Yakutia is one of the coldest permanently inhabited regions in the world, characterized by a subarctic climate with average January temperatures near -40 °C and the minimum below -60 °C. Recently, we demonstrated accelerated epigenetic aging of the Yakutian population in comparison to their Central Russian counterparts, residing in a considerably milder climate. In this paper, we analyzed these cohorts from the inflammaging perspective and addressed two hypotheses: a mismatch in the immunological profiles and accelerated inflammatory aging in Yakuts. We found that the levels of 17 cytokines displayed statistically significant differences in the mean values between the groups (with minimal p-value = 2.06 × 10-19), and 6 of them are among 10 SImAge markers. We demonstrated that five out of these six markers (PDGFB, CD40LG, VEGFA, PDGFA, and CXCL10) had higher mean levels in the Yakutian cohort, and therefore, due to their positive chronological age correlation, might indicate a trend toward accelerated inflammatory aging. At the same time, a statistically significant biological age acceleration difference between the two cohorts according to the inflammatory SImAge clock was not detected because they had similar levels of CXCL9, CCL22, and IL6, the top contributing biomarkers to SImAge. We introduced an explainable deep neural network to separate individual inflammatory profiles between the two groups, resulting in over 95% accuracy. The obtained results allow for hypothesizing the specificity of cytokine and chemokine profiles among people living in extremely cold climates, possibly reflecting the effects of long-term human (dis)adaptation to cold conditions related to inflammaging and the risk of developing a number of pathologies.
Keywords: Yakutia; climate; cold environment; deep neural network; explainable artificial intelligence; inflammatory profile.