ORCID Profile
0000-0002-5069-3197
Current Organisation
King's College London
Does something not look right? The information on this page has been harvested from data sources that may not be up to date. We continue to work with information providers to improve coverage and quality. To report an issue, use the Feedback Form.
Publisher: Springer Science and Business Media LLC
Date: 13-06-2017
DOI: 10.1038/NCOMMS15833
Abstract: The persistence of common, heritable psychiatric disorders that reduce reproductive fitness is an evolutionary paradox. Here, we investigate the selection pressures on sequence variants that predispose to schizophrenia, autism, bipolar disorder, major depression and attention deficit hyperactivity disorder (ADHD) using genomic data from 150,656 Icelanders, excluding those diagnosed with these psychiatric diseases. Polygenic risk of autism and ADHD is associated with number of children. Higher polygenic risk of autism is associated with fewer children and older age at first child whereas higher polygenic risk of ADHD is associated with having more children. We find no evidence for a selective advantage of a high polygenic risk of schizophrenia or bipolar disorder. Rare copy-number variants conferring moderate to high risk of psychiatric illness are associated with having fewer children and are under stronger negative selection pressure than common sequence variants.
Publisher: Informa UK Limited
Date: 09-2016
Publisher: Wiley
Date: 23-12-2017
DOI: 10.1002/OBY.21715
Publisher: Elsevier BV
Date: 05-2012
Publisher: Wiley
Date: 07-2014
DOI: 10.1002/UOG.13306
Abstract: To determine the effect of different ultrasound machine-probe combinations on nuchal translucency (NT) measurements and to assess how this impacts on the accuracy of the NT-derived component of first-trimester screening for trisomy 21. Sixteen different ultrasound machine-probe combinations were used for axial measurement of 2.0-, 3.0- and 4.0-mm spaced targets in an ultrasound phantom. Differences between the measured and known values were determined. The mean of the axial measurements was used to calculate adjusted risks for trisomy 21, given specific clinical scenarios. Differences observed using different machine-probe combinations for the 2.0-mm target ranged from 1.8-2.2 mm for the 3.0-mm target, 2.7-3.2 mm and for the 4-mm target, 3.7-4.3 mm, and exceeded those due to intraobserver variability. For a fetal crown-rump length of 50.0 mm and NT measurement of 2.0 mm, the maximum/minimum measurements in the fetus of a 40-year-old woman led to derived risks ranging from 1 in 32 (NT, 2.2 mm) to 1 in 189 (NT, 1.8 mm) and in the fetus of a 20-year-old with an NT of 3.0 mm these ranged from 1 in 102 (NT, 3.2 mm) to 1 in 160 (NT, 2.7 mm). We have described the effect of machine-probe combinations on small but very precise ultrasound measurements. Such machine-probe combinations led to greater variability than those ascribed to intraobserver differences, and significantly affected the screening risk for the same fixed measurement. This finding has implications for Down syndrome screening algorithms and audit of ultrasound operators. Furthermore, most ultrasound machines are neither calibrated nor specified for measurements of tenths of a mm.
Publisher: Oxford University Press (OUP)
Date: 13-09-2011
DOI: 10.1093/BIOSTATISTICS/KXR030
Abstract: Relative risks (RRs) are generally considered preferable to odds ratios in prospective studies. However, unlike logistic regression for odds ratios, the standard log-binomial model for RR regression does not respect the natural parameter constraints and is therefore often subject to numerical instability. In this paper, we develop a reliable and flexible method for fitting log-binomial models. We use an Expectation–Maximization (EM) algorithm where the multiplicative event probability is viewed as the joint probability for a collection of latent binary outcomes. This gives a simple iterative scheme that provides stable convergence to the maximum likelihood estimate. In addition to reliability, the method offers some flexible generalizations, including models with unspecified isotonic regression functions. We examine the method's performance using simulations and data analyses of the age-specific RR of mortality following heart attack. These analyses demonstrate the potential for numerical instability in RR regression and show how this can be overcome using the proposed approach. Source code to implement the method in R is provided as supplementary material available at Biostatistics online.
Publisher: Wiley
Date: 19-04-2022
DOI: 10.1002/GEPI.22449
Abstract: Substantial advances have been made in identifying genetic contributions to depression, but little is known about how the effect of genes can be modulated by the environment, creating a gene-environment interaction. Using multivariate reaction norm models (MRNMs) within the UK Biobank (N = 61294-91644), we investigate whether the polygenic and residual variance components of depressive symptoms are modulated by 17 a priori selected covariate traits-12 environmental variables and 5 biomarkers. MRNMs, a mixed-effects modelling approach, provide unbiased polygenic-covariate interaction estimates for a quantitative trait by controlling for outcome-covariate correlations and residual-covariate interactions. A continuous depressive symptom variable was the outcome in 17 MRNMs-one for each covariate trait. Each MRNM had a fixed-effects model (fixed effects included the covariate trait, demographic variables, and principal components) and a random effects model (where polygenic-covariate and residual-covariate interactions are modelled). Of the 17 selected covariates, 11 significantly modulate deviations in depressive symptoms through the modelled interactions, but no single interaction explains a large proportion of phenotypic variation. Results are dominated by residual-covariate interactions, suggesting that covariate traits (including neuroticism, childhood trauma, and BMI) typically interact with unmodelled variables, rather than a genome-wide polygenic component, to influence depressive symptoms. Only average sleep duration has a polygenic-covariate interaction explaining a demonstrably nonzero proportion of the variability in depressive symptoms. This effect is small, accounting for only 1.22% (95% confidence interval: [0.54, 1.89]) of variation. The presence of an interaction highlights a specific focus for intervention, but the negative results here indicate a limited contribution from polygenic-environment interactions.
Publisher: Cold Spring Harbor Laboratory
Date: 04-08-2021
DOI: 10.1101/2021.08.02.21261499
Abstract: Substantial advances have been made in identifying genetic contributions to depression, but little is known about how the effect of genes can be modulated by the environment, creating a gene-environment interaction. Using multivariate reaction norm models (MRNMs) within the UK Biobank (N=61294-91644), we investigate whether the polygenic and residual variation of depressive symptoms are modulated by 25 a-priori selected covariate traits: 12 environmental variables, 5 biomarkers and polygenic risk scores for 8 mental health disorders. MRNMs provide unbiased polygenic-covariate interaction estimates for a quantitative trait by controlling for outcome-covariate correlations and residual-covariate interactions. Of the 25 selected covariates, 11 significantly modulate depressive symptoms, but no single interaction explains a large proportion of phenotypic variation. Results are dominated by residual-covariate interactions, suggesting that covariate traits (including neuroticism, childhood trauma and BMI) typically interact with unmodelled variables, rather than a genome-wide polygenic score, to influence depressive symptoms. Only average sleep duration has a polygenic-covariate interaction explaining a demonstrably non-zero proportion of the variability in depressive symptoms. This effect is small, accounting for only 1.22% (95% CI [0.54,1.89]) of variation. The presence of an interaction highlights a specific focus for intervention, but the negative results here indicate a limited contribution from polygenic-environment interactions.
Publisher: Public Library of Science (PLoS)
Date: 06-2017
Publisher: Elsevier BV
Date: 10-2017
DOI: 10.1016/J.PREGHY.2017.06.007
Abstract: To investigate longitudinal fetal growth and growth velocity for commonly measured biometric parameters in women with chronic hypertension. Two centre retrospective European study of women with chronic hypertension ascertained at pregnancy booking. Ultrasound measurements of head circumference (HC), abdominal circumference (AC) and femur length (FL) were used to derive longitudinal fetal growth charts derived using functional linear discriminant analysis (FLDA). These were compared to existing cross sectional and longitudinal charts, as was birthweight. 282 women with a median of 3 third trimester ultrasound examinations were included. Gestation at delivery was 37.5weeks (SD 2.68), birthweight 3049g (SD 785). Birthweight 90th percentile 20.2%. Fetal size curves derived from women with chronic hypertension were no different to cross sectional and longitudinal charts for a normal population. Compared to a standard longitudinal biometry chart, growth velocity (mm/day) in chronic hypertension was higher for AC and FL at 30-32weeks (AC 1.447vs 1.357 p<0.05 FL 0.296vs 0.269 p<0.01) and 34-36weeks (AC 1.325vs 1.140 p<0.01 FL 0.248vs 0.198 p<0.01). In women with chronic hypertension there is an excess of both SGA and LGA babies compared to population standards. Growth velocity of the AC and FL was greater after 30weeks compared to a normal population.
Publisher: Wiley
Date: 02-2014
DOI: 10.1002/UOG.13212
Location: United Kingdom of Great Britain and Northern Ireland
Location: United Kingdom of Great Britain and Northern Ireland
Location: United Kingdom of Great Britain and Northern Ireland
No related grants have been discovered for Alexandra Gillett.