Project outline:

Study of Learning Ability, Development and Genes

Background

Intellectual disability (ID) in childhood is common with a prevalence rate of around 2.6% (Emerson, 2003). Children with ID have substantially higher rates of persistent comorbid psychiatric disorders (ADHD, ASD, Conduct Disorders, Anxiety Disorders), around 4-7 times higher than the general population (e.g., Einfeld et al., 2006; Emerson, 2003; Emerson & Hatton, 2007; Emerson & Einfeld, 2010). The most strongly associated childhood diagnoses are ADHD (odds ratio=10 compared with non-ID controls) and autistic spectrum disorder (ASD; odds ratio=75; Emerson, 2003). Comorbid psychopathology in children with ID is important because it increases parent stress more than the severity of ID, raises the cost of care and has a long term impact on adult outcomes (Herring et al., 2006).

The majority of children with mild-moderate ID have no clearly identifiable cause for their ID (Einfeld et al., 2006). Although multiple common gene variants, comorbid neurological illness and psychosocial disadvantage contribute to the aetiology, there is growing evidence that the rate of subtle chromosomal anomalies (deletions and duplications), known as copy number variants (CNVs), is elevated in those with previously unexplained ID (Girirajan, et al., 2010; Ropers, 2008; Sagoo et al., 2009). As a result of this work, some UK clinical genetics services are offering array CGH diagnostic services and a recent consensus statement recommends using chromosomal microarrays in the first tier of investigating developmental delay (Miller et al., 2010).

In a recently completed study of ADHD (Williams et al., 2010) we found a significantly elevated rate of CNVs in those with ID + ADHD (36%) than amongst controls (7%) or those with ADHD alone (11%). This is actually a higher CNV rate than previously reported for idiopathic ID (around 10%-12% for significant CNVs) and leads us to postulate that in individuals with idiopathic ID, the presence of large, rare CNVs in ID is clinically important because it might increase the risk of comorbid neuropsychiatric disorder. We now propose to explicitly examine whether the presence of large, rare CNVs in children with idiopathic ID is associated with a broad range of psychopathology (not ADHD only).

Aims

* To investigate in children with mild/moderate Intellectual Disability (IQ test score less than 70) whether subtle chromosomal anomalies (CNVs) are associated with;

- The presence of psychopathlogy

- Higher levels of ADHD and autism scores

- Higher levels of aggression

- Lower levels of adaptive functioning

As such the primary outcome measures is the child's psychopathology and the main predictor is the presence of a Copy Number Variant (CNV).

Hypotheses

Children with Intellectual Disability and CNVs will have higher rates of psychopathology and lower levels of functioning

.

Aims

To identify methylation variation mediating lipid pathways.

The current project is an initial step aimed at uncovering genetic aetiology of psychotic experiences in a population based study of adolescents. The aims of the current study are three-fold. First, to examine potential associations between previously identified and reviewed Schizophrenia associating single nucleotide polymorphisms (SNPs) and dimension-specific psychotic experiences in a community based sample, using Twin Early Development Study (TEDS;Oliver & Plomin, 2006) data. Psychotic experiences are measured in TEDS using a quantitative dimension-specific Psychotic Experiences Questionnaire. Second, to create a gene score made up of previously identified risk alleles and to test for associations with psychotic experiences. Third, to replicate significant associations found in stage one of the analyses using Avon Longitudinal Study of Parents and Children (ALSPAC) sample.

It is hypothesised that some of the previously identified Schizophrenia associating SNPs will be also associating with the dimension-specific psychotic experiences. It is further expected that there will be an association between the polygenic scores and psychotic experiences. Finally, it is hypothesised that any statistically significant findings made using TEDS sample will be then replicated using ALSPAC.

The study will employ allelic and genotypic association analysis between SNPs of interest and psychotic-like symptoms as well as polygenic risk analysis. Sex, age and SES will be treated as potential covariates.

Background

ALSPAC is a vital resource in the study of child development and has contributed to much influential work on the topic of child development and paediatric disease.

Age-related macular degeneration (AMD) is the most frequent cause of blindness in Europe and accounts for severe visual loss in around 25 million people worldwide.(1) It is characterised by reduction in visual acuity and contrast sensitivity with subsequent loss of central vision. There are two categories of AMD: wet (neovascular and exudative) and dry. Wet AMD accounts for 90% of blindness due to AMD (2) and if untreated results in an average loss of approximately 4 lines of visual acuity with 2 years of disease onset.(3) Timely identification and treatment of those with neovascular ARMD improves visual outcomes.(4) The economic burden of AMD in terms of direct and indirect costs is significant.(5)

A meta analysis by Chakravarthy et al in 2010 found that risk factors for AMD include major risk factors (with a strong, consistent association): increasing age, current cigarette smoking, previous cataract surgery and family history of AMD. Risk factors with a moderate and consistent association include history of cardiovascular disease, hypertension, raised plasma fibrinogen and high body mass index. Weak, inconsistent association was found between AMD and gender, ethnicity, diabetes, iris colour, history of cerebrovascular disease and serum total and HDL cholesterol and triglyceride levels.(4) The same study identified 73 potential risk factors for late age-related macular degeneration including nine genetic factors. Increased sun exposure and low dietary intake of zinc and vitamin E were also highlighted by Smith et al using data pooled from three continents.(6)

More recently the link between genetics, environment and AMD has become far better understood with the discovery of potential AMD susceptibility genes. One such gene is Complement Factor H (CFH), a variant of which is rs1061170 (T1277C, Y402H) which is characterised by substitution of histidine for tyrosine at codon 402 on the long arm of chromosome 1, region 31. This is a missense mutation, which results in altered binding of the CFH protein to sulphated polysaccharides and may contribute to the pathogenesis of AMD through complement dysregulation. In individuals of European descent the Y402H variant was found to confer a 2-fold higher risk of late AMD per copy.(7,8)In homozygous individuals, those with neovascular AMD may have a decreased treatment response, particularly to anti-VEGF agents, when compared to individuals who are homozygous for the T-allele.(9) Along with another common variant (rs1410996), Y402H is thought to explain around 17% of AMD liability.(10)

There is evidence that chronic low-grade inflammation is key in the pathogenesis of AMD.(11) Raised serum inflammatory markers have been identified as possible indicators of increased risk of developing AMD, including C-reactive protein (CRP), interleukins (IL)-2 and -6, tumour necrosis factor (TNF)-a, soluble intercellular adhesion molecule (sICAM)-1 and C3a-desArg as have raised levels of urinary pro-inflammatory cytokines, including transforming growth factor (TGF)-b1 (OR 1.24 CI 1.02-1.50; Pless than 0.031) and macrophage chemoattractant protein (MCP)-1 (OR 1.07 CI 1.02-1.12 pless than 0.008) in early AMD and MCP-1 (OR 1.10 CI 1.03-1.17 pless than 0.003) in geographic atrophy.(11,12) Linked to this is the finding of the Blue Mountains Eye study that there is reduced renal function in AMD demonstrated by decreased estimated glomerular filtration rate and creatinine clearance.(13) Urinary MCP-1 levels above median levels were 2.5 times more likely to be found in individuals who were heterozygous or homozygous for the C-allele (Y402H) (Pless than 0.040).

The study team are experienced in working with ALSPAC data and among them have a wealth of epidemiological, genetic and ophthalmological clinical experience. The existing ALSPAC data contains snps directly associated with the Complement factor H and we will therefore be able to address the research question with existing data. This is a proposal for a small pilot study of the potential association between complement factor H genotype and AMD-related traits, which will inform the direction of future work.

Background:

We have previously found a negative association between umbilical cord iron concentration and early childhood eczema in ALSPAC[1], but associations between maternal anaemia and childhood atopic outcomes have not been reported. Other studies have reported that maternal hypertension in pregnancy is associated with impaired infant lung function[2] and, along with preeclampsia, with an increased risk of early childhood wheezing[3], but data are lacking on whether these conditions predict asthma and impaired lung function later in childhood. It would be interesting to know whether any effects of maternal hypertension/preeclampsia on these outcomes are mediated through lower birth weight and whether preeclampsia might explain any of the well-known birth order/parity association with atopy. Clearly, if anaemia and maternal hypertension/preeclampsia are causes of childhood asthma and atopy there may be important opportunities for primary prevention.

Aims:

To analyse the associations between maternal anaemia and hypertensive disorders of pregnancy and respiratory and atopic outcomes in the offspring.

Hypotheses:

* Anaemia is associated with increased risk of atopic disease

* Hypertension and preeclampsia are associated with increased risk of asthma and impaired lung function.

Variables needed for primary analyses:

Derived exposure variables of interest:

* Hypertensive disorder of pregnancy (gestational hypertension plus preeclampsia)

* Preeclampsia

* Gestational hypertension

* Pre-existing hypertension

* Early pregnancy Hb

Confounders and outcomes: We have all potential confounder variables with the exception of gestational weight gain. Outcomes will include asthma, hayfever, eczema, atopy and total IgE at 7 years, and lung function at 8-9 years (we have these variables too).

If we confirm associations with hypertension it might be of interest to carry out secondary analyses to explore associations with changes in blood pressure at different stages of gestation using the appropriate derived variables.

APOE is a gene found on chromosome 19 in humans, which encodes a protein (ApoE) which is involved in lipid transport. ApoE is the main lipid transport protein in the brain.There are 3 known variants of APOE: epsilon2, epsilon3 and epsilon4. epsilon4 is thought to be the ancestral allele of APOE(Mahley et al 2000), but in the UK population the expected frequencies of epsilon3, epsilon4 and epsilon2 are 0.78, 0.14 and 0.08 respectively.(Abdollahi et al 2006) These variants are the single of single nucleotide polymorphisms at two locations: T2060C(Cys112Arg) and C2198T (Arg158Cys).(Ensembl) The epsilon3 allele has cysteine and arginine in these two positions, the epsilon2 allele has cysteine in both and the epsilon4 allele has arginine in both.(Weisgraber et al 1981)

Possession of an e4 allele has been robustly shown to increase the risk of late onset Alzheimer's disease (Corder et al 1993). It is now known that ApoE is involved in neuronal repair, with E4 being the least efficient isoform (White et al 2001) and that APOE genotype influences outcome following head injury (Teasdale et al 1997).

Positive pleiotropy of the e4 allele has been investigated over the last 10 years with mixed results (e.g. Hubacek et al 2001, Bunce et al 2011). There is no effect of e4 on IQ (Taylor et al 2011), but several fMRI studies have been reported as showing more efficient memory activation in young adult e4 carriers (e.g. Scarmeas et al 2005). Working memory and e4 has been little studied, but there are suggestions from small imaging studies that e4 may be beneficial for working memory in young adults (Rusted et al 2013). Conversely studies of middle aged and older adults have found a deleterious effect of e4 on working memory (Reinvang et al 2010, Greenwood et al 2005)) A study of 445 people of mixed ages found no effect (Alexander et al 2007)

Defects in working memory have been reported as one of the earliest deficits seen in Alzheimer's disease (McKhann et al 1984). I therefore aim to study the relationship between APOE genotype and working memory in a large community sample of young adults.

Hypotheses

1) Null hypothesis: APOE genotype does not influence working memory performance

2) APOE genotype influences working memory performance, either positively or negatively.

Exposure Variable

APOE genotype

Outcome Variables

Working memory performance, specifically performance on the n-back test aged 18 with accuracy across the 2 and 3 back levels being the primary outcome.

Confounding Variables

In genetic studies it is often difficult to identify confounders as they have to have an independent relationship with both the genotype and the outcome of interest. Head injury outcome is known to be associated with APOE genotype and such injuries can have a wide range of adverse neuropsychological effects. APOE genotype is known to influence LDL levels. I would like to use this as a proxy means of checking genotyping accuracy as is routinely done in this field. Higher LDL levels have also been linked to increased cognitive problems in later life. I would like to investigate a number of other variables likely to modify working memory performance to see if they have any effect on my results.

Depression is one of the most commonly occuring mental illnesses, with a lifetime prevalence of 9.4%

(Kessler, 2012). The World Health Organization has ranked depression among the top 4 leading causes of

disability worldwide (WHO, 2008). Rates of depression also appear to be universally higher for women

than men, with women being roughly at 2 times greater risk of developing depression than males

(Kessler, 2003; Weissman & Klerman, 1977). This sex difference has been attributed to differences in

seeking out rewarding behaviours (Ryba & Hopko, 2012), response toward emotionally salient stimuli

(Bradley, Codispoti, Sabatinelli, & Lang, 2001) and differences in underlying neurocircuitry including the

inferior frontal gyrus and left amygdala (Stevens & Hamann, 2012). Children of parents diagnosed with

major depressive disorder (MDD) are also 2-3 times more likely to develop depression (Johnstone,

Lawrie, & Cosway, 2002; Lieb, Isensee, Hofler, & Wittchen, 2002; Weissman et al., 2006) and 2-3 times

more likely to develop other anxiety or behavioural disorders (Lieb et al., 2002; Mars et al., 2012;

Weissman et al., 2006). This risk suggests the influence of either familial genetic loading or

environmental factors. The most commonly reported risk allele is a low-expressing variant of the

serotonin transporter gene (5HTTLPR) (for more information please see appendix 1b).

Maternal behaviours towards their own children can also be affected by depression. Mothers become less

emotionally responsive to their infants as a result of depression and this has noticable effects on the longterm

behavioural and social outcome of the offspring. In infancy, these children are more likely to show

avoidance or disorganized attachment behaviours (to strange situation tasks) (Madigan, Moran, &

Schuengel, 2007; Martins & Gaffan, 2000) and higher rates of internalizing/ externalizing behavioural

problems (Garai et al., 2009) than children of non-depressed mothers. Children of depressed mothers have

poorer scores on the Child Behavioural Checklist (CBCL) when children were exposed to maternal

depression during their first year of life (Bagner, Pettit, Lewinsohn, & Seeley, 2010). They also show

lower scores on the Peabody picture vocabulary test; a measure of verbal fluency (Brennan et al., 2000),

lower scores on the Rey Auditory Verbal Learning task; a measure of declarative recall memory (Mannie,

Barnes, Bristow, Harmer, & Cowen, 2009), but, executive functioning does not seem to be affected by

parental depression (Micco et al., 2009). However, emotional labeling deficits remain the most commonly

reported deficit among offspring of depressed parents, suggesting difficulties in the emotional processing

network is part of the core psychopathology among at-risk children.

The purpose of this study is to determine the extent to which childhood emotional face labeling ability

mediates risk for depression (by having one or more parent with depression) and the psychiatric outcome

in adolescence. We predict that errors in emotional labeling (measured at 8 years of age) in childhood will

be significantly associated with (1) exposure to maternal depression in childhood, and (2) depression or

other psychiatric effects in adolescence. We predict this emotional face labeling ability will account for

part of the measured association between maternal depression severity and youth depression in

adolescence. Furthermore, we predict that the child's genetic risk for depression (identified by the lowexpressing

serotonin allele status) will be a moderator between maternal depression exposure and their

emotional face labeling ability. We predict that short allele status in the children will increase the effect of

exposure to maternal depression and this effect with be seen as greater difficulty on the

The aim of this proposal is to specify formulae for allelic scores that index levels of expression, methylation and the metabolome and subsequently construct these scores in a large sample of AS cases and controls with GWAS data, and test to see whether the scores correlate with affection status, potentially identifying interesting biological pathways.

This analysis involves two distinct stages:(i) Genome-wide association analysis of genome-wide methylation, expression and metabolomic data. This will involve association analysis of both common and low frequency variants derived from genome-wide SNP chip platforms (Illumina 550K or 660K) and low density genome-wide sequencing, as well as imputed variants from the thousand genomes and UK10K projects. Because of computational constraints associated with the extremely large number of phenotypes modelled, statistical analysis will proceed using SNPTEST on a cleaned set of 8365 unrelated individuals of confirmed British ancestry. (ii) Construction of allelic scores to serve as genetic proxies for these molecular variables. I will investigate how allelic scores can best be constructed using a variety of approaches from simple weighted counts of strongly associated variants, to genome-wide allelic scores- an approach which I have pioneered in the context of individual risk prediction, to more sophisticated methods involving machine learning and lasso regression. The predictive validity of these measures will be assessed by cross validation, and/or using them to predict the same outcome in cohorts with similar data (e.g. TwinsUK in the case of genome-wide expression; Kettunen et al. (2012) in the case of metabolomic data). Please note, that this part of the project is pretty much identical to that proposed in MRC Unite Programme 4. Once the formulae for generating allelic scores has been validated, these scores will be derived and correlated with a large sample AS cases and controls with GWAS data.

The impact of adolescent sleep disorders upon neurocognitive function and the development of chronic pain conditions.

Childhood and adolescence is a critical period for developing homeostatic sleep-wake patterns. There is a growing body of evidence to suggest that childhood sleep problems may precipitate difficulties in later life. For example, sleep restriction in children can result in memory and verbal function deficits (Randazzo et al, 1998). Objective actigraphy monitoring also indicates that children (between ages 9 and 12) with higher levels of sleep fragmentation show diminished neurocognitive function on tests such as the continuous performance test and symbol digit substitution (Sadeh, Gruber and Raviv, 2002). Furthermore, Gregory et al (2009) found that sleep problems at ages 5 and 9 could predict neuropsychological deficits at age 13; specifically, tasks requiring a high demand of mental flexibility and working memory. Similar deficits in attentional capacity, memory and cognitive function have also been found in children with sleep-related breathing disorders (Blunden, Lushington and Kennedy, 2001; Beebe, 2006; Lewin et al, 2002). Overall, these studies highlight the importance of managing sleep disorders in childhood.

Pain is an interruptive but adaptive stimulus alerting us to potential threat or danger. Chronic pain patients, however, find it difficult to disengage from the constant threat of pain. Therefore, it is common for pain-related cognitive biases to develop as patients become increasingly vigilant to pain stimuli. Such maladaptive attention processes can unfortunately, prolong pain states. Neuropsychological complaints are common amongst chronic pain patients (Hart et al, 2001) and are assumed to stem from the attention-demanding interference of pain. It may be that pre-existing neuropsychological impairments can precipitate the transition from acute to chronic pain disorders however. If childhood sleep disorders can result in diminished neurocognitive capacity, this may induce a cognitive vulnerability wherein attention towards pain will already be enhanced due to problems with distraction and inhibition.

Psychosocial measures such as depression and negative beliefs about pain have been shown to increase one's probability of transitioning from an acute to a chronic pain condition (Casey, Greenburg et al, 2008; Pincus, Burton, Vogel and Field, 2002). Whether neuropsychological factors pose a risk for the development of chronic pain has yet to be addressed. If such risk factors are identified, early psychological interventions could be critical in preventing the development of certain pain conditions.

Aim and Objectives

To consider whether sleep problems in childhood and adolescence impact upon neurocognitive function and the subsequent development of chronic pain disorders. Using regression methods, results from the study will highlight whether poor sleep in adolescence will contribute increased rates of chronic pain conditions in later life.

It is expected that those with higher incidences of sleep problems will show reduced neurocognitive function. Links between these diminished functions and the development of chronic pain conditions will then be explored.

Hypotheses

Poor sleep will be associated with an increased level of chronic pain reports. Poor sleep will also result in diminished neurocognitive function. The association between sleep and chronic pain will attenuate after the control of neurocognitive function.

Exposure Variables

Early life/adolescent sleep problems

Outcome Variables

Chronic pain condition (y/n)?

Confounding Variables

Gender, socio-economic status, depression, neurocognitive function.