On the power of epigenome-wide association studies using a disease-discordant twin design

Li, Weilong; Christiansen, Lene; Hjelmborg, Jacob; Baumbach, Jan; Tan, Qihua

doi:10.1093/bioinformatics/bty532

Cited by 38 publications

(34 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since our interest is in the slope not the intercept, influence of the complexity in mean level of X-linked methylation is minimized. Moreover, the use of twins in the association analyses helped to control for genetic confounding and to achieve enriched statistical power [16]. As a result, our analysis on monozygotic twins yielded significant age-related methylation changes for CpGs escaping XCI, as well as those under XCI in females.…”

Section: Discussionmentioning

confidence: 91%

“…Then we examined the X-linked DNA methylation levels in the two sexes to characterize the different methylation patterns in relation to XCI in females and analysed the methylation levels as a function of age for different patterns in male and female samples separately. The use of Danish twins helped to control for confounding from unmeasured genetic and rearing environmental factors with enriched statistical power [6,16]. Discovery and replication were carried out in the two Danish twin cohorts and replicated results were then validated in the Scottish LBC1921 birth cohort.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Exploratory analysis of age and sex dependent DNA methylation patterns on the X-chromosome in whole blood samples

Lund

Christensen

et al. 2020

Genome Med

Self Cite

View full text Add to dashboard Cite

Background: Large numbers of autosomal sites are found differentially methylated in the aging genome. Due to analytical difficulties in dealing with sex differences in X-chromosome content and X-inactivation (XCI) in females, this has not been explored for the X chromosome. Methods: Using data from middle age to elderly individuals (age 55+ years) from two Danish cohorts of monozygotic twins and the Scottish Lothian Birth Cohort 1921, we conducted an X-chromosome-wide analysis of age-associated DNA methylation patterns with consideration of stably inferred XCI status. Results: Through analysing and comparing sex-specific X-linked DNA methylation changes over age late in life, we identified 123, 293 and 55 CpG sites significant (FDR < 0.05) only in males, only in females and in both sexes of Danish twins. All findings were significantly replicated in the two Danish twin cohorts. CpG sites escaping XCI are predominantly de-methylated with increasing age across cohorts. In contrast, CpGs highly methylated in both sexes are methylated even further with increasing age. Among the replicated sites in Danish samples, 16 (13%), 24 (8.2%) and 3 (5.5%) CpGs were further validated in LBC1921 (FDR < 0.05). Conclusions: The X-chromosome of whole blood leukocytes displays age-and sex-dependent DNA methylation patterns in relation to XCI across cohorts.

show abstract

Section: Discussionmentioning

confidence: 91%

Section: Introductionmentioning

confidence: 99%

Exploratory analysis of age and sex dependent DNA methylation patterns on the X-chromosome in whole blood samples

Lund

Christensen

et al. 2020

Genome Med

Self Cite

View full text Add to dashboard Cite

show abstract

“…We think, however, explanation of the phenomenon concerns whether the observed age-related hypermethylation change is the cause or response to aging. In this regard, the longitudinal design that links methylation change at an individual level with risk of death should provide direct estimates of the association and help with causation inference (Li et al 2018). Together with the result from Lund et al (2019), we postulate that the observed age-associated hypermethylation on the sex chromosomes in males could mainly represent an active response to the aging process that helps to maintain male survival.…”

Section: Discussionmentioning

confidence: 99%

Age and mortality associated DNA methylation patterns on the X-chromosome in male and female samples

Lund

Baumbach

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

BackgroundMultiple epigenetic association studies on human aging have been performed reporting large numbers of sites differentially methylated across ages on the autosomal chromosomes. The Xchromosome has been studied little, due to analytical difficulties in dealing with sex differences in X-chromosome content and X-inactivation in females. Based on large collections of genome-wide DNA methylation data on two Danish cohorts of identical twins (mean ages, 66 and 79 years) and the Lothian Birth Cohort 1921 (mean age 79 years), we conducted a chromosome-wide association analysis on male and female samples separately with equal sample sizes to discover age-dependent X-linked DNA methylation patterns using chromosome 20 with about same number of CpGs analysed as an autosomal reference, and compare the age-related changes in DNA methylation between the two sexes. In addition, age-related methylation sites were assessed for their associations with mortality. ResultsWe identified more age-related DNA methylation sites (FDR<0.05) in females than in males.Among them, predominantly more sites were hypermethylated in the older as compared with the younger cohorts, a pattern similar to that observed on chromosome 20. Among the age-related sites, 13 CpGs in males and 24 CpGs in females were found significant (FDR<0.05) in all cohorts.Survival analysis showed that there are more age-methylated CpGs that contribute to reduce mortality than those that increase mortality in male but not in female samples. ConclusionThe X-chromosome displays significant age-and sex-dependent methylation patterns which might be differentially associated with mortality in the two sexes.

show abstract

“…Twin and family studies robustly demonstrate that genetic factors play a role in risk for MD, with heritability estimates of roughly 35% for MD and 45% for early-onset MD 9 ; these heritability estimates indicate that MD is suitable for epigenetic study using twins 10 . Moreover, a large number of genetic loci have been identified for MD, supporting the role of genetic factors in the etiology of MD 11 , 12 .…”

Section: Introductionmentioning

confidence: 87%

An epigenome-wide association study of early-onset major depression in monozygotic twins

et al. 2020

View full text Add to dashboard Cite

Major depression (MD) is a debilitating mental health condition with peak prevalence occurring early in life. Genomewide examination of DNA methylation (DNAm) offers an attractive complement to studies of allelic risk given it can reflect the combined influence of genes and environment. The current study used monozygotic twins to identify differentially and variably methylated regions of the genome that distinguish twins with and without a lifetime history of early-onset MD. The sample included 150 Caucasian monozygotic twins between the ages of 15 and 20 (73% female; Mage = 17.52 SD = 1.28) who were assessed during a developmental stage characterized by relatively distinct neurophysiological changes. All twins were generally healthy and currently free of medications with psychotropic effects. DNAm was measured in peripheral blood cells using the Infinium Human BeadChip 450 K Array. MD associations with early-onset MD were detected at 760 differentially and variably methylated probes/regions that mapped to 428 genes. Genes and genomic regions involved neural circuitry formation, projection, functioning, and plasticity. Gene enrichment analyses implicated genes related to neuron structures and neurodevelopmental processes including cell-cell adhesion genes (e.g., PCDHA genes). Genes previously implicated in mood and psychiatric disorders as well as chronic stress (e.g., NRG3) also were identified. DNAm regions associated with early-onset MD were found to overlap genetic loci identified in the latest Psychiatric Genomics Consortium meta-analysis of depression. Understanding the time course of epigenetic influences during emerging adulthood may clarify developmental phases where changes in the DNA methylome may modulate individual differences in MD risk.

show abstract

On the power of epigenome-wide association studies using a disease-discordant twin design

Abstract: Supplementary data are available at Bioinformatics online.

Cited by 38 publications

References 17 publications

Exploratory analysis of age and sex dependent DNA methylation patterns on the X-chromosome in whole blood samples

Exploratory analysis of age and sex dependent DNA methylation patterns on the X-chromosome in whole blood samples

Age and mortality associated DNA methylation patterns on the X-chromosome in male and female samples

An epigenome-wide association study of early-onset major depression in monozygotic twins

Contact Info

Product

Resources

About