Interest in immortality dates back thousands of years, but longevity—or living longer—has arguably never been as topical as it is today. In this article, we walk through a recent study that uncovered eight key lifestyle factors for longevity.
Introduction
One of the earliest examples of interest in longevity—or in this case, immortality—can be traced back to the Epic of Gilgamesh, an ancient poem written roughly 3,000–4,000 years ago. In this epic, Gilgamesh, the king of Uruk (an ancient city located in what is now known as Iraq), discovers that ‘Life, which you look for, you will never find…’. In essence, Gilgamesh is confronted with the news that lifespan is not within your own hands.
Fast forward to the present day, and we have tech-billionaires like Bryan Johnson picking up where Gilgamesh left off, and seeking to push the boundaries of human lifespan. While Johnson’s approach is shrouded in secrecy, is very extreme, and has been criticised as unproven, there are legitimate, science-based lifestyle approaches that can help us to live longer. To understand the extent to which these factors may prolong our lifespan, we turn to a study published in January this year (2024) in the American Journal of Clinical Nutrition (1).
Bryan Johnson reportedly spends $2 million per year trying to extend his lifespan (Image credit: Magda Wosinska/Magdalena Wosinska, via The Guardian).
Study design
Participants
The Million Veteran Program is a prospective cohort study of US military veterans that aimed to understand the genetic and non-genetic determinants of chronic disease. Recruitment began in 2011, with the goal of recruiting one million veterans. As of 2020, the program had 819,417 participants; however, after excluding participants who did not fill out all relevant lifestyle surveys or who did not fit the inclusion criteria, a total of 276,132 participants were included in the current study (which is still a lot of people!).
Methods
The eight lifestyle factors (or behaviours) were based on the concept of preventive lifestyle medicine, which proposes six low-risk lifestyle behaviours, i.e., nutrition, physical activity, stress management, restorative sleep, avoidance of risky substances, and social connections. The Million Veteran Program collected data on smoking, alcohol use, and opioid use, so combining these three risky behaviours with the others forms the eight behaviours measured in this study.
Once the researchers measured these behaviours they assigned participants a score of either one or zero based on whether they achieved these lifestyle behaviours or not. They defined meeting these behaviours—and thus achieving a score of one—in the following ways:
Nutrition: eating a healthful plant-based dietary pattern (defined as in the upper 40% of scores of the healthful plant-based diet index (2))
Physical activity: being leisurely active for roughly 150 minutes per week or more
Stress management: scoring low on questionnaires assessing stress
Restorative sleep: sleeping 7–9 hours per day
Smoking: never smoking
Alcohol use: never drinking or typically having no more than four drinks in a day
Opioid use: having no opioid use disorder
Finally, researchers estimated the effect of these behaviours on risk of death. Importantly, the researchers attempted to account for other factors that could influence the risk of death (i.e., age, sex, race/ethnicity, education level, income, marriage status, and body mass index (BMI)) in their statistical adjustment model. This was done to try to isolate the specific effect of the behaviours of interest on risk of death.
Life’s Essential 8, from the American Heart Association, has some overlap with the lifestyle behaviours discussed in this article. The eight recommendations from Life’s Essential 8 are to: eat a healthy diet, be more active, quit tobacco, get healthy sleep, manage weight, control cholesterol, manage blood sugar, and manage blood pressure.
Results
Performing any one of these lifestyle behaviours was associated with a significant reduction in risk of death (compared to not doing so). In addition, each additional behaviour added years to an individual’s estimated life expectancy (see Figure 1 and Figure 2, below). Astonishingly, a male individual’s life expectancy was 24.0 years greater if they performed all eight behaviours compared to zero (Figure 1), whereas this figure was 21.5 years for females (Figure 2). Importantly, when the researchers tested the robustness of their results through a series of further analyses, their results were largely unchanged.
Figure 1. Estimated prolonged male life expectancy at age 40 attributable to performance of each successive lifestyle behaviour (compared to none) (1).
Figure 1. Estimated prolonged female life expectancy at age 40 attributable to performance of each successive lifestyle behaviour (compared to none) (1).
Summary
This study shows the possible years of life attributable to healthy lifestyle behaviours, while also showcasing a linear relationship between healthy lifestyle behaviours and a reduced risk of death. In addition, because the study participants were nationally representative, these findings may apply to the general US population (and potentially other Western populations).
Conclusion
Although genetics are the number one driver of longevity (3), healthy lifestyle behaviours clearly play a large role. Armed with this information, you can focus on achieving the behaviours you struggle with if longevity is the goal.
If you would like some tips to change your behaviour for the better, check out the latest episode of the Training121 Podcast where we walk through the COM-B model and how to use it to guide behaviour change. And if you would like supplemental football training to improve your skills, reach out to our team of expert coaches at [email protected] to get started.
Thanks for reading!
Patrick Elliott, BSc, MPH
Health and Nutrition Science Communication Officer at Training121
Instagram: @just.health.info
Twitter: @PatrickElliott0
References
(1) Nguyen XT, Li Y, Wang DD, Whitbourne SB, Houghton SC, Hu FB, Willett WC, Sun YV, Djousse L, Gaziano JM, Cho K, Wilson PW; VA Million Veteran Program. Impact of 8 lifestyle factors on mortality and life expectancy among United States veterans: The Million Veteran Program. Am J Clin Nutr. 2024;119(1):127–35. Available at: https://www.sciencedirect.com/science/article/pii/S000291652366280X
(2) Satija A, Bhupathiraju SN, Rimm EB, Spiegelman D, Chiuve SE, Borgi L, Willett WC, Manson JE, Sun Q, Hu FB. Plant-Based Dietary Patterns and Incidence of Type 2 Diabetes in US Men and Women: Results from Three Prospective Cohort Studies. PLoS Med. 2016 Jun;13(6):e1002039. Available at: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1002039
(3) Gavrilov LA, Gavrilova NS. New Developments in the Biodemography of Aging and Longevity. Gerontology. 2015;61(4):364–71. Available at: https://karger.com/ger/article/61/4/364/148099/New-Developments-in-the-Biodemography-of-Aging-and
Technical Terms
Prospective cohort study: This type of study enrols a group of people (the cohort) and measures a number of baseline characteristics (e.g., age, sex, diet, lifestyle, etc.). Then, the researchers track this cohort over time and oftentimes, they repeatedly measure characteristics at different time intervals (e.g., every 2–4 years). This is done to track health behaviours (like diet) over time. Eventually, if the study participants experience health conditions (e.g., a heart attack) or if they die, researchers can use a variety of statistical procedures to see if any of their characteristics (e.g., sex, lifestyle behaviours, diet, etc.) may have impacted the risk of their health condition or death, and so on. This research design is very important to understand the role of diet and lifestyle for longevity or disease risk.
Statistical adjustment model: Statistical adjustment is a method to account for the effects of variables that may be associated with an outcome, but are not the focus of the study, so that researchers can better identify the independent effect of the variables that are the focus of the study on their outcome of interest. For example, if I wanted to understand the effect of diet on the risk of death, I would add other variables like age, sex, smoking status, and so on, to the statistical model so that the effect I observe for diet is independent of these other variables (known as confounding variables). Without doing this, I could inadvertently think that eating a healthy diet lowers the risk of death, but on further inspection, it could actually be that the people eating a healthy diet were also younger and didn’t smoke (so the effect of diet was not independent of age and smoking, and thus diet may not have a ‘real’ effect on risk of death). Therefore, having a strong statistical adjustment model reduces the risk of bias in effect estimates.
Nationally representative: This means that the study participants reflected the US population in terms of demographic factors like sex, race/ethnicity, education level, income, and so on. The benefit of using nationally representative data is that you can state with more certainty that the findings in the study may apply to the wider US population at large (and even other Western countries, to an extent).