association between degenerative disc disease (DDD) and multiple sclerosis (MS). The purpose of the current study is to determine the presence and extent of cervical DDD in young patients (age <35) with MS, an age cohort that is less well studied for these changes. Methods: Retrospective chart review of consecutive patients aged <35 referred from the local MS clinic who were MRI scanned between May 2005 and November 2014. 80 patients (51 female and 29 male) with MS of any type ranging between 16 and 32 years of age (average 26) were included. Images were reviewed by 3 raters and assessed for presence and extent of DDD, as well as cord signal abnormalities. Interrater agreement was assessed using Kendall’s W and Fleiss’ Kappa statistics. Results: Substantial to very good interrater agreement was observed using our novel DDD grading scale. At least some degree of DDD was found in over 91% of patients. The majority scored mild (grade 1, 30-49%) to moderate (grade 2, 39-51%) degenerative changes. Cord signal abnormality was seen in 56-63%. Cord signal abnormality, when present, occurred exclusively at degenerative disc levels in only 10-15%, significantly lower than other distributions (P < .001 for all pairwise comparisons). Conclusions: MS patients demonstrate unexpected cervical DDD even at a young age. Future study is warranted to investigate the underlying etiology, such as altered biomechanics. Furthermore, cord lesions were found to occur independently of DDD.

PMID:37339165 | DOI:10.1177/08465371231180815

JMIR Cardio. 2023 Jun 20;7:e45352. doi: 10.2196/45352.

ABSTRACT

BACKGROUND: The prediction of posttransplant health outcomes for pediatric heart transplantation is critical for risk stratification and high-quality posttransplant care.

OBJECTIVE: The purpose of this study was to examine the use of machine learning (ML) models to predict rejection and mortality for pediatric heart transplant recipients.

METHODS: Various ML models were used to predict rejection and mortality at 1, 3, and 5 years after transplantation in pediatric heart transplant recipients using United Network for Organ Sharing data from 1987 to 2019. The variables used for predicting posttransplant outcomes included donor and recipient as well as medical and social factors. We evaluated 7 ML models-extreme gradient boosting (XGBoost), logistic regression, support vector machine, random forest (RF), stochastic gradient descent, multilayer perceptron, and adaptive boosting (AdaBoost)-as well as a deep learning model with 2 hidden layers with 100 neurons and a rectified linear unit (ReLU) activation function followed by batch normalization for each and a classification head with a softmax activation function. We used 10-fold cross-validation to evaluate model performance. Shapley additive explanations (SHAP) values were calculated to estimate the importance of each variable for prediction.

RESULTS: RF and AdaBoost models were the best-performing algorithms for different prediction windows across outcomes. RF outperformed other ML algorithms in predicting 5 of the 6 outcomes (area under the receiver operating characteristic curve [AUROC] 0.664 and 0.706 for 1-year and 3-year rejection, respectively, and AUROC 0.697, 0.758, and 0.763 for 1-year, 3-year, and 5-year mortality, respectively). AdaBoost achieved the best performance for prediction of 5-year rejection (AUROC 0.705).

CONCLUSIONS: This study demonstrates the comparative utility of ML approaches for modeling posttransplant health outcomes using registry data. ML approaches can identify unique risk factors and their complex relationship with outcomes, thereby identifying patients considered to be at risk and informing the transplant community about the potential of these innovative approaches to improve pediatric care after heart transplantation. Future studies are required to translate the information derived from prediction models to optimize counseling, clinical care, and decision-making within pediatric organ transplant centers.

PMID:37338974 | DOI:10.2196/45352

Stat Methods Med Res. 2023 Jun 20:9622802231181220. doi: 10.1177/09622802231181220. Online ahead of print.

ABSTRACT

In clinical research, it is important to study whether certain clinical factors or exposures have causal effects on clinical and patient-reported outcomes such as toxicities, quality of life, and self-reported symptoms, which can help improve patient care. Usually, such outcomes are recorded as multiple variables with different distributions. Mendelian randomization (MR) is a commonly used technique for causal inference with the help of genetic instrumental variables to deal with observed and unobserved confounders. Nevertheless, the current methodology of MR for multiple outcomes only focuses on one outcome at a time, meaning that it does not consider the correlation structure of multiple outcomes, which may lead to a loss of statistical power. In situations with multiple outcomes of interest, especially when there are mixed correlated outcomes with different distributions, it is much more desirable to jointly analyze them with a multivariate approach. Some multivariate methods have been proposed to model mixed outcomes; however, they do not incorporate instrumental variables and cannot handle unmeasured confounders. To overcome the above challenges, we propose a two-stage multivariate Mendelian randomization method (MRMO) that can perform multivariate analysis of mixed outcomes using genetic instrumental variables. We demonstrate that our proposed MRMO algorithm can gain power over the existing univariate MR method through simulation studies and a clinical application on a randomized Phase III clinical trial study on colorectal cancer patients.

PMID:37338962 | DOI:10.1177/09622802231181220

Stat Methods Med Res. 2023 Jun 20:9622802231181224. doi: 10.1177/09622802231181224. Online ahead of print.

ABSTRACT

Censored data frequently appeared in applications across a variety of different areas like epidemiology or medical research. Traditionally statistical inference on this data mechanism was based on some pre-assigned models that will suffer from the risk of model-misspecification. This article proposes a two-folded shrinkage procedure for simultaneous structure identification and variable selection of the semiparametric accelerated failure additive model with right-censored data, in which the nonparametric functions are addressed by spline approximation. Under some regularity conditions, the consistency of model structure identification is theoretically established in the sense that the proposed method can automatically separate the linear and zero components from the nonlinear ones with probability approaching to one. Detailed issues in computation and turning parameter selection are also discussed. Finally, we illustrate the proposed method by some simulation studies and two real data applications to the primary biliary cirrhosis data and skin cutaneous melanoma data.

PMID:37338958 | DOI:10.1177/09622802231181224

J Bone Miner Res. 2023 Jun 20. doi: 10.1002/jbmr.4871. Online ahead of print.

ABSTRACT

Postoperative hypoparathyroidism (PO-hypoPT) is an uncommon complication of total thyroidectomy in thyroid cancer patients. While long-term hypoPT causes characteristic changes in bone metabolism, the risk of fractures in hypoPT remains inconclusive. We investigated the risk of fractures in Korean thyroid cancer patients with PO-hypoPT. This was a retrospective cohort study using data from the Korea Central Cancer Registry and Korean National Health Insurance Service. We analyzed 115,821 thyroid cancer patients aged ≥18 years, who underwent total thyroidectomy between 2008 and 2016. The risk of any fractures, including vertebral, hip, humerus, and wrist fractures, according to parathyroid function after total thyroidectomy was analyzed using the multivariable Cox proportional hazard model. The PO-hypoPT and preserved parathyroid function groups included 8,789 (7.6%) and 107,032 (92.4%) patients, respectively. Over a mean follow-up duration of 4.8 years, 159 (1.8%) and 2,390 (2.2%) fractures occurred in the PO-hypoPT and preserved parathyroid function groups, respectively. The risk of any fractures was significantly lower in the PO-hypoPT group than in the preserved parathyroid function group (HR, 0.83; 95% CI, 0.70-0.98; P=0.037) after adjusting for confounders. Regarding the fracture site, only the risk of vertebral fractures was significantly lower in the PO-hypoPT group compared to the preserved parathyroid function group (HR, 0.67; 95% CI, 0.47-0.96; P=0.028) after adjusting for confounders. Subgroup analyses showed that bone mineral density measurements and calcium supplementation interacted with the relationship between PO-hypoPT and the risk of any fractures (P for interactions=0.010 and 0.017, respectively). PO-hypoPT was associated with a lower risk of fractures in thyroid cancer patients, especially at the vertebra. The relatively low bone turnover caused by PO-hypoPT and appropriate management for PO-hypoPT with active vitamin D and calcium may prevent the deterioration of skeletal health in thyroid cancer patients who can easily be exposed to long-term overtreatment with levothyroxine. This article is protected by copyright. All rights reserved.

PMID:37338940 | DOI:10.1002/jbmr.4871