Consistent effects of the genetics of happiness across the lifespan … – Nature.com

Cohorts, genotyping and phenotyping Adolescent brain cognitive development (ABCD) ABCD cohort description

The Adolescent Brain Cognitive Development (ABCD) cohort is a longitudinal study of brain development and child health7. Investigators at 21 sites around the USA conducted repeated assessments of brain maturation in the context of social, emotional, and cognitive development, as well as a variety of health and environmental outcomes. We analysed data from release 3.0. At the time of the survey questions, the children ranged in age from 9 to 12 years. Informed written consent was provided by parents and assent was provided by children. The ABCD research protocol approved was approved by the Institutional Review Board of University of California San Diego (IRB# 160091)16.

Data used in the preparation of this article were obtained from the Adolescent Brain Cognitive DevelopmentSM (ABCD) Study (https://abcdstudy.org), held in the NIMH Data Archive (NDA). This is a multisite, longitudinal study designed to recruit more than 10,000 children age 910 and follow them over 10 years into early adulthood. The ABCD Study is supported by the National Institutes of Health and additional federal partners under award numbers U01DA041048, U01DA050989, U01DA051016, U01DA041022, U01DA051018, U01DA051037, U01DA050987, U01DA041174, U01DA041106, U01DA041117, U01DA041028, U01DA041134, U01DA050988, U01DA051039, U01DA041156, U01DA041025, U01DA041120, U01DA051038, U01DA041148, U01DA041093, U01DA041089, U24DA041123, U24DA041147. A full list of supporters is available at https://abcdstudy.org/federal-partners.html. A listing of participating sites and a complete listing of the study investigators can be found at https://abcdstudy.org/consortium_members/. ABCD consortium investigators designed and implemented the study and/or provided data but did not necessarily participate in the analysis or writing of this report. This manuscript reflects the views of the authors and may not reflect the opinions or views of the NIH or ABCD consortium investigators.

The ABCD data repository grows and changes over time. The ABCD data used in this report came from https://doi.org/10.15154/1526432) DOIs can be found at https://dx.doi.org/10.15154/1526432. All methods were carried out in accordance with relevant guidelines and regulations.

DNA was extracted from saliva samples of the ABCD participants17. These samples were genotyped on the Affymetrix NIDA SmokeScreen Array (Affymetrix, Santa Clara, CA, USA). The QC procedures are described in full at the following URL: https://doi.org/10.15154/1503209.

ABCD genetic principal components (GPCs) were created using genotyped only SNPs using plink-pca flag.

A set of questions taken from the ABCD Youth NIH Toolbox Positive Affect Items was used. These questions measured aspects of positive emotions and affective well-being in the past week, specifically being attentive, delighted, calm, relaxed, enthusiastic, interested, confident, energetic and able to concentrate. Responses were measured as not true, somewhat true or very true. Each item was analysed separately as well as a combined score that was the sum of responses to the individual questions. In addition to the happiness PGS the models were adjusted for age, sex, and principal genetic components (PGCs) 18.

As the initial UK Biobank GWAS was run in the white British sub-group, testing was performed firstly in the white (as defined by ABCD) participants and secondly in the whole sample, with ancestry treated as a factor variable. The other ancestral backgrounds of this cohort as defined by ABCD are; White, Black, Hispanic, Asian, and Other (Table S6).

Creation of the derived MRI variables from the ABCD cohort has been described in detail elsewhere18. For the purposes of this study, total frontal lobe volume was derived by summing the 22 frontal lobe subsection variables of the left and right hemisphere19. Additionally, we looked at total grey and white matter volume and left and right hippocampus volume. The hippocampal body and tail regions and white matter hyperintensity volume were not available for replication. All outcomes were transformed into z scores and all models were adjusted for the happiness PGS, age, sex, PGCs 18, and MRI site. For models that included participants from different ancestries, a factor variable for ancestry was included (Table S7). Models were weighted to match the American community survey (ACS) data by the weighting variable acs raked propensity score. Relationship filtering was also performed removing one individual at random from any pair of participants with valid phenotypes, who were determined to be related by ABCD.

Add Health is a nationally representative cohort study of more than 20,000 adolescents from the USA who were aged 1219 years at baseline assessment in 199495. They have been followed through adolescence and into adulthood with five in-home interviews in five waves (IV) conducted in 1995, 1996, 20012002, 20082009 and 20162018. In this analysis, participants ranged from 24.3 to 34.7 years old, 53% were female and 62% were non-Hispanic white. The study was approved by the University of California San Diego Institutional Review Board (IRB #190002XX). Informed consent was obtained from all subjects.

Saliva samples were obtained as part of the Wave IV data collection. Two Illumina arrays were used for genotyping, with approximately 80% of the sample genotyped with the Illumina Omni1-Quad BeadChip and the remainder of the group genotyped with the Illumina Omni2.5-Quad BeadChip. After quality control, genotyped data were available for 9974 individuals (7917 from the Omni1 chip and 2057 from the Omni2 chip) on 609,130 SNPs present on both genotyping arrays20. Imputation was performed separately for European ancestry (imputed using the HRC reference panel) and non-European ancestry samples (imputed using the 1000 Genomes Phase 3 reference panel)21. For more information on the genotyping and quality control procedures see the Add Health GWAS QC report online at: https://addhealth.cpc.unc.edu/wp-content/uploads/docs/user_guides/AH_GWAS_QC.pdf.

Add Health Genetic Principal components (variable name pspcN, where N is the number of the PC) were derived centrally by Add Health. To prevent identification of individuals they are randomly reordered in sets of 5, i.e. PCs 15 were reordered so PC1 was may not be the PC with the largest variance. We adjusted models for the first 2 sets of PCs i.e. GPCs 110.

The outcome happiness variable was collected during the at-home interview of Wave IV and was derived from the response to the question: How often was the following true during the past seven days? You felt happy. Responses were given as: never or rarely; sometimes; a lot of the time; most of the time or all of the time; refused; don't know. Those who responded with the latter two options were excluded. Remaining categories were coded from never=0 to all of the time=3.

Ancestry in Add Health is defined in the psancest variable as European, African, Hispanic and East Asian (Table S8). Additionally, Add Health provides a weighting variable to make the results reflective of the US population. In these analyses the models were weighted by the Wave IV variable gswgt4_2.

UK Biobank is a cohort of over half a million UK residents, aged from approximately 4070 years at baseline. It was created to study environmental, lifestyle and genetic factors in middle and older age22. Baseline assessments occurred over a 4-year period, from 2006 to 2010, across 22 UK centres. These assessments were comprehensive and included social, cognitive, lifestyle and physical health measures.

UK Biobank obtained informed consent from all participants, and this study was conducted under generic approval from the NHS National Research Ethics Service (approval letter dated 29 June 2021, Ref 21/NW/0157) and under UK Biobank approvals for application #71392 Investigating complex relationships between genetics, exposures, biomarkers, endophenotypes and cardiometabolic, inflammatory, immune and brain-related health outcomes (PI Rona Strawbridge; GWAS)#17689 (PI Donald Lyall; imaging).

In March 2018, UK Biobank released genetic data for 487,409 individuals, genotyped using the Affymetrix UK BiLEVE Axiom or the Affymetrix UK Biobank Axiom arrays (Santa Clara, CA, USA) containing over 95% common content. Pre-imputation quality control, imputation and post-imputation cleaning were conducted centrally by UK Biobank (described in the UK Biobank release documentation)23.

Several structural and functional brain MRI measures are available in UK Biobank as imaging derived phenotypes (IDPs)24. The brain imaging data, as of January 2021, were used (N=47,920). Participants were excluded if they had responded to either of the happiness questions used for the GWAS meta-analysis, were missing more than 10% of their genetic data, if their self-reported sex did not match their genetic sex, if they were determined by UK Biobank to be heterozygosity outliers, and if they were not of white British ancestry (classified by UK Biobank based on self-report and genetic principal components)23.

Brain imaging data used here were processed and quality-checked by UK Biobank and we made use of the IDPs25,26. Details of the UK Biobank imaging acquisition and processing, including structural segmentation and white matter diffusion processing, are freely available from three sources: the UK Biobank protocol: http://biobank.ctsu.ox.ac.uk/crystal/refer.cgi?id=2367 and documentation: http://biobank.ctsu.ox.ac.uk/crystal/refer.cgi?id=1977 and in protocol publications (https://biobank.ctsu.ox.ac.uk/crystal/docs/brain_mri.pdf).

We investigated key imaging substrates previously associated with psychological health e.g., mood disorder, cognitive health. Total white matter hyperintensity volumes were calculated on the basis of T1 and T2 fluid-attenuated inversion recovery, derived by UK Biobank. White matter hyperintensity volumes were log-transformed due to a positively skewed distribution. We constructed general factors of white matter tract integrity using principal component analysis. The two separate unrotated factors used were fractional anisotropy (FA), gFA, and mean diffusivity (MD), gMD, previously shown to explain 54% and 58% of variance, respectively27. We constructed a general factor of frontal lobe grey matter volume using 16 subregional volumes as per Ferguson et al.27. Total grey matter and white matter volumes were corrected for skull size (by UK Biobank). Models were adjusted for the happiness PGS, age, sex, PGCs 18.

LDpred28 established the LD structure of the genome using a reference panel of 1000 unrelated white British UK Biobank participants (the PGS training set). These participants had not been used in the discovery GWAS or have valid MRI data and passed the same QC as described above. SNPs were excluded if they had MAF<0.01, had HWE P<1 106 or had imputation score<0.8. Scores were then created in the validation set using an infinitesimal model. Models using polygenic scores (PGS) derived using LDpred were adjusted for age, sex, genotyping array and the first eight GPCs.

Due to the lower cohort size of ABCD and Add Health, it would not have been possible to remove 1000 participants from the analyses to use as a training set without markedly reducing the power of the analyses. Therefore, we used the same 1000 unrelated UK Biobank participants as the training set to establish LD and this was used to generate the PGS for the participants in these datasets29. The only additional step was to find the SNPs that were found in both the training (UK Biobank) and validation (ABCD and Add Health) datasets and passed the same SNP filtering criteria in both datasets, with an additional filter that MAF threshold was set at>0.0130. The number of SNPs in each LDpred PGS can be found in supplementary table (S9).

For each pair of related individuals (as determined by ABCD using variables genetic paired subjected 14) one participant was excluded at random. Models were adjusted for age at interview, sex and the first 10 GPCs. For multi-ancestry models, ancestry was treated as a factor variable.

p values for analyses were false discovery rate (FDR)-adjusted31.

View original post here:
Consistent effects of the genetics of happiness across the lifespan ... - Nature.com