Category: Statistics

Sleep. 2021 May 26:zsab027. doi: 10.1093/sleep/zsab027. Online ahead of print.

ABSTRACT

Current approaches to automated sleep staging from the electroencephalogram (EEG) rely on constructing a large labeled training and test corpora by aggregating data from different individuals. However, many of the subjects in the training set may exhibit changes in the EEG that are very different from the subjects in the test set. Training an algorithm on such data without accounting for this diversity can cause underperformance. Moreover, test data may have unexpected sensor misplacement or different instrument noise and spectral responses. This work proposes a novel method to learn relevant individuals based on their similarities effectively. The proposed method embeds all training patients into a shared and robust feature space. Individuals that share strong statistical relationships and are similar based on their EEG signals are clustered in this feature space before being passed to a deep learning framework for classification. Using 994 patient EEGs from the 2018 Physionet Challenge (≈ 6,561 hours of recording), we demonstrate that the clustering approach significantly boosts performance compared to state-of-the-art deep learning approaches. The proposed method improves, on average, a precision score from 0.72 to 0.81, a sensitivity score from 0.74 to 0.82, and a Cohen’s Kappa coefficient from 0.64 to 0.75 under 10-fold cross-validation.

PMID:34038560 | DOI:10.1093/sleep/zsab027

Ann Neurol. 2021 May 26. doi: 10.1002/ana.26131. Online ahead of print.

ABSTRACT

OBJECTIVE: Polygenic risk scores (PRSs) assess the individual genetic propensity to a condition by combining sparse information scattered across genetic loci, often displaying small effect sizes. Most PRSs are constructed in European-ancestry populations, limiting their use in other ethnicities. Here we constructed and validated a PRS for late-onset Alzheimer’s Disease (LOAD) in Caribbean Hispanics (CH).

METHODS: We employed a CH discovery (N=4,312) and independent validation sample (N=1,850) to construct an ancestry-specific PRS (“CH-PRS”) and evaluated its performance alone and with other predictors using the area under curve (AUC) and logistic regression (strength of association with LOAD and statistical significance). We tested if CH-PRS predicted conversion to LOAD in a subsample with longitudinal data (N=1,239). We also tested the CH-PRS in an independent replication CH cohort (N=200) and brain autopsy cohort (N=33). Finally, we tested the effect of ancestry on PRS by employing European and African American discovery cohorts to construct alternative PRSs (“EUR-PRS”, “AA-PRS”).

RESULTS: The full model (LOAD ~ CH-PRS + sex + age + APOE-&ip.eop;4), achieved an AUC=74% (OR_CH-PRS =1.51 95%CI=1.36-1.68), raising to >75% in APOE-&ip.eop;4 non-carriers. CH-PRS alone achieved an AUC=72% in the autopsy cohort, raising to AUC=83% in full model. Higher CH-PRS was significantly associated with clinical LOAD in the replication CH cohort (OR=1.61, 95%CI=1.19-2.17) and significantly predicted conversion to LOAD (HR=1.93, CI=1.70-2.20) in the longitudinal subsample. EUR-PRS and AA-PRS reached lower prediction accuracy (AUC=58% and 53%, respectively).

INTERPRETATION: Enriching diversity in genetic studies is critical to provide an effective PRS in profiling LOAD risk across populations. This article is protected by copyright. All rights reserved.

PMID:34038570 | DOI:10.1002/ana.26131

J Refract Surg. 2021 Apr;37(4):256-262. doi: 10.3928/1081597X-20210209-01. Epub 2021 Apr 1.

ABSTRACT

PURPOSE: To evaluate the clinical outcomes including patient-reported outcome measures in a sample of eyes undergoing bilateral cataract surgery with implantation of a new model of presbyopia-correcting intraocular lens (IOL).

METHODS: This non-randomized prospective case series enrolled 206 eyes of 103 patients undergoing phacoemulsification cataract surgery with bilateral implantation of the TECNIS Synergy IOL (Johnson & Johnson Vision). High and low contrast visual acuity, refractive, defocus curve, and patient-reported visual performance (Catquest-9SF questionnaire) outcomes were evaluated during a 3-month follow-up.

RESULTS: A total of 96.1% (99 of 103) and 91.3% (94 of 103) of patients achieved binocular postoperative uncorrected distance (UDVA) and near visual acuity (UNVA) of 0.00 logMAR (20/20), respectively. Mean postoperative mesopic UNVA for both eyes was 0.14 ± 0.03 logMAR. Likewise, mean binocular UDVA and UNVA were 0.00 ± 0.03 and 0.04 ± 0.02 logMAR. An almost flat mean defocus curve was obtained, with visual acuities between 0.00 and 0.10 logMAR for most defocus levels in both eyes. A reduction of contrast led to a limited but statistically significant change in UNVA in both eyes (P < .001). The Rasch calibrated scoring of item 2 and the Rasch calibrated mean score of the Catquest-9SF questionnaire increased significantly with surgery (P < .001).

CONCLUSIONS: This new presbyopia-correcting IOL provides a continuous range of functional focus, with a limited deterioration under mesopic conditions, which is perceived as a satisfactory outcome by the patient if proper patient selection is performed. [J Refract Surg. 2021;37(4):256-262.].

PMID:34038663 | DOI:10.3928/1081597X-20210209-01

Int Health. 2021 May 26:ihab020. doi: 10.1093/inthealth/ihab020. Online ahead of print.

ABSTRACT

BACKGROUND: Factors affecting the adoption and use of eHealth applications have been extensively researched from different perspectives in different regions. However, with the changing circumstances (e.g. the coronavirus disease 2019 pandemic), new influencing factors might evolve and can influence the attitudes of consumers towards using eHealth applications. The purpose of this study was to identify and evaluate the evolving factors affecting consumer attitudes towards the use of eHealth applications and provide implications for the future of Neom.

METHODS: An online survey questionnaire was used to collect data from 976 eHealth consumers in Saudi Arabia, which included 527 male and 449 female participants. Findings were analysed using the statistical means and standard deviations for each item in the questionnaire to analyse the role of each factor in depth. Statistical t-tests were used to identify significant differences between the groups categorised by age and gender.

RESULTS: ‘Necessity but not interest’ (mean 4.5 [standard deviation {SD} 1.12]) and fear (mean 4.5 [SD 1.13]) and psychological factors including depression (mean 4.4 [SD 1.54]), stress (mean 4.2 [SD 1.09]) and anxiety (mean 4.3 [SD 1.61]) were identified to be major evolving influencing factors, while other factors including performance expectancy, ease of use, enjoyment and incentives were identified to be comparatively less influential.

CONCLUSIONS: Increasing adoption of eHealth mainly due to necessity but not out of interest can have serious implications for patients and the adoption of eHealth technologies in the future.

PMID:34038551 | DOI:10.1093/inthealth/ihab020

J Cosmet Dermatol. 2021 May 26. doi: 10.1111/jocd.14255. Online ahead of print.

ABSTRACT

BACKGROUND: Monopolar radiofrequency(MRF) for facial rejuvenation works by diffusely heating the dermis and subdermal fibro-septal network. Immediate collagen contraction and long-term neo-collagenesis produce anti-aging effects.

AIM: To lift and repose drooping fat planes along their vectors of descent to restore natural, youthful contours using MRF and to objectively measure, quantify and analyse the anti-aging effects of MRF on the backdrop of functional facial anatomy.

METHODS: This was a prospective study on 30 subjects aged 25-65 years who underwent three sessions of MRF (Exilis,BTL Aesthetics) with a modified protocol at weekly intervals. Clinical evaluation at baseline and 4 weeks after the final session was done according to the Global drooping and wrinkle classification & scoring system(GDWCSS) and statistically evaluated. Two patients each were analyzed with Ultrasound and Vectra-H1 imaging for change in subcutaneous fat depth and volume over different facial areas, respectively.

RESULTS: There was improvement in all parameters (volume alteration and wrinkles) of the GDWCSS, of which improvement in cases that had baseline severe(grade-4) temporal fat loss, malar fat loss, periorbital hollowing and puffiness was statistically significant. The results of imaging supported the clinical findings.

CONCLUSION: MRF targets facial collagen as deep as the facial fat, thereby achieving ‘reversal’ of the grooves and bulges of aging with subtle volume re-distribution and favorable ‘remoulding’ of fat pads. The result is simultaneous skin tightening, lifting and contouring in manner that is natural, regenerative and corrective. Given its versatility, efficacy, safety and tolerability, it is emerging as a power-player in the dermatologist’s anti-aging armamentarium.

PMID:34038038 | DOI:10.1111/jocd.14255

Neotrop Entomol. 2021 May 26. doi: 10.1007/s13744-021-00883-1. Online ahead of print.

ABSTRACT

This work presents a simulation for the trajectory of a locust swarm of the species Schistocerca cancellata (Audinet-Serville), between May 22nd and July 29th, 2020, in Argentina. To obtain the directions, temperature, and intensities of the wind, used to determine the daily traveled distance of the insects, the data of weather forecast from Weather Research and Forecasting (WRF) model are used. A statistical analysis shows the effectiveness of the forecast model used in comparison with the real data given by SENASA, which provides latitude and longitude coordinates for the cited period. The results found for the movement of the cloud were satisfactory, they matched with the real data, identifying that temperature and wind speed have a great influence on the movement of locust swarms. The methodology used allows monitoring in real-time their movement, predicting the trajectory and making it possible to plan actions by government control agencies with pesticides in convenient areas.

PMID:34037969 | DOI:10.1007/s13744-021-00883-1

Fam Process. 2021 May 26. doi: 10.1111/famp.12666. Online ahead of print.

ABSTRACT

The working alliance has been shown to be a robust predictor of couple therapy outcomes. However, there are still questions regarding the best way to conceptualize and analyze the association between the alliance and outcomes in the couple therapy context. This study presents results from a relatively novel analytic approach for evaluating the alliance and therapy outcomes via the shared influence model (i.e., common fate statistical methodology) with 93 couples who had at least three sessions. Examining the alliance in couple therapy this way allows for nuance in conceptualizing the process of the alliance among partners and the therapist in couple therapy. Empirically, the question that remains is whether the shared experience of the alliance between partners is a stronger predictor of therapy outcome as compared to the unique individual experiences of the alliance. This approach has not yet been utilized to model the couple alliance-outcome association. Results revealed that the shared influence model demonstrated a significant association between the alliance and therapy outcome. Specifically, the couple-level alliance construct was adequately represented from the partners’ alliance ratings with the therapist, and couples that report higher shared alliance experiences were also likely to have better therapy outcomes. These findings suggest that conceptualizing and analyzing the alliance at the couple level may be strongly advantageous. Implications for research and practice are provided.

PMID:34037992 | DOI:10.1111/famp.12666

J Gen Intern Med. 2021 May 26. doi: 10.1007/s11606-021-06915-1. Online ahead of print.

ABSTRACT

BACKGROUND: Few studies have investigated the relationship between industry funding/conflicts of interest and authors’ positions in opinion pieces on drug safety. Harmful effects of varenicline, a treatment for smoking cessation, have been highly contested.

OBJECTIVE: To examine the association between pharmaceutical industry funding/authors’ financial conflicts of interest and position on varenicline in opinion articles, especially in relation to the minimization of harms; to assess whether opinion pieces on drug safety issues written by authors with conflicts of interest are more frequently cited in the news or social media.

DESIGN: Cross-sectional analysis.

PARTICIPANTS: English language opinion pieces and narrative reviews about varenicline published between May 2006 and February 2019.

MAIN MEASURES: Odds ratios and 95% confidence intervals; the Mann-Whitney two-sample statistic was used to test for differences in Altmetric scores, a measure of media attention.

KEY RESULTS: Of the 221 included articles, 30.3% (67) disclosed the funding source and 62.9% (139) disclosed authors’ conflicts of interest. Authors of opinion pieces on varenicline who reported financial ties to the pharmaceutical industry (as a conflict of interest or funding source) were more likely to minimise the cardiovascular and psychiatric risk of varenicline compared to those without conflicts of interest or industry funding (OR: 4.00; 95% CI: 1.32 to 12.16 for cardiovascular risk; OR: 8.51; 95% CI: 3.79 to 19.11 for psychiatric risk). These associations persisted in sensitivity analyses. No statistically significant difference in Altmetric score was found between articles with (mean 15.83, median 3) and without (mean 11.90, median 1) conflicts of interest, indicating similar media attention (p-value=0.11).

CONCLUSIONS: We found that authors with financial ties to drug companies were more likely to publish opinion pieces that minimised harms of varenicline. These results raise questions about journals’ editorial policies to accept reviews of treatments from authors with financial relationships with manufacturers.

PMID:34037923 | DOI:10.1007/s11606-021-06915-1

Can J Public Health. 2021 May 26. doi: 10.17269/s41997-021-00527-2. Online ahead of print.

ABSTRACT

OBJECTIVES: Poor housing conditions and household crowding have been identified as important health concerns for Indigenous populations in many countries but have not been explored in relation to adverse birth outcomes in these populations. We investigated housing conditions and adverse birth outcomes in a nationally representative sample of Indigenous people in Canada.

METHODS: Data were from a cohort of births between May 2004 and May 2006 created by linking birth and infant death registration data with the 2006 Canadian census. Log-binomial regression was used to examine associations between housing variables (persons per room and needed household repairs) and three adverse birth outcomes: preterm birth (PTB), small-for-gestational-age (SGA) birth, and infant mortality. Separate regression models were run for First Nations, Métis and Inuit mothers, with adjustment for parity and parental socio-economic variables.

RESULTS: Need for major household repairs was associated with a slightly increased risk of PTB among First Nations and Métis mothers (adjusted RRs 1.12 and 1.13, respectively; 95% CI 0.94-1.34 and 0.89-1.44, respectively) and a moderately increased risk of infant death in all three groups (aRR = 1.69, 95% CI 1.00-2.85). Household crowding was also associated with a slightly elevated risk of PTB in all three groups (aRR = 1.10, 95% CI 0.95-1.29) and with an increased risk of infant mortality among First Nations (aRR = 1.57, 95% CI 0.97-2.53).

CONCLUSION: This study highlights the need to improve understanding of links between housing conditions and perinatal health outcomes in Indigenous populations, including examining cause-specific infant mortality in relation to housing characteristics.

PMID:34037966 | DOI:10.17269/s41997-021-00527-2

Bull Math Biol. 2021 May 26;83(7):80. doi: 10.1007/s11538-021-00912-5.

ABSTRACT

Mechanical and mechanochemical models of pattern formation in biological tissues have been used to study a variety of biomedical systems, particularly in developmental biology, and describe the physical interactions between cells and their local surroundings. These models in their original form consist of a balance equation for the cell density, a balance equation for the density of the extracellular matrix (ECM), and a force-balance equation describing the mechanical equilibrium of the cell-ECM system. Under the assumption that the cell-ECM system can be regarded as an isotropic linear viscoelastic material, the force-balance equation is often defined using the Kelvin-Voigt model of linear viscoelasticity to represent the stress-strain relation of the ECM. However, due to the multifaceted bio-physical nature of the ECM constituents, there are rheological aspects that cannot be effectively captured by this model and, therefore, depending on the pattern formation process and the type of biological tissue considered, other constitutive models of linear viscoelasticity may be better suited. In this paper, we systematically assess the pattern formation potential of different stress-strain constitutive equations for the ECM within a mechanical model of pattern formation in biological tissues. The results obtained through linear stability analysis and the dispersion relations derived therefrom support the idea that fluid-like constitutive models, such as the Maxwell model and the Jeffrey model, have a pattern formation potential much higher than solid-like models, such as the Kelvin-Voigt model and the standard linear solid model. This is confirmed by the results of numerical simulations, which demonstrate that, all else being equal, spatial patterns emerge in the case where the Maxwell model is used to represent the stress-strain relation of the ECM, while no patterns are observed when the Kelvin-Voigt model is employed. Our findings suggest that further empirical work is required to acquire detailed quantitative information on the mechanical properties of components of the ECM in different biological tissues in order to furnish mechanical and mechanochemical models of pattern formation with stress-strain constitutive equations for the ECM that provide a more faithful representation of the underlying tissue rheology.

PMID:34037880 | DOI:10.1007/s11538-021-00912-5