In this large cohort of UK Biobank participants, we first identified KDM-BAacceland PhenoAgeAccel were significantly associated with prevalent severe MASLD, as well as liver cirrhosis and cancer. Particularly, we observed that the adverse effects of advanced biological aging on these three CLDs were mostly stronger in males than in females. KDM-BAaccel and PhenoAgeAccel showed better performance than CA in predicting MASLD. Moreover, there were joint effects but no interactions of BA acceleration and genetic risk in severe MASLD, liver cirrhosis and cancer incidence. In addition, participants who at a high genetic risk level had the greatest 10-year absolute risk reduction of severe MASLD (6.74 per 1000 person-years) if decreasing the PhenoAgeAccel. Our findings indicated that alleviating biological aging is important for preventing serious liver-related diseases and could offset the deleterious effect of inherent genetic risk.
Most previous studies focused on the role of clinical biological aging indicators in the occurrence of several health outcomes. Mak et al. conducted a prospective cohort study to investigate the association between three clinical qualified biological aging indicators and the risk of five common cancers. They found that age-adjusted KDM-BA and PhenoAge were correlated with an increased risk of lung and colorectal cancers, while PhenoAge was additionally linked to increased risk of breast cancer [19]. Another prior study has suggested that PhenoAge was an independent risk factor for lung cancer and might serve as a potential biomarker for prediction of lung cancer [20]. In consistence with our findings, Xia et al. disclosed that participants with accelerated DNA methylation age were found to have a higher risk of incident MASLD than those without accelerated DNA methylation age [21]. Another research identified MASH patients exhibited accelerated epigenetic age that links with increased liver fibrosis [11]. A recent study also reported that long telomere length was associated with reduced risk of MASLD incidence and a positive addictive interaction between high genetic risk score and low telomere length [22]. Intriguingly, we found the deleterious effects of advanced biological aging on of CLDs in males were larger than in females, which was the first time reported by population-based study. Sex differences in MASLD prevalence was partly attributed to estrogens [23]. In both mice and humans, aging could enhance the susceptibility of alcohol-induced liver injury by modulating the SIRT1-C/EBPα-miR-223 axis in neutrophilic [24]. Generally, male drink more frequently than female, which might be a potential reason for our observed gender difference.
There are several potential biological mechanisms might mediate the link between BA and severe liver-related diseases. In the process of aging, the liver cells may develop changes such as telomere shortening, nuclear area increase, mitochondrial DNA damage, and induce the secretion of pro-inflammatory cytokines, thereby resulting in liver damage [25]. Animal experiments have shown that aging could impact liver microcirculatory function and sinusoidal phenotype [26]. Moreover, the accumulation of senescent cells promotes hepatic steatosis and lipid accumulation, which lead to liver damage by inducing inflammation, cell death, fibrosis and promoting organ-specific toxicity [27–28]. Aging can not only directly lead to liver damage, but also induce the progression of MASLD to its more serious stage. Hepatic steatosis might cause mitochondrial dysfunction, hepatic stele cell activation, and hepatic fibrosis, that promote MASLD to the development of HCC from MASLD [1, 29]. The exact biological mechanism underlying the aging and MASLD-associated liver cirrhosis and cancer remains to be further studied.
Since aging is a multi-factorial process, a single BA predictor is not sufficient to monitor the risk of various age related disease phenotypes. Several existing BA predictors including epigenetic clocks, telomere length, transcriptomic, proteomic and metabolomic markers have been widely applied to predicted health outcomes [8]. Given each predictor reflects specific aspect of the aging process, a combination of these indicators appears to be an ideal predictor to predict MASLD risk. Aging rate is not the same in different tissues, and it is not realistic to get a comprehensive marker from various tissues. Compared to those epigenetic and omics biomarkers, routine clinical biomarkers with characteristics of convenient detection and economic advantage are more suitable for large scale population-based screening and dynamic monitoring. In this work, both PhenoAge and KDM-BA yielded better performance than CA in MASLD prediction. The accuracy of PRS for fat-related liver conditions seems not ideal compared with other reported parameters with a mean value exceeded 0.60 [30–32]. With regard to the numbers of known genetic components, more susceptibility loci for fat-related liver diseases are needed to be discovered in future. Additionally, we identified individuals at highest genetic level and BA had the highest standardized 10-year absolute risk of severe MASLD and its serious outcomes. Thus, individuals at a high genetic risk level should be more attention to the accelerated biological aging.
A few limitations should be acknowledged in current study. First, the definition of the new-onset severe MASLD (including MASH) was according to the ICD-10 codes from hospital records, which was more accuracy but might underestimate the real prevalence and incidence. Indeed, it is unpractical to monitor the development of MASLD in a large scale cohort with almost half of million participants. On the other hand, we also confirmed BA acceleration predisposed to incident its serious outcome including liver cirrhosis and cancer. Second, the clinical parameters of each participant were detected at baseline, we could not evaluate the impact of biological aging trajectory on the risk of liver outcomes. Further longitudinal population-based study with repeated measurement of these clinical biomarkers at various times are warranted. Third, the majority of participants in UK Biobank cohort were middle-aged or elderly Europeans, the effect of accelerated BA on liver-related diseases needed to be evaluated in younger groups. Fourth, the unfavorable role of BA on CLD risk are warrant to be verified in different type of BA predictors in future.
In conclusion, Accelerated BA quantified by clinical biomarkers was associated with an increased risk of severe MASLD, as well as its advanced outcomes. Implementing interventions that slow biological aging could serve as an effective preventive strategy for this growing public health problem.