This project is a MRC IEU PhD student mini project.

We have previously shown that an early onset of menarche is associated with increased levels of depression in 14 year old girls and we found evidence that this effect it causal. It is unclear, however, whether this effect persists beyond adolescence.

Common genetic variants in the oxytocin receptor gene (OXTR) have previously been associated with social behaviours, personality and mood (Li et al., 2015; Saphire-Bernstein et al., 2011). Emerging evidence has demonstrated that genetic variants such as rs53576 and rs2254298 may play a role in social behaviour and interpersonal relationships, with individuals carrying homozygote G alleles having greater sociality compared to those carrying A alleles (Li et al., 2015). These individuals are also reported to have higher empathy and higher sensitivity to social cues (Connelly et al., 2014). There is also evidence to suggest that variation in OXTR is also associated with lower symptoms of depression and anxiety (Brune, 2012; Saphire-Bernstein et al., 2011). More recently, evidence has shown an association between OXTR variants and pathways to addictive behaviour such as alcohol and drug misuse, via an interaction with early environmental stressors (Buisman-Pijlman et al., 2014). Research in this area may uncover mechanisms underlying addiction and poor mental health. Identifying these mechanisms is crucial for improving and developing new treatments for these disorders.

Despite evidence linking OXTR with mood, personality and social behaviour, some of this research shows small effect sizes or contrasting results. For instance, a meta-analysis by Bakermans-Kranenburg and van Ijzendoorn, (2014) failed to find any association between rs53576 polymorphisms and sociality. However, a follow up meta-analysis by Li et al., (2015) found the opposite with modest effect sizes between this SNP and sociality.

One possible reason for contrasting findings could be the role of the environment and this interaction between the genetics and early environmental exposures. It has been established that early environmental factors play a role in developmental pathways (Buisman-Pijlman et al., 2014; Stingaris et al., 2014), yet not all research considers environmental factors playing a role in later mental health outcomes. Research exploring the role of the 5-HTTLPR genotype alluded to the importance of early life stressors and maternal depression on later childhood symptomology (Araya et al., 2008). Consequently, the role of the early environment must be explored when examining causes for maladaptive developmental pathways. A further point is that many of the studies use very small sample sizes that often fail to detect any of these complex and important effects. To further clarify the role of OXTR in personality, mood, social behaviour and addiction, research has to maximise all available output. Using from data from cohort studies provides a rich background of data and allows researchers to examine developmental pathways from an epidemiological approach, rather than opportunity sampling. Another bonus is that using cohort data allows for increased sample sizes that can help determine true effects. Using both phenotypic and genetic data from the ALSPAC study, this project will examine the association between OXTR and mood, personality characteristics, sociality and addictive behaviours. By using the ALSPAC database, this project can examine any interactions between genetics and early environmental factors (e.g., early life stress, parenting styles, socioeconomic status and parental depression) with these outcomes.

Difficulties learning to read are often attributed to difficulties processing speech. This study examines the comorbidity of parent reported speech and hearing difficulties associated with otitis media with effusion in the early years and the impact of these difficulties on later literacy development and, in particular, prevalence of dyslexia.

The goal of the proposed research is to examine the range of behavioral and cognitive profiles associated with strongly predisposing, rare genetic events for autism. Rare protein trucating variants have been associated with these conditions in clinical ASD populations, and we would like to complement the case-control findings with a general population consideration of their effects. This gene-first approach would aid in elucidating the nature of the noted risk relationships, most specifically the penetrance and phenotypic range of rare variant influence.

It has been suggested that cognitive vulnerability that interacts with negative life events leads to depression. The cognitive vulnerability is an important target for interventions since it might be possible to modify the response to adverse events and thereby prevent depression. Little is known about the influences on cognitive vulnerability but previous research in ALSPAC (Pearson et al., 2013) has found that maternal cognitive vulnerability is associated with later child cognitive vulnerability independently of depressive symptoms in the mother or the child. It is possible this association results from the child’s modeling of the mothers’ cognitive vulnerability. We are not aware though of any previous research that has investigated the influence of paternal cognitive vulnerability on the cognitive vulnerability of the offspring. This study aims to investigate the relationship between paternal cognitive vulnerability and offspring’s cognitive vulnerability at 18 years using the ALSPAC data.

Short-term health benefits of breastfeeding are clear. There is also evidence of long-term breastfeeding effects, including lower blood pressure and total cholesterol, reduced risk of obesity and type-2 diabetes, and higher performance in intelligence tests.
To have long-term effects, it is necessary that having been breastfed or not (as well as other aspects, such as quality and duration) gets “imprinted” in the body and such imprinting lasts until later life periods. In this regard, epigenetics (the study of specific regulators of where, when and how much genes – regions of the DNA that contain information to generate proteins and other products by the cell – are on or off) is of potential interest. Importantly, epigenetic modifications normally pass from one cell to another after cell division, so epigenetic modifications are not easily erased throughout cell cycles over time.
Previous studies using ALSPAC data reported that maternal smoking during pregnancy was associated with methylation (a type of epigenetic marker, consisting of a chemical modification in specific regions of the DNA) patterns in the offspring that sustained until later life stages. Regarding breastfeeding, a few studies looking at specific regions of the methylome (set of all regions of the DNA where methylation can occur) provide some evidence that epigenetic factors might be involved with breastfeeding effects. However, no study to date scanned the entire methylome in relation to breastfeeding or evaluated the maintenance of one or more breastfeeding-associated methylation patterns over time.
Regarding intelligence, an additional potential mechanism relates to breast milk being a source of long-chain polyunsaturared fatty acids (LC-PUFAs) including docosahexaenoic acid (DHA), which have been implicated in brain development. Indeed, many milk substitutes currently include DHA and other LC-PUFAs. One possibility to test this hypothesis is to evaluate if the association between breastfeeding and intelligence differs according to the individual’s capacity to synthetize DHA from metabolic precursors. Special attention has been given to genetic variation in the FADS2 gene, which encodes a protein involved in desaturation processes required for endogenous synthesis of LC-PUFAS from shorter chain fatty acids. However, two published studies (including one study using data from the Avon Longitudinal Study of Parents and Children – ALSPAC) in this regarding successfully detected a FADS2-breastfeeding interaction, but the coefficients were in inconsistent. Three additional twin studies did not detect any strong interaction. Therefore, there is no firm conclusion regarding a potential FADS2-breastfeeding interaction to date. Such potential interaction also suggests that other metabolic processes might play important roles in the health effects of breastfeeding, although a comprehensive evaluation (eg, using metabolimic data) has not been performed to date. We will also analyse the rs66698963 polymorphism, which is a 22 bp insert-deletion within FADS2. This polymorphism has recently been associated with LC-PUFA levels. Moreover, signs of positive selection have also been detected, suggesting that this SNP may be functional.
In general, genetic aspects of breastfeeding have been poorly studied. However, narrow-sense heritability estimates of breastfeeding behaviour (ever vs. never) 44%-85% that have been reported in different populations. This suggests that genetic variants account for a substantial proportion of variability in breastfeeding behaviour in these populations and are, therefore, important aspects of breastfeeding. It also raises the possibility that genetic factors underlie at least some of the association of breastfeeding with health outcomes through pleiotropic effects (either mediated by or independent on breastfeeding). Moreover, exploring breastfeeding genetics might also lead to the identification of genetic variants robustly associated with breastfeeding, which could be used as instrumental variables to strengthen causal inferences through the Mendelian randomization design.
The present project aims at investigating biological mechanisms underlying breastfeeding health effects by studying the FADS2-breastfeeding interaction and additional genetic, epigenetic and metabolic aspects of breastfeeding.

A recent HEFCE report has noted that mental health problems suffered by those in higher education rose from approximately 8,000 in 2008-2009 to 18,000 in 2012-2013. Screen time has also increased over this time point, and has led to speculation that the two may be associated. However, understanding whether this is causal is notoriously difficult, as other factors could influence both screen time and anxiety levels. This research aims to investigate these associations, whilst trying to take in to account the factors that could distort the relationship.

This project will examine the effect of mother's thyroid function during pregnancy on child DNA methylation in samples of umbilical cord blood. The effects of child DNA methylation on neurocognitive outcome will then be investigated.

Our research group has recently used data from the Avon Longitudinal study to discover that although psychotic experiences, depressive symptoms and anxiety symptoms are treated as separate, they may be symptoms of the same underlying mental illness. We termed this illness 'common mental distress' (Stochl et al., 2014). We did this study in ALSPAC when children were 13 years of age. We formed a 'new' variable (common mental distress) using existing measures of depressive symptoms and psychotic like experiences which we would now like to investigate further. We think it would be of benefit to identify early life risk factors (prior to the age of 13) that increase the chance that adolescents will score highly for common mental distress at age 13. We would also like to know whether adolescents who scored highly on our common mental distress variable when they were 13, continue to have mental health problems in late adolescence and early adulthood.