Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Dec 1;143(12):3763-3775.
doi: 10.1093/brain/awaa334.

Recalibrating the epigenetic clock: implications for assessing biological age in the human cortex

Gemma L Shireby  1 , Jonathan P Davies  1 , Paul T Francis  1   2 , Joe Burrage  1 , Emma M Walker  1 , Grant W A Neilson  1 , Aisha Dahir  1 , Alan J Thomas  3 , Seth Love  4 , Rebecca G Smith  1 , Katie Lunnon  1 , Meena Kumari  5 , Leonard C Schalkwyk  6 , Kevin Morgan  7 , Keeley Brookes  8 , Eilis Hannon  1 , Jonathan Mill  1
Affiliations

Recalibrating the epigenetic clock: implications for assessing biological age in the human cortex

Gemma L Shireby et al. Brain. .

Abstract

Human DNA methylation data have been used to develop biomarkers of ageing, referred to as 'epigenetic clocks', which have been widely used to identify differences between chronological age and biological age in health and disease including neurodegeneration, dementia and other brain phenotypes. Existing DNA methylation clocks have been shown to be highly accurate in blood but are less precise when used in older samples or in tissue types not included in training the model, including brain. We aimed to develop a novel epigenetic clock that performs optimally in human cortex tissue and has the potential to identify phenotypes associated with biological ageing in the brain. We generated an extensive dataset of human cortex DNA methylation data spanning the life course (n = 1397, ages = 1 to 108 years). This dataset was split into 'training' and 'testing' samples (training: n = 1047; testing: n = 350). DNA methylation age estimators were derived using a transformed version of chronological age on DNA methylation at specific sites using elastic net regression, a supervised machine learning method. The cortical clock was subsequently validated in a novel independent human cortex dataset (n = 1221, ages = 41 to 104 years) and tested for specificity in a large whole blood dataset (n = 1175, ages = 28 to 98 years). We identified a set of 347 DNA methylation sites that, in combination, optimally predict age in the human cortex. The sum of DNA methylation levels at these sites weighted by their regression coefficients provide the cortical DNA methylation clock age estimate. The novel clock dramatically outperformed previously reported clocks in additional cortical datasets. Our findings suggest that previous associations between predicted DNA methylation age and neurodegenerative phenotypes might represent false positives resulting from clocks not robustly calibrated to the tissue being tested and for phenotypes that become manifest in older ages. The age distribution and tissue type of samples included in training datasets need to be considered when building and applying epigenetic clock algorithms to human epidemiological or disease cohorts.

Keywords: DNA methylation; age; brain; clock; cortex.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Comparison of chronological age with DNA methylation age derived using four DNA methylation age clocks. Shown are comparisons of chronological age with predicted age in (A) the testing dataset (n = 350 cortical samples) and (B) the independent test dataset (n = 1221 cortical samples). DNAm age was predicted using four DNA methylation age clocks: (i) our novel DNAmClockCortical; (ii) Horvath’s DNAmClockMulti; (iii) Zhang’s DNAmClockBlood and (iv) Levine’s DNAmClockPheno. The x-axis represents chronological age (years) and the y-axis represents predicted age (years). Each point on the plot represents an individual sample. Our cortical clock out-performed the three alternative DNAm clocks across all accuracy statistics. DNA methylation age estimates derived using the DNAmClockMulti [A(ii) testing and B(ii) independent test] and the DNAmClockBlood [A(iii) testing and B(iii) independent test] appear to have a non-linear relationship with chronological age.
Figure 2
Figure 2
The cortical DNA methylation age clock has elevated accuracy in human cortex samples across the lifespan. Shown is the distribution of the error (DNA methylation age − chronological age) for each age decile in (A) the testing dataset (n = 350 cortical samples), and (B) the independent test dataset (n = 1221 cortical samples) for each of the four DNA methylation age clocks: (i) our novel DNAmClockCortical; (ii) Horvath’s DNAmClockMulti; (iii) Zhang’s DNAmClockBlood and (iv) Levine’s DNAmClockPheno. Deciles were calculated by assigning chronological age into 10 bins and are represented along the x-axis by the numbers 1 to 10, followed by parentheses, which display the age range included in each decile. The ends of the boxes are the upper and lower quartiles of the errors, the horizontal line inside the box represents the median deviation and the two lines outside the boxes extend to the highest and lowest observations. Outliers are represented by points beyond these lines. The red horizontal line represents perfect prediction (zero error). Our novel DNAmClockCortical [A(i) testing and B(i) independent test] consistently had the smallest error across the age groups, shown by the tightness of the box plot distributions along the zero-error line. The DNAmClockMulti over-predicted younger ages [deciles 1–5 in A(ii)], shown by box plot distributions that are above the zero-error line, and under predicted older ages [deciles 8–10 in A(ii) and deciles 1–10 in B(ii)], shown by box plot distributions below the zero-error line. The DNAmClockBlood [A(iii) testing and B(iii) independent test] and the DNAmClockPheno [A(iv) testing and B(iv) independent test] consistently under predicted age, with under prediction of DNA methylation age increasing with chronological age.
Figure 3
Figure 3
The blood-based DNA methylation clock performs best in data derived from whole blood samples.(A) Shown is a comparison of DNA methylation age estimates against chronological age in a large whole blood dataset (n = 1175), where DNAm age derived using four DNA methylation age clocks: (i) our novel DNAmClockCortical; (ii) Horvath’s DNAmClockMulti; (iii) Zhang’s DNAmClockBlood and (iv) Levine’s DNAmClockPheno. The x-axis represents chronological age (years), the y-axis represents predicted age (years). Each point on the plot represents an individual in the whole blood dataset. Our novel clock does not predict as well in blood compared to the cortex, although it has a similar predictive ability to Horvath’s clock. The distribution of the error (DNA methylation age − chronological age) is presented in B for each decile for each of the four DNA methylation clocks. Deciles were calculated by assigning chronological age into 10 bins and are represented along the x-axis by the numbers 1 to 10, followed by parentheses, which display the age range included in each decile. The ends of the boxes are the upper and lower quartiles of the errors, the horizontal line inside the box represents the median deviation and the two lines outside the boxes extend to the highest and lowest observations. Outliers are represented by points beyond these lines. The red horizontal line represents perfect prediction (zero error).
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Baker DJ, Wijshake T, Tchkonia T, LeBrasseur NK, Childs BG, van de Sluis B, et al.Clearance of p16Ink4a-positive senescent cells delays ageing-associated disorders. Nature 2011; 479: 232–6. - PMC - PubMed
    1. Bell JE, Alafuzoff I, Al-Sarraj S, Arzberger T, Bogdanovic N, Budka H, et al.Management of a twenty-first century brain bank: experience in the BrainNet Europe consortium. Acta Neuropathol 2008; 115: 497–507. - PubMed
    1. Bernstein BE, Meissner A, Lander ES.. The mammalian epigenome. Cell 2007; 128: 669–81. - PubMed
    1. Buck N, McFall S. Understanding Society: design overview. Longitudinal and Life Course Studies 2011; 3: 5–17.
    1. Chen Y, Lemire M, Choufani S, Butcher DT, Grafodatskaya D, Zanke BW, et al.Discovery of cross-reactive probes and polymorphic CpGs in the Illumina Infinium HumanMethylation450 microarray. Epigenetics 2013; 8: 203–9. - PMC - PubMed

Publication types

MeSH terms