In the present study, for the first time, we examined the association of genetically predicted leukocyte TL and facial aging in a large population-based cohort, the UK Biobank, using the MR method. By leveraging several MR estimation approaches, we found that longer genetically predicted leukocyte TL was associated with a lower likelihood of facial aging. Our study, corroborating previous experimental studies of skin cells, provides further evidence to support a causal role of leukocyte TL in facial aging.
To the best of our knowledge, no studies have been published to investigate the role of TL in facial aging using population-based cohorts until now. A previous study examined TL in cells of the sun-protected and sun-exposed skins . Further studies employed ultraviolet-exposed skin cells to study the roles of TL in photoaging [18,19,20,21]. These studies, taken together, were suggestive of the functions of TL in skin aging at the cellular level [5, 22]. Further evidence at the population level, however, is lacking. The limited number of studies on this topic at the population level is partly due to the difficulties in facial aging measurements. For example, although the three-dimensional human facial morphology assessment instruments , among others , can offer objective and comprehensive features for facial aging, they have not been widely available for aging researchers because of infeasibility and high cost.
To overcome the challenges in facial aging measurement and to make the best use of questionnaire-based instruments, a subjective assessment of facial aging could be valuable in this regard measurement [25,26,27]. In this study, we examined the role of TL by taking advantage of questionnaire-based measurement for facial aging in the UK Biobank. An advantage of this type of data is that the large sample size in the UK Biobank could outweigh the concerns in measurement error for questionnaire-based facial aging assessment. Indeed, we obtained consistent and significant estimates by using various MR approaches in this study. Further studies are warranted to explore the biological mechanisms of TL in facial aging.
Despite the advantages of the large sample size, our study is prone to several limitations. First, the MR assumptions, particularly the no-pleiotropy assumption, must be satisfied in order to yield a valid estimate. The intercept of MR-Egger regression analysis approaches zero suggesting no strong evidence for directional pleiotropy. Second, the MR methods used in this study also assume a linear relationship between TL and facial aging. We cannot examine further if there was a non-linear relationship. Third, as facial aging GWAS was analyzed using mixed linear models, the effect sizes cannot be interpreted using a more intuitive way. The results could rather be used for testing purposes. Fourth, as alluded above, the facial aging was measured using a single question in the UK Biobank, which could lead to measurement error. Future studies are encouraged in order to develop a more detailed questionnaire-based instrument to measure diverse dimensions of facial aging. Fifth, in this study, facial aging was assessed only using qualitative data and not objectively assessed using skin turgor  or facial wrinkles per area of skin . Lastly, TL was measured in leukocyte rather than skin cells. However, a previous study demonstrated that the correlation of TL in leukocyte and skin cells was high (correlation coefficient r = 0.83) , and a recent study found that TL varies by tissue type but is generally correlated among tissue types .
In summary, our study provided novel evidence to support a causal role of genetically predicted leukocyte TL in facial aging. Further studies are warranted to explore the biological mechanisms of TL in facial aging.