Categories
Nevin Manimala Statistics

Pitfalls in brain age analyses

Hum Brain Mapp. 2021 Jun 30. doi: 10.1002/hbm.25533. Online ahead of print.

ABSTRACT

Over the past decade, there has been an abundance of research on the difference between age and age predicted using brain features, which is commonly referred to as the “brain age gap.” Researchers have identified that the brain age gap, as a linear transformation of an out-of-sample residual, is dependent on age. As such, any group differences on the brain age gap could simply be due to group differences on age. To mitigate the brain age gap’s dependence on age, it has been proposed that age be regressed out of the brain age gap. If this modified brain age gap is treated as a corrected deviation from age, model accuracy statistics such as R2 will be artificially inflated to the extent that it is highly improbable that an R2 value below .85 will be obtained no matter the true model accuracy. Given the limitations of proposed brain age analyses, further theoretical work is warranted to determine the best way to quantify deviation from normality.

PMID:34190372 | DOI:10.1002/hbm.25533

Categories
Nevin Manimala Statistics

Integration of survival data from multiple studies

Biometrics. 2021 Jun 30. doi: 10.1111/biom.13517. Online ahead of print.

ABSTRACT

We introduce a statistical procedure that integrates datasets from multiple biomedical studies to predict patients’ survival, based on individual clinical and genomic profiles. The proposed procedure accounts for potential differences in the relation between predictors and outcomes across studies, due to distinct patient populations, treatments and technologies to measure outcomes and biomarkers. These differences are modeled explicitly with study-specific parameters. We use hierarchical regularization to shrink the study-specific parameters towards each other and to borrow information across studies. The estimation of the study-specific parameters utilizes a similarity matrix, which summarizes differences and similarities of the relations between covariates and outcomes across studies. We illustrate the method in a simulation study and using a collection of gene expression datasets in ovarian cancer. We show that the proposed model increases the accuracy of survival predictions compared to alternative meta-analytic methods.

PMID:34190337 | DOI:10.1111/biom.13517

Categories
Nevin Manimala Statistics

Varying-coefficient regression analysis for pooled biomonitoring

Biometrics. 2021 Jun 30. doi: 10.1111/biom.13516. Online ahead of print.

ABSTRACT

Human biomonitoring involves measuring the accumulation of contaminants in biological specimens (such as blood or urine) to assess individuals exposure to environmental contamination. Due to the expensive cost of a single assay, the method of pooling has become increasingly common in environmental studies. The implementation of pooling starts by physically mixing specimens into pools, and then measures pooled specimens for the concentration of contaminants. An important task is to reconstruct individual-level statistical characteristics based on pooled measurements. In this article, we propose to use the varying-coefficient regression model for individual-level biomonitoring and provide methods to estimate the varying-coefficients based on different types of pooled data. Asymptotic properties of the estimators are presented. We illustrate our methodology via simulation and with application to pooled biomonitoring of a brominated flame retardant provided by the National Health and Nutrition Examination Survey (NHANES). This article is protected by copyright. All rights reserved.

PMID:34190334 | DOI:10.1111/biom.13516

Categories
Nevin Manimala Statistics

Synthesizing Indirect Effects in Mediation Models With Meta-Analytic Methods

Alcohol Alcohol. 2021 Jun 30:agab044. doi: 10.1093/alcalc/agab044. Online ahead of print.

ABSTRACT

AIMS: A mediator is a variable that explains the underlying mechanism between an independent variable and a dependent variable. The indirect effect indicates the effect from the predictor to the outcome variable via the mediator. In contrast, the direct effect represents the predictor’s effort on the outcome variable after controlling for the mediator.

METHODS: A single study rarely provides enough evidence to answer research questions in a particular domain. Replications are generally recommended as the gold standard to conduct scientific research. When a sufficient number of studies have been conducted addressing similar research questions, a meta-analysis can be used to synthesize those studies’ findings.

RESULTS: The main objective of this paper is to introduce two frameworks to integrating studies using mediation analysis. The first framework involves calculating standardized indirect effects and direct effects and conducting a multivariate meta-analysis on those effect sizes. The second one uses meta-analytic structural equation modeling to synthesize correlation matrices and fit mediation models on the average correlation matrix. We illustrate these procedures on a real dataset using the R statistical platform.

CONCLUSION: This paper closes with some further directions for future studies.

PMID:34190317 | DOI:10.1093/alcalc/agab044

Categories
Nevin Manimala Statistics

HIV infection and risk of COVID-19 mortality: A meta-analysis

Medicine (Baltimore). 2021 Jul 2;100(26):e26573. doi: 10.1097/MD.0000000000026573.

ABSTRACT

BACKGROUND: This meta-analysis aimed to estimate the association of human immunodeficiency virus (HIV) infection and risk of coronavirus disease 2019 (COVID-19) mortality.

METHODS: We systematically retrieved articles published on HIV infection and risk of COVID-19 mortality through PubMed, EMBase, China National Knowledge Infrastructure, WanFang, and Chongqing VIP databases using a predefined search strategy from December 1, 2019 to January 31, 2021. Newcastle-Ottawa Scale (NOS) was used to assess the quality of the included studies. Cochran Q test and I2 statistics were quantified to measure heterogeneity. Odds ratio (OR) and 95% confidence intervals (CI) were computed and displayed in the form of forest plots. Subgroup analysis was performed to explore the source of heterogeneity. Funnel plot, Begg test, and Egger test were used to assess potential publication bias. Stata software version 11.0 was used to analyze all the statistical data.

RESULTS: We included 10 studies with 18,122,370 COVID-19 patients, of whom 41,113 were with HIV infection and 18,081,257 were without HIV infection. The pooled overall results suggested that people living with HIV infection had a higher risk of mortality from COVID-19 than those without HIV infection (OR = 1.252, 95% CI 1.027-1.524). Subgroup analysis showed that people living with HIV infection had a higher risk of COVID-19 mortality than those without HIV infection in the United States (OR = 1.520, 95% CI 1.252-1.845) and in South Africa (OR = 1.122, 95% CI 1.032-1.220); however, no significant association was found in the United Kingdom (OR = 0.878, 95% CI 0.657-1.174).

CONCLUSION: Patients with HIV infection should be the emphasis population to prevent the risk of mortality during the clinical treatment of COVID-19 patients.

PMID:34190201 | DOI:10.1097/MD.0000000000026573

Categories
Nevin Manimala Statistics

Association study of the serum 25(OH)D concentration and myopia in Chinese children

Medicine (Baltimore). 2021 Jul 2;100(26):e26570. doi: 10.1097/MD.0000000000026570.

ABSTRACT

To analyze the serum 25 hydroxyvitamin D (25[OH]D) concentration in Chinese children with myopia and explore its correlation with myopia.From July to September in 2019, myopic children were collected from the Myopia Influencing Factors Survey Project. The basic information and vision related behaviors of the subjects were collected by questionnaire. The diopter of the children without dilated pupils was measured by the computerized refractometer. Meanwhile, 5 ml fasting venous blood samples were collected for the determination of serum 25(OH)D concentration.A total of 186 children were included in this study, including 90 males and 96 females, with an average age of 8 ± 3.26 years. The detection rate of serum 25(OH)D deficiency in myopic children was 65.59% (122/186). There was statistical significance in the detection rate of serum 25(OH)D deficiency in children with different myopic degrees (χ2 = 6.635, P = .010). The average serum 25(OH)D concentration in myopic children was 14.86 (10.67-18.96) ng/ml, and the difference of serum 25(OH)D concentration in children with different myopia degrees was statistically significant (Z = 20.23, P < .001). Logistic regression analysis showed that after controlling for gender, parental myopia, after-school class, and outdoor activities, the prevalence of developing moderate and high myopia was 2.051 times (95% confidence interval: 1.272-3.724) higher in the serum 25(OH)D deficiency group than in the serum 25(OH)D sufficiency group. There is a positive correlation between serum 25(OH)D concentration and the equivalent spherical degree of myopic children.The study found that serum 25(OH)D concentration is closely related to the prevalence of myopia in Chinese children. The results further support the conclusion that children with a higher level of serum 25(OH)D have a lower prevalence of moderate to high myopia. The results of this study provide a basis for further research into the relationship between vitamin D and visual development in children and its mechanisms.

PMID:34190200 | DOI:10.1097/MD.0000000000026570

Categories
Nevin Manimala Statistics

Climate pacing of millennial sea-level change variability in the central and western Mediterranean

Nat Commun. 2021 Jun 29;12(1):4013. doi: 10.1038/s41467-021-24250-1.

ABSTRACT

Future warming in the Mediterranean is expected to significantly exceed global values with unpredictable implications on the sea-level rise rates in the coming decades. Here, we apply an empirical-Bayesian spatio-temporal statistical model to a dataset of 401 sea-level index points from the central and western Mediterranean and reconstruct rates of sea-level change for the past 10,000 years. We demonstrate that the mean rates of Mediterranean industrial-era sea-level rise have been significantly faster than any other period since ~4000 years ago. We further highlight a previously unrecognized variability in Mediterranean sea-level change rates. In the Common Era, this variability correlates with the occurrence of major regional-scale cooling/warming episodes. Our data show a sea-level stabilization during the Late Antique Little Ice Age cold event, which interrupted a general rising trend of ~0.45 mm a-1 that characterized the warming episodes of the Common Era. By contrast, the Little Ice Age cold event had only minor regional effects on Mediterranean sea-level change rates.

PMID:34188029 | DOI:10.1038/s41467-021-24250-1

Categories
Nevin Manimala Statistics

An Investigation of Mental Health Status Among Medical Staff Following COVID-19 Outbreaks: A Cross-Sectional Study

Med Sci Monit. 2021 Jun 30;27:e929454. doi: 10.12659/MSM.929454.

ABSTRACT

BACKGROUND COVID-19 (coronavirus disease 2019) broke out in China. This study was to investigate the situation of mental health status among medical staff following COVID-19. MATERIAL AND METHODS A cross-sectional study was conducted through structured questionnaires to collect the demographical information of the participating medical staff via WeChat following COVID-19 crisis. The Center for Epidemiologic Studies-Depression Scale (CES-D), impact of events scale revised (IES-R), and Pittsburgh Sleep Quality Index (PSQI) scale were used to evaluate depression, post-traumatic stress disorder (PTSD) symptoms, and sleep quality, respectively. 95% confidence intervals (CI) were calculated. RESULTS A total of 597 medical staff’s information was included for the statistical analysis, and found 45.23% of subjects had PTSD symptoms, the mean PSQI score was 6.320±3.587. The results of multivariable analysis implied that medical workers who did not participate in the Hubei aid program (ß=4.128; 95% CI, 0.983-7.272; P=0.010) and PTSD symptoms (ß=7.212; 95% CI, 4.807-9.616; P<0.001) were associated with a higher tendency to depression. The PSQI score was linearly related to the CES-D score (ß=1.125; 95% CI, 0.804-1.445; P<0.001). Subgroup analysis showed that medical workers who did not participate in the Hubei aid program, no traumatic experience before COVID-19 outbreaks, and PTSD symptoms may affect the tendency to depression in females, but not in males. PSQI score was linearly related to the CES-D score both in males and females. CONCLUSIONS The medical staff with PTSD symptoms and higher PSQI score may have a higher tendency to depression following COVID-19 outbreaks.

PMID:34188013 | DOI:10.12659/MSM.929454

Categories
Nevin Manimala Statistics

Reply

J Hypertens. 2021 Aug 1;39(8):1726-1727. doi: 10.1097/HJH.0000000000002877.

NO ABSTRACT

PMID:34188008 | DOI:10.1097/HJH.0000000000002877

Categories
Nevin Manimala Statistics

Cardiovascular outcomes in patients at high cardiovascular risk with previous myocardial infarction or stroke

J Hypertens. 2021 Aug 1;39(8):1602-1610. doi: 10.1097/HJH.0000000000002822.

ABSTRACT

BACKGROUND: Guidelines recommend to start blood pressure (BP)-lowering drugs also according to cardiovascular risk including history of cardiovascular events. We hypothesized that in patients with a history of myocardial infarction (MI), stroke, both or none of those, the index events predict the next event and have different SBP risk associations to different cardiovascular outcomes.

DESIGN AND MEASUREMENTS: In this pooled posthoc, nonprespecified analysis, we assessed outcome data from high-risk patients aged 55 years or older with a history of cardiovascular events or proven cardiovascular disease, randomized to the Ongoing Telmisartan Alone and in Combination with Ramipril Global Endpoint Trial and to Telmisartan Randomized Assessment Study in ACE Intolerant Subjects with Cardiovascular Disease Trial investigating telmisartan, ramipril and their combination with a median follow-up of 56 months. Standardized office BP was measured every 6 months. Associations of mean achieved BP on treatment were investigated on MI, stroke and cardiovascular death. We identified patients with previous MI (N = 13 487), stroke (N = 4985), both (N = 1509) or none (N = 10 956) of these index events. Analyses were done by Cox regression, analysis of variance and Chi2-test. 30 937 patients with complete data were enrolled between 1 December 2001 and 31 July 2003, and followed until 31 July 2008. Data of both trials were pooled as the outcomes were similar.

RESULTS: Patients with MI as index event had a higher risk to experience a second MI [hazard ratio 1.42 (confidence interval (CI) 1.20-1.69), P < 0.0001] compared with patients with no events but no increased risk for a stroke as a next event [hazard ratio 0.95 (CI 0.73-1.23), n.s.]. The risk was roughly doubled when they had both, MI and stroke before [hazard ratio 2.07 (CI 1.58-2.71), P < 0.0001]. Patients with a stroke history had a roughly three-fold higher likelihood to experience a second stroke [hazard ratio 2.89 (CI 2.37-3.53) P < 0.0001] but not MI [hazard ratio 1.07 (CI 0.88-1.32), n.s.]. Both types of index events increased roughly three-fold the risk of a second stroke compared with no previous events. The SBP-risk relationship was not meaningfully altered by the event history. After MI and stroke the risk for subsequent events and cardiovascular death was increased over the whole SBP spectrum. A J-shape relationship between BP and outcome was only observed for cardiovascular death.

CONCLUSION: Previous MI and previous stroke are associated with increased risk for the same event in the future, independent of achieved SBP. Thus, secondary prevention may also be chosen according to the event history of patients.

CLINICAL TRIAL REGISTRATION: http://clinicaltrials.gov. Unique identifier: NCT00153101.

PMID:34188004 | DOI:10.1097/HJH.0000000000002822