Generation Scotland


Research publications from 2017.

Balancing the Local and the Universal in Maintaining Ethical Access to a Genomics Biobank. Heeney, C and Kerr, S.M. (2017). BMC Medical Ethics.

There are compelling public goods arguments for sharing research data as widely and as openly as possible, but in medical research there is also a responsibility to ensure individual privacy and to operate within the consent given at the time of recruitment. This paper uses the working practice of Generation Scotland, which operates a managed access policy, to discuss these issues.


Protein-altering variants associated with body mass index implicate pathways that control energy intake and expenditure in obesity. Turcot, V., Lu, Y., Highland, H.M., Schurmann, C., Justice, A.E., Fine, R.S… Loos, R.J.F. (2017). Nature Genetics 50, 26–41.

This study combined data from multiple cohorts, including GS, to look for evidence of gene coding variants that affect body mass index. Data from over 700,000 participants were analysed and 13 genes carrying damaging mutations identified. This may seem like a poor return compared to the 250 gene variants previously reported by GWAS, but each of these newly found gene damaging variants have effect sizes 10 times that of GWAS ‘hits’. In the case of MC4R mutation carriers, this equated with a 7kg weight increase over non-carriers.


Association analysis in over 329,000 individuals identifies 116 independent variants influencing neuroticism. Luciano, M., Hagenaars, S.P., Davies, G., Hill, W.D., Clarke, T.K., Shirali, M… Deary, I.J. (2017). Nature Genetics 50(1):6-11

In this major study, GS acted as a replication cohort for the findings made in the much larger UK Biobank study. 329,000 participants had measures of neuroticism and genetic data. 116 independent genetic influences on this universal human personality trait were found. These were in turn linked to brain processes that were shared with risk of depression.


Association between exposure to second-hand smoke and telomere length: cross-sectional study of 1303 non-smokers. Lu, L., Johnman, C., McGlynn, L., Mackay, D., Shiels, P., Pell, J. (2014). International Journal of Epidemiology.

GS collected data on smoking and exposure to smoke. This study showed that high exposure to second hand smoke accelerated normal biological ageing, measured by decreased telomere length.  These findings support efforts to protect the public from second hand smoke. 


Genome-wide haplotype-based association analysis of major depressive disorder in Generation Scotland and UK Biobank. Howard, D. M., Hall, L. S., Hafferty, J. D., Zeng, Y., Adams, M. J., Clarke, T-K., ... McIntosh, A. M. (2017). Translational Psychiatry, 7.

Most studies looking for genetic associations with a disease trait consider common genetic variants one at a time. This is the conventional GWAS approach.  Here, a modified approach was taken to try to capture all of the genetic variation region by region, using a haplotype approach, which has the potential to highlight the ancient origin of the causal genetic variation. This approach proved successful in finding two hitherto unrecognised genetic risk loci for major depressive disorder in GS, one of which was replicated in the UK Biobank cohort and coincided with a locus previously linked to bipolar disorder.


Genetic risk of major depressive disorder: the moderating and mediating effects of neuroticism and psychological resilience on clinical and self-reported depression. Navrady, L.B., Adams, M.J., Chan, S.W.Y.,   Major Depressive Disorder Working Group of the Psychiatric Genomics ConsortiumS., Ritchie, J., McIntosh, A.M. (2017). Psychological Medicine.

Much of psychology and psychiatry depends upon self-reports. What this study finds is that neuroticism increases and psychological resilience reduces self-reported depression. 


Association Between Schizophrenia-Related Polygenic Liability and the Occurrence and Level of Mood-Incongruent Psychotic Symptoms in Bipolar Disorder. Allardyce, J., Leonenko, G., Hamshere, M., Pardinas, A. F., Forty, L., Gordon-Smith, K… Escott-Price, V. (2017). JAMA Psychiatry.

The overlap in genetic risk for schizophrenia and bipolar disorder is now well established, but this study takes an important step forwards by showing a gradient of genetic effect that correlates with a gradient of symptom severity and overlap between these notionally distinct psychiatric conditions.


Biological and clinical insights from genomic analysis of plasma lipids in >300,000 individuals. Liu, D.J., Peloso, G.M., Yu, H., Butterworth, A.S., Wang, X., Mahajan, A… Kathiresan, S. (2017). Nature Genetics. 49, pages 1758–1766.

GS contributed exome chip data to this massive study (>300,000 participants) of genetic association with cholesterol. 444 independent variants in 250 loci were found. Various trait specific and functional studies were undertaken to test and validate the findings.  


Self-Reported Medication Use Validated Through Record Linkage to National Prescribing Data. Hafferty, J. D., Campbell, A. I., Navrady, L. B., Adams, M. J., MacIntyre, D., Lawrie, S. M… McIntosh, A. M. (2017). Journal of Clinical Epidemiology pii: S0895-4356(17)30306-2.

Much hope rests on the promise of data science applied to health research. This study set out to test the reliability of self-reported medication use in GS participants through linkage to routine NHS Scotland prescribing data. Self-reported antidepressant and antihypertensive usage showed good correlation with national prescribing data, but mood stabilisers less so, suggesting that record linkage may provide a more reliable indicator of past illness and a means to monitor present and future onset of illness.


Electronic health record and genome-wide genetic data in Generation Scotland participants. Kerr, S. M., Campbell, A., Marten, J., Vitart, V., Mcintosh, A. M., Porteous, D. J., Hayward, C (2017). Wellcome Open Research

Much hope rests on the promise of data science applied to health research. This study provides the first formal demonstration that routine NHS Scotland biochemistry data, in this case for levels of blood urate, can substitute for clinic based measures and replicate genetic findings. This has two important impacts, first by avoiding clinic based costs for sample collection and analysis and second by providing a means to follow subjects over time to follow the course of illness. In short, the Electronic Health Record is a rich resource of real world data that can be used in research to characterise the health trajectory of participants, available at low cost and a high degree of timeliness.


Haplotype-based association analysis of general cognitive ability in Generation Scotland, the English Longitudinal Study of Ageing, and UK Biobank. Howard, D. M., Adams, M. J., Clarke, T-K., Wigmore, E. M., Zeng, Y., Hagenaars, S. P… Mcintosh, A. M. (2017). Wellcome Open Research.

Applying the same haplotype approach as for major depressive disorder (Howard et al, 2017 Translational Psychiatry), this study looked for evidence of genetic association with general cognitive ability in a combined data set of 48,000 participants from GS and two other cohorts. A couple of tantalising observations were made, but none matched stringent statistical significance, perhaps reflecting the available sample size.   


Genome-wide association study of alcohol consumption and genetic overlap with other health-related traits in UK Biobank (N=112 117).  Clarke, T-K., Adams, M.J., Davies, G., Howard, D.M., Hall, L.S., Padmanabhan, S., ... McIntosh, A.M. (2017). Molecular Psychiatry. 

Alcohol misuse is a major and global health problem linked to many serious conditions. The UK Biobank and GS asked participants about their alcohol consumption. This study made new genetic discoveries linked to alcohol consumption and, importantly, showed that there was genetic overlap with years in schooling, cholesterol, smoking and weight. 


Genome-wide meta-analysis associates HLA-DQA1/DRB1 and LPA and lifestyle factors with human longevity. Joshi, P. K., Pirastu, N., Kentistou, K. A., Fischer, K., Hofer, E., Schraut, K.E… Wilson, J.F. (2017). Nature Communications. 8:910 

What makes for a long and healthy life? Where does the balance lie between nature (genes) and nurture (environment)? By using GS and other cohorts with genetic data and for whom the age of death of parents was known, this study was able to replicate 5 findings on longevity from previous studies and add 2 more. Moreover, it was possible to show a positive effect on lifespan from giving up smoking, educational attainment, openness to new experience and cholesterol levels and a negative effect of risk of heart disease, cigarette smoking, lung cancer, diabetes and body fat.  On average, a body mass index increase of one unit reduces lifespan by 7 months, while 1 extra year of education adds 11 months to expected lifespan.


Regional variation in health is predominantly driven by lifestyle rather than genetics. Amador, C., Xia, C., Nagy, R., Campbell, A., Porteous, D., Smith, B.H… Haley, C.S. (2017). Nature Communications. 8:801 

As we recruited participants in family groups from different regions of Scotland, this study was able to ask the important questions of what best explains the difference in health and lifespan between different cities. For example, mean life expectancy in Edinburgh is several years higher than in Glasgow, despite being just 50 miles apart. The study found that regional variation for most obesity traits was associated with lifestyle and socioeconomic variables, such as smoking, diet and deprivation which are potentially modifiable. This has important implications for healthcare policies, suggesting that inequalities can be tackled with appropriate social and economic interventions.


CNV-association meta-analysis in 191,161 European adults reveals new loci associated with anthropometric traits. Macé, A., Tuke, M. A., Deelen, P., Kristiansson, K., Mattsson, H., Nõukas … Kutalik Zoltán. (2017). Nature Communications. 8: 744.

The frequent source of individual genetic differences are single base differences (SNPs), but there are also regions of the genome where longer tracts can vary in the number of copies any one individual carries, with some being deleted and others duplicated (copy number variations, CNVs). CNVs have the potential to alter the effective dose of entire genes and so are important clues as to where genes involved in human traits and disease may reside.  This study looked at all of the GS cohort and others, just over 190,000 samples in total, for evidence of CNVs affecting body measurements. Genes previously identified by the SNP method were identified independently by the CNV method and new findings, not yet seen by the SNP method, were also found. Although relatively rare, the effect of the CNVs is more profound on carriers than the corresponding SNP effects.


Chronic pain in families: A cross-sectional study of shared social, behavioural, and environmental influences. Campbell, P., Jordan, K. P., Smith, B. H., Scotland, G., Dunn, K. M. (2017). Pain.

Chronic pain is common and debilitating, but often poorly managed. The reasons why some but not others suffer from chronic pain, when all other things seem to be equal, is important to understand.  Since we asked all GS participants to tell us about pain, we could look to see if there were common factors in relatives and households. Household context did seem to be important, but mostly it was down to personal circumstances. 


Inherited chromosomally integrated human herpesvirus 6 genomes are ancient, intact and potentially able to reactivate from telomeres. Zhang, E., Bell, A.J., Wilkie, G.S., Suárez, N.M., Batini, C., Veal, C.D… Royle, N.J. (2017). Journal of Virology 91(18)

About 1 in a 100 people worldwide are carriers of chromosomally integrated human herpes virus 6 genomes. Using samples from GS with known HHV6 integrations and for others samples from around the world, this study came to the surprising conclusion that most integrations are ancient, going back many generations in time, and that all of the viral genes are usually intact and could potentially be reactivated.  


Do regional brain volumes and major depressive disorder share genetic architecture? A study of Generation Scotland (n=19 762), UK Biobank (n=24 048) and the English Longitudinal Study of Ageing (n=5766). Wigmore, M.E., Clarke, T-K., Howard, D.M., Adams, M.J., Hall, L.S.,  Zeng, Z (2017). Translational Psychiatry.

Brain imaging has been used to identify unusual brain anatomy in cases of schizophrenia and more subtle differences in bipolar disorder. Little is known about major depression.  This study asked whether genes that affect regional brain volumes overlap with those for major depressive disorder, The study drew upon brain imaging data from Generation Scotland (n=19 762), UK Biobank (n=24 048) and the English Longitudinal Study of Ageing (n=5766). There was a modest association with the volume of the hippocampus, the working memory centre of the brain, but no other areas.


Novel Urinary Peptidomic Classifier Predicts Incident Heart Failure. Zhang, Z.-Y., Ravassa, S., Nkuipou-Kenfack, E., Yang, W.-Y., Kerr, S., Koeck, T, … Staessen, J. A.(2017). Journal of American Heart Assosciation 6(8).

There is great interest in finding measures that can be easily made in urine and which detect organ damage. This study identified a set of 96 proteins that did a good job of detecting heart failure. For clinical utility, this heart failure predictor needs to be refined for improved sensitivity and specificity and validated in larger, independent samples.


Novel blood pressure locus and gene discovery using GWAS and expression datasets from blood and the kidney. Wain, L. V., Vaez, A., Jansen, R., Joehanes, R., van der Most, P. J., Mesut Erzurumluoglu, A., … Ehret, G. B. (2017). Hypertension 70(3).

Elevated blood pressure is a major risk factor for cardiovascular disease and has a substantial genetic contribution. Genes that influence blood pressure could identify new drug targets for the treatment of hypertension. This massive study looked for blood pressure risk factors in over 150,000 individuals and checked these in an independent set of over 200,000, including members of GS. 6 new gene signals were found and a previous 2 confirmed.  48 genes showed evidence for involvement in blood pressure regulation.  3 new genes for kidney function were also identified.


Cohort Profile: Stratifying Resilience and Depression Longitudinally (STRADL): a questionnaire follow-up of the Generation Scotland Scottish Family Health Study (GS:SFHS). Navrady, L.B., Wolters, M.K., Macintyre, D.J., Clarke, T-K., Campbell, A.I., Murray, A.D., … McIntosh, A.M. (2017).  International Journal of Epidemiology.

‘Stratifying Resilience and Depression Longitudinally (STRADL)’ is an ambitious 5-year research programme that builds directly on GS:SFHS. It is led by the GS Mental Health Expert Working Group, led by Professor Andrew McIntosh, and supported by GS Executive members Professors David Porteous and Alison Murray, and others. This paper provides a detailed description of the STRADL research aims, study design and protocol.


Intelligence and neuroticism in relation to depression and psychological distress: Evidence from two large population cohorts. Navrady, L. B., Ritchie, S. J., Chan, S. W. Y., Kerr, D., Adams, M. J., Hawkins, E., … McIntosh, A. M. (2017). European Psychiatry.

Neuroticism is a well measured personality trait that seems to correlate with a number of physical and behavioural conditions. By combining data from the Generation Scotland: Scottish Family Health Study (GS:SFHS, n=19,200) and UK Biobank (n=90,529), this study was able to show that neuroticism was strongly associated with increased risk for depression and higher psychological distress.  Interestingly, higher intelligence was found to act as a protective factor against the effects of neuroticism on psychological distress but did not protect against a diagnosis of depression in those high in neuroticism.


Validation of Surrogates of Urine Osmolality in Population Studies. Youhanna, S., Bankir, L., Jungers, P., Porteous, D., Polasek, O., Bochud, M., Hayward, C., Devuyst, O (2017). American Journal of Nephrology, 46(1):26-36.

We asked GS participants to donate a sample of urine at the clinic visit as it is possible to run assays on many components of urine to study kidney, liver and heart function. This paper used GS samples to test and validate a new way to assay urine concentration as an indicator of kidney function and chronic kidney disease.


Genome-wide physical activity interactions in adiposity ― a meta-analysis of 200,452 adults. Graff, M., Scott, R.A., Justice, A.E., Young, K.L., Feitosa, M.F., Barata L, … Kilpeläinen, T.O. (2017). PLoS Genetics 13(4):e1006528.

This study of over 200,000 individuals, including the GS cohort, looked to see if there was any genetic evidence for physical activity modifying the risk of obesity. Eleven new gene findings for obesity were found.


Genome-Wide Meta-Analysis of 241,258 Adults Accounting for Smoking Behavior Identifies Novel Loci for Obesity Traits. Justice, A.E., Winkler, T.W., Feitosa, M.F., Graff, M., Fisher, V.A., Young, K., … Cupples, L.A. (2017). Nature Communications 8:14977

In this study of over 50,000 smokers and nearly 200,000 non-smokers, including the GS cohort, 23 new genes for obesity were found, 9 of which were strongly influenced by smoking status.


Genomic analyses identify hundreds of variants associated with age at menarche and support a role for puberty timing in cancer risk. Day, F. R., Thompson, D. J., Helgason, H., Chasman, D. I., Finucane, H., Sulem, P., … Perry, J. R. B. (2017).  Nature Genetics 49, 834–841

The timing of puberty is variable with early puberty associated with various adult diseases. We asked female GS participants when they had their first period. We combined our data with several other large studies, totalling 370,000 subjects. This very large study confirmed a strong genetic effect on age at menarche, with close to 400 independent signals that suggest a link to cancer susceptibilty. 


Assessing the presence of shared genetic architecture between Alzheimer's disease and major depressive disorder using genome-wide association data. Gibson, J., Russ, T.C., Adams, M.J., Clarke, T.K., Howard, D.M., Fernandez-Pujals, A.M., … McIntosh, A.M. (2017). Translational Psychiatry 7(4):e1094.

Depression is a common feature of Alzheimer’s disease. This study looked to see whether, with the current state of knowledge, there was any evidence for genetic risk sharing between Alzheimer’s disease and major depressive disorder. The answer was no, but the results should be interpreted with caution as we currently have an incomplete picture of the genetic factors that influence these conditions.


Genetic effects influencing risk for major depressive disorder in China and Europe. Bigdeli, T. B., Ripke, S., Peterson, R. E., Trzaskowski, M., Bacanu, S-A., Abdellaoui, A., Andlauer, T. F. M., … Kendler, K. S. (2017). Translational Psychiatry (2017) 7, e1074; doi:10.1038/tp.2016.292.

Major depressive disorder is a common condition with a strong gene influence. We have good measures of depression in GS and have shared these with the Psychiatric Genetics Consortium. In this study, we asked whether the same genes are associated with depression in European populations as in Chinese populations. The evidence is weak, but that reflects in part the incomplete picture we have of the genetic risk factors in each population.


A qualitative study of participants' views on re-consent in a longitudinal biobank. Dixon-Woods, M., Kocman, D., Brewster, L., Willars, J., Laurie, G., Tarrant Carolyn (2017). BMC Medical Ethics.

GS put significant time, resources and effort into the early planning, consultation, ethics and consent processes, but how did participants feel 10 years down the line? In this independent study, GS partipicants were recontacted to discuss their views on consent and attitudes to possible research uses of their samples and data. The study provides broad support for the way in which GS was set up and the information provided to participants, it is good guidance for other biobanking studies. It also raises some questions and suggestions about evolving issues around consent and re-consent.


Exploration of haplotype research consortium imputation for genome-wide association studies in 20,032 Generation Scotland participants. Nagy, R., Boutin, T. S., Marten, J., Huffman, J. E., Kerr, S. M., Campbell, A., ... Hayward, C. (2017). Genome Medicine.

Despite the fact GS has health and genetic data on over 20,000 participants, many studies have needed much higher numbers, hence the emphasis on joining forces with other similar efforts world-wide. However, by comparison to many of these other cohorts, GS has richer and deeper measures of health and can link the clinic data to routine medical record data. Furthermore, unlike many other studies, GS is family-based and so can compare related individuals to get more information on inheritance patterns. This important proof-of-principle study shows that these features can more than compensate for absolute size. New findings, previously hidden in other larger studies, were made in relation to measures of diabetes, heart disease and gout. 


Genetic stratification to identify risk groups for Alzheimer's disease. Marioni, R. E., Campbell, A., Hagenaars, S. P., Nagy, R., Amador, C., Hayward, C., … Deary, I. J. (2017). Journal of Alzheimer's Disease.

The strongest single gene effect for risk of Alzheimer's Disease is APOE4, but there are now many other genes known to have smaller, aggregate effects. This study specifically excluded the APOE4 effect to look closer at the effect of these other genetic determinants in 60-70 year old GS participants, before the typical age of onset for dementia. The study showed a small effect on measures of thinking skills, suggesting that this approach may be useful for predicting those at higher or lower overall risk. 


Genetic epidemiology of motor neuron disease-associated variants in the Scottish population. Black, H. A., Leighton, D. J., Cleary, E. M., Rose, E., Stephenson, L., Colville, S., … Chandran, S. (2017). Neurobiology of Aging, 51, 178.e11–178.e20.

Motor neuron disease is a complex genetic condition that is relatively common in the Scottish population. When new candidate genes are proposed it is important to check whether these replicate and/or whether the same gene can be mutated in other ways in independent cases. GS provided the well matched control samples to ensure that the conclusions from this study were sound and estimate the frequency of MND mutations in the Scottish population.


Chronic pain, depression and cardiovascular disease linked through a shared genetic predisposition: analysis of a family-based cohort and twin study. van Hecke, O., Hocking, L. J., Torrance, N., Campbell, A., Padmanabhan, S., Porteous, D. J., … Smith, B. H. (2017).  PLoS ONE, 22 Feb 2017.

Chronic pain, depression and risk of heart disease tend to go together, but is not clear whether this is cause or effect and whether they have a common genetic basis. This study combined GS data with data from a large twin study, Twins UK, to test this. An increased co-occurrence of chronic pain, depression and cardiovascular disease in both cohorts suggesting a shared genetic contribution over and above any environmental influences. These findings suggest that what might appear to be quite separate conditions (depression, pain and heart disease) may share biological mechanisms and that each should be taken into account when studying the other.


Genome-wide association analyses for lung function and chronic obstructive pulmonary disease identify new loci and potential druggable targets.  Wain, L. V., Shrine, N., Soler Artigas, M., Mesut Erzurumluoglu, A., Noyvert, B., Bossini-Castillo, L.. … Tobin, M. D. (2017). Nature Genetics.

This study builds on earlier pioneering work from the same group looking at the genetic determinants of lung function and their relations to chronic obstructive lung disease. The study took advantage of newly available data from UK Biobank to identify putative risk genes. They then looked for confirmatory evidence in previously studied cohorts including GS. New genes were implicated. Their known function pointed to biological processes, such as the elastic fibres lining the lung, as possible new drug targets.  


Genome-wide association analysis identifies novel blood pressure loci and offers biological insights into cardiovascular risk. Warren, H. R., Evangelou, E., Cabrera, C. P., Gao, H., Ren, M., Mifsud, B., … Elliott, P. (2017). Nature Genetics.

This study took advantage of newly available data from UK Biobank to test for genetic determinants of blood pressure. GS and a large set of other cohorts were used to test for replication of the findings, 107 in all. Further computer and lab based analyses highlighted biological mechanisms that may offer ways to intervene early to reduce the risk of high blood pressure and hypertension.  


Evidence for Large-scale Gene-by-smoking Interaction Effects on Pulmonary Function. Aschard, H., Tobin, M. D., Hancock, D. B., Skurnik, D., Sood, A., James, A., ... Kraft, P. (2017). International Journal of Epidemiology.

GS has supported early gene finding studies with this group in relation to lung function. The environmental effect of smoking and lung function is strong and very well established. This study asked whether those at highest genetic risk are more severely affected by smoking. The answer seems to be yes, but larger studies are needed to firm up these findings. It begs the question as to a possible increased risk of passive smoking on more genetically vulnerable individuals.