HorizonBio

Every cell in your body carries the same DNA, yet a liver cell behaves nothing like a neuron. The key to this biological flexibility is epigenetics, and at its heart lies DNA methylation: one of the best-studied molecular switches in human physiology.

What is DNA methylation?

DNA methylation is the addition of a methyl group (-CH₃) to a cytosine base, almost always where it is followed by a guanine in the sequence: sites called CpG dinucleotides. This chemical tag does not alter the underlying genetic code but profoundly affects whether a gene is expressed. Heavily methylated promoter regions are typically silenced; unmethylated regions are accessible to transcription machinery.

Crucially, methylation patterns are dynamic. Diet, exercise, stress, sleep, and environmental exposures all shift them across the lifespan, which means your epigenome is a running record of the choices your body has made.

The link to metabolic health

Research published over the past decade has established clear associations between aberrant DNA methylation and the hallmarks of metabolic disease: insulin resistance, dyslipidaemia, and type 2 diabetes:

ABCG1 hypermethylation is consistently associated with increased risk of type 2 diabetes in prospective cohort studies.¹
Methylation at TXNIP loci correlates with fasting glucose and HbA1c levels and with future type 2 diabetes risk, independently of established clinical risk factors.¹
Differentially methylated regions near the PPARGC1A gene (encoding PGC-1α) are linked to altered mitochondrial biogenesis and energy expenditure in human skeletal muscle and islets.^2,3
Epigenetic clocks such as Horvath, PhenoAge, and GrimAge estimate biological age and predict mortality and healthspan outcomes more accurately than chronological age alone - though "metabolic age" is not yet a single standardised clinical construct.^4,5,6

Your epigenome is not your destiny. It is a dial, and the right interventions can turn it in your favour.

How HorizonBio reads the dial

Our predictive engine analyses methylation signatures across targeted CpG sites using a random forest classifier trained on population-scale datasets. Rather than returning a single risk score, the platform highlights the individual sites driving the classification, giving clinicians and users a transparent, site-level explanation.

Age-bias correction is applied to every sample: because methylation patterns naturally drift with chronological age, raw scores must be normalised before meaningful inter-individual comparisons can be made. The result is a residualised metabolic health score that separates biological from chronological ageing.

What this means for prevention

Unlike genetic variants, methylation marks are reversible. Intervention studies have shown that caloric restriction, aerobic exercise, and Mediterranean-style diets can produce measurable shifts in metabolic methylation patterns - though the magnitude, loci involved, and timeframes vary by study design and tissue.

HorizonBio's goal is to make this science accessible: not just as a diagnostic snapshot, but as a longitudinal tool that empowers individuals and their care teams to measure, intervene, and verify.

Key takeaways

DNA methylation is a reversible epigenetic mark that controls gene expression without changing the DNA sequence.
Metabolic disease is associated with reproducible methylation changes at specific CpG loci, including ABCG1, TXNIP, and PPARGC1A.
Validated epigenetic clocks can stratify biological ageing and mortality risk earlier and more precisely than chronological age alone.
Lifestyle and dietary interventions can shift methylation patterns, making epigenomics a candidate tool for monitoring as well as predicting.