Epigenetic age predictors in community-dwelling adults with high impact knee pain

Abstract

Gerontological research reveals considerable interindividual variability in aging phenotypes, and emerging evidence suggests that high impact chronic pain may be associated with various accelerated biological aging processes. In particular, epigenetic aging is a robust predictor of health-span and disability compared to chronological age alone. The current study aimed to determine whether several epigenetic aging biomarkers were associated with high impact chronic pain in middle to older age adults (44-78 years old). Participants (n = 213) underwent a blood draw, demographic, psychosocial, pain and functional assessments. We estimated five epigenetic clocks and calculated the difference between epigenetic age and chronological age, which has been previously reported to predict overall mortality risk, as well as included additional derived variables of epigenetic age previously associated with pain. There were significant differences across Pain Impact groups in three out of the five epigenetic clocks examined (DNAmAge, DNAmPhenoAge and DNAmGrimAge), indicating that pain-related disability during the past 6 months was associated with markers of epigenetic aging. Only DNAmPhenoAge and DNAmGrimAge were associated with higher knee pain intensity during the past 48 h. Finally, pain catastrophizing, depressive symptomatology and more neuropathic pain symptoms were significantly associated with an older epigenome in only one of the five epigenetic clocks (i.e. DNAmGrimAge) after correcting for multiple comparisons (corrected p's < 0.05). Given the scant literature in relation to epigenetic aging and the complex experience of pain, additional research is needed to understand whether epigenetic aging may help identify people with chronic pain at greater risk of functional decline and poorer health outcomes.

Keywords: Biological aging; aging biomarker; chronic pain; epigenetic aging; pain biomarker.

Figure 1.

Bar plot of predicted epigenetic age difference (predicted epigenetic age—chronological age) with respect to the pain impact (a) DNAmAge; (b) DNAmAgeHannum; (c) DNAmAgeSkinBloodClock; (d) DNAmPhenoAge, and (e) DNAmGrimAge. p-values were obtained by ANCOVAs adjusting for age, sex, study site, and chronological age as covariates. Standard errors were marked on the bar plot.

Figure 2.

Scatter plot of predicted epigenetic age difference (predicted epigenetic age—chronological age) with respect to WOMAC pain. (a) DNAmAge; (b) DNAmAgeHannum; (c) DNAmAgeSkinBloodClock; (d) DNAmPhenoAge, and (e) DNAmGrimAge. p-values were obtained by linear regression model adjusting for age, sex, study site, and chronological age as covariate.

Figure 3.

Heatmap of the partial correlation coefficients between predicted epigenetic age and variables across five domains, including clinical pain, cognitive function, experimental pain, physical function, and psychosocial function. Each row is a clinical variable and each column is a type of the predicted epigenetic age difference (predicted epigenetic age - chronological age), including DNAmAge, DNAmPhenoAge, and DNAmGrimAge. Chronological age, sex, race and study site were adjusted as covariates. The partial correlation coefficient was indicated by a red-blue color scale, where red indicated positive correlation, blue indicated negative correlation. Significant partial correlations (i.e. p ≤ 0.05) were indicated by *.