Background: Obesity is now considered to be a global epidemic and gene-environment interaction studies are crucial to understanding the genetic architecture of this disease. The objectives of this research were to (1) review the current evidence of gene-environment interactions in the field of obesity, (2) examine the interactions between obesity predisposing gene variants and physical activity using precise physical activity data and (3) analyze a novel gene-environment interaction between obesity predisposing gene variants and multiple pregnancies. Methods: The data for the gene-environment interaction analyses were collected from the EpiDREAM study: a prospective cohort including participants of six ethnic backgrounds from 21 countries worldwide. A subset of 17 423 participants with complete genotype and phenotype information was included in the analysis. Obesity predisposing single nucleotide polymorphisms were analyzed independently and as a genetic risk score. General linear models were used to analyze all main effects and interactions. Results: Physical activity interacted with FTO rs9939609 to modulate BMI (Pinteraction=0.032) and BAI (Pinteraction=3.26 x 10-4). Increased physical activity attenuated the impact of FTO on obesity. Four SNPs displayed significant associations with physical activity: NTRK2 rs1211166 (P=0.015), BDNF rs6265 (P=0.007), BDNF rs1401635 P=0.003) and NPC1 rs1805081 (P=3.52 x 10-4). Multiple pregnancies was significantly associated with BMI (Pinteraction=1.17 x 10-5) BAI (Pinteraction=3.47 x 10-7) and also interacted with FTO rs9939609 to modulate BMI (Pinteraction=0.014). The impact of FTO on BMI was accentuated by multiple pregnancies in the EpiDREAM cohort. Discussion: Both physical activity and parity have a significant impact on obesity measures and these effects appear to be relevant on a global scale. Our results confirm the physical activity x FTO interaction in a multi-ethnic context and indicate that parity may also interact with FTO polymorphisms. / Thesis / Master of Science (MSc)

Background: Obesity is a multifactorial disease caused by the interplay of environmental and genetic risk factors. With the prevalence of obesity more than doubling since 1980, this disease has become a global epidemic. The objectives of this research were to (1) review the current evidence of gene-environment interactions (GEI) in the field of obesity, (2) investigate novel GEI involving sedentary behaviour, sleep duration and alcohol consumption, (3) assess GEI using a cumulative environmental risk score, and (4) provide an overview of methodological weaknesses in GEI studies and provide suggestions for future directions. Methods: The data for the gene-environment interaction analyses were collected from the EpiDREAM study: a cohort study including participants of six ethnic backgrounds from 17 countries worldwide. A subset of 17 423 participants with complete genotype and phenotype information was included in the analysis. Twenty-three obesity predisposing single nucleotide polymorphisms (SNPs) were analyzed independently and as a genetic risk score (GRS). Linear regression models were used to analyze these interactions. Results: Heritability, monogenic and polygenic obesity studies provide converging evidence that obesity-predisposing genes interact with a variety of environmental exposures including physical activity and diet patterns. In the EpiDREAM cohort, we found that increased sedentary time did not interact with obesity predisposing SNPs or the GRS to modulate BMI. The interaction between sedentary time and physical activity was also not significant. We observed a U-shaped association between sleep duration and BMI and sleep duration did not appear to moderate the impact of the obesity predisposing SNPs or the GRS. However, we did observe an alcohol x FTO rs1421085 interaction, whereby increased alcohol consumption attenuated the impact of FTO rs1421085 variation on BMI. We also found that the combined effect of several environmental risk factors significantly modified the effect of FTO rs3751812 on BMI. Specifically, we found that the effect of the FTO rs3751812 SNP on BMI was over two times greater among those in the highest quartile of environmental risk compared to those in the lowest quartile. The GRS did not interact with any of the exposures tested. Discussion: Our results indicate that sedentary behaviour did not moderate the impact of obesity predisposing genes, while alcohol consumption decreased the impact of variation in FTO rs3751812 on BMI. We also observed that variation in FTO rs3751812 interacted with a cumulative environmental risk score to moderate BMI. The growing body of GEI evidence has provided a deeper understanding of obesity aetiology and may have tremendous applications in the emerging field of personalized medicine and individualized lifestyle recommendations. Although the number of gene-environment interaction analyses has increased rapidly across multiple disciplines, addressing methodological concerns such as statistical modeling, confounding, biological assumptions and measurement precision will be necessary to fully exploit the potential of the GEI field. With the development of new methodological and measurement techniques such as hypothesis-free genome wide interaction studies and deep phenotyping, it may be possible to translate the information from GEI studies into public health policy and personalized medicine for obesity and other complex human diseases. / Thesis / Doctor of Philosophy (PhD)

No description available.

The frontostriatal dopamine system modulates brain function and is affected by both genetic and environmental factors. Dysfunction of this system is associated with many pathological states, including schizophrenia. The enzyme catechol-O- methyltransferase (COMT) metabolises dopamine and its gene contains a polymorphism (Val¹⁵⁸Met) that affects enzyme activity. Delta-9- tetrahydrocannabinol (THC), the main psychoactive component of cannabis, has been suggested to interact with this polymorphism to increase the risk for psychosis and cognitive impairments. Dopaminergic mechanisms are a plausible candidate for mediating this interaction. I used microdialysis coupled with high performance liquid chromatography (HPLC) to examine the effects of THC on extracellular dopamine and its metabolites in the nucleus accumbens, dorsal striatum and medial prefrontal cortex (mPFC) in freely moving mice. Following acute COMT inhibition with tolcapone, THC increased extracellular dopamine levels in the nucleus accumbens in tolcapone-, but not in vehicle-, treated mice. The introduction of the low activity Met allele into the COMT gene produced a highly specific, novel mouse model of the Val158Met polymorphism. In contrast to the effects of acute COMT inhibition, the Met allele protected against THC-induced changes in accumbal dopamine. No interactive neurochemical effects were observed in the dorsal striatum (pharmacological and genetic study) or in a preliminary study of the mPFC (genetic study only). On a progressive ratio task measuring motivational salience, the direction of the interactive effect between COMT genotype and THC differed between 2 independent cohorts and provided tentative leads that stress/arousal-dependent effects on COMT may have a confounding effect. My data provide evidence that COMT activity modulates the effect of THC on accumbal dopamine function, and suggest the mechanism through which this interaction is mediated differs between acute and lifelong reduction in COMT activity. Through the interactive effect on the dopaminergic system, the data provide a potential mechanism for the reported interaction between COMT and cannabis/THC in determining psychosis risk and cognitive impairments.

In recent years, screening approaches known as two-step methods have been proposed to detect gene-environment interactions for genome-wide association studies (GWAS). Genetic and environmental factors are believed to affect disease outcome as well as various quantitative traits such as height and blood pressure. The performance of the two-step methods has not been demonstrated in the quantitative trait setting. This thesis examines the method proposed by Wang and Abbott (2008) for generating genotyped markers in linkage disequilibrium (LD) and takes this approach in simulating data pertaining to a quantitative trait. The simulation results demonstrate that the two-step methods maintain type I error and have power to detect the quantitative trait locus. In this setting, the EG method (Murcray et al., 2009) is influenced by the strength and structure of the gene-environment dependency, the sample type, and the disease model. As such, the power of the EG method can fluctuate depending on the type of data while the DG method (Kooperberg and LeBlanc, 2008) remains fairly robust across a wide range of scenarios. The performance of the combined two-step approaches (EDGE (Gauderman et al., 2013) and H2 (Murcray et al., 2011) methods) tends to favour the more powerful underlying method. The power of the EDGE method can be improved if DG and EG demonstrates similar power while the H2 method can be made more powerful by choosing the appropriate parameters. / Thesis / Master of Science (MSc)

No description available.

The present thesis is based on four studies focusing on alcohol consumption among Swedish adolescents, and therewith related psychosocial and genetic factors. One main objective was to study the reasons for drinking alcohol among different population - representative samples of adolescents in order to identify motives for drinking. Relationships between these drinking motives, alcohol consumption, and alcohol - related problems were also investigated. Three motives emerged from this study: social - enhancement, coping and dominance. The association with alcohol consumption and alcohol - related problems was positive for social - enhancement and coping motives, but negative for the dominance motive. A significant heritability of alcohol use disorders has been demonstrated by family, adoption and twin studies. Environmental influences have also been acknowledged to play an important role in the development of alcohol use disorders. Moreover, the interaction between genetic and environmental factors is likely to influence the risk - resilience for alcohol use disorders. In view of this knowledge, plausible candidate polymorphisms were considered in gene - environment interaction models. An effect of the genetic polymorphisms was only present when a G x E model was considered. A genetic variant of the clock gene Period2, in an interaction with sleep problems, was studied in relation to alcohol consumption among adolescents. High alcohol consumption was associated with the AA genotype of the PER2 SNP10870 polymorphism, in an interaction with several and frequent sleep problems, among adolescent boys. A genetic variant in the opioid µ receptor 1 gene, in an interaction with alcohol consumption, was studied in relation to depressive symptoms. Depressive symptoms were predicted by the G allele of the OPRM1 A118G polymorphism, in an interaction with high alcohol consumption, among adolescent girls. Additionally, the PER2 SNP10870 and the OPRM1 A118G polymorphisms were studied in a sample of severely alcoholic females. Furthermore, alcohol consumption was assessed by using different instruments, such as biomarkers and surveys. Comparisons were carried out to identify the most suitable method to assess alcohol consumption among adolescents. Questionnaire and interview seemed more suitable tools than biomarkers in this regard.The results eventually support the importance of psychosocial and genetic influences, and their interaction effect on alcohol consumption among adolescents.

<p>Background: Previous research suggests that early life socioeconomic stress and certain genetic polymorphisms may be partly associated with increased adiposity; however, research on both genetic and environmental predictors fail to account for the dramatic increase in obesity over that last several decades. Hypothesis: It was hypothesized that a GxE interaction between DRD2-related SNPs and parental education would predict trunk and total fat mass. This same interaction would also predict total calories from a 24-hour diet recall, which would mediate its effect on trunk and total fat mass. Sample: The current study analyzed genetic and psychosocial data from 697 participants collected for the Family Heart Study, an investigation examining the relationship between psychosocial behaviors and cardiovascular risk factors. Methods: Interactions were assessed between four single nucleotide polymorphisms (SNPs) in the D2 receptor and ANKK1 genes and tertiles of parental education predicting DXA-scan-measured trunk and total body fat mass. A measure of total calories, as assessed by a 24-hour diet recall, was tested as a mediator of this effect. Results: An interaction between mother's education and RS1116313 SNP predicted trunk fat (F(4,191)=2.94, p=0.022) and total body fat (F(4, 191)=3.94, p=0.004). The effects were driven by a reduction in trunk and total fat mass among C/C or T/T homozygotes with a high mother's education, which was not observed among C/T heterozygotes. Father's education was neither an interactive nor a main effect predictor in any models. Interactions predicting total calories were also non-significant, and no support for mediation was found. Post-hoc analyses revealed that leisure activity was also not a mediator. Alternatively, certain dietary components were predicted by the interactions between mother's education and RS1124492 and between mother's education and RS1800498. Conclusions: Trunk and total body fat composition are predicted by an interaction between mother's education and the RS1116313 SNP. This effect does not appear to be mediated by total calories or leisure activity. Other SNPs associated with the D2 receptor interact with mother's education to predict dietary components.</p> / Dissertation

The aetiology of Parkinson's disease (PD) is still unclear. Research findings suggest that both environmental and genetic factors may contribute to its development. The interactions between genes and the environment might exist and play a key role. Cigarette smoking was found to be one of the few factors exhibiting a protective effect. If chemical compounds found in cigarette smoke influence PD risk, the difference in the ability of certain individuals in metabolising these substances might alter their susceptibility to the risk of developing PD. Many metabolic enzyme genes exhibit polymorphic traits with alteration of gene function. These might be associated with an altered susceptibility of individuals to PD. Few studies have examined the hypothesis that metabolic enzyme gene polymorphisms might modulate the effect of smoking on PD risk. However, it is crucial to consider these potential interactions when we try to elucidate the aetiology of PD. Even if each factor only contributes a slight variation and influences a small portion of the whole population, non-linear and unpredictable interactions may account for a high proportion of the aetiological fraction. Previous studies have not been strictly designed to examine the interactions between smoking and metabolic enzyme genetic polymorphisms. These studies have not been able to elucidate the extent of the interaction. Therefore, this PhD project attempted to examine whether genetic factors, operating in the phase one and phase two metabolic pathways, interact with smoking to influence the development of PD. This is the first genetic epidemiological study of PD specifically addressing this issue. The research aids in further understanding the aetiology of PD and may be useful for identifying people at higher risk. A case-only design was chosen for this project for two reasons: first, PD is a relatively rare disease and the case-only design is much more efficient at detecting gene-environment interactions; second, the PD cases for the project were recruited over the past few years and represent a prevalence series, for which an appropriate comparison group for the cases is difficult to identify and recruit. In a case-only study, only cases are used to investigate the multiplicative effects of the exposures and susceptible genotypes of interest, while non-case subjects (traditionally controls) are solely used to test the independence between the exposure and the susceptible genotype. Therefore, this approach avoids the challenges of control selection, a major limitation inherent in the case-control approach. This thesis comprised of three independent studies: the first study investigated the interactions between genetic polymorphisms of GSTM1, P1, T1 and Z1 and smoking in PD; the second study examined the interactions between genetic polymorphisms of CYP2E1 and smoking in PD; and the third study examined the interactions between genetic polymorphisms of CYP2D6 and smoking in PD. The first two studies recruited 400 white Caucasian PD cases from both hospital wards and private neurology clinics (230 men and 170 women). The third study further included 142 white Caucasian PD cases newly recruited from the same sources (542 in total, 321 men, and 221 women). The mean age of cases was 67 years with the average onset age at 60 years. GSTM1, GSTP1, GSTT1, GSTZ1 AND CYP2E1 genotyping processes were performed using protocols previously published with minor modification, whereas CYP2D6 genotyping methods were mainly developed by me with assistance from associate supervisor Dr. George Mellick. Reliability and validity of the PCR and RFLP methods were assessed through re-conducting the genotype assays using at least a 10% sample of our DNA samples. The results for all re-assessments were 100% concordant. Crude bivariate analyses were adjusted for potential confounding effects of the variables, including age at onset, gender, family history of PD and pesticide exposures. Among our unaffected, aged subjects (mean age: 63.9 years, sd: 11.4 years), the genotype frequencies at each locus were similar to those reported in other Caucasian populations. The first study showed that the proportion of carriers of the GSTP1-114Val allele (mutant) increased with increasing smoking dose from 0 to > 30 pack-years. Homozygotes of the 114Ala allele (wild-type) decreased with increasing smoking dose (trend test: p=0.02). This trend existed both in male and female cases. This dose-effect relationship was most significant in the group of cases with late-onset PD (i.e., age at onset > 55 years) with the ORicase-only values of 1.88 (95%CI: 0.65-5.48) and 2.63 (95%CI: 1.07-6.49) for > 0-10 and > 10 pack-years, respectively. No similar trend was found among our unaffected, aged subjects (p=0.42). Haplotype analyses revealed significant differences for GSTP1 haplotypes between smoking and non-smoking PD cases (ORicase-only for *C haplotype=2.00 (95%CI: 1.11-3.60), p=0.03). In this case, smoking-exposed PD cases were more likely to posses the *C haplotype defined by A to G base-pair transition at nucleotide +313 and C to T base-pair transition at nucleotide +341 (at amino acid level, valine at both positions 105 and 114). The second study found no difference in CYP2E1 genotype frequencies between PD cases who ever smoked compared to those who never smoked (odds ratio for interaction (ORi) = 1.00 (95% CI: 0.39-2.51, p=0.99)). No CYP2E1 gene-smoking interactions were detected in relation to age at onset of PD. The third study found that among cases without regular pesticide exposures, CYP2D6 PMs who smoked more than 5 pack-years had a later mean age at disease onset (68.6 years) than those with extensive metaboliser phenotypes (EMs) (61.1 years, p=0.02) and non-smokers (60.5 years, p=0.01). Analysis of aged subjects without PD confirmed that neither smoking status nor CYP2D6 PM status was associated with age itself. Our data suggest: 1. smoking exposure is independent of GSTM1, P1, T1, Z1 and CYP2E1 genotypes; 2. smoking may be, to some extent, associated with CYP2D6 genotypes; 3. there are no multiplicative interactive effects linking smoking and GSTM1, T1, Z1 or CYP2E1 genotypes with the risk for PD; 4. there is a multiplicative interactive effect between smoking and GSTP1 haplotype - particularly for genotypes carrying the 114Val allele; and 5. there is a multiplicative interactive effect between smoking and CYP2D6 PMs - particularly for people who ever smoked cigarettes more than 5 pack-years. In general, this thesis provides a model for exploring the gene-smoking interactions in PD. Further studies need to consider the recruitment of a large number of population-based and randomly-selected samples and to pay more attention to measurement of environmental exposures. Further studies also need to examine simultaneously the impact of smoking, pesticide exposures and other potential risk factors on PD. These studies will build evidence for interactions contributing to this common neurological movement disorder.

abstract: The present study utilized longitudinal data from a high-risk community sample (n=254, 52.8% female, 47.2% children of alcoholics, 74% non-Hispanic Caucasian) to test questions concerning the effects of genetic risk, parental knowledge, and peer substance use on emerging adult substance use disorders (SUDs). Specifically, this study examined whether parental knowledge and peer substance use mediated the effects of parent alcohol use disorder (AUD) and genetic risk for behavioral undercontrol on SUD. The current study also examined whether genetic risk moderated effects of parental knowledge and peer substance use on risk for SUD. Finally, this study examined these questions over and above a genetic "control" which explained a large proportion of variance in the outcome, thereby providing a stricter test of environmental influences. Analyses were performed in a path analysis framework. To test these research questions, the current study employed two polygenic risk scores. The first, a theory-based score, was formed using single-nucleotide polymorphisms (SNPs) from receptor systems implicated in the amplification of positive effects in the presence of new/exciting stimuli and/or pleasure derived from using substances. The second, an empirically-based score, was formed using a data-driven approach that explained a large amount of variance in SUDs. Together, these scores allowed the present study to test explanations for the relations among parent AUD, parental knowledge, peer substance use, and SUDs. Results of the current study found that having parents with less knowledge or an AUD conferred greater risk for SUDs, but only for those at higher genetic risk for behavioral undercontrol. The current study replicated research findings suggesting that peer substance use mediated the effect of parental AUD on SUD. However, it adds to this literature by suggesting that some mechanism other than increased behavioral undercontrol explains relations among parental AUD, peer substance use, and emerging adult SUD. Taken together, these findings indicate that children of parents with AUDs comprise a particularly risky group, although likelihood of SUD within this group is not uniform. These findings also suggest that some of the most important environmental risk factors for SUDs exert effects that vary across level of genetic propensity. / Dissertation/Thesis / Doctoral Dissertation Psychology 2015