ORCID Profile
0000-0002-2460-0508
Current Organisations
University College London
,
University of Bristol
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: Oxford University Press (OUP)
Date: 12-2016
DOI: 10.1093/IJE/DYX028
Publisher: Springer Science and Business Media LLC
Date: 19-09-2015
Publisher: Center for Open Science
Date: 04-03-2021
Abstract: Anorexia nervosa (AN) polygenic liability has been associated with mental health traits, eating problems, and body mass index (BMI) in adolescence and adulthood, but little is known about its manifestations in early childhood. We explore AN polygenic score (PGS) associations with six childhood domains: BMI, eating problems, neurodevelopment, emotional problems, disruptive/aggressive behaviour, and temperament ersonality in 15,205 children from the Norwegian Mother, Father and Child Cohort Study. Results did not support associations between AN PGS and developmental phenotypes in girls. For boys, we observed an association between AN PGS and higher temperamental fussiness at 6 months, (b= 0.036 [95% CI=0.010,0.061]). Our results suggest that genetic risk for AN as indexed by the PGS has few observable manifestations in key neurodevelopmental domains in the first 8 years of life. Future studies with more powerful PGS that track children longer may aid in understanding how and when genetic risk for AN manifest developmentally.
Publisher: BMJ
Date: 22-05-2019
DOI: 10.1136/BMJ.L1855
Abstract: To investigate the role of body mass index (BMI), systolic blood pressure, and smoking behaviour in explaining the effect of education on the risk of cardiovascular disease outcomes. Mendelian randomisation study. UK Biobank and international genome-wide association study data. Predominantly participants of European ancestry. Educational attainment, BMI, systolic blood pressure, and smoking behaviour in observational analysis, and randomly allocated genetic variants to instrument these traits in mendelian randomisation. The risk of coronary heart disease, stroke, myocardial infarction, and cardiovascular disease (all subtypes all measured in odds ratio), and the degree to which this is mediated through BMI, systolic blood pressure, and smoking behaviour respectively. Each additional standard deviation of education (3.6 years) was associated with a 13% lower risk of coronary heart disease (odds ratio 0.86, 95% confidence interval 0.84 to 0.89) in observational analysis and a 37% lower risk (0.63, 0.60 to 0.67) in mendelian randomisation analysis. As a proportion of the total risk reduction, BMI was estimated to mediate 15% (95% confidence interval 13% to 17%) and 18% (14% to 23%) in the observational and mendelian randomisation estimates, respectively. Corresponding estimates were 11% (9% to 13%) and 21% (15% to 27%) for systolic blood pressure and 19% (15% to 22%) and 34% (17% to 50%) for smoking behaviour. All three risk factors combined were estimated to mediate 42% (36% to 48%) and 36% (5% to 68%) of the effect of education on coronary heart disease in observational and mendelian randomisation analyses, respectively. Similar results were obtained when investigating the risk of stroke, myocardial infarction, and cardiovascular disease. BMI, systolic blood pressure, and smoking behaviour mediate a substantial proportion of the protective effect of education on the risk of cardiovascular outcomes and intervening on these would lead to reductions in cases of cardiovascular disease attributable to lower levels of education. However, more than half of the protective effect of education remains unexplained and requires further investigation.
Publisher: Springer Science and Business Media LLC
Date: 27-04-2019
Publisher: Springer Science and Business Media LLC
Date: 19-11-2019
DOI: 10.1038/S41467-019-12424-X
Abstract: Alcohol use is correlated within spouse-pairs, but it is difficult to disentangle effects of alcohol consumption on mate-selection from social factors or the shared spousal environment. We hypothesised that genetic variants related to alcohol consumption may, via their effect on alcohol behaviour, influence mate selection. Here, we find strong evidence that an in idual’s self-reported alcohol consumption and their genotype at rs1229984, a missense variant in ADH1B , are associated with their partner’s self-reported alcohol use. Applying Mendelian randomization, we estimate that a unit increase in an in idual’s weekly alcohol consumption increases partner’s alcohol consumption by 0.26 units (95% C.I. 0.15, 0.38 P = 8.20 × 10 −6 ). Furthermore, we find evidence of spousal genotypic concordance for rs1229984, suggesting that spousal concordance for alcohol consumption existed prior to cohabitation. Although the SNP is strongly associated with ancestry, our results suggest some concordance independent of population stratification. Our findings suggest that alcohol behaviour directly influences mate selection.
Publisher: Springer Science and Business Media LLC
Date: 30-07-2019
Publisher: IEEE
Date: 11-2016
Publisher: American Association for the Advancement of Science (AAAS)
Date: 17-04-2020
Abstract: In this paper, we review and demonstrate how population phenomena can inflate genotype-phenotype associations.
Publisher: Springer Science and Business Media LLC
Date: 14-07-2020
DOI: 10.1038/S41467-020-17117-4
Abstract: Estimates from Mendelian randomization studies of unrelated in iduals can be biased due to uncontrolled confounding from familial effects. Here we describe methods for within-family Mendelian randomization analyses and use simulation studies to show that family-based analyses can reduce such biases. We illustrate empirically how familial effects can affect estimates using data from 61,008 siblings from the Nord-Trøndelag Health Study and UK Biobank and replicated our findings using 222,368 siblings from 23andMe. Both Mendelian randomization estimates using unrelated in iduals and within family methods reproduced established effects of lower BMI reducing risk of diabetes and high blood pressure. However, while Mendelian randomization estimates from s les of unrelated in iduals suggested that taller height and lower BMI increase educational attainment, these effects were strongly attenuated in within-family Mendelian randomization analyses. Our findings indicate the necessity of controlling for population structure and familial effects in Mendelian randomization studies.
Publisher: Cold Spring Harbor Laboratory
Date: 09-04-2019
DOI: 10.1101/602516
Abstract: Mendelian randomization (MR) is a widely-used method for causal inference using genetic data. Mendelian randomization studies of unrelated in iduals may be susceptible to bias from family structure, for ex le, through dynastic effects which occur when parental genotypes directly affect offspring phenotypes. Here we describe methods for within-family Mendelian randomization and through simulations show that family-based methods can overcome bias due to dynastic effects. We illustrate these issues empirically using data from 61,008 siblings from the UK Biobank and Nord-Trøndelag Health Study. Both within-family and population-based Mendelian randomization analyses reproduced established effects of lower BMI reducing risk of diabetes and high blood pressure. However, while MR estimates from population-based s les of unrelated in iduals suggested that taller height and lower BMI increase educational attainment, these effects largely disappeared in within-family MR analyses. We found differences between population-based and within-family based estimates, indicating the importance of controlling for family effects and population structure in Mendelian randomization studies.
Publisher: The Royal Society
Date: 03-2019
DOI: 10.1098/RSOS.181049
Abstract: Schizophrenia is a debilitating and heritable mental disorder associated with lower reproductive success. However, the prevalence of schizophrenia is stable over populations and time, resulting in an evolutionary puzzle: how is schizophrenia maintained in the population, given its apparent fitness costs? One possibility is that increased genetic liability for schizophrenia, in the absence of the disorder itself, may confer some reproductive advantage. We assessed the correlation and causal effect of genetic liability for schizophrenia with number of children, age at first birth and number of sexual partners using data from the Psychiatric Genomics Consortium and UK Biobank. Linkage disequilibrium score regression showed little evidence of genetic correlation between genetic liability for schizophrenia and number of children ( r g = 0.002, p = 0.84), age at first birth ( r g = −0.007, p = 0.45) or number of sexual partners ( r g = 0.007, p = 0.42). Mendelian randomization indicated no robust evidence of a causal effect of genetic liability for schizophrenia on number of children (mean difference: 0.003 increase in number of children per doubling in the natural log odds ratio of schizophrenia risk, 95% confidence interval (CI): −0.003 to 0.009, p = 0.39) or age at first birth (−0.004 years lower age at first birth, 95% CI: −0.043 to 0.034, p = 0.82). We find some evidence of a positive effect of genetic liability for schizophrenia on number of sexual partners (0.165 increase in the number of sexual partners, 95% CI: 0.117–0.212, p = 5.30×10 −10 ). These results suggest that increased genetic liability for schizophrenia does not confer a fitness advantage but does increase mating success.
Publisher: Springer Science and Business Media LLC
Date: 18-01-2019
DOI: 10.1038/S41467-018-08219-1
Abstract: Large studies use genotype data to discover genetic contributions to complex traits and infer relationships between those traits. Co-incident geographical variation in genotypes and health traits can bias these analyses. Here we show that single genetic variants and genetic scores composed of multiple variants are associated with birth location within UK Biobank and that geographic structure in genotype data cannot be accounted for using routine adjustment for study centre and principal components derived from genotype data. We find that major health outcomes appear geographically structured and that coincident structure in health outcomes and genotype data can yield biased associations. Understanding and accounting for this phenomenon will be important when making inference from genotype data in large studies.
Publisher: Center for Open Science
Date: 13-07-2021
Abstract: Dietary factors are assumed to play an important role in cancer risk, apparent in consensus recommendations for cancer prevention that promote nutritional changes. However, the evidence in this field has been generated predominantly through observational studies, which may result in biased effect estimates because of confounding, exposure misclassification, and reverse causality. With major geographical differences and rapid changes in cancer incidence over time, it is crucial to establish which of the observational associations reflect causality and to identify novel risk factors as these may be modified to prevent the onset of cancer and reduce its progression. Mendelian randomization (MR) uses the special properties of germline genetic variation to strengthen causal inference regarding potentially modifiable exposures and disease risk. MR can be implemented through instrumental variable (IV) analysis and, when robustly performed, is generally less prone to confounding, reverse causation and measurement error than conventional observational methods and has different sources of bias (discussed in detail below). It is increasingly used to facilitate causal inference in epidemiology and provides an opportunity to explore the effects of nutritional exposures on cancer incidence and progression in a cost-effective and timely manner.Here, we introduce the concept of MR and discuss its current application in understanding the impact of nutritional factors (e.g., any measure of diet and nutritional intake, circulating biomarkers, patterns, preference or behaviour) on cancer aetiology and, thus, opportunities for MR to contribute to the development of nutritional recommendations and policies for cancer prevention. We provide applied ex les of MR studies examining the role of nutritional factors in cancer to illustrate how this method can be used to help prioritise or deprioritise the evaluation of specific nutritional factors as intervention targets in randomised controlled trials. We describe possible biases when using MR, and methodological developments aimed at investigating and potentially overcoming these biases when present. Lastly, we consider the use of MR in identifying causally relevant nutritional risk factors for various cancers in different regions across the world, given notable geographical differences in some cancers. We also discuss how MR results could be translated into further research and policy. We conclude that findings from MR studies, which corroborate those from other well-conducted studies with different and orthogonal biases, are poised to substantially improve our understanding of nutritional influences on cancer. For such corroboration, there is a requirement for an interdisciplinary and collaborative approach to investigate risk factors for cancer incidence and progression.
Publisher: Center for Open Science
Date: 13-03-2021
Abstract: Estimating effects of parental and sibling genotypes (indirect genetic effects) can provide insight into how the family environment influences phenotypic variation. There is growing molecular genetic evidence for effects of parental phenotypes on their offspring (e.g. parental educational attainment), but the extent to which siblings affect each other is currently unclear.Here we used data from s les of unrelated in iduals, without (singletons) and with biological full-siblings (non-singletons), to investigate and estimate sibling effects. Indirect genetic effects of siblings increase (or decrease) the covariance between genetic variation and a phenotype. It follows that differences in genetic association estimates between singletons and non-singletons could indicate indirect genetic effects of siblings.We used UK Biobank data to estimate polygenic risk score (PRS) associations for height, BMI and educational attainment in singletons (N = 50,143) and non-singletons (N = 328,549). The educational attainment PRS association estimate was 12% larger (95% C.I. 3%, 21%) in the non-singleton s le than in the singleton s le, but the height and BMI PRS associations were consistent. Birth order data suggested that the difference in educational attainment PRS associations was driven by in iduals with older siblings rather than firstborns. The relationship between number of siblings and educational attainment PRS associations was non-linear PRS associations were 24% smaller in in iduals with 6 or more siblings compared to the rest of the s le (95% C.I. 11%, 38%). We estimate that a 1 SD increase in sibling educational attainment PRS corresponds to a 0.025 year increase in the index in idual’s years in schooling (95% C.I. 0.013, 0.036).Our results suggest that older siblings influence the educational attainment of younger siblings, adding to the growing evidence that effects of the environment on phenotypic variation partially reflect social effects of germline genetic variation in relatives.
Publisher: American Psychiatric Association Publishing
Date: 07-2019
Publisher: Springer Science and Business Media LLC
Date: 05-2022
DOI: 10.1038/S41588-022-01062-7
Abstract: Estimates from genome-wide association studies (GWAS) of unrelated in iduals capture effects of inherited variation (direct effects), demography (population stratification, assortative mating) and relatives (indirect genetic effects). Family-based GWAS designs can control for demographic and indirect genetic effects, but large-scale family datasets have been lacking. We combined data from 178,086 siblings from 19 cohorts to generate population (between-family) and within-sibship (within-family) GWAS estimates for 25 phenotypes. Within-sibship GWAS estimates were smaller than population estimates for height, educational attainment, age at first birth, number of children, cognitive ability, depressive symptoms and smoking. Some differences were observed in downstream SNP heritability, genetic correlations and Mendelian randomization analyses. For ex le, the within-sibship genetic correlation between educational attainment and body mass index attenuated towards zero. In contrast, analyses of most molecular phenotypes (for ex le, low-density lipoprotein-cholesterol) were generally consistent. We also found within-sibship evidence of polygenic adaptation on taller height. Here, we illustrate the importance of family-based GWAS data for phenotypes influenced by demographic and indirect genetic effects.
Publisher: Cold Spring Harbor Laboratory
Date: 07-05-2019
DOI: 10.1101/630426
Abstract: Genetic associations and correlations are perceived as confirmation that genotype influences one or more phenotypes respectively. However, genetic correlations can arise from non-biological or indirect mechanisms including population stratification, dynastic effects, and assortative mating. In this paper, we outline these mechanisms and demonstrate available tools and analytic methods that can be used to assess their presence in estimates of genetic correlations and genetic associations. Using educational attainment and parental socioeconomic position data as an exemplar, we demonstrate that both heritability and genetic correlation estimates may be inflated by these indirect mechanisms. The results highlight the limitations of between-in idual estimates obtained from s les of unrelated in iduals and the potential value of family-based studies. Use of the highlighted tools in combination with within-sibling or mother-father-offspring trio data may offer researchers greater opportunity to explore the underlying mechanisms behind genetic associations and correlations and identify the underlying causes of estimate inflation.
Publisher: Wiley
Date: 08-10-2016
DOI: 10.1002/IJC.30436
Publisher: Springer Science and Business Media LLC
Date: 12-11-2020
DOI: 10.1038/S41467-020-19478-2
Abstract: Numerous observational studies have attempted to identify risk factors for infection with SARS-CoV-2 and COVID-19 disease outcomes. Studies have used datasets s led from patients admitted to hospital, people tested for active infection, or people who volunteered to participate. Here, we highlight the challenge of interpreting observational evidence from such non-representative s les. Collider bias can induce associations between two or more variables which affect the likelihood of an in idual being s led, distorting associations between these variables in the s le. Analysing UK Biobank data, compared to the wider cohort the participants tested for COVID-19 were highly selected for a range of genetic, behavioural, cardiovascular, demographic, and anthropometric traits. We discuss the mechanisms inducing these problems, and approaches that could help mitigate them. While collider bias should be explored in existing studies, the optimal way to mitigate the problem is to use appropriate s ling strategies at the study design stage.
Publisher: Informa UK Limited
Date: 18-07-2023
Publisher: Cold Spring Harbor Laboratory
Date: 02-03-2020
Publisher: Springer Science and Business Media LLC
Date: 12-11-2015
DOI: 10.1038/SREP16509
Abstract: We investigated the role of common genetic variation in educational attainment and household income. We used data from 5,458 participants of the National Child Development Study to estimate: 1) the associations of rs9320913, rs11584700 and rs4851266 and socioeconomic position and educational phenotypes and 2) the univariate chip-heritability of each phenotype and the genetic correlation between each phenotype and educational attainment at age 16. The three SNPs were associated with most measures of educational attainment. Common genetic variation contributed to 6 of 14 socioeconomic background phenotypes and 17 of 29 educational phenotypes. We found evidence of genetic correlations between educational attainment at age 16 and 4 of 14 social background and 8 of 28 educational phenotypes. This suggests common genetic variation contributes both to differences in educational attainment and its relationship with other phenotypes. However, we remain cautious that cryptic population structure, assortative mating and dynastic effects may influence these associations.
Publisher: Ovid Technologies (Wolters Kluwer Health)
Date: 11-2019
Publisher: Public Library of Science (PLoS)
Date: 07-07-2022
DOI: 10.1371/JOURNAL.PGEN.1010247
Abstract: Estimating effects of parental and sibling genotypes (indirect genetic effects) can provide insight into how the family environment influences phenotypic variation. There is growing molecular genetic evidence for effects of parental phenotypes on their offspring (e.g. parental educational attainment), but the extent to which siblings affect each other is currently unclear. Here we used data from s les of unrelated in iduals, without (singletons) and with biological full-siblings (non-singletons), to investigate and estimate sibling effects. Indirect genetic effects of siblings increase (or decrease) the covariance between genetic variation and a phenotype. It follows that differences in genetic association estimates between singletons and non-singletons could indicate indirect genetic effects of siblings if there is no heterogeneity in other sources of genetic association between singletons and non-singletons. We used UK Biobank data to estimate polygenic score (PGS) associations for height, BMI and educational attainment in self-reported singletons (N = 50,143) and non-singletons (N = 328,549). The educational attainment PGS association estimate was 12% larger (95% C.I. 3%, 21%) in the non-singleton s le than in the singleton s le, but the height and BMI PGS associations were consistent. Birth order data suggested that the difference in educational attainment PGS associations was driven by in iduals with older siblings rather than firstborns. The relationship between number of siblings and educational attainment PGS associations was non-linear PGS associations were 24% smaller in in iduals with 6 or more siblings compared to the rest of the s le (95% C.I. 11%, 38%). We estimate that a 1 SD increase in sibling educational attainment PGS corresponds to a 0.025 year increase in the index in idual’s years in schooling (95% C.I. 0.013, 0.036). Our results suggest that older siblings may influence the educational attainment of younger siblings, adding to the growing evidence that effects of the environment on phenotypic variation partially reflect social effects of germline genetic variation in relatives.
Publisher: Oxford University Press (OUP)
Date: 09-06-2023
DOI: 10.1093/IJE/DYAD079
Abstract: Previous Mendelian randomization (MR) studies using population s les (population MR) have provided evidence for beneficial effects of educational attainment on health outcomes in adulthood. However, estimates from these studies may have been susceptible to bias from population stratification, assortative mating and indirect genetic effects due to unadjusted parental genotypes. MR using genetic association estimates derived from within-sibship models (within-sibship MR) can avoid these potential biases because genetic differences between siblings are due to random segregation at meiosis. Applying both population and within-sibship MR, we estimated the effects of genetic liability to educational attainment on body mass index (BMI), cigarette smoking, systolic blood pressure (SBP) and all-cause mortality. MR analyses used in idual-level data on 72 932 siblings from UK Biobank and the Norwegian HUNT study, and summary-level data from a within-sibship Genome-wide Association Study including & 000 in iduals. Both population and within-sibship MR estimates provided evidence that educational attainment decreased BMI, cigarette smoking and SBP. Genetic variant–outcome associations attenuated in the within-sibship model, but genetic variant–educational attainment associations also attenuated to a similar extent. Thus, within-sibship and population MR estimates were largely consistent. The within-sibship MR estimate of education on mortality was imprecise but consistent with a putative effect. These results provide evidence of beneficial in idual-level effects of education (or liability to education) on adulthood health, independently of potential demographic and family-level confounders.
Publisher: Public Library of Science (PLoS)
Date: 04-11-2021
DOI: 10.1371/JOURNAL.PGEN.1009883
Abstract: Spousal comparisons have been proposed as a design that can both reduce confounding and estimate effects of the shared adulthood environment. However, assortative mating, the process by which in iduals select phenotypically (dis)similar mates, could distort associations when comparing spouses. We evaluated the use of spousal comparisons, as in the within-spouse pair (WSP) model, for aetiological research such as genetic association studies. We demonstrated that the WSP model can reduce confounding but may be susceptible to collider bias arising from conditioning on assorted spouse pairs. Analyses using UK Biobank spouse pairs found that WSP genetic association estimates were smaller than estimates from random pairs for height, educational attainment, and BMI variants. Within-sibling pair estimates, robust to demographic and parental effects, were also smaller than random pair estimates for height and educational attainment, but not for BMI. WSP models, like other within-family models, may reduce confounding from demographic factors in genetic association estimates, and so could be useful for triangulating evidence across study designs to assess the robustness of findings. However, WSP estimates should be interpreted with caution due to potential collider bias.
Publisher: Springer Science and Business Media LLC
Date: 05-11-2018
DOI: 10.1038/S41562-018-0461-X
Abstract: Success in school and the labour market relies on more than high intelligence. Associations between "non-cognitive" skills in childhood, such as attention, self-regulation, and perseverance, and later outcomes have been widely investigated. In a systematic review of this literature, we screened 9553 publications, reviewed 554 eligible publications, and interpreted results from 222 better quality publications. Better quality publications comprised randomised experimental and quasi-experimental studies (EQIs), and observational studies that made reasonable attempts to control confounding. For academic achievement outcomes there were 26 EQI publications but only 14 were available for meta-analysis with effects ranging from 0.16 to 0.37SD. However, within sub-domains effects were heterogeneous. The 95% prediction interval for literacy was consistent with negative, null and positive effects (-0.13 to 0.79). Similarly heterogeneous findings were observed for psychosocial, cognitive and language, and health outcomes. Funnel plots of EQIs and observational studies showed asymmetric distributions and potential for small study bias. There is some evidence that non-cognitive skills associate with improved outcomes. However, there is potential for small study and publication bias that may over-estimate true effects, and heterogeneity of effect estimates spanned negative, null and positive effects. The quality of evidence from EQIs under-pinning this field is lower than optimal and more than a third of observational studies made little or no attempt to control confounding. Interventions designed to develop children's non-cognitive skills could potentially improve outcomes. The inter-disciplinary researchers interested in these skills should take a more strategic and rigorous approach to determine which interventions are most effective.
Publisher: F1000 Research Ltd
Date: 29-07-2019
DOI: 10.12688/WELLCOMEOPENRES.15334.1
Abstract: Mendelian randomization (MR) uses genetic information to strengthen causal inference concerning the effect of exposures on outcomes. This method has a broad range of applications, including investigating risk factors and appraising potential targets for intervention. MR-Base has become established as a freely accessible, online platform, which combines a database of complete genome-wide association study results with an interface for performing Mendelian randomization and sensitivity analyses. This allows the user to explore millions of potentially causal associations. MR-Base is available as a web application or as an R package . The technical aspects of the tool have previously been documented in the literature. The present article is complimentary to this as it focuses on the applied aspects. Specifically, we describe how MR-Base can be used in several ways, including to perform novel causal analyses, replicate results and enable transparency, amongst others. We also present three use cases, which demonstrate important applications of Mendelian randomization and highlight the benefits of using MR-Base for these types of analyses.
Publisher: Cold Spring Harbor Laboratory
Date: 02-09-2020
DOI: 10.1101/2020.08.30.20184846
Abstract: Alzheimer’s disease (AD) and open angle glaucoma (OAG) are common age-related neurodegenerative disorders with shared pathological features, leading to the hypothesis that glaucoma may represent a type of “ocular Alzheimer’s disease”. However, no causal relationship has yet been established. To test for a causal relationship, bi-directional two-s le Mendelian randomization analyses were performed using summary data from the largest available genome-wide association studies of AD and OAG. The effect on AD risk from exposure to genetically predicted OAG was measured using 24 single nucleotide polymorphisms (SNPs). In the reverse direction, the effect on glaucoma risk from exposure to genetically predicted AD was measured using 25 SNPs. Additionally, the relationship between AD and measurements of optic disc morphology (vertical cup:disc ratio (VCDR), optic cup area, optic disc area) and intraocular pressure (IOP) were investigated. People with congenitally larger optic discs, a phenotype not regarded to be related to glaucoma, had a lower risk of AD (OR=0.80 per mm 2 increase in disc area 95%CI=0.66,0.97 P =0.02) and people with genetically predicted AD had smaller optic disc sizes (−0.03 standard deviation change in mm 2 optic disc area per doubling odds of AD, 95%CI=-0.05,0.00 P -value=0.03). However, there was little evidence that exposure to genetically predicted OAG affected AD risk (OR=1.00 per doubling odds of OAG, 95%CI=0.98,1.03 P =0.83). Nor did genetically predicted IOP, VCDR or optic cup area influence AD risk. In the reverse direction, there was little evidence that genetically predicted AD had a causal effect on risk of OAG, IOP, VCDR or optic cup area. Genetic analyses show that congenital optic disc area influences AD risk but provide little support for a causal relationship between OAG and AD, suggesting that previous observed associations between OAG and AD may be due to reverse causation, confounding or other forms of bias. Glaucoma refers to a heterogenous group of neurodegenerative conditions characterised by progressive optic nerve head cupping and visual field loss. Primary open-angle glaucoma (POAG) is the commonest age-related glaucoma, accounting for 2/3 of all glaucoma cases. Elevated intraocular pressure (defined as IOP mmHg), age, myopia (negative refractive error), and family history are the main risk factors for POAG. Indeed, POAG is usually diagnosed on the basis of elevated IOP or diurnal spikes in IOP combined with progressive optic nerve head cupping and visual field loss. Furthermore, normal neuroretinal rim width (Figure 1) follows the ISNT rule (inferior superior nasal temporal) and so vertical optic cup:disc ratio (VCDR) is used clinically to distinguish pathological glaucomatous cupping from physiological cupping. The caveat is that congenitally larger optic discs tend to have larger physiological optic cups, and so optic cup area needs to be adjusted for optic disc area (Figure 1). While glaucoma is characterised by progressive increases in optic cup size, optic disc area does not change over a lifetime. IOP is the only modifiable risk factor for glaucoma and surgical, laser and medical interventions which lower IOP have been proven to slow down the progression of glaucomatous optic disc cupping and visual field loss. In contrast, elevated IOP without glaucomatous optic disc cupping or visual field loss is defined as ocular hypertension, not glaucoma. Measurements of IOP by applanation tonometry can be influenced by central corneal thickness, resistance, and hysteresis, and need to be corrected for these factors, e.g. by using the Ocular Response Analyzer. Nevertheless, the probability of converting from untreated ocular hypertension to POAG is ~2-3% per year. Some people develop progressive optic nerve head cupping and visual field loss despite IOP mmHg): so-called normal tension glaucoma (NTG). POAG and NTG are widely considered to represent a continuum in open angle glaucoma (OAG), and they are strongly correlated genetically (Figure 9). Nonetheless, risk factors other than IOP appear to be more important to the pathogenesis of NTG. For ex le, migraine, vasospasm, systemic hypotension and primary vascular dysregulation have all been associated with NTG (Figure 1). Many have also been linked to Alzheimer’s dementia. Many GWAS do not distinguish between POAG and NTG yet, this may be important to studies conducted in different populations since the prevalence of NTG varies widely depending on ethnicity. For ex le, ~30% of open angle glaucoma cases of European-descent have normal IOP, whereas % of open angle glaucoma cases in Japan have normal IOP. Furthermore, several GWAS have identified new risk loci for glaucoma from related phenotypes: IOP, corneal thickness, resistance and hysteresis, VCDR, and optic cup area (Figure 1). These phenotypes make powerful quantitative traits in GWAS and they are highly heritable across populations, but they are not sufficient in idually to meet the diagnostic definition of glaucoma. The importance is that some of these phenotypes, e.g. optic disc cupping, are not specific to glaucoma. Indeed, the differential diagnosis for optic disc cupping includes compressive lesions (e.g. pituitary macroadenoma), ischaemic, demyelinating, inflammatory, infiltrative, infectious, congenital and inherited disorders of the optic nerve. Therefore, signs of neurodegeneration in the eye, like optic disc cupping, can arise from a variety of aetiologies and not just glaucoma (Figure 1). Several epidemiological studies from the US, France, Germany, Australia, South Korea, Taiwan, and Japan have reported that open angle glaucoma (OAG) is more prevalent among people with Alzheimer’s disease (AD) or that AD is more prevalent among people with OAG. However, other studies have reported no association. We searched PubMed for studies published between database inception and 28 January 2020 that had investigated the relationship between Alzheimer’s disease and glaucoma using the search terms (“Alzheimer” or “Alzheimer’s” and glaucoma”). Papers in English and other languages were included, if there was an English abstract for assessment. We found the putative relationship between AD and glaucoma was the subject of several reviews and two meta-analyses. The first meta-analysis of 8 observational studies (6870 AD cases) concluded people with OAG have an increased risk of AD (RR=1.52 95% CI: 1.41-1.63 I 2 =97%, p .001). A positive association was found when analyses were restricted to Asia (RR=2.03 95%CI: 1.02-4.07) but not when they were restricted to America (RR=0.91 95%CI: 0.89-0.94). The second systematic review of 10 studies found that people with AD (RR=0.92 95% 95%CI:0.89-0.94 I 2 =89%, p .001) or dementia (RR=0.94 95%CI: 0.92-0.96 I 2 = 89.4%, p .001) had a lower risk of OAG. The studies cited in both reviews differed in case definition, ascertainment and population ethnicity, and were highly heterogenous in study design: results varied from large positive associations in small studies to negative or null estimates in cohort and record-linkage studies. Mechanistically, many authors have suggested that OAG is linked to intracranial pressure (ICP). It is reported that ICP is lower in people with OAG than healthy controls and ICP is lower in people with normal tension glaucoma (NTG) than people with OAG associated with elevated intraocular pressure (IOP). Turnover of cerebrospinal fluid (CSF) halves from birth to old age and is significantly reduced in people with AD and normal pressure hydrocephalus (NPH). In addition, people with NPH who receive ventriculoperitoneal shunts have increased risk of NTG. Hence, some authors have hypothesised that raised translaminar pressure gradient (the difference between IOP and ICP across the laminar cribrosa of the optic nerve head) may be responsible for the pathogenesis of glaucoma because this hypothesis would explain why it is both possible for some people to develop progressive optic nerve head cupping and visual field loss despite IOPs in the normal range, while raised IOP alone is insufficient to cause glaucoma in others (i.e. glaucoma is caused by low ICP). As low ICP is associated with AD, low ICP may be the mechanistic link between AD and OAG (Figure 2). An alternative hypothesis is that the production, circulation and absorption of intraocular fluid shares similar mechanisms to that of cerebrospinal fluid, and the failure of these mechanisms leads to the build-up of neurotoxins in OAG and AD. This hypothesis might explain why Aβ proteins and tau are detected in the retina of people affected by OAG and AD. In support of this hypothesis, there is substantial overlap in gene expression in the ciliary body compared with the choroid plexus, e.g. ion and water channels and transporters, and the renin-angiotensin system. There may also be a role for the glymphatic system, which provides a mechanism for the clearance of soluble waste products, e.g. Aβ protein in ocular fluid or CSF. The influx and efflux of CSF into the glymphatic system occurs via periarterial and perivenous spaces respectively, which finally drain into dural and cervical lymphatic vessels. The optic nerve is surrounded by CSF and is thought to have its own specialised glymphatic network. Here, the lamina cribrosa provides an additional barrier to fluid transport from inside the eye to the optic nerve in a manner that is dependent on translaminar pressure gradient. In models of glaucoma, defects in the lamina cribrosa are associated with the redirection of potentially harmful solutes, e.g. Aβ protein, from the intra-axonal compartment and glymphatic system of the optic nerve to its extracellular spaces. This leads to the build-up of potentially harmful solutes within the optic nerve and the degeneration of retinal ganglion cell axons. Though not well understood, it is possible that similar abnormalities in the glymphatic system of the brain might exist in AD leading to the accumulation of neurotoxins. There is also evidence that trans-synaptic neurodegeneration in the eye, e.g. from glaucoma, causes secondary neurodegeneration in functionally connected subcortical and cortical structures in the brain. Likewise, neurodegenerative processes in the brain, e.g. from dementia, can cause secondary neurodegeneration of the optic nerve. Tau pathology in AD is known to spread over time and could conceivably cause secondary neurodegeneration in the eye. Furthermore, amyloid microangiopathy can affect retinal and choroidal vasculature as well as cerebral blood flow in AD. Although amyloid microangiopathy has not been investigated in glaucoma, vascular dysfunction and genes involved in vascular endothelial morphology and genesis are consistently implicated in glaucoma. Hence, the neurodegenerative processes and vascular abnormalities common to both disorders might explain why they are both associated with visual field defects and the degeneration of retinal ganglion cells (RGCs). Studies using optical coherence tomography (OCT) imaging of the fundus have shown that thinning of the retinal nerve fibre layer (RNFL: the layer composed of RGC axons) and ganglion cell-inner plexiform layer (GC-IPL: composed of RGC bodies and dendrites) associate with future cognitive decline and dementia diagnosis. People with known glaucoma diagnoses were excluded from these analyses. However, the optic nerve is composed of RGC axons, which means that thinning of the RNFL and/or GC-IPL are signs of optic neuropathy from any cause, not just OAG. Hence, signs of neurodegeneration have been detected in the eye using OCT that are associated with future cognitive decline, but previous studies have not been designed to show whether these changes represent early OAG or whether they result from other unrelated neurogenerative processes affecting RGCs that are causally related to AD. Prior to this study, it was not clear whether there was a causal relationship between OAG and AD, whether their shared pathological features were merely non-specific signs of neurodegeneration, or whether some people were coincidentally affected by both disorders because of their high prevalence in older adults. Using the largest available population cohorts to maximize our statistical power (International Genomics of Alzheimer’s Project (IGAP), Alzheimer’s disease working group of the Psychiatric Genomics Consortium (PGC-ALZ), Alzheimer’s Disease Sequencing Project (ADSP), National Eye Institute Glaucoma Human Genetics Collaboration Heritable Overall Operational Database (NEIGHBORHOOD) consortium and UK Biobank) we compared the results of observational epidemiological studies with causal estimates from Mendelian Randomization (MR) analyses to make causal inferences about the biological relationship between AD and OAG. In contrast to previous observational reports of an association between OAG and AD, we found weak evidence of a causal relationship between AD and OAG in either direction. Nor did we find strong evidence of a causal relationship between intraocular pressure (IOP), vertical cup:disc ratio (VCDR), and optic cup area with AD. Our data did, however, suggest that larger congenital optic disc size has a protective effect on AD risk. The genetic evidence in this study does not provide support for a causal relationship between AD and OAG or any related glaucoma phenotype, suggesting that the observed associations in previous studies were due to reverse causation, confounding and other types of bias. One possible source is collider bias, which occurs when two variables, e.g. AD and raised IOP, can independently cause a third collider variable, e.g. optic disc cupping and/or other signs of RGC degeneration: signs that are generally used to diagnose OAG. Collider bias is also an issue in studies of phenotypes related to glaucoma that are also used to define case-control status in the same cohort, e.g. IOP. In other words, conditioning on phenotypes which are also used to ascertain case-control status will bias the analyses. An additional source of bias may be caused by methods of ascertainment. GWAS generally define OAG based on a threshold for optic disc cupping +/-raised IOP, but non-progressive optic disc cupping is not glaucoma. Moreover, other possible causes of optic disc cupping, e.g. compressive or congenital (Panel), would not be excluded by a single anterior segment examination, but would require further investigation, e.g. MRI head scan, and serial measurements over time. Evidence that larger optic disc area is protective against AD or that people born with smaller optic discs have a greater risk of AD in future might support the idea of “cognitive reserve”, i.e. people with larger optic nerves and other correlated neuronal structures may be more resilient to age-related neurodegenerative processes. The links between specific genetic variants, e.g. APOE , optic disc size and educational attainment with AD suggest there may be several biological pathways that are causally related to AD (Figure 10). In summary, clinicians and scientists should be aware there is little evidence for a causal relationship between AD and OAG. OAG is widely considered to be an IOP-driven disease indeed, IOP mmHg is often used to diagnose POAG. However, this definition of glaucoma can lead to bias. Neurodegenerative changes affecting the eye can arise from multiple aetiologies and it is possible that IOP, AD and other unknown factors are independent risk factors that cause a similar pattern of RGC degeneration. Without strong evidence of a causal relationship, we predict little benefit in repurposing drugs developed for AD in clinical trials for OAG, except where they target common downstream pathways of neurodegeneration.
Publisher: F1000 Research Ltd
Date: 07-11-2019
DOI: 10.12688/WELLCOMEOPENRES.15334.2
Abstract: Mendelian randomization (MR) estimates the causal effect of exposures on outcomes by exploiting genetic variation to address confounding and reverse causation. This method has a broad range of applications, including investigating risk factors and appraising potential targets for intervention. MR-Base has become established as a freely accessible, online platform, which combines a database of complete genome-wide association study results with an interface for performing Mendelian randomization and sensitivity analyses. This allows the user to explore millions of potentially causal associations. MR-Base is available as a web application or as an R package . The technical aspects of the tool have previously been documented in the literature. The present article is complementary to this as it focuses on the applied aspects. Specifically, we describe how MR-Base can be used in several ways, including to perform novel causal analyses, replicate results and enable transparency, amongst others. We also present three use cases, which demonstrate important applications of Mendelian randomization and highlight the benefits of using MR-Base for these types of analyses.
Publisher: Wiley
Date: 19-03-2016
DOI: 10.1002/CAM4.695
Publisher: Oxford University Press (OUP)
Date: 11-12-2018
DOI: 10.1093/IJE/DYY273
Publisher: Springer Science and Business Media LLC
Date: 31-10-2017
DOI: 10.1038/MP.2017.210
Location: United Kingdom of Great Britain and Northern Ireland
Location: Hong Kong
Location: United Kingdom of Great Britain and Northern Ireland
Start Date: 2016
End Date: 2019
Funder: Economic and Social Research Council
View Funded Activity