Introduction
Adiposity is a worldwide major public health problem with an alarmingly increasing prevalence over the past 2 decades. Forty-two percent of adults in the United States are obese, and a total of 1.9 billion adults worldwide are overweight or obese, which continue to climb, represent a major health and economic burden [
1]. During obesity, skin-resident and systemic adipose tissue derives secreted factors and increases inflammation and has detrimental impact on facial aging [
2‐
4]. The consequences of obesity in the skin are underestimated. Increased body mass index puts great disserve to skin physiology, skin barrier, collagen structure, and wound healing process, triggering the possibility of skin cancer [
5]. Skin, as a sociologically meaningful interface, has psychological implications different from other organs, which makes facial rejuvenation and skin health care particularly attractive in the context of the global population aging. With the popularization of health care knowledge and the improvement of living conditions, the life span of the general population has extended and the global population has aged, which leads to a major paradigm shift in the focus of medical research to preventing and managing aging and optimizing the overall healthy life span [
6,
7]. The society is paying increasing attention to the treatment and early intervention of skin aging.
Growing data suggest that facial aging is linked to obesity, lowering quality of life, and compromising mental health [
8,
9]. In 2007, Yosipovitch et al. highlighted the association between adiposity and dermatologic conditions [
10]. Subsequently, systematic reviews and meta-analyses have validated these putative correlations between adiposity (mostly BMI used as an indicator) and face aging [
6,
11]. However, previous studies are based on observational epidemiological designs, which are prone to reverse causation and unmeasured confounding [
12]. Specifically, the independent roles of adipose tissue from different depots in the relationship between adiposity and facial aging are still unclear, which is crucial to improve the rejuvenation modalities. Mendelian randomization (MR) provides an alternative approach to investigate causality by using genetic variants as instrumental variables and thereby accounting for observational study bias [
13‐
15]. Here, we provide an MR analysis of the relationship between face aging and obesity along with a more precise indicator.
Methods
UK Biobank is a large-scale biomedical database and research resource that contains detailed genetic and health data from over half a million UK individuals [
16]. In UK Biobank, Participants were recruited from National Health Service central registries across the UK, with provided informed permission and the Northwest Multicenter Research Ethics Committee granted ethical approval. Considering the genome-wide association studies (GWAS) used to identify genetic instruments were predominantly based on European samples, our analysis exclusively included European subjects.
Adiposity measures were ascertained following standardized protocols [
6,
17]. Weight was measured without shoes and heavy outer clothing and BF% were measured via a Tanita BC-418 MA body composition analyser. Height was measured with a Seca 202 height measure, and waist circumferences were obtained with a tape measure [
6,
17]. BMI was computed using National Institutes of Health standard (
https://www.nhlbi.nih.gov/health/educational/lose_wt/BMI/bmicalc.htm). BF% and WC’s GWAS data were output from GWAS pipeline using Phesant derived variables from UK Biobank [
18]. As genetic instruments for BMI, BF%, and WC, we employed single nucleotide polymorphisms (SNPs) detected from the largest European descendant GWAS to date (458, 395, 374 near-independent SNPs, respectively) [
17,
19].
Facial aging was conducted using data from the population-based UK Biobank. This deposit from which facial aging domain was obtained, provides full details of the genome-wide association study (GWAS) pipeline developed by the MRC-IEU for the full UK Biobank (version 3, March 2018) genetic data [
20]. The method of this domain to determine perceived age based on questionnaire has been described previously [
21,
22]. Subjective evaluations of facial aging were measured by the question: do people say that you look younger than you are, older than you are, about your age, do not know, and prefer not to answer?
MR analysis
The data on the necessary variables and the facial aging GWAS were harmonized by the separate chromosomes and positions. The five MR methods applied in the research have been described previously [
23,
24]. The random-effects inverse-variance weighted (IVW) method was used as the primary estimator for the MR analysis [
25], for it has a higher statistical power with the assumption that all SNPs are valid instrumental variables. The weighted median approach [
26], MR-Egger regression [
27], weighted mode method [
28] and simple mode method were used as reciprocal analysis. To investigate whether any single SNP had a disproportionate effect on the overall results, IVW analyses were re-performed leaving out SNPs once a time. To investigate if the SNPs affected facial ageing but not through adiposity, horizontal pleiotropy analysis was applied and identified SNPs of heterogeneity were extracted separately for literature background check. The causal estimates would be regarded as heterogeneous if the
p value for Cochran’s
Q test was above 0.05 and
I2 was above 0.25. Scatter plots, leave-one-out analysis plots, forest plots and funnel plots were created to visually assess the results.
All statistical analyses were conducted in R 3.6 and TwoSampleMR package [
29]. R version 3.4.0 and RStudio 2022.07.2 both for MacOS, were used for all statistical analyses. The weighted median method was used to supplement the MR-Egger, in order to gain a further dependable estimation of the unproductive effect.
The present study only used publicly available summary-level statistics. No individual-level data were analyzed. Ethical approval therefore was not required.
Sensitivity analyses
Sensitivity analyses were applied, including heterogeneity and pleiotropy analysis, to rule out the potential that these SNPs influenced the Mendelian randomization studies. A leave-one-out analysis was also completed to determine whether any SNP-driven relationships existed.
Discussion
In the present study, for the first time, we examined the association of genetically predicted adiposity and facial aging in a large population-based cohort, the UK Biobank (as well as IEU OpenGWAS project applied in supplementary materials). By leveraging several MR estimation approaches, we found that longer genetically predicted obesity was associated with a more severe likelihood of facial aging. Our study, corroborating previous observational and experimental studies, provides further evidence to support a causal role of adiposity in facial aging. And to the best of our knowledge, no studies have used population-based cohorts to investigate the topic.
The difficulty in measuring face aging is one reason why there are so few research on this topic at the population level. For instance, although three-dimensional human face morphology evaluation devices [
35], among others [
36], can deliver objective and comprehensive characteristics for facial aging, aging researchers have not had widespread access to them due to their impracticability and high cost. A subjective evaluation of facial aging may prove helpful in overcoming the difficulties associated with measuring face aging and in maximizing the use of questionnaire-based tools [
37]. That’s why we finally chose to use questionnaire-based measurements from the UK Biobank [
38,
39]. The UK Biobank and IEU OpenGWAS projects’ high sample sizes have the potential to allay worries about measurement error for the evaluation of face aging based on questionnaires, which is a benefit of this sort of data.
Simultaneously, when discussing the potential risks of adiposity, people frequently refer to the involvement of adipose tissue. SAT plays an essential role in the regulation of skin homeostasis and rejuvenation [
40,
41]. Laboratory studies showed that age-dependent regulatory cells (ARCs) in SAT increase in abundant after middle age and display high levels of pro-inflammatory markers, which may accelerate facial aging [
42]. However, as limited publicly available databases, the facial ageing’s casualty with subcutaneous adipose tissue (SAT) cannot be detected with only two SNPs harvested from GWAS (Additional file
2).
Then we shifted our gaze to visceral adipose tissue (VAT). An UK Biobank-based MR research identified a link between genetically determined VAT accumulation and longevity, showing that visceral adiposity may be detrimental to systemic aging [
43]. Based on this, we investigated VAT’s effects on the more localized facial aging. More detailed causal relationship was revealed in the negative results of multivariable MR analysis and two sample MR analysis between VAT and facial aging (Additional file
1). Among the three adiposity indicators, WC has a better casual association with facial aging (Additional file
1: Table S1) Then, a two sample MR analysis was implemented between VAT and facial aging. We observed that the OR per SD increase of VAT on facial ageing using the IVW method was 1.047 (95% CI − 0.002, 0.093) and
P-value is greater than 0.05 among the five MR methods tested (Additional file
1: Table S2).
The primary and secondary effects of SAT and VAT on facial aging have not yet been verified. Although VAT has no anatomic advantage over SAT, considering that the increase of WC is more closely related to the accumulation of VAT, we suggested that the excessive VAT accumulation might also be involved in the aggravation of facial aging through multiple pathophysiological processes [
46]. One the other hand, several researchers considered that white adipose tissue (WAT) in SAT can be a beneficial therapeutic target for facial aging. Subcutaneous WAT have a thick ‘skin’ confronting with the down-regulating behavior of aging-regulated RNA [
47,
48]. Among a pairwise age group comparisons, only a few immune- and inflammation-associated genes are found to be differentially regulated in the skin compared to the brain and blood, indicating that subcutaneous WAT performs a better immunomodulatory capacity with typical inflammaging signs [
49].
In addition, factors affecting facial aging are also very diverse. Novel results revealed a detrimental effect of heavier smoking on facial aging [
50]. Another nonlinear analysis indicated a potential threshold relationship between alcohol intake and telomere length [
51]. But there is still no MR evidence showing alcohol consumption or other unhealthy lifestyles have any direct causality with facial aging. There is also little statistical evidence suggesting whether factors of low-quality life (like poverty, disability, long-term UV exposure or air pollution exposure), and mental health (including depression and anxiety) trigger facial aging. We will continue to conduct studies on these topics.
Limitations
Despite the advantages of the large sample size, our study is prone to several limitations. First and foremost, in this study we performed reliable and meaningful estimations using a variety of MR methods. Adiposity can be used as a mediator to further investigate the underlying molecular causes of face aging. Simultaneously, it is also necessary to analyze data in more nuanced manners, such as grouping data according to race, gender and living habits, or applying more complex MR models to verify their direct or indirect associations. Second, although we have assumed a linear relationship between adiposity and facial aging in this study, we still do not know whether this is a direct linear causal relationship or indirect nonlinear relationship. Further research needs to be conducted by applying more complex MR models to verify their direct or indirect correlations. Third, considering there is few basic research with high hierarchy of evidence revealing the relationship between adiposity and facial aging or between subcutaneous adipose tissue and facial aging, our present findings need further replication in multicenter cohorts and basic research contained high-throughput sequencing analysis of human samples. Last but not least, as previously stated, the UK Biobank and IEU OpenGWAS research quantified face age using a single query, which could lead to measurement error. Future research is encouraged in order to create a more comprehensive questionnaire-based approach for measuring diverse aspects of facial aging. In this study, we performed reliable and meaningful estimations using a variety of MR methods. Adiposity can be used as a mediator to further investigate the underlying molecular causes of face aging. Simultaneously, it is also necessary to analyze data in more nuanced manners, such as grouping data according to race, gender and living habits, or applying more complex MR models to verify their direct or indirect associations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.