What this is
- This research examines the potential causal relationship between epigenetic age acceleration and various cancers using .
- It focuses on four : HannumAge, Horvath Intrinsic Age, PhenoAge, and GrimAge.
- Findings suggest that GrimAge acceleration may increase the risk of colorectal cancer, while associations with other cancers are less consistent.
Essence
- GrimAge acceleration is associated with an increased risk of colorectal cancer, with an odds ratio of 1.12 per year increase. Other showed inconsistent associations with cancer risk.
Key takeaways
- GrimAge acceleration increases colorectal cancer risk, with an odds ratio of 1.12 per year increase. This finding is supported by multiple analyses and is consistent across different datasets.
- The effect of GrimAge acceleration on colon cancer is stronger (odds ratio of 1.15) compared to rectal cancer (odds ratio of 1.05), indicating a potential specificity in cancer type.
- Other , including HannumAge and PhenoAge, did not show consistent evidence of causality with cancer, suggesting that not all epigenetic measures are equally predictive of cancer risk.
Caveats
- The study is limited by the small number of genetic instruments for GrimAge, which may affect statistical power and increase the risk of bias due to horizontal pleiotropy.
- Findings for other cancers were inconsistent and did not pass multiple testing corrections, indicating that further research is needed to confirm these associations.
- The reliance on summary-level data limits the ability to account for potential confounders and effect modifiers, such as sex and lifestyle factors.
Definitions
- epigenetic clocks: Heritable indicators of biological aging derived from DNA methylation patterns at specific sites.
- Mendelian randomization: A method using genetic variants as instrumental variables to infer causality between an exposure and an outcome.
AI simplified
Introduction
DNA methylation (DNAm) at specific cytosine-phosphate-guanine (CpG) sites has been found to be strongly correlated with chronological age. Biological age, as predicted by DNAm patterns at specific CpG sites, may differ from chronological age on an individual basis. Observational evidence suggests that epigenetic age acceleration (i.e. when an individual’s biological age is greater than their chronological age) may be associated with an increased risk of mortality and age-related diseases, including cancer (Fransquet et al., 2019).
Epigenetic clocks are heritable indicators of biological ageing derived from DNAm data. Each clock is based on DNAm levels measured at a different set of CpG sites, which capture distinctive features of epigenetic ageing (Liu et al., 2020). ‘First-generation’ epigenetic clocks, such as HannumAge (Hannum et al., 2013) and Intrinsic HorvathAge (Horvath, 2013), have been derived from DNAm levels at CpG sites found to be strongly associated with chronological age. HannumAge is trained on 71 age-related CpGs found in blood (Hannum et al., 2013), while Intrinsic HorvathAge is based on 353 age-related CpGs found in several human tissues and cell types, and is further adjusted for blood cell counts (Horvath, 2013). More recently, ‘second-generation’ epigenetic clocks, such as, PhenoAge (Levine et al., 2018) and GrimAge (Lu et al., 2019a), have been developed to predict age-related morbidity and mortality. PhenoAge incorporates data from 513 CpGs associated with mortality and nine clinical biomarkers (i.e. albumin, creatinine, serum glucose, C-reactive protein, lymphocyte percentage, mean corpuscular volume, red cell distribution width, alkaline phosphatase and leukocyte count) (Levine et al., 2018), and GrimAge includes data from 1,030 CpGs associated with smoking pack-years and seven plasma proteins (i.e. cystatin C, leptin, tissue inhibitor metalloproteinases 1, adrenomedullin, beta-2-microglobulin, growth differentiation factor 15, and plasminogen activation inhibitor 1 (PAI-1)) (Lu et al., 2019a). Due to differences in their composition, HannumAge and Intrinsic HorvathAge are better predictors of chronological age (Hannum et al., 2013; Horvath, 2013), while PhenoAge and GrimAge stand out for their ability to predict health and lifespan (Levine et al., 2018; Lu et al., 2019a; McCrory et al., 2021).
Several studies suggest that HannumAge, Intrinsic HorvathAge, PhenoAge and GrimAge acceleration are associated with cancer risk (Levine et al., 2018; Ambatipudi et al., 2017; Levine et al., 2015; Dugue et al., 2021; Kresovich et al., 2019b; Kresovich et al., 2019a; Zheng et al., 2016). In contrast, others indicate that evidence in support of this claim is weak or non existent (Dugué et al., 2018; Hillary et al., 2020; Durso et al., 2017; Wang et al., 2021). This lack of consensus could be explained by biases that often affect observational research, such as reverse causation (e.g. cancer influencing the epigenome and not the other way around) and residual confounding (e.g. unmeasured, or imprecisely measured confounders of the association between epigenetic age acceleration and cancer) (Relton and Davey Smith, 2012).
The strength of the associations between epigenetic age acceleration and different cancers has also been found to vary across epigenetic clocks. For instance, positive associations between epigenetic age acceleration and colorectal cancer seem to be much stronger when biological age is estimated using second-generation clocks (i.e. PhenoAge and GrimAge) (Dugue et al., 2021) rather than first-generation clocks (i.e. HannumAge and Intrinsic HorvathAge) (Dugué et al., 2018; Durso et al., 2017). Lack of consensus across epigenetic clocks could be explained by differences in their algorithms (which may reflect different mechanisms of biological ageing), as well as heterogeneity in study designs (Fransquet et al., 2019). Furthermore, even if there were a consensus, it would still be unclear whether age-related DNA methylation plays a causal role in cancer risk or if it merely acts as a non-causal prognostic biomarker.
Mendelian randomization (MR), a method that uses genetic variants as instrumental variables to infer causality between a modifiable exposure and an outcome, is less likely to be affected by residual confounding and reverse causation than traditional observational methods (Davey Smith and Ebrahim, 2003). A recent genome-wide association study (GWAS) meta-analysis has revealed 137 genetic loci associated with epigenetic age acceleration (as measured by six epigenetic biomarkers) that may be used within an MR framework (McCartney et al., 2021).
McCartney et al., 2021 used IVW MR, MR-Egger, weighted median and weighted mode methods to explore the genetically predicted effects of HannumAge, Intrinsic HorvathAge, PhenoAge and GrimAge acceleration on breast, ovarian, and lung cancer. Here, we extend this analysis to include colorectal and prostate cancer (two of the most common cancers worldwide Sung et al., 2021) and use additional methods and datasets to verify the robustness of our findings.
The aim of this two-sample MR study was to examine the genetically predicted effects of epigenetic age acceleration (as measured by HannumAge Hannum et al., 2013, Horvath Intrinsic Age Horvath, 2013, PhenoAge Levine et al., 2018 and GrimAge Lu et al., 2019a) on multiple cancers (i.e., breast, prostate, colorectal, ovarian and lung cancer) using summary genetic association data from (1) McCartney et al. (N = 34,710) (McCartney et al., 2021), (2) the UK Biobank (N cases = 2671–13,879; N controls = 173,493–372,016), (3) FinnGen (N cases = 719–8401; N controls = 74,685–174,006) and (4) several international cancer genetic consortia (N cases = 11,348–122,977; N controls = 15,861–105,974).
Materials and methods
Reporting guidelines
This study has been reported according to the STROBE-MR guidelines (Skrivankova et al., 2021; Supplementary file 2).
Genetic instruments for epigenetic age acceleration
We obtained summary genetic association estimates for epigenetic age acceleration measures of HannumAge (Hannum et al., 2013), Intrinsic HorvathAge (Horvath, 2013), PhenoAge (Levine et al., 2018), and GrimAge (Lu et al., 2019a) from a recent GWAS meta-analysis of biological ageing (McCartney et al., 2021), which included 34,710 participants of European ancestry. Across the 28 European ancestry studies considered in the analysis, 57.3% of participants were female. A detailed description of the methods that were used can be found in the publication by McCartney et al., 2021. In short, the Horvath epigenetic age calculator software (https://dnamage.genetics.ucla.edu↗) or standalone scripts were used to calculate age adjusted DNAm estimates. Outlier samples with clock methylation estimates of +/−5 s.d. from the mean were excluded from further analysis. SNPs were genotyped and imputed independently for each cohort included in the meta-analysis. Genotypes were imputed using either the HRC or the 1000 Genomes Project Phase 3 reference panels in all cohorts but the Sister Study (which did not have imputed data at the time of analysis) and the Genetics of Lipid Lowering Drugs and Diet Network Study (which used whole-genome sequencing data). GWAS summary statistics were obtained in each cohort using additive linear models adjusted for sex and genetic principal components, and they were later processed and harmonised using the ‘EasyQC’ R package. Fixed effect meta-analyses were performed using the METAL software (Willer et al., 2010).
We used the clump_data function in the ‘TwoSampleMR’ R package to select GWAS-significant SNPs (P < 5 × 10−8) for each epigenetic age acceleration measure and perform linkage disequilibrium (LD) clumping (r2 <0.001) using the European reference panel from the 1000 Genomes Project Phase 3 v5.
We identified 9 independent SNPs for HannumAge, 24 for Intrinsic HorvathAge, 11 for PhenoAge and 4 for GrimAge (Supplementary file 1 — Table 1). The proportions of trait variance explained by genetic instruments (R2) and instrument strength (F-statistic) were calculated using the following formulae: R2 = (2β2×MAF×(1-MAF))/(2β2×MAF×(1-MAF) + 2 N × MAF × (1-MAF)×SE2) and F = (R2×(N-2))/(1-R2) (where MAF = effect allele frequency, β = effect estimate of the SNP in the exposure GWAS, SE = standard error, N = sample size) (Palmer et al., 2012). The genetic instruments for HannumAge, Intrinsic HorvathAge, PhenoAge and GrimAge acceleration explained 1.48%, 4.41%, 1.86%, and 0.47% of the trait variance, respectively. All the selected SNPs had F-statistics greater than 10 (HannumAge median 38 and range 31–99, Intrinsic HorvathAge median 47 and range 31–240, PhenoAge median 45 and range 32–89, GrimAge median 36 and range 31–45).
| Cancer type | Source | N cases (%) 1 | N controls |
|---|---|---|---|
| Breast | BCAC | 122,977 (53.7%) | 105,974 |
| UK Biobank | 13,879 (6.5%) | 198,523 | |
| FinnGen | 8401 (7.8%) | 99,321 | |
| Ovarian | OCAC | 25,509 (38.4%) | 40,941 |
| UK Biobank | 1218 (0.6%) | 198,523 | |
| FinnGen | 719 (0.7%) | 99,321 | |
| Prostate | PRACTICAL | 79,148 (56.4%) | 61,106 |
| UK Biobank | 9132 (5.0%) | 173,493 | |
| FinnGen | 6311 (7.8%) | 74,685 | |
| Lung | ILCCO | 11,348 (41.7%) | 15,861 |
| UK Biobank | 2671 (0.7%) | 372,016 | |
| FinnGen | 1681 (1.0%) | 173,933 | |
| Colorectal | GECCO | 58,131 (46.3%) | 67,347 |
| UK Biobank | 5657 (1.5%) | 372,016 | |
| FinnGen | 3022 (1.7%) | 174,006 |
Genetic Association Data sources for cancer outcomes
We obtained summary-level genetic association data for cancer outcomes from the UK Biobank, FinnGen and several international cancer genetic consortia: the Breast Cancer Association Consortium (BCAC), the Ovarian Cancer Association Consortium (OCAC), the Consortium of Investigators of Modifiers of BRCA1/2 (CIMBA), the Prostate Cancer Association Group to Investigate Cancer Associated Alterations in the Genome (PRACTICAL), the International Lung Cancer Consortium (ILCCO) and the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO) (Table 1). Further details of the studies and the data obtained are described in Appendix 1.
We extracted genetic association data for the selected SNPs from each cancer GWAS (for breast, prostate, colorectal, ovarian and lung cancers). LD proxies (r2 >0.8) were used when the SNPs of interest were missing from the cancer GWAS dataset. The proxies were located using the MR-Base platform, which calculates LD using the European subset of individuals from the 1000 Genomes Project reference panel as above (Hemani et al., 2018). The ‘LDlinkR’ R package version 1.1.2 was used to find proxies for cancer data that were not included in the MR-Base platform. The exposure and outcome datasets were then harmonised to ensure the genetic associations reflect the same effect allele. Palindromic SNPs with minor allele frequencies (MAF) <0.3 were aligned, while those with MAF ≥0.3 or mismatching strands were excluded.
Power calculations
Statistical power was calculated using an online calculator for MR available at: https://shiny.cnsgenomics.com/mRnd/↗. Calculations were performed separately for each clock-cancer combination. They were based on a type one error rate of 0.05, the proportion of phenotypic variance explained by genetic variants (R2) for each measure of epigenetic age acceleration, and the total number of cases and controls included in the meta-analysis for each cancer. Across combinations of the four epigenetic clock acceleration and five cancer measures, we had 80% power to detect ORs as small as 1.04–1.39 (Supplementary file 1 — Table s2).
Statistical analysis
We estimated the genetically predicted effects of epigenetic age acceleration (as measured by HannumAge Hannum et al., 2013, Horvath Intrinsic Age Horvath, 2013, PhenoAge Levine et al., 2018 and GrimAge Lu et al., 2019a) on multiple cancers (i.e. breast, prostate, colorectal, ovarian, and lung cancer) using a two-sample MR framework (Figure 1).
Flowchart summarising study methods. Abbreviations: BCAC, Breast Cancer Association Consortium; OCAC, Ovarian Cancer Association Consortium; PRACTICAL, Prostate Cancer Association Group to Investigate Cancer Associated Alterations in the Genome; ILCCO, International Lung Cancer Consortium; GECCO, Genetics and Epidemiology of Colorectal Cancer Consortium; LD, linkage disequilibrium; IVW, inverse variance weighted; MR, Mendelian randomization; FDR, false discovery rate; GWAS, genome-wide association study; CAUSE, Causal Analysis Using Summary Effect estimates, SNP, single-nucleotide polymorphism.
Main analyses
Main analyses were performed using multiplicative random effects inverse variance weighted (IVW) MR, a method that combines the genetically predicted effect of epigenetic age acceleration on cancer across genetic variants (Burgess et al., 2013). This is the default IVW MR method in the ‘TwoSampleMR’ R package, as it accounts for excess heterogeneity across SNP-specific estimates (as opposed to the fixed effect IVW method) and it does not affect the relative weights of individual SNP estimates (in contrast to the additive random effects IVW method) (Bowden et al., 2017).
We used fixed effect meta-analysis to pool results across studies (i.e. UK Biobank, FinnGen and international consortia). For colorectal cancer, we only pooled FinnGen and GECCO estimates, since UK Biobank participants were already included in GECCO. I2 statistics and their corresponding confidence intervals were used to estimate heterogeneity across study estimates (von Hippel, 2015). A Benjamini-Hochberg false discovery rate (FDR) < 5% was used to correct the pooled main IVW results for multiple testing (Benjamini and Hochberg, 1995). This correction was applied considering a total of 20 independent statistical tests (4 clocks x 5 cancers = 20).
Sensitivity analyses
MR assumes genetic instruments for epigenetic age acceleration are (1) associated with epigenetic age acceleration (relevance assumption), (2) independent of confounders of the association between the instruments and cancer (independence assumption), and (3) only associated with cancer through their effect on epigenetic age acceleration (exclusion restriction assumption) (Didelez and Sheehan, 2007; Davies et al., 2018).
As a sensitivity analysis and to test for potential violations of the relevance assumption, we calculated F-statistics and the R2 for each measure of epigenetic age acceleration (Burgess and Thompson, 2011). Other sensitivity analyses included MR-Egger (Bowden et al., 2015), weighted median (Bowden et al., 2016) and weighted mode (Hartwig et al., 2017) methods, which are robust to some of the assumptions of the IVW approach (described in Appendix 1). These results were also pooled across studies, as explained above. Consistency across different MR methods would suggest that it is less likely that the independence and exclusion restriction assumptions are violated.
We further assessed the validity of the independence assumption by conducting MR analyses using negative control outcomes (i.e. skin colour, ease of skin tanning). Evidence of causality between our genetic instruments for epigenetic age acceleration and these negative control outcomes would suggest potential bias due to population stratification that has not been fully accounted for through adjustments in the GWAS (Sanderson et al., 2021). We also assessed the genetically predicted effect of epigenetic age acceleration on cancer risk factors (i.e. body mass index, waist circumference, pack years of smoking, time spent doing vigorous physical activity, age completed full time education, years of schooling, and alcohol intake frequency) to detect potential violations of the exclusion restriction assumption. GWAS data for negative control outcomes and cancer risk factors were obtained using the University of Bristol’s IEU OpenGWAS API (for more details, see Appendix 1).
Where associations between genetically predicted epigenetic age acceleration and cancer were identified, we additionally performed single-SNP two-sample MR analysis to assess whether the effects were likely to be driven by a single SNP. We used the METAL software (Willer et al., 2010) to conduct a GWAS meta-analysis of cancer genetic association data obtained from the UK Biobank, FinnGen and international cancer genetic consortia. We then used these meta-analysed summary statistics in two-sample MR analyses. Scatter plots showing the effects of genetic instruments on epigenetic clock acceleration against their effects on cancer were created using the ‘TwoSampleMR’ R package. Additionally, Cochran’s Q statistics were used to quantify global heterogeneity across SNP-specific MR estimates (Bowden et al., 2019) and MR-Egger intercept tests were performed to detect horizontal pleiotropy (Bowden et al., 2015).
We also used Causal Analysis using Summary Effect Estimates (CAUSE) (Morrison et al., 2020), a method that uses genome-wide summary statistics to disentangle causality (i.e. SNPs are associated with cancer through their effect on epigenetic age acceleration) from correlated horizontal pleiotropy (i.e. SNPs are associated with epigenetic age acceleration and cancer through a shared heritable factor), while taking into account uncorrelated horizontal pleiotropy (i.e. SNPs are associated with epigenetic age acceleration through separate mechanisms). It uses Bayesian modelling to assess whether the sharing model (i.e. model that fixes the causal effect at zero) fits the data at least as well as the causal model (i.e. model that allows a causal effect different from zero).
Secondary analyses
As a secondary analysis, we conducted two-sample MR of epigenetic age acceleration and cancer subtypes (i.e. breast cancer: ER+, ER-, triple negative, luminal B/HER2-negative-like, HER2-enriched-like, luminal A-like, luminal B-like, BRCA1 and BRCA2; ovarian cancer: high-grade serous, low-grade serous, invasive mucinous, clear cell, endometrioid, BRCA1 and BRCA2; prostate cancer: advanced, advanced [vs non-advanced], early onset, high risk [vs low risk], and high risk [vs low and intermediate risk]; lung cancer: adenocarcinoma and squamous cell; colorectal cancer: colon-specific, proximal colon-specific, distal colon-specific, rectal-specific, male and female) (Appendix 1).
We also performed two-sample MR analyses of epigenetic age acceleration and parental history of cancer in the UK Biobank for breast, prostate, lung and bowel cancer (Appendix 1). Data on parental history of ovarian cancer were not available in UK Biobank. Family history data correlate with combined hospital record and questionnaire data and it has been suggested that they provide better power to detect GWAS-significant associations for some phenotypes in the UK Biobank (DeBoever et al., 2020). Therefore, we expected these results to be consistent with those obtained in the main analyses.
MR results were reported as the odds ratio (OR) of site-specific cancer per one year increase in genetically predicted epigenetic age acceleration. These did not require any scale transformations, as the GWAS of biological ageing (McCartney et al., 2021) reported epigenetic age acceleration in years.
LD Score regression (Bulik-Sullivan et al., 2015) was used to identify genome-wide genetic correlations between epigenetic age acceleration and cancer. Genetic correlations were estimated using full GWAS summary statistics for the epigenetic clocks and cancer, as well as the 1000 Genomes Project European LD reference panel. Traits with mean heritability chi-square values < 1.02 were excluded from the analyses.
Finally, bidirectional MR analyses were conducted to assess the causality and directionality of the link between epigenetic clock acceleration and telomere length, another measure of biological ageing that has been shown to influence cancer risk in prior MR studies (Telomeres Mendelian Randomization Collaboration et al., 2017; Gao et al., 2020; Kuo et al., 2019). The MR Steiger test of directionality was used to confirm the assumption that the exposure causes the outcome is valid (Hemani et al., 2017). We also corroborated our findings by rerunning the analyses using data that had undergone Steiger filtering to remove SNPs that explained more variance in the outcome than in the risk factor. Genetic association data for measured telomere length were obtained from Codd et al., 2021, the largest GWAS of telomere length available through the OpenGWAS API at the time of analysis (N = 472,174, for more details, see Appendix 1).
All MR analyses were performed using R software version 4.0.2. Two sample MR analyses were conducted using the ‘TwoSampleMR’ package version 0.5.5. Meta-analyses of IVW results were performed using the ‘meta’ package version 4.18. GWAS meta-analyses used to perform single-SNP MR analyses were done using the METAL software (Willer et al., 2010). CAUSE analyses were conducted using the ‘cause’ package version 1.2.0. Forest plots were created using the ‘ggforestplot’ package version 0.1.0. LD Scores were computed using the ‘ldsc’ command line tool version 1.0.1. The code used in this study is available at: https://github.com/fernandam93/epiclocks_cancer↗.
Results
Breast cancer
We did not find strong evidence of causality between epigenetic age acceleration and breast cancer (GrimAge IVW OR = 0.98, 95% CI 0.95–1.00, p = 0.08; PhenoAge IVW OR = 0.99, 95% CI 0.98–1.01, p = 0.23; HannumAge IVW OR = 0.99, 95% CI 0.97–1.02, p = 0.63; and Intrinsic HorvathAge IVW OR = 0.99, 95% CI 0.98–1.00, p = 0.13) (Figure 2, Appendix 2—figure 1, Appendix 2—figure 2, Appendix 2—figure 3, Appendix 2—figure 4, Appendix 2—figure 5, Appendix 2—figure 6, Appendix 2—figure 7, Appendix 2—figure 8, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6).
Fixed effect meta-analysis of inverse-variance weighted Mendelian randomization estimates for genetically predicted effects of epigenetic age acceleration on multiple cancers. Odds ratios and 95% confidence intervals are reported per 1 year increase in (A) GrimAge acceleration, (B) PhenoAge acceleration, (C) HannumAge acceleration and (D) Intrinsic HorvathAge acceleration. GrimAge, PhenoAge, HannumAge and Intrinsic HorvathAge acceleration were instrumented by 4, 11, 9, and 24 genetic variants, respectively. All meta-analysis estimates were calculated using data from UK Biobank, FinnGen and international consortia, except for colorectal cancer estimates, which exclude UK Biobank data to avoid double counting.
Ovarian cancer
There was also limited evidence of causality between epigenetic age acceleration and ovarian cancer (GrimAge IVW OR = 0.99, 95% CI 0.93–1.06, p = 0.78; PhenoAge IVW OR = 0.98, 95% CI 0.96–1.01, p = 0.24; HannumAge IVW OR = 1.00, 95% CI 0.96–1.04, p = 0.95; and Intrinsic HorvathAge IVW OR = 1.00, 95% CI 0.97–1.02, p = 0.89) (Figure 2, Appendix 2—figure 1, Appendix 2—figure 2, Appendix 2—figure 3, Appendix 2—figure 4, Appendix 2—figure 5, Appendix 2—figure 6, Appendix 2—figure 7, Appendix 2—figure 8, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6).
Prostate cancer
Meta-analysed IVW MR findings suggested that genetically predicted GrimAge acceleration decreased the risk of prostate cancer (OR = 0.93 per year increase in GrimAge acceleration, 95% CI 0.87–0.99, p = 0.02) (Figure 2, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6). Although the direction of the genetically predicted effect was consistent across main and sensitivity MR analyses (i.e. MR-Egger, weighted median and weighted mode) (Appendix 2—figure 1, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6), the main IVW result for GrimAge and prostate cancer did not withstand multiple testing correction (FDR p = 0.16) (Supplementary file 1 — Table s6).
We did not find consistent evidence of causality between other measures of epigenetic age acceleration and prostate cancer (PhenoAge IVW OR = 1.01, 95% CI 0.99–1.03, p = 0.31; HannumAge IVW OR = 0.98, 95% CI 0.95–1.02, p = 0.39; and Intrinsic HorvathAge IVW OR = 0.99, 95% CI 0.98–1.01, p = 0.42) (Figure 2, Appendix 2—figure 1, Appendix 2—figure 2, Appendix 2—figure 3, Appendix 2—figure 4, Appendix 2—figure 5, Appendix 2—figure 6, Appendix 2—figure 7, Appendix 2—figure 8, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6).
Lung cancer
Meta-analysed IVW MR findings suggested that genetically predicted Intrinsic HorvathAge acceleration decreased the risk of lung cancer (OR = 0.97 per year increase in Intrinsic HorvathAge acceleration, 95% CI 0.93–1.00, p = 0.03) (Figure 2, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6). However, these results did not survive multiple testing correction (FDR p = 0.21) and were not strongly supported by sensitivity analyses (Appendix 2—figure 1, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6).
We did not find evidence of causality between other measures of epigenetic age acceleration and lung cancer (GrimAge IVW OR = 1.00, 95% CI 0.91–1.09, p = 0.96; PhenoAge IVW OR = 0.97, 95% CI 0.94–1.00, p = 0.06; and HannumAge IVW OR = 0.99, 95% CI 0.95–1.04, p = 0.82) (Figure 2, Appendix 2—figure 1, Appendix 2—figure 2, Appendix 2—figure 3, Appendix 2—figure 4, Appendix 2—figure 5, Appendix 2—figure 6, Appendix 2—figure 7, Appendix 2—figure 8, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6).
Colorectal cancer
Meta-analysed IVW MR findings suggested that genetically predicted GrimAge acceleration increased the risk of colorectal cancer (OR = 1.12 per year increase in GrimAge acceleration, 95% CI 1.04–1.20, p = 0.002) (Figure 2, Supplementary file 1 — Table s3, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6). These results survived multiple testing correction (FDR p = 0.04) and there was little evidence of heterogeneity across FinnGen and GECCO estimates (I2 = 0%, 95% CI ‘NA’, p = 0.61). Additionally, the direction of the genetically predicted effect was consistent across main and sensitivity MR analyses (i.e. MR-Egger, weighted median, and weighted mode) (Figure 3, Supplementary file 1 — Table s3, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6) and results were consistent when using UK Biobank data alone (IVW OR = 1.15, 95% CI 1.04–1.28, p = 0.007) (Figure 2, Supplementary file 1 — Table s4).
We did not find evidence of residual population stratification in MR analyses using negative control outcomes (Appendix 2—figure 9, Supplementary file 1 — Table s7), nor did we find evidence of horizontal pleiotropy via potential colorectal cancer risk factors (Appendix 2—figure 10, Supplementary file 1 — Table s8).
Single-SNP analysis revealed that the effect was not driven by a single SNP (Supplementary file 1 — Table s9). Figure 4 shows the effect of genetic instruments on GrimAge acceleration against their effect on colorectal cancer. Moreover, there was no detectable evidence of uncorrelated horizontal pleiotropy (MR-Egger intercept = –0.13, 95% CI –0.33–0.07, p = 0.33), or heterogeneity across individual SNP estimates (Cochran’s Q 7.12, p = 0.07) (Supplementary file 1 — Table s10). We further explored the genetically predicted effect of GrimAge on colorectal cancer using GECCO data only and found no evidence against bias due to correlated pleiotropy (CAUSE OR = 1.00, 95% credible intervals 0.96–1.04, p = 1.00; shared q = 4%, 95% credible intervals 0–24%) (Appendix 2—figure 11).
Among subtypes, we found strong evidence for a causal relationship between GrimAge acceleration and colon cancer (IVW OR = 1.15, 95% CI 1.09–1.21, p = 0.006). In contrast, we did not find such evidence for rectal cancer (IVW OR = 1.05, 95% CI 0.97–1.13, p = 0.24). After further stratification, the magnitude of the genetically predicted effect of GrimAge acceleration on colon cancer was the same for distal (IVW OR = 1.16, 95% CI 1.03–1.29, p = 0.01) and proximal colon cancer (IVW OR = 1.16, 95% CI 0.97–1.40, p = 0.11). Also, sex-stratified results suggest that GrimAge acceleration may influence colorectal cancer in both males (IVW OR = 1.12, 95% CI 1.00–1.25, p = 0.05) and females (IVW OR = 1.14, 95% CI 1.04–1.26, p = 0.008) (Figure 5, Supplementary file 1 Table s11).
These findings were further supported by evidence of a positive association between GrimAge acceleration and parental history of colorectal cancer (OR = 1.06, 95% CI 1.00–1.12, p = 0.03) (Figure 6, Supplementary file 1 — Table s12). Additionally, LD Score regression coefficients for GrimAge acceleration and colorectal cancer were also in the expected direction (GECCO rg = 0.28, p < 0.001; UK Biobank rg = 0.15, p = 0.21; FinnGen rg = 0.27, p = 0.29) (Appendix 2—figure 12, Supplementary file 1 — Table s13).
We did not find consistent evidence of causality between other measures of epigenetic age acceleration and colorectal cancer (PhenoAge IVW OR = 1.00, 95% CI 0.97–1.02, p = 0.73; HannumAge IVW OR = 1.02, 95% CI 0.97–1.08, p = 0.37; and Intrinsic HorvathAge IVW OR = 1.00, 95% CI 0.97–1.02, p = 0.79) (Figure 2, Appendix 2—figure 1, Appendix 2—figure 2, Appendix 2—figure 3, Appendix 2—figure 4, Appendix 2—figure 5, Appendix 2—figure 6, Appendix 2—figure 7, Appendix 2—figure 8, Supplementary file 1 — Table s3, Supplementary file 1 — Table s4, Supplementary file 1 — Table s5, Supplementary file 1 — Table s6).
Fixed effect meta-analysis of Mendelian randomization estimates for genetically predicted effects of GrimAge acceleration on multiple cancers. Odds ratios and 95% confidence intervals are reported per 1 year increase in GrimAge acceleration. GrimAge acceleration was instrumented by four genetic variants. Results were obtained using inverse variance weighted MR (dark blue), weighted median (sky blue) and weighted mode (turquoise) methods. All meta-analysis estimates were calculated using data from UK Biobank, FinnGen and international consortia, except for colorectal cancer estimates, which exclude UK Biobank data to avoid double counting.
Scatter plot showing the effect of genetic instruments on GrimAge acceleration against their effect on colorectal cancer. FinnGen and Genetics and Epidemiology of Colorectal Cancer (GECCO) genome-wide association estimates for colorectal cancer were meta-analysed using the METAL software. UK Biobank estimates were not included in the meta-analysis to avoid double counting participants included in the GECCO consortium. Results were obtained using inverse variance weighted MR (light blue), weighted median (dark blue) and weighted mode (light green) methods.
Mendelian randomization estimates for genetically predicted effects of GrimAge acceleration on colorectal cancer subtypes. Odds ratios and 95% confidence intervals are reported per 1 year increase in GrimAge acceleration. GrimAge acceleration was instrumented by four genetic variants. Results were obtained using inverse variance weighted MR (dark blue), weighted median (sky blue) and weighted mode (turquoise) methods. Data source: GECCO.
Mendelian randomization estimates for genetically predicted effects of GrimAge acceleration on parental history of multiple cancers. Odds ratios and 95% confidence intervals are reported per 1 year increase in GrimAge acceleration. GrimAge acceleration was instrumented by four genetic variants. Results were obtained using inverse variance weighted MR (dark blue), weighted median (sky blue) and weighted mode (turquoise) methods. Data source: UK Biobank.
Telomere length
In bidirectional MR analyses, we found evidence that genetically predicted GrimAge acceleration may be a cause of telomere shortening (IVW beta coefficient = −0.07 per year increase in GrimAge acceleration, 95% CI –0.09 to –0.05, p < 0.001) and that genetically predicted longer telomere length may increase Intrinsic HorvathAge acceleration (IVW beta coefficient = 0.57 per standard deviation increase in telomere length, 95% CI 0.39–0.77, p = 0.002) (Appendix 2—figure 13, Supplementary file 1 — Table s14).
Steiger filtering showed that all genetic instruments for GrimAge acceleration were stronger predictors of GrimAge acceleration than telomere length. In contrast, it identified 20 genetic instruments for telomere length that were better predictors of Intrinsic HorvathAge acceleration than telomere length (Supplementary file 1 — Table s15). After removing these SNPs from the analyses, the results were still suggestive of an effect of telomere length on Intrinsic HorvathAge acceleration (IVW beta coefficient = 0.71 per standard deviation increase in telomere length, 95% CI 0.57–0.85, p < 0.001) (Appendix 2—figure 14, Supplementary file 1 — Table s16).
There was little evidence of causality between other measures of epigenetic age acceleration and telomere length (Appendix 2—figure 13, Supplementary file 1 — Table s14).
Discussion
In this comprehensive two-sample MR study of epigenetic age acceleration and multiple cancers, we found evidence to suggest that genetically predicted GrimAge acceleration may increase the risk of colorectal cancer in both males and females. Among subtypes, effects appeared to be stronger in relation to colon than rectal cancer. Our MR results also suggested that genetically predicted GrimAge acceleration may decrease the risk of prostate cancer and that genetically predicted Intrinsic HorvathAge acceleration may be protective against lung cancer. Nevertheless, these did not pass multiple testing correction. Finally, we found no consistent evidence for other measures of epigenetic age acceleration and cancers.
Our MR estimates for the association between GrimAge and colorectal cancer were consistent with those reported in Dugue et al., 2021, an observational nested case-control study in the Melbourne Collaborative Cohort Study (RR = 1.04 per year increase in GrimAge acceleration, 95% CI 1.01–1.07, p = 0.02). However, our findings contrast with those highlighted in Hillary et al., 2020, an observational cohort study that used Generation Scotland data. The latter authors observed no evidence of an association between GrimAge acceleration and colorectal cancer after correcting for multiple testing. Nevertheless, it is possible that their analyses were underpowered, as their sample only included 63 colorectal cancer cases (0.66%). More importantly, the direction of the reported estimate is consistent with our findings and those presented in Dugue et al., 2021.
Observational evidence for the association between other measures of epigenetic ageing and cancer is inconclusive (the pre-existing evidence has been summarised in Supplementary file 1 — Table s17). For instance, epigenetic clock acceleration has been positively associated with breast (Ambatipudi et al., 2017; Kresovich et al., 2019b; Kresovich et al., 2019a) and lung cancer (Levine et al., 2018; Levine et al., 2015; Dugue et al., 2021) in some studies. However, (Durso et al., 2017, Hillary et al., 2020) and (Dugué et al., 2018) did not find strong evidence to support this. In some cases, observational evidence is stronger for some clocks than it is for others. For example, for colorectal cancer, evidence of a positive association is much stronger for second-generation clocks (Dugue et al., 2021) than for first-generation clocks (Dugué et al., 2018; Durso et al., 2017). In the case of prostate cancer, as in our study, apart from weak evidence of an inverse association with GrimAge, no other associations have been observed (Dugue et al., 2021; Dugué et al., 2018). To date, the association between epigenetic age acceleration and ovarian cancer has not been explored observationally. Although our findings were less susceptible to biases that often influence observational research, they still provide no compelling evidence of a causality between several measures of epigenetic clock acceleration and cancer.
This MR study had several strengths. For instance, we pooled results from multiple sources using fixed effect meta-analysis to improve the precision of the MR estimates presented in McCartney et al., 2021. We also conducted extra sensitivity analyses, such as MR of negative control outcomes, MR of cancer risk factors, single-SNP MR and CAUSE analyses, to assess the validity of the MR assumptions. Moreover, we performed subtype-specific MR analyses and sought to corroborate our results using UK Biobank GWAS data on parental history of cancer and LD Score regression. Additionally, our findings contribute to the identification of modifiable targets for future interventions aimed at reversing epigenetic ageing for the prevention of cancer. Compared to clinical trials, MR provides a cheaper, quicker, and ethical way of assessing the long-term impact of interventions on epigenetic ageing. This is especially relevant while attempts to develop interventions which reverse epigenetic ageing are still in early stages (Fahy et al., 2019; Fitzgerald et al., 2021; Gensous et al., 2020; Chen et al., 2019).
The findings from this study should be interpreted in light of its limitations. We only identified four genetic instruments for GrimAge acceleration, which explained 0.47% of the variance in the trait. This could lead to two issues: low statistical power and horizontal pleiotropy. First, our GrimAge analyses were underpowered to detect ORs < 1.20 for colorectal cancer. Therefore, it is possible that our findings do not reflect a true effect (we identified an OR = 1.12 for colorectal). Similarly, our study was underpowered to detect genetically predicted effects of GrimAge acceleration on cancer subtypes and cancers with smaller sample sizes (i.e. ovarian and lung cancer). Some of our sensitivity analyses, such as the MR-Egger intercept test used to detect uncorrelated horizontal pleiotropy, also had low power, resulting in imprecise estimates. The weighted mode method may also be misleading in this context, as its use is limited in the presence of very few SNPs. Although these limitations potentially undermine the validity of our results, it is reassuring that point estimates for the genetically predicted effect of GrimAge acceleration on colorectal cancer were consistent across MR methods and study populations. However, since CAUSE analyses did not provide evidence against confounding by correlated horizontal pleiotropy, it is possible that the genetically predicted effects identified are attributed to correlated pleiotropy (whereby SNPs are associated with epigenetic age acceleration and cancer through a shared heritable factor) rather than a causal effect of GrimAge on cancer risk.
One could argue that because the results for GrimAge acceleration were inconsistent with those obtained for other measures of epigenetic age acceleration, chance and horizontal pleiotropy are more likely explanations for our findings. However, inconsistencies across epigenetic ageing measures do not necessarily invalidate our results. They may simply reflect differences in how clocks were trained (i.e. they were trained on different outcomes, tissues, and populations). Different clocks may capture information on distinct underlying biological ageing mechanisms (Liu et al., 2020). For example, GrimAge was trained on mortality and smoking (factors which are closely related to cancer risk), which may explain why it outperforms other measures of epigenetic ageing in predicting time-to-cancer (Lu et al., 2019a).
Although little is known about the underlying mechanisms, GrimAge may plausibly influence cancer risk through hormonal, inflammatory and metabolic processes (Yu et al., 2020; Bottazzi et al., 2018; Lau and Robinson, 2021). In bidirectional MR analyses, we found evidence that genetically predicted GrimAge acceleration may be a cause of telomere shortening, another marker of biological ageing. Shorter telomeres have been shown to lower cancer risk in prior MR analyses (Telomeres Mendelian Randomization Collaboration et al., 2017; Gao et al., 2020; Kuo et al., 2019), so it is plausible that GrimAge acceleration decreases cancer risk, at least in part, via its effect on telomere length. GrimAge acceleration may still increase cancer risk via pathways other than those related to cellular division. The positive effect of GrimAge acceleration on cancer via these other pathways may counteract the negative effects mediated via telomere length, resulting in null MR results for GrimAge acceleration and breast, ovarian, prostate and lung cancer, and positive MR results for GrimAge acceleration and colorectal cancer. To better understand the biology of ageing, future studies should consider running MR analyses using data on DNAm-predicted telomere length, since DNAm telomere length is independent of telomerase activity and has been more strongly associated with age than measured telomere length (Lu et al., 2019b).
Although promising in terms of consistency and biological plausibility, further research is required to confirm our findings. For example, multivariable MR (Burgess and Thompson, 2015; Sanderson et al., 2019) could be used to disentangle the causal effects of GrimAge acceleration on cancer from shared heritable factors such as and blood cell composition. Additionally, our analyses could be replicated using other large independent cancer datasets. Although we conducted MR analyses on parental history of cancer, their effect estimates are not directly comparable to those obtained in the main analyses due to cases in the GWAS-by-proxy of parental endpoints being defined as either or both parents reportedly having a type of cancer. Furthermore, it would also be useful to replicate our analyses once a larger GWAS of epigenetic ageing with more genetic instruments for GrimAge acceleration is available. This would allow for a more rigorous assessment of horizontal pleiotropy and may be used to assess clustering of genetic variants to reveal distinct biological mechanisms underlying the effects (Foley et al., 2021). In spite of these suggestions, we acknowledge that it may be challenging to get access to suitable datasets for replication purposes in the short term.
The selection of ‘super controls’ (e.g. in UK Biobank, FinnGen and GECCO), with no other cancers, related lesions (i.e. benign, in situ, uncertain or unspecified behaviour neoplasms) or reported family history of cancer, could have inflated cancer GWAS effect sizes (and our MR estimates), because ‘super controls’ are healthier than the general population and are less likely to be genetically predisposed to develop cancer.
Another limitation is that we did not have access to individual level data. Therefore, we were unable to stratify the analyses by potential effect modifiers, such as sex, smoking, and menopausal status. Moreover, we did not have sex-specific instruments for sex-specific cancers. However, it is unlikely that the genetic architecture of epigenetic clock acceleration differs across sexes, as DNAm levels at individual clock CpGs are highly correlated between males and females (Grodstein et al., 2020; Tajuddin et al., 2019).
Finally, to reduce bias due to population stratification, this study was conducted using data from participants of European ancestry only. The GWAS data used for the analyses had been adjusted for the top genetic principal components for the same reason. Furthermore, our MR of negative control outcomes suggests that our MR results are unlikely to be biased by residual population stratification. Despite this, confounding due to population stratification, dynastic effects and assortative mating cannot be ruled out completely, as it is not possible to test the second MR assumption (i.e. independence assumption). Furthermore, more research is required to see if our results could translate to other ancestries.
From a public health perspective, our work provides potentially relevant findings. Observational and Mendelian randomization studies suggest that GrimAge acceleration may be influenced by several cancer risk factors, such as obesity and smoking (Lu et al., 2019a; McCartney et al., 2021). If GrimAge acceleration is a causal mediator between these risk factors and colorectal cancer, the GrimAge clock may be a treatable intermediary when targeting the underlying risk factors is not feasible or too difficult to accomplish. It could also be targeted in populations at high-risk of colorectal cancer. Nevertheless, we think it may be too early to make claims regarding the clinical utility of our findings. The GrimAge clock has only recently been created and very few studies have assessed its association with colorectal cancer. More research is required to corroborate our results and to evaluate whether GrimAge acceleration can be modified through lifestyle or clinical interventions.
In conclusion, our findings suggest that genetically predicted GrimAge acceleration may increase the risk of colorectal cancer. Findings were less consistent for other epigenetic clocks and cancers. Further work is required to investigate the potential mechanisms underlying the genetically predicted effects identified in this study.