Research Report: Causal Effects of Mental Well-being on Healthy Aging

Research Background

With the significant increase in average life expectancy, aging issues have become increasingly prominent. Challenges such as comorbidity, disability, and the overall social stability of healthcare services and finances have become more severe. Achieving healthy aging, which means maintaining a high quality of health even with significantly extended lifespan, has emerged as a pressing issue. Mental well-being is considered to play a significant role in various lifestyle habits and diseases and is a key factor in healthy aging. Although some surveys and cohort studies have found links between mental well-being and ideal physical health, better functional abilities, or increased survival rates, the causal relationship between mental well-being and healthy aging has not been confirmed due to the inherent limitations of observational studies, such as confounding bias and reverse causality.

Additionally, mental well-being and aging trajectories are intricately linked with socioeconomic status (SES), which further complicates research efforts. Therefore, it remains unclear whether and how mental well-being causally influences healthy aging and whether this association is independent of SES.

Researchers hypothesize that certain lifestyle factors, behaviors, physical functions, and diseases might mediate the impact of mental well-being on healthy aging. Using Mendelian Randomization (MR) to study the causal relationships of these mediating pathways could help in formulating more practical health policies to reduce the health disparities caused by poor mental well-being as people age.

Research Source

The paper for this study is titled “Mendelian Randomization Evidence for the Causal Effect of Mental Well-being on Healthy Aging” and was published in the journal Nature Human Behaviour. The paper was authored by a group of scientists from multiple research institutions and was published online in 2024.

Research Methods

This study applied a two-step, two-sample Mendelian Randomization (MR) method to explore the impacts of mental well-being, independently of SES, on aging phenotypes and dissected the mediating roles of lifestyle, behaviors, physical functions, and diseases in this association. The research design was divided into two main stages:

1. Stage One: Assessing the causal relationship between mental well-being traits and aging phenotypes and determining whether these causal effects are independent of SES.
2. Stage Two: Identifying potential mediators and quantifying their mediating roles between the spectrum of mental well-being and aging gene-environment interaction phenotypes (aging-gip).

Detailed Description of Research Process

Data Sources and Samples:

• The study utilized large-scale genome-wide association study (GWAS) data from European ancestry populations, with total sample sizes ranging from 80,852 to 2,370,390.
• The analysis estimated the causal effects of mental well-being (including life satisfaction, positive affect, neuroticism, and depressive symptoms) on genetic aging phenotypes and their components (including resilience, self-rated health, health-span, parental lifespan, and longevity), with sample sizes ranging from 36,745 to 1,012,240.
• All data were sourced from the largest GWAS of European ancestry populations.

Step 1: UVMR and MVMR Analysis

• Univariate MR (UVMR) and Multivariable MR (MVMR) methods were used to assess the causal effects of each mental well-being trait and each mediator on aging phenotypes.
• UVMR and MVMR estimated the causal effects of each SES indicator on mental well-being traits to determine which causal effects are independent of the other two.
• Finally, MVMR was used to estimate the direct effects of the mental well-being spectrum on aging phenotypes after adjusting for SES indicators, to determine whether the causal effects of mental well-being on aging phenotypes are independent of SES.

Step 2: Mediating MR Analysis

• First, UVMR was used to estimate the causal effects of the mental well-being spectrum on mediators.
• Reverse MR analysis determined the bidirectionality between mediators and the mental well-being spectrum to ensure the validity of the mediator model.
• Next, MVMR was used to estimate the causal effects of each mediator on aging phenotypes, provided that the mediator had a causal link with aging phenotypes in UVMR.
• The Delta method calculated the mediating proportion of each mediator between the mental well-being spectrum and aging phenotypes.

Research Results

Total Effects of Mental Well-being on Aging Phenotypes:

• Linkage Disequilibrium Score (LDSC) regression analysis found genetic correlations between all mental well-being traits and aging phenotypes (aging-gip).
• Using UVMR and MVMR analyses, results showed that better mental well-being was positively associated with higher aging-gip and its components (resilience, self-rated health, health-span, and parental lifespan), independent of SES indicators such as income, education, and occupation.

Effects of Mental Well-being on Aging Phenotypes Independent of SES:

• MVMR analysis results showed that income, rather than education or occupation, played a major role in SES’s effects on mental well-being.
• MR-LAP analysis confirmed the causality of the causal links between the mental well-being spectrum, mediators, and aging phenotypes, eliminating biases from sample overlap, winner’s curse, and weak instrumental variables.

Mediation MR Analysis:

• Thirty-three mediators were identified, including lifestyle factors, behaviors, physical functions, and diseases.
• Among lifestyle factors, TV viewing time, smoking (initiation age and daily number of cigarettes), and intake of cheese and fresh fruits mediated the relationship between the mental well-being spectrum and aging-gip.
• In the behavior domain, the use of antihypertensive drugs and non-steroidal anti-inflammatory drugs, cognitive performance, and age at menarche also served as mediators.
• In terms of physical functions, body fat, blood lipids, muscle mass, and inflammation all played mediating roles in the relationship.
• Cardiovascular diseases (CVDs) were the main mediators among diseases, with heart failure, stroke, coronary atherosclerosis, and ischemic heart disease each mediating more than 5% of the total effect of the mental well-being spectrum on aging-gip.

Robustness and Limitations of Causal Inference:

• The study ensured the robustness of causal estimates through various sensitivity analyses and validated results using multiple methods.
• However, as with all MR analyses, despite weak instrumental variables, horizontal pleiotropy, outliers, and sample overlap not significantly affecting causal estimates, results should still be interpreted cautiously.

Research Significance and Value

This study reveals the causal impact of mental well-being on aging phenotypes independent of SES and outlines various mediating pathways. The findings emphasize the importance of prioritizing mental well-being in aging-related health policies and propose targeted intervention strategies to reduce health disparities caused by poor mental well-being in aging.

Highlights:

• This study is the first to comprehensively evaluate the causal effects of mental well-being on aging phenotypes using a two-step, two-sample Mendelian Randomization method.
• There is a positive causal relationship between mental well-being independent of socioeconomic status and aging phenotypes.
• Multiple lifestyle factors, behaviors, physical functions, and diseases mediate the impact of mental well-being on aging, providing a theoretical foundation and direction for further studies.

This research provides new evidence for the importance of mental well-being in achieving healthy aging and guides future public health policy-making and intervention measures.

Research Methodology and Data Availability

All GWAS data used in the study are publicly available. The study used several packages (twosamplemr, mvmr, mrpresso, and mrlap) in R software for data analysis. Detailed descriptions of the research methods and custom code used can be found on GitHub.