The Association of Physical Activity Behaviors and Patterns With Aging Acceleration: Evidence From the UK Biobank

Jianwei Zhu; Yao Yang; Yu Zeng; Xin Han; Wenwen Chen; Yao Hu; Yuanyuan Qu; Huazhen Yang; Unnur A Valdimarsdóttir; Fang Fang; Huan Song

doi:10.1093/gerona/glad064

. 2023 Feb 23;78(5):753–761. doi: 10.1093/gerona/glad064

The Association of Physical Activity Behaviors and Patterns With Aging Acceleration: Evidence From the UK Biobank

Jianwei Zhu ^1,^#, Yao Yang ^2,^3,^#, Yu Zeng ^4,⁵, Xin Han ^6,⁷, Wenwen Chen ⁸, Yao Hu ^9,¹⁰, Yuanyuan Qu ^11,¹², Huazhen Yang ^13,¹⁴, Unnur A Valdimarsdóttir ^15,¹⁶, Fang Fang ¹⁷, Huan Song ^18,^19,^✉

Editor: David Le Couteur

¹ Department of Orthopedic Surgery and West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

² West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

³ Med-X Center for Informatics, Sichuan University, Chengdu, China

⁴ West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

⁵ Med-X Center for Informatics, Sichuan University, Chengdu, China

⁶ West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

⁷ Med-X Center for Informatics, Sichuan University, Chengdu, China

⁸ Division of Nephrology, Kidney Research Institute, State Key Laboratory of Biotherapy and Cancer Center, West China Hospital, Sichuan University, Chengdu, China

⁹ West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

¹⁰ Med-X Center for Informatics, Sichuan University, Chengdu, China

¹¹ West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

¹² Med-X Center for Informatics, Sichuan University, Chengdu, China

¹³ West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

¹⁴ Med-X Center for Informatics, Sichuan University, Chengdu, China

¹⁵ Center of Public Health Sciences, Faculty of Medicine, University of Iceland, Reykjavík, Iceland

¹⁶ Institute of Environmental Medicine, Karolinska Institute, Stockholm, Sweden

¹⁷ Institute of Environmental Medicine, Karolinska Institute, Stockholm, Sweden

¹⁸ West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China

¹⁹ Med-X Center for Informatics, Sichuan University, Chengdu, China

^✉

Address correspondence to: Huan Song, MD, PhD, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Guo Xue Lane 37, Chengdu, China. E-mail: songhuan@wchscu.cn

These authors contributed equally.

Roles

David Le Couteur: MBBS, FRACP, PhD, Decision Editor

PMCID: PMC10172984 PMID: 36815559

Abstract

Prior evidence suggests that physical activity may reduce the risk of multiple diseases and mortality. However, whether and how physical activity affects the aging process remains largely unexplored. We included 284 479 UK Biobank participants and computed leukocyte telomere length (LTL) deviation (ie, the difference between genetically determined and observed LTL) and biological age acceleration (defined as the discrepancy between the phenotypic age of a person and the average phenotypic age in the cohort of individuals with the same age and sex) as the indexes for aging acceleration. Linear and logistic models were used to estimate the associations of self-reported physical activity items and patterns (identified by principal component analysis), as well as accelerometer-assessed physical activity, with aging acceleration. Analyses of physical activity patterns indicated, a higher level of adherence to activity patterns predominated by strenuous sports, other exercises, walking for pleasure, heavy and light housework, and public transportation use was associated with a lower risk of aging acceleration, whereas a higher level of adherence to patterns predominated by job-related activities was associated with a higher risk of aging acceleration. Analysis among 62 418 participants with accelerometer-measured physical activity corroborated these results. Physical activity, such as strenuous sports and other exercises in leisure time and the use of public transportation, was associated with reduced biological aging. Besides highlighting the importance of engaging in physical activity for healthy aging, our results provide further evidence for the beneficial effect of physical activity on the telomere attrition process.

Keywords: Biological aging, Leukocyte telomere length, Physical activity, Polygenic risk score, Principal component analysis

Physical activity, as a major component of a healthy lifestyle (1), has been shown to have an inverse association with the occurrence of numerous medical conditions including metabolic diseases (2) and cardiovascular disease (3). In healthy individuals, enhanced aerobic fitness and skeletal muscle strength resultant from habitual physical exercise may explain such effects through, for instance, improved glucose homeostasis, insulin sensitivity (4), and hepatic fat metabolism (5,6). Likewise, randomized clinical trials indicate that, for individuals with cardiometabolic diseases, exercise-based programs benefit in terms of optimizing metabolic parameters (7), improving quality of life, and reducing health care use (8). In addition, despite the inconsistent results (9), most prior investigations suggested reduced all-cause mortality (10,11) and disease-specific mortality (11,12) among individuals who engaged in recommended aerobic and muscle-strengthening activities, compared with individuals not engaged in such activities (10). The underlying mechanism linking physical activity to an increased life expectancy, however, remains largely unexplored. Moreover, due to the heterogeneity of physical activity measurements and the understudied interactions between multiple modes of physical activity (eg, physically active people tend to have high levels of many types of physical exercise), existing literature provides little information regarding the optimal patterns of physical activity on health and aging.

Biological age has been proposed as an index of overall body fitness, focusing on the degree of accumulated pathophysiological changes that contribute toward mortality over time. For instance, phenotypic age is one of the widely used biological measures of aging, estimated by 9 multisystem clinical chemistry biomarkers (13). It was established using data from 11 617 adults at 20 or above in the U.S. National Health and Nutrition Examination Survey, showing a good predictive value for multiple aging-related diseases and overall mortality (13,14). However, most biological age measures are considered suboptimal as the algorithms were mainly established based on the association between composite biomarkers and age in the training data set (14). Alternatively, the shortening of telomere length (TL) has also been postulated as a biological measure of human aging (15). In vitro and in vivo studies suggest that TL might reflect the degree of cellular senescence and oxidative stress (16), whereas evidence from epidemiological studies is less conclusive (17). Several longitudinal studies indicate that TL measured in leukocyte (LTL) is associated with age-related diseases and mortality (18), while others failed to do so (19). In addition to methodological limitations (eg, small sample size and cross-sectional design), the ignorance of individual variability in LTL, namely the genetic determinants of LTL (30%–40%) and its shortening rate (20), might also contribute partly to the observed discrepancy in previous results.

In the present study, leveraging data from the UK Biobank including rich phenotypic information on physical activity and other lifestyle factors, measurement of LTL, and individual-level genotyping data, we aimed to examine the association of self-reported physical activity behaviors and patterns with aging acceleration, measured by LTL deviation (ie, the difference between expected and measured LTL) after considering genetic determinants of LTL, gender, and age. We further used objectively measured physical activity by the accelerometer to determine the association to aging acceleration (ie, LTL deviation and biological age acceleration [BAA]).

Method

Study Population and Design

UK Biobank is a prospective community-based study that recruited half a million participants aged 40–69 years across the United Kingdom. Details of this study have been described elsewhere (21). Briefly, information on sociodemographic characteristics, lifestyle (including physical activity), and environmental factors was collected through touchscreen questionnaires or verbal interviews at recruitment. Physical measurements and biosample collection were also conducted during the participant’s first visit to an assessment center (ie, baseline). With the content of having their health and survival status monitored, data of UK Biobank participants were linked with multiple national health registers. The inpatient hospital data include data extracted from Hospital Episode Statistics from England, Patient Episode Database from Wales, and the Scottish Morbidity Record from Scotland, with full coverage of UK Biobank participants since 1997. Primary care data were obtained from multiple general practice data system suppliers, available for approximately 45% of the study participants. Genotyping data from blood samples were released for 484 325 participants using 2 genotyping arrays specifically designed for UK Biobank with 95% shared marker content (ie, UK Biobank Axiom Array or Affymetrix UK BiLEVE Axiom Array) (22). LTL, derived from blood samples collected at baseline, was measured for 474 074 participants, based on multiplex qPCR where the LTL was determined as the relative telomere to single copy gene (T/S) ratio after adjusting for the technical parameters (ie, enzyme batch, temperature, operator, primer batch, and humidity) (23).

In the present study, we first excluded UK Biobank participants who withdrew their inform consent (n = 48, see Figure 1). In addition, to assess the nongenetically determined LTL deviation (ie, the primary measure of aging acceleration), we removed 139 186 participants with no data on LTL measurement (n = 29 882), non-European ancestry (n = 76 444), or no eligible genotyping data (ie, genotyping rate <99% or kinship coefficient ≥0.044, n = 32 860). Furthermore, to alleviate the concern that physical activity was influenced by pre-existing somatic diseases, we calculated the Charlson Comorbidity Index (CCI) (24) based on medical diagnoses from self-reported, hospital inpatient, and primary care data, as a proxy of baseline physical fitness (a list of relevant diagnoses and their corresponding International Classification of Diseases [ICD] 10th edition [ICD-10] codes are shown in Supplementary Table 1). Individuals with CCI ≥ 1 (n = 78 789) or no self-reported data on all physical activity items (n = 5) were then removed, leaving 284 479 eligible participants in the analyses.

The UK Biobank has full ethical approval from the NHS National Research Ethics Service (16/NW/0274) and informed consent was obtained before data collection from each participant. The present study was also approved by the biomedical research ethics committee of West China Hospital (reference number: 2019-1171).

Identification of Physical Activity Behaviors and Patterns

At recruitment, physical activity levels were assessed by the validated short self-reported International Physical Activity Questionnaire using a touch-screening questionnaire, focusing on 4 major categories including leisure-time activities, housework-related activities, job-related activities, and daily transportation-related activities (detailed questions are shown in Supplementary Table 2). Specifically, leisure-time activities included strenuous sports (ie, sports that make you sweat or breathe hard), other exercises, and walking for pleasure. Housework-related activities referred to conditions of heavy (eg, weeding, lawn mowing, carpentry, and digging) and light do-it-yourself (DIY; eg, pruning and watering the lawn). For these categories, participants reported their attendance (yes or no), frequency (days/week), and duration (minutes/day). Items of job-related activities asked if the participant’s job involved mainly walking/standing, heavy manual/physical work, or shift work, as well as length/frequency of working per week. Daily transportation-related activities included the modes of transportations (car/motor vehicle, cycle, walk, or public transportation), and the total hours of driving. For each item, participants who chose “prefer not to answer” and “do not know” were categorized into the “unknown” group.

We then studied patterns of physical activity in the study population by taking into account the correlation and interplay between the above physical activity items, using the principal component analysis (PCA) (25) with “promax” rotation based on data of 154 787 participants who had no missing data on any of the 25 items. Through this analysis, we aimed to identify distinct physical activity patterns of major relevance, representing 8 principal components (PCs) with eigenvalue >1.0 (accounting for in total 73.1% of variance, Supplementary Figure 1). According to variables with top weights (ie, loading values) in each PC, these components could be named as patterns predominated by strenuous sports, other exercises, heavy DIY, light DIY, walking for pleasure, high-intensity work, public transport, and time spent on working (Supplementary Figure 1). We then calculated PC scores by integrating all included variables weighted by their factor loadings and determined the level of exposure to each pattern (ie, each PC) according to the tertile distribution of the corresponding PC score (ie, from low to high: <1st tertile, 1st–2nd tertile, and >2nd tertile).

Besides self-reported physical activity, we extracted data on objectively measured physical activity among a subsample of UK Biobank participants who joined a follow-up survey with physical activity recorded via a wrist-worn accelerometer between 2013 and 2015. A validation study revealed that the total volume of accelerometer-assessed physical activity explained 44%–47% of the variance in activity energy expenditure (26).

Indexes of Aging Acceleration

The primary measure of aging acceleration in the present study was LTL deviation, which was quantified in 2 steps (see formulas in Supplementary Methods). First, we derived the residual (ie, the difference between predicted and observed LTL) from a linear model on LTL measured at baseline, integrating the impact of age, sex, and polygenetic risk score (PRS) for LTL. Second, the residual was standardized by the predicted value of LTL. Thus, the generated LTL deviation can be used to determine the presence of aging acceleration (ie, LTL deviation >0), with a unit of increase representing a 1% increase in biological aging at a individual level. Specifically, the PRS, as a proxy of genetically determined LTL, was computed using penalized regression (Least Absolute Shrinkage and Selection Operator, LASSO) (20) based on publicly available summary statistics of Genome-Wide Association Studies from independent samples (20), using an analytic data set containing 5 067 039 single-nucleotide polymorphisms for the 284 479 eligible UK Biobank participants. As a result, the LTL deviation, generated after the consideration of both demographic and genetic determinants of LTL, stands for nongenetically determined LTL potentially altered by nongenetic factors such as physical activity.

As an alternative index of aging acceleration, the BAA was also assessed, defined as the difference (in years) between the phenotypic age (ie, a measure of biological aging estimated by 9 clinical chemistry biomarkers, see Supplementary Methods) of a person and the average phenotypic age in the cohort of individuals of the same age and sex (27). With easy accessibilities of those involved biomarkers, BAA has been widely used in previous studies (28,29), with a demonstrated high consistency with healthspan-related characteristics (29) and satisfactory accuracy for predicting mortality hazards (27). For both indexes, a positive value (>0) indicated the presence of aging acceleration (ie, biologically older than expected), while a negative value indicated the presence of biological fitness (ie, biologically younger than expected).

Covariates

Data on demographic (eg, birth year, sex, and ethnicity), socioeconomic (eg, annual household income), and lifestyle (eg, alcohol drinking and smoking status) factors were collected through a questionnaire, whereas anthropometric variables were measured at baseline assessment center visit. Body mass index (BMI) was calculated from height and weight. We calculated the Townsend deprivation index (TDI), as a proxy of area-based deprivation (30). To study the potential modifying effect of mental health on the interested associations, we obtained information on the history of psychiatric disorder, defined as a diagnosis of any psychiatric disorder (see ICD-10 codes in Supplementary Table 1) based on self-report, inpatient hospital, or primary care data before recruitment.

Statistical Analysis

We estimated the associations of 25 individual items of physical activity, as well as 8 physical activity patterns derived from PCA, with both indexes of aging acceleration based on linear (for LTL deviation and BAA as a continuous variables) and logistic (for the presence of aging acceleration [LTL deviation > 0 or BAA > 0], yes or no) regression models. The models were partially or fully adjusted for age (as continuous variable), sex (male or female), smoking and drinking status (never, previous, current, or unknown), annual household income (<£18 000, £18 000–30 999, £31 000–51 999, £52 000–100 000, >£100 000, or unknown), and TDI (as a continuous variable). In the analyses of physical activity patterns, we also stratified the analysis by sex (male and female), history of psychiatric disorder (yes or no), and level of PRS for LTL (<1st tertile, 1st–2nd tertile, and >2nd tertile).

To test if the association between physical activity and aging acceleration could be modified by BMI, we additionally adjusted for or stratified the analysis by BMI (<25, 25–30, or ≥30.0 kg/m²) in a sensitivity analysis. Furthermore, there might be a concern of collinearity between items related to attendance, frequency, and duration of the same activity mode in the PCA, which might preclude the identification of PCs reflecting the complex interplays between items of different activity modes. We, therefore, repeated the PCA using only items related to attendance for each category (see items of “attendance” in Figure 2). Last, to test the robustness of results to the self-reported physical activity, we repeated the analyses among participants with available data on accelerometer-assessed physical activity (measured as average vector magnitude in milli-gravity units, as a continuous or categorical variable) and LTL deviation (n = 62 418) or BAA (n = 52 193). A 2-sided p < .05 was considered statistically significant. All analyses were done using R software, version 4.0.2.

Figure 2. — Association between different physical activity items and the presence of aging acceleration.^a This figure illustrates the association between different physical activity-related items and the presence of aging acceleration (A, by LTL deviation; B, by BAA) using heatmap. The horizontal gradient bar presents point estimations of ORs, ranging from 0.7 to 1.3. LTL = leukocyte telomere length; BAA = biological age acceleration; OR = odds ratio. ^aThe case experiencing aging acceleration was defined when an individual had LTL deviation>0 (A) or BAA > 0 (B). ^bLTL deviation (in %) was calculated in 2 steps. First, we derived the residual (ie, the difference between observed and predicted values) from a linear model integrating impact of age, sex, and polygenetic risk score (PRS) for LTL on LTL measured at baseline. Second, the residual was standardized by the predicted value of LTL (ie, LTL deviation = −residuals/predicted LTL × 100). ^cBAA was defined as the difference (in years) between the phenotypic age (ie, a measure of biological aging estimated by 9 clinical chemistry biomarkers) of a person and the average phenotypic age in the cohort of individuals of the same age and sex. ^dORs were derived from logistic regression models, adjusted for age, sex, smoking and drinking status, annual household income, and Townsend deprivation index. The reference group included the individuals with the absence of studied exposure (for binary variables), or those with the lowest exposure level (for ordinal variables). ^eNot attendance: Participant travelling from home to workplace less than once a week. ^fNot attendance: Participant spending less than 28 h/wk (the first quartile) at work.

Results

A total of 284 479 participants were included in the analyses of the 25 items of physical activity, with a mean age at recruitment of 56.60 years and a sex ratio (male:female) of 0.82:1 (Table 1). Fifty-three percent(150 784) participants were considered to have an aging acceleration based on the presence of LTL deviation >0. We found that all baseline characteristics were largely comparable between participants with and without such aging acceleration.

Table 1.

Characteristics of the Qualified Study Participants at Recruitment

Characteristics	Total, N = 284 479	LTL deviation^* > 0, N = 150 784	LTL deviation^* ≤ 0, N = 133 695
Age at baseline (y), mean (SD)	56.60 (7.94)	56.57 (7.98)	56.65 (7.88)
Sex, n (%)
Female	155 642 (0.55)	82 433 (0.55)	73 209 (0.55)
Male	128 837 (0.45)	68 351 (0.45)	60 486 (0.45)
Townsend deprivation index, mean (SD)	−1.66 (2.87)	−1.62 (2.89)	−1.70 (2.85)
Average household income (%)
<£18 000	49 495 (0.17)	26 892 (0.18)	22 603 (0.17)
£18 000–£30 999	62 975 (0.22)	33 736 (0.22)	29 239 (0.22)
£31 000–£51 999	67 084 (0.24)	35 258 (0.23)	31 826 (0.24)
£52 000–£100 000	52 724 (0.19)	27 402 (0.18)	25 322 (0.19)
>£100 000	13 748 (0.05)	6 928 (0.05)	6 820 (0.05)
Unknown	38 453 (0.14)	20 568 (0.14)	17 885 (0.13)
Education level (%)
College or university degree	91 160 (0.32)	46 430 (0.31)	44 730 (0.33)
A levels/AS levels or equivalent	32 738 (0.12)	16 984 (0.11)	15 754 (0.12)
O levels/GCSEs or equivalent	64 020 (0.23)	34 323 (0.23)	29 697 (0.22)
CSEs or equivalent	15 920 (0.06)	8 814 (0.06)	7 106 (0.05)
NVQ or HND or HNC or equivalent	18 459 (0.06)	10 047 (0.07)	8 412 (0.06)
Other professional qualifications	14 416 (0.05)	7 620 (0.05)	6 796 (0.05)
Unknown	47 766 (0.17)	26 566 (0.18)	21 200 (0.16)
Body mass index, kg/m²(%)
<25	94 688 (0.33)	48 964 (0.32)	45 724 (0.34)
25–30	121 498 (0.43)	64 367 (0.43)	57 131 (0.43)
≥30	64 088 (0.23)	35 176 (0.23)	28 912 (0.22)
Unknown	4 205 (0.01)	2 277 (0.02)	1 928 (0.01)
Smoking status (%)
Never	159 548 (0.56)	83 431 (0.55)	76 117 (0.57)
Previous	96 442 (0.34)	51 230 (0.34)	45 212 (0.34)
Current	27 584 (0.10)	15 627 (0.10)	11 957 (0.09)
Unknown	905 (0.00)	496 (0.00)	409 (0.00)
Alcohol status (%)
Never	8 276 (0.03)	4 390 (0.03)	3 886 (0.03)
Previous	8 399 (0.030	4 449 (0.03)	3 950 (0.03)
Current	267 584 (0.94)	141 825 (0.94)	125 759 (0.94)
Unknown	220 (0.00)	120 (0.00)	100 (0.00)
History of psychiatric disorders (%)
No	250 693 (0.88)	132 257 (0.88)	118 436 (0.89)
Yes	33 786 (0.12)	18 257 (0.12)	15 259 (0.11)

Open in a new tab

Notes: LTL = leukocyte telomere length; GCEs = General certificate of secondary educations; CSEs = Certificate of Secondary Educations; NVQ = National Vocational Qualification; HND = Higher National Diploma; HNC = Higher National Certificate.

*LTL deviation (in %) was calculated by 2 steps: first, we derived the residual (ie, the difference between observed and predicted values of data) from a linear model integrating impact of age, sex, and polygenetic risk score (PRS) for LTL on LTL measured at baseline. Then, the residual was standardized by the predicted value of LTL (ie, LTL deviation = −residuals/predicted LTL × 100).

Analyses on the associations between items of physical activity and the indexes of aging acceleration revealed that all items related to leisure-time activities and housework-related activities were associated with a lower value of LTL deviation as well as a lower risk of having LTL deviation >0 (Supplementary Table 3 and Figure 2A). The strongest associations were noted for attendance of strenuous sports (β = −0.45 [95% confidence interval {CI} −0.63, −0.27] for LTL deviation and odds ratio [OR] = 0.95 [95% CI 0.93, –0.97] for LTL deviation >0). Conversely, most job-related activities were associated with a higher value of LTL deviation and a higher risk of aging acceleration. We observed identical result patterns for BAA, with generally greater point estimates (Figure 2B and Supplementary Table 3).

The analyses of derived physical activity patterns by PCA among 154 787 participants with complete physical activity data consolidated the above findings, in both partially adjusted and fully adjusted models (Supplementary Table 4). In brief, a higher level of adherence to physical activity patterns predominated by leisure-time activities (ie, strenuous sports, other exercises, and walking for pleasure), housework-related activities (ie, heavy and light DIY), or use of public transportation was associated with a lower risk of aging acceleration (ORs ranged 0.92–0.98 for LTL deviation >0 and 0.83–0.95 for BAA >0, Figure 3), whereas a higher level of adherence to patterns predominated by job-related activities (high-intensity work and time spent working) was associated with a higher risk of aging acceleration (ORs ranged 1.02–1.06 for LTL deviation >0 and 1.06–1.12 for BAA >0, Figure 3). We observed sex disparity in these associations, with the associations being stronger among females than males (Table 2). In addition, subgroup analyses indicated the associations between physical activity patterns and aging acceleration were not modified by the history of psychiatric disorders (for both LTL deviation and BAA, Supplementary Table 5) or level of PRS for LTL (for LTL deviation, Supplementary Table 6).

Figure 3. — Associations between identified physical activity patterns and the indexes of aging acceleration. The forest plot shows the associations between identified physical activity patterns and aging acceleration indexed by 2 different methods and physical activity patterns and aging acceleration indexed by 2 different methods and physical. LTL = leukocyte telomere length; BAA = biological age acceleration; OR = odds ratio. ^a LTL deviation (in %) was calculated in 2 steps. First, we derived the residual (ie, the difference between observed and predicted values of data) from a linear model integrating impact of age, sex, and polygenetic risk score (PRS) for LTL on LTL measured at baseline. Second, the residual was standardized by the predicted value of LTL (ie, LTL deviation = −residuals/predicted LTL × 100). ^bBAA was defined as the difference (in years) between the phenotypic age (ie, a measure of biological aging estimated by 9 clinical chemistry biomarkers) of a person and the average phenotypic age in the cohort of individuals of the same age and sex. ^cβ (95%CIs) were derived from linear regression models, adjusted for age, sex, smoking and drinking status, annual household income, and Townsend deprivation index. ^dORs were derived from logistic regression models, adjusted for age, sex, smoking and drinking status, annual household income, and Townsend deprivation index. The reference group included the individuals with the absence of studied exposure (for binary variables), or those with the lowest exposure level (for ordinal variables).

Table 2.

Associations Between Identified Physical Activity Patterns and the Indexes of Aging Acceleration, Subgrouped by Sex

Principal components	Female				Male
	LTL deviation^* (N = 79 342)		BAA^† (N = 66 488)		LTL deviation^* (N = 75 445)		BAA^† (N = 64 761)
	As continuous variable, β (95% CI)^‡	As binary variable (>0 vs ≤0), OR (95% CI) ^§	As continuous variable, β (95% CI) ^‡	As binary variable (>0 vs ≤0), OR (95% CI)^§	As continuous variable, β (95% CI)^‡	As binary variable (>0 vs ≤0), OR (95% CI)^§	As continuous variable, β (95% CI)^c	As binary variable (>0 vs ≤0), OR (95% CI)^d
PC1_predominated by “strenuous sports”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	−0.15 (−0.40, 0.09)	0.97 (0.94, 1.01)	−0.13 (−0.21, −0.05)	0.97 (0.93, 1.00)	−0.12 (−0.39, 0.16)	0.99 (0.96, 1.03)	−0.15 (−0.24, −0.07)	0.94 (0.90, 0.98)
>2nd tertile	−0.57 (−0.84, −0.30)	0.93 (0.90, 0.97)	−0.39 (−0.48, −0.30)	0.88 (0.84, 0.92)	−0.32 (−0.58, −0.06)	0.98 (0.95, 1.02)	−0.52 (−0.60, −0.44)	0.81 (0.78, 0.85)
PC2_predominated by “other exercises”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	−0.29 (−0.55, −0.03)	0.98 (0.95, 1.02)	−0.25 (−0.34, −0.17)	0.91 (0.88, 0.95)	−0.18 (−0.43, 0.08)	0.98 (0.95, 1.02)	−0.37 (−0.45, −0.29)	0.86 (0.83, 0.90)
>2nd tertile	−0.71 (−0.98, −0.44)	0.93 (0.89, 0.96)	−0.44 (−0.53, −0.35)	0.85 (0.82, 0.89)	−0.21 (−0.46, 0.05)	0.99 (0.95, 1.03)	−0.52 (−0.60, −0.44)	0.81 (0.78, 0.84)
PC3_predominated by “Heavy DIY”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	−0.61 (−0.85, −0.37)	0.93 (0.90, 0.96)	−0.25 (−0.33, −0.17)	0.91 (0.88, 0.95)	−0.02 (−0.30, 0.26)	0.99 (0.95, 1.03)	−0.29 (−0.38, −0.21)	0.91 (0.87, 0.95)
>2nd tertile	−0.39 (−0.66, −0.12)	0.96 (0.92, 0.99)	−0.23 (−0.32, −0.14)	0.93 (0.89, 0.97)	0.03 (−0.24, 0.29)	1.00 (0.96, 1.04)	−0.25 (−0.33, −0.17)	0.91 (0.87, 0.95)
PC4_predominated by “Light DIY”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	−0.41 (−0.67, −0.15)	0.95 (0.92, 0.99)	0.00 (−0.09, 0.08)	1.00 (0.96, 1.04)	−0.27 (−0.52, −0.01)	0.97 (0.94, 1.01)	−0.15 (−0.23, −0.07)	0.96 (0.92, 1.00)
>2nd tertile	−0.47 (−0.73, −0.20)	0.94 (0.91, 0.98)	−0.05 (−0.13, 0.04)	0.98 (0.94, 1.02)	−0.10 (−0.35, 0.15)	0.98 (0.95, 1.02)	−0.19 (−0.27, −0.12)	0.92 (0.89, 0.96)
PC5_predominated by “Walking for pleasure”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	−0.17 (−0.43, 0.10)	0.99 (0.96, 1.03)	−0.25 (−0.34, −0.17)	0.90 (0.86, 0.94)	−0.20 (−0.45, 0.05)	0.97 (0.94, 1.01)	−0.20 (−0.27, −0.12)	0.91 (0.87, 0.95)
>2nd tertile	−0.27 (−0.53, −0.01)	0.98 (0.95, 1.02)	−0.43 (−0.52, −0.35)	0.86 (0.82, 0.89)	−0.31 (−0.56, −0.06)	0.97 (0.94, 1.01)	−0.36 (−0.44, −0.29)	0.85 (0.82, 0.89)
PC6_predominated by “High-intensity work”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	0.05 (−0.20, 0.30)	0.98 (0.95, 1.02)	−0.05 (−0.13, 0.03)	1.02 (0.98, 1.06)	0.29 (0.03, 0.55)	1.04 (1.00, 1.08)	0.02 (−0.06, 0.10)	1.00 (0.96, 1.05)
>2nd tertile	0.45 (0.18, 0.72)	1.03 (0.99, 1.07)	0.05 (−0.04, 0.13)	1.04 (0.99, 1.08)	0.88 (0.61, 1.16)	1.10 (1.06, 1.15)	0.18 (0.10, 0.27)	1.09 (1.04, 1.14)
PC7_predominated by “Public transportation”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	−0.32 (−0.58, −0.06)	0.96 (0.92, 0.99)	−0.11 (−0.20, −0.03)	0.98 (0.94, 1.02)	0.02 (−0.24, 0.27)	1.00 (0.97, 1.04)	−0.24 (−0.32, −0.16)	0.91 (0.87, 0.95)
>2nd tertile	−0.71 (−0.98, −0.45)	0.91 (0.87, 0.94)	−0.18 (−0.27, −0.09)	0.94 (0.90, 0.98)	−0.54 (−0.79, −0.29)	0.94 (0.91, 0.98)	−0.34 (−0.41, −0.26)	0.89 (0.85, 0.93)
PC8_predominated by “Time spend working”
≤1st tertile	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
1st–2nd tertile	0.21 (−0.03, 0.45)	1.03 (1.00, 1.06)	0.25 (0.17, 0.33)	1.09 (1.05, 1.14)	0.22 (−0.09, 0.52)	1.02 (0.97, 1.06)	0.26 (0.17, 0.36)	1.11 (1.05, 1.16)
>2nd tertile	−0.12 (−0.41, 0.18)	0.99 (0.95, 1.03)	0.20 (0.11, 0.30)	1.09 (1.04, 1.14)	0.34 (0.05, 0.63)	1.04 (1.00, 1.08)	0.34 (0.25, 0.43)	1.15 (1.10, 1.21)

Open in a new tab

Notes: Sex of participant was acquired from central registry at recruitment, which contain a mixture of the sex the national health service had recorded for the participant and self-reported sex. N = The total number of participants according to different sex and different aging indexes; BAA = biological age acceleration; LTL = leukocyte telomere length; OR = odds ratio.

^†BAA was defined as the difference (in years) between the phenotypic age (ie, a measure of biological aging estimated by 9 clinical chemistry biomarkers) of a person and the average phenotypic age in the cohort of individuals of the same age and sex.

^‡β (95% CIs) were derived from linear regression models, adjusted for age, smoking and drinking status, annual household income, and Townsend deprivation index.

^§ORs were derived from logistic regression models, adjusted for age, smoking and drinking status, annual household income, and Townsend deprivation index. The reference group was the individuals with the absence of studied exposure items (for binary items), or those with the lowest exposure level (for ordinal items).

These results attenuated but remained statistically significant after additional adjustment for BMI measured at baseline (Supplementary Table 7). In the stratified analysis by BMI, we observed slightly stronger associations of physical activity patterns with LTL deviation among individuals with lower BMI (eg, when comparing above second to below first tertile of activity pattern predominated by other exercises, β = −0.70 [95% CI −1.02, −0.38] was among individuals with BMI < 25, whereas −0.26 [95% CI −0.66, 0.15] among individuals with BMI ≥ 30.0 kg/m², respectively; Supplementary Table 8). This was, however, not the case for BAA (eg, the corresponding β = −0.08 [95% CI −0.18, −0.02]) and −0.52 ([95% CI −0.65, −0.39], respectively; Supplementary Table 8).

The PCA using 15 binary items (yes or no, for the attendance of different activity modes) identified 6 activity patterns. In contrast to the results of the original PCA, PCs predominated by “strenuous sports” and “other exercise” were combined as one PC, PCs predominated by “Heavy DIY” and “Light DIY” were combined as one PC, while the other PCs remained similar (Supplementary Figure 2). Importantly, the analyses on the associations of these newly identified patterns with the indexes of aging acceleration revealed identical results, implying negative associations of patterns related to leisure-time activity (ie, “strenuous sports and other exercise,” “heavy/light DIY,” and “walking for pleasure”), but positive associations of work-related activity patterns (ie, “high-intensity work” and “time spent working”), with aging acceleration (Supplementary Table 9). Restricting the analysis to individuals with objectively measured physical activity, we detected a linear association between increasing level of physical activity (ie, the overall acceleration average, as a continuous variable) with decreasing aging acceleration (β = −0.15 [95% CI −0.27, −0.03] for LTL deviation and β = −0.34 [95% CI −0.37, −0.30] for BAA, Supplementary Table 10). The analysis of the level of physical activity intensity (according to tertile distribution) revealed a U-shaped relationship, with the reduced aging acceleration observed for individuals with moderate intensity of physical activity, compared with those with the lowest or highest intensity of physical activity. Analyses on the presence of aging acceleration (ie, LTL deviation > 0 or BAA > 0) showed consistent result patterns (Supplementary Table 10).

Discussion

Our study on about 284 000 participants of the community-based UK Biobank cohort indicated that multiple types of physical activity were associated with a reduced rate of biological aging, as well as a decreased risk of experiencing aging acceleration, particularly among females. In addition, the analyses of identified physical activity patterns through PCA, which enabled the consideration of the complexity and interplays between individual items of physical activity, highlighted a possible protective role of several specific physical activities, including strenuous sports, other exercises, and use of public transportation, on aging as a consistent link was noted between a higher adherence level of these activities to a lower value of aging acceleration indexes (ie, LTL deviation and BAA). Furthermore, the findings based on self-reported physical activity were consolidated using data on accelerometer-measured physical activity showing a U-shaped association between physical activity and aging acceleration, underscoring the necessity of engaging in physical activity in leisure time of daily life, in terms of decelerating aging process and maintaining healthy aging. Importantly, as the primary measure of aging acceleration in the present study, that is, LTL deviation, represents by definition the nongenetically determined LTL shortening, our findings demonstrate that the beneficial effect of physical activity on longevity might be partly explained by its key influence on the telomere attrition process.

Despite their methodological diversity (eg, the difference in physical activity measurements and study designs), previous investigations have demonstrated that physical activity is associated with a wide range of health benefits, including reduced risk of numerous diseases (31,32) and all-cause as well as cause-specific mortality (12). For instance, a US prospective cohort of 11 351 individuals aged 45–64 years with repeated self-reported data on physical activity, reported that both maintaining recommended activity levels and initiating or increasing physical activity levels during later life were associated with a lower risk of developing heart failure (31). Evidence from randomized controlled trials also supported the role of regular aerobic exercise in improving cognition (33), memory function, and age-related brain atrophy (34). Furthermore, the dose-response relationship between leisure-time physical activity and mortality was studied in a pooled analysis of 6 population-based prospective cohorts in United States and Europe using self-reported data, indicating a benefit threshold at approximately 3–5 times the recommended leisure-time physical activity meeting the 2008 Physical Activity Guidelines for Americans minimum (35). As all these observations showed a consistent link between physical activity and various aging-related health conditions, it is plausible that physical activity might also impact the human aging process in general. However, the existing evidence for such is limited. Despite the small sample size (approximately 500) and short follow-up period (1 year), clinical trials conducted among high-school students suggested a positive impact of weekly physical activity on biological age, measured by a set of biomarkers (36,37). With regard to biological aging measured by TL, a recent systematic review meta-analyzed data from 11 studies including 2 210 individuals, and concluded that the evidence supporting a role of physical activity on TL was weak mainly due to the high risk of publication bias, different study populations, differences in physical activity and TL measurements, and insufficient control for important confounders such as comorbidities, other lifestyle factors, and genetic determinants of TL (38). Exercise intervention studies (with 20–200 participants) were also considered to have very low-quality and controversial findings. Some revealed that TL did not differ between exercise intervention and control groups (38), whereas 1 randomized controlled trial showed that endurance and interval training protocol could increase telomerase activity and TL (39).

Based on a large-scale community-based cohort, the results of the present study are the first to significantly advance the knowledge in this area by clearly demonstrating a protective role of physical activity, measured subjectively and objectively, on biological aging in general, and telomere-driven aging specifically. The major strength of the study is nested in its large sample size (approximately 284 000 participants), the meticulous effort to distinguish the effect of distinct physical activity behaviors and patterns, the consistent findings for 2 different measures of aging acceleration (ie, nongenetically determined LTL and BAA), and the mutually-verified results derived from self-reported and objectively measured data on physical activity. Our novel finding that physical activity might directly modulate the noninherited telomere shortening, independent of BMI and the presence of previous psychiatric disorders, gains support from prior animal and human experimental studies, which show that persistent exercise might directly function on telomere preservation through the increasing activity of telomerase (ie, an enzyme that adds nucleotides to the telomeric ends), upregulating telomere-stabilizing proteins, and reducing apoptosis and cell-cycle arrest (40,41). In addition, because the inevitable telomere shortening could be hastened by other biological processes and components, such as inflammation, oxidative stress (42), and reduction of skeletal muscle satellite cells (43), it is also possible that physical activity indirectly protects telomeres by inducing antioxidant activity, improving inflammatory balance, and increasing the number of satellite cells (43).

In addition, our finding of positive associations between the highest exposure level to job-related activities and aging acceleration, independently of leisure-time physical activities, is consistent with previous reports (44). For instance, 1 Finnish cohort study of 604 participants indicated an association between high-strain jobs (eg, blue-collar jobs in comparison to white-collar jobs), as well as long working hours (ie, more than 40 hours per week), with accelerated epigenetic aging (45). The possible explanations of these results include work-related psychological stress (46), unfavorable lifestyles (eg, higher prevalence of sleep problems (47)), as well as adverse biological alterations (eg, worse cardiometabolic and inflammatory profile (48)) among high-strain or high-loading workers.

One of the limitations of the present study is the cross-sectional nature of the analyses, namely that both physical activity and LTL deviation and BAA were assessed using data collected at the same time (recruitment). The lack of clear temporal order limited our ability to explore the causality of the observed associations. Also, physical activity data with detailed information on type, duration, and frequency were collected merely at recruitment, potentially influenced by recall bias and might not reflect the longitudinal pattern. In addition, although the prior study showed an acceptable consistency between LTL measured at baseline and after an average of 5 years among 1 351 participants of the UK Biobank with paired samples (23), future studies, with ideally dynamic surveillance on both physical activity and LTL deviation over time, are warranted to provide a more accurate assessment on the effect of physical activity on TL shortening and possible contributions to causal inference. Besides, the calculation of other aging measurements, such as the DNAm PhenoAge as a powerful epigenetic index with the potentially superior capability of capturing risks for a variety of aging outcomes (49), was not feasible (ie, due to the absence of data on DNA methylation), underscoring the necessity of further mechanistic explorations, for validating our findings. Last, it is also worth noting that the UK Biobank participants are not representative of the U.K. general population (50). The generalization of these findings to the whole United Kingdom or other populations needs to be done with caution or further validation.

Our study based on data from 284 000 UK Biobank participants clearly demonstrated that physical activity, such as strenuous sports and other exercises in leisure time and use of public transportation, was associated with a reduced biological aging. Besides highlighting the importance of engaging in physical activity among the general population for healthy aging, our results provide the strongest evidence to date on the beneficial effect of physical activity on the telomere attrition process, after taking into consideration the genetically determined LTL shortening.

Supplementary Material

glad064_suppl_Supplementary_Material

Click here for additional data file.^{(1.6MB, pdf)}

Acknowledgments

We express our sincere thanks to all the sponsors and team members involved in West China Biomedical Big Data Center and Med-X Center for Informatics, Sichuan University. This research was done using the UK Biobank Resource under Application 54803.

Contributor Information

Jianwei Zhu, Department of Orthopedic Surgery and West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China.

Yao Yang, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China; Med-X Center for Informatics, Sichuan University, Chengdu, China.

Yu Zeng, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China; Med-X Center for Informatics, Sichuan University, Chengdu, China.

Xin Han, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China; Med-X Center for Informatics, Sichuan University, Chengdu, China.

Wenwen Chen, Division of Nephrology, Kidney Research Institute, State Key Laboratory of Biotherapy and Cancer Center, West China Hospital, Sichuan University, Chengdu, China.

Yao Hu, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China; Med-X Center for Informatics, Sichuan University, Chengdu, China.

Yuanyuan Qu, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China; Med-X Center for Informatics, Sichuan University, Chengdu, China.

Huazhen Yang, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China; Med-X Center for Informatics, Sichuan University, Chengdu, China.

Unnur A Valdimarsdóttir, Center of Public Health Sciences, Faculty of Medicine, University of Iceland, Reykjavík, Iceland; Institute of Environmental Medicine, Karolinska Institute, Stockholm, Sweden.

Fang Fang, Institute of Environmental Medicine, Karolinska Institute, Stockholm, Sweden.

Huan Song, West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China; Med-X Center for Informatics, Sichuan University, Chengdu, China.

Funding

This work was supported by 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan University (grant no. ZYYC21005 to H.S.), and by the National Natural Science Foundation of China (81971262 to H.S.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflict of Interest

All authors listed on a manuscript considered for publication have approved its submission. All authors have completed the Unified Competing Interest form (available on request from the corresponding author) and declare: no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years, no other relationships or activities that could appear to have influenced the submitted work.

Author Contributions

H.S., J.Z., and Y.Y. are responsible for the study’s concept and design. H.Y., Y.H., Y.Z., and Y.Q. did the data and project management. Y.Y., Y.Z., W.C., and X.H. did the data cleaning and analysis. J.W., Y.Y., X.H., F.F., U.A.V., and H.S. interpreted the data. J.Z., Y.Y., Y.Z., F.F., U.A.V., and H.S. drafted the manuscript. All the authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

Data Availability

Data from UK Biobank are available per the researchers request (https://www.ukbiobank.ac.uk/enable-your-research/apply-for-access).

Ethical Approval and Consent to Participate

The UK Biobank has full ethical approval from the NHS National Research Ethics Service (16/NW/0274) and informed consent were obtained before data collection from each participant. The present study was also approved by the biomedical research ethics committee of West China Hospital (reference number: 2019-1171).

References

1.US Department of Health and Human Services. Physical Activity Guidelines for Americans. 2nd ed. US Department of Health and Human Services; 2018. [Google Scholar]
2. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012;380(9838):219–229. doi: 10.1016/s0140-6736(12)61031-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Lavie CJ, Ozemek C, Carbone S, Katzmarzyk PT, Blair SN. Sedentary behavior, exercise, and cardiovascular health. Circ Res. 2019;124(5):799–815. doi: 10.1161/circresaha.118.312669 [DOI] [PubMed] [Google Scholar]
4. Egan B, Zierath JR. Exercise metabolism and the molecular regulation of skeletal muscle adaptation. Cell Metab. 2013;17(2):162–184. doi: 10.1016/j.cmet.2012.12.012 [DOI] [PubMed] [Google Scholar]
5. Hannukainen JC, Borra R, Linderborg K, et al. Liver and pancreatic fat content and metabolism in healthy monozygotic twins with discordant physical activity. J Hepatol. 2011;54(3):545–552. doi: 10.1016/j.jhep.2010.07.029 [DOI] [PubMed] [Google Scholar]
6. Stefan N. Causes, consequences, and treatment of metabolically unhealthy fat distribution. Lancet Diabetes Endocrinol. 2020;8(7):616–627. doi: 10.1016/S2213-8587(20)30110-8 [DOI] [PubMed] [Google Scholar]
7. Church TS, Blair SN, Cocreham S, et al. Effects of aerobic and resistance training on hemoglobin A1c levels in patients with type 2 diabetes: a randomized controlled trial. JAMA. 2010;304(20):2253–2262. doi: 10.1001/jama.2010.1710 [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Zelko A, Bukova A, Kolarcik P, et al. A randomized controlled trial to evaluate utilization of physical activity recommendations among patients of cardiovascular healthcare centres in Eastern Slovakia: study design and rationale of the AWATAR study. BMC Public Health. 2018;18(1):454. doi: 10.1186/s12889-018-5349-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Zhao G, Li C, Ford ES, et al. Leisure-time aerobic physical activity, muscle-strengthening activity and mortality risks among US adults: the NHANES linked mortality study. Br J Sports Med. 2014;48(3):244–249. doi: 10.1136/bjsports-2013-092731 [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Zhao MV, Magnussen CG, Xi B. Recommended physical activity and all cause and cause specific mortality in US adults: prospective cohort study. BMJ. 2020;370:m2031. doi: 10.1136/bmj.m2031 [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Schoenborn CA, Stommel M. Adherence to the 2008 adult physical activity guidelines and mortality risk. Am J Prev Med. 2011;40(5):514–521. doi: 10.1016/j.amepre.2010.12.029 [DOI] [PubMed] [Google Scholar]
12. Kamada M, Shiroma EJ, Buring JE, Miyachi M, Lee IM. Strength training and all-cause, cardiovascular disease, and cancer mortality in older women: a cohort study. J Am Heart Assoc. 2017;6(11):e007677. doi: 10.1161/JAHA.117.007677 [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Liu Z, Kuo PL, Horvath S, Crimmins E, Ferrucci L, Levine M. A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: a cohort study. PLoS Med. 2018;15(12):e1002718. doi: 10.1371/journal.pmed.1002718 [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Wu JW, Yaqub A, Ma Y, et al. Biological age in healthy elderly predicts aging-related diseases including dementia. Sci Rep. 2021;11(1):15929. doi: 10.1038/s41598-021-95425-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Lopez-Otin C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging. Cell. 2013;153(6):1194–1217. doi: 10.1016/j.cell.2013.05.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Shay JW. Role of telomeres and telomerase in aging and cancer. Cancer Discov. 2016;6(6):584–593. doi: 10.1158/2159-8290.CD-16-0062 [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Sanders JL, Newman AB. Telomere length in epidemiology: a biomarker of aging, age-related disease, both, or neither? Epidemiol Rev. 2013;35:112–131. doi: 10.1093/epirev/mxs008 [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Mons U, Muezzinler A, Schottker B, et al. Leukocyte telomere length and all-cause, cardiovascular disease, and cancer mortality: results from individual-participant-data meta-analysis of 2 large prospective cohort studies. Am J Epidemiol. 2017;185(12):1317–1326. doi: 10.1093/aje/kww210 [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Bischoff C, Petersen HC, Graakjaer J, et al. No association between telomere length and survival among the elderly and oldest old. Epidemiology. 2006;17(2):190–194. doi: 10.1097/01.ede.0000199436.55248.10 [DOI] [PubMed] [Google Scholar]
20. Li C, Stoma S, Lotta LA, et al. Genome-wide association analysis in humans links nucleotide metabolism to leukocyte telomere length. Am J Hum Genet. 2020;106(3):389–404. doi: 10.1016/j.ajhg.2020.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Biobank U. Data Providers and Dates of Data Availability . https://biobank.ctsu.ox.ac.uk/crystal/exinfo.cgi?src=Data_providers_and_dates
22. Welsh S, Peakman T, Sheard S, Almond R. Comparison of DNA quantification methodology used in the DNA extraction protocol for the UK Biobank cohort. BMC Genomics. 2017;18(1):26. doi: 10.1186/s12864-016-3391-x [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Codd V, Denniff M, Swinfield C, et al. Measurement and initial characterization of leukocyte telomere length in 474,074 participants in UK Biobank. Nat Aging. 2022;2(2):170–179. doi: 10.1038/s43587-021-00166-9 [DOI] [PubMed] [Google Scholar]
24. Thygesen SK, Christiansen CF, Christensen S, Lash TL, Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson Comorbidity Index conditions in the population-based Danish National Registry of Patients. BMC Med Res Methodol. 2011;11:83. doi: 10.1186/1471-2288-11-83 [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Jannasch F, Riordan F, Andersen LF, Schulze MB. Exploratory dietary patterns: a systematic review of methods applied in pan-European studies and of validation studies. Br J Nutr. 2018;120(6):601–611. doi: 10.1017/s0007114518001800 [DOI] [PMC free article] [PubMed] [Google Scholar]
26. White T, Westgate K, Hollidge S, et al. Estimating energy expenditure from wrist and thigh accelerometry in free-living adults: a doubly labelled water study. Int J Obes (Lond). 2019;43(11):2333–2342. doi: 10.1038/s41366-019-0352-x [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Ying K, Zhai R, Pyrkov TV, et al. Genetic and phenotypic analysis of the causal relationship between aging and COVID-19. Commun Med. 2021;1(1):35. doi: 10.1038/s43856-021-00033-z [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Parker DC, Bartlett BN, Cohen HJ, et al. Association of blood chemistry quantifications of biological aging with disability and mortality in older adults. J Gerontol A Biol Sci Med Sci. 2020;75(9):1671–1679. doi: 10.1093/gerona/glz219 [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Hastings WJ, Shalev I, Belsky DW. Comparability of biological aging measures in the National Health and Nutrition Examination Study, 1999-2002. Psychoneuroendocrinology. 2019;106:171–178. doi: 10.1016/j.psyneuen.2019.03.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Asmundson GJG, Paluszek MM, Landry CA, Rachor GS, McKay D, Taylor S. Do pre-existing anxiety-related and mood disorders differentially impact COVID-19 stress responses and coping? J Anxiety Disord. 2020;74:102271. doi: 10.1016/j.janxdis.2020.102271 [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Florido R, Kwak L, Lazo M, et al. Six-year changes in physical activity and the risk of incident heart failure: ARIC study. Circulation. 2018;137(20):2142–2151. doi: 10.1161/circulationaha.117.030226 [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Yoon M, Yang PS, Jin MN, et al. Association of physical activity level with risk of dementia in a nationwide cohort in Korea. JAMA Netw Open. 2021;4(12):e2138526. doi: 10.1001/jamanetworkopen.2021.38526 [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Pisani S, Mueller C, Huntley J, Aarsland D, Kempton MJ. A meta-analysis of randomised controlled trials of physical activity in people with Alzheimer’s disease and mild cognitive impairment with a comparison to donepezil. Int J Geriatr Psychiatry. 2021;36(10):1471–1487. doi: 10.1002/gps.5581 [DOI] [PubMed] [Google Scholar]
34. Erickson KI, Voss MW, Prakash RS, et al. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci USA. 2011;108(7):3017–3022. doi: 10.1073/pnas.1015950108 [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Arem H, Moore SC, Patel A, et al. Leisure time physical activity and mortality: a detailed pooled analysis of the dose-response relationship. JAMA Intern Med. 2015;175(6):959–967. doi: 10.1001/jamainternmed.2015.0533 [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Thuc DC. The influence of physical activities on biological age parameters of females from 17 to 18 years old. J Sports Med Ther. 2018;3:75–79. doi: 10.29328/journal.jsmt.1001030 [DOI] [Google Scholar]
37. Prysiazhniuk SP, Iermakov S, Oleniev D, Eider J, Mazurok N. Influence of weekly physical exercises on indicators of biological age of student’s youth. Phys Educ Stud. 2021;25(1):58–66. doi: 10.15561/20755279.2021.0108 [DOI] [Google Scholar]
38. Valente C, Andrade R, Alvarez L, Rebelo-Marques A, Stamatakis E, Espregueira-Mendes J. Effect of physical activity and exercise on telomere length: systematic review with meta-analysis. J Am Geriatr Soc. 2021;69(11):3285–3300. doi: 10.1111/jgs.17334 [DOI] [PubMed] [Google Scholar]
39. Werner CM, Hecksteden A, Morsch A, et al. Differential effects of endurance, interval, and resistance training on telomerase activity and telomere length in a randomized, controlled study. Eur Heart J. 2019;40(1):34–46. doi: 10.1093/eurheartj/ehy585 [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Werner C, Hanhoun M, Widmann T, et al. Effects of physical exercise on myocardial telomere-regulating proteins, survival pathways, and apoptosis. J Am Coll Cardiol. 2008;52(6):470–482. doi: 10.1016/j.jacc.2008.04.034 [DOI] [PubMed] [Google Scholar]
41. Ludlow AT, Witkowski S, Marshall MR, et al. Chronic exercise modifies age-related telomere dynamics in a tissue-specific fashion. J Gerontol A Biol Sci Med Sci. 2012;67(9):911–926. doi: 10.1093/gerona/gls002 [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Barnes RP, Fouquerel E, Opresko PL. The impact of oxidative DNA damage and stress on telomere homeostasis. Mech Ageing Dev. 2019;177:37–45. doi: 10.1016/j.mad.2018.03.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Kadi F, Ponsot E. The biology of satellite cells and telomeres in human skeletal muscle: effects of aging and physical activity. Scand J Med Sci Sports. 2010;20(1):39–48. doi: 10.1111/j.1600-0838.2009.00966.x [DOI] [PubMed] [Google Scholar]
44. Ahola K, Sirén I, Kivimäki M, et al. Work-related exhaustion and telomere length: a population-based study. PLoS One. 2012;7(7):e40186. doi: 10.1371/journal.pone.0040186 [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Freni-Sterrantino A, Fiorito G, D’Errico A, et al. Work-related stress and well-being in association with epigenetic age acceleration: a Northern Finland birth cohort 1966 study. Aging. 2022;14(3):1128–1156. doi: 10.18632/aging.203872 [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Harvey SB, Modini M, Joyce S, et al. Can work make you mentally ill? A systematic meta-review of work-related risk factors for common mental health problems. Occup Environ Med. 2017;74(4):301–310. doi: 10.1136/oemed-2016-104015 [DOI] [PubMed] [Google Scholar]
47. Virtanen M, Ferrie JE, Gimeno D, et al. Long working hours and sleep disturbances: the Whitehall II prospective cohort study. Sleep. 2009;32(6):737–745. doi: 10.1093/sleep/32.6.737 [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Virtanen M, Magnusson Hansson L, Goldberg M, et al. Long working hours, anthropometry, lung function, blood pressure and blood-based biomarkers: cross-sectional findings from the CONSTANCES study. J Epidemiol Community Health. 2019;73(2):130–135. doi: 10.1136/jech-2018-210943 [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Levine ME, Lu AT, Quach A, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging. 2018;10(4):573–591. doi: 10.18632/aging.101414 [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Sudlow C, Gallacher J, Allen N, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12(3):e1001779. doi: 10.1371/journal.pmed.1001779 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

glad064_suppl_Supplementary_Material

Click here for additional data file.^{(1.6MB, pdf)}

Data Availability Statement

Data from UK Biobank are available per the researchers request (https://www.ukbiobank.ac.uk/enable-your-research/apply-for-access).

[CIT0001] 1.US Department of Health and Human Services. Physical Activity Guidelines for Americans. 2nd ed. US Department of Health and Human Services; 2018. [Google Scholar]

[CIT0002] 2. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012;380(9838):219–229. doi: 10.1016/s0140-6736(12)61031-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0003] 3. Lavie CJ, Ozemek C, Carbone S, Katzmarzyk PT, Blair SN. Sedentary behavior, exercise, and cardiovascular health. Circ Res. 2019;124(5):799–815. doi: 10.1161/circresaha.118.312669 [DOI] [PubMed] [Google Scholar]

[CIT0004] 4. Egan B, Zierath JR. Exercise metabolism and the molecular regulation of skeletal muscle adaptation. Cell Metab. 2013;17(2):162–184. doi: 10.1016/j.cmet.2012.12.012 [DOI] [PubMed] [Google Scholar]

[CIT0005] 5. Hannukainen JC, Borra R, Linderborg K, et al. Liver and pancreatic fat content and metabolism in healthy monozygotic twins with discordant physical activity. J Hepatol. 2011;54(3):545–552. doi: 10.1016/j.jhep.2010.07.029 [DOI] [PubMed] [Google Scholar]

[CIT0006] 6. Stefan N. Causes, consequences, and treatment of metabolically unhealthy fat distribution. Lancet Diabetes Endocrinol. 2020;8(7):616–627. doi: 10.1016/S2213-8587(20)30110-8 [DOI] [PubMed] [Google Scholar]

[CIT0007] 7. Church TS, Blair SN, Cocreham S, et al. Effects of aerobic and resistance training on hemoglobin A1c levels in patients with type 2 diabetes: a randomized controlled trial. JAMA. 2010;304(20):2253–2262. doi: 10.1001/jama.2010.1710 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0008] 8. Zelko A, Bukova A, Kolarcik P, et al. A randomized controlled trial to evaluate utilization of physical activity recommendations among patients of cardiovascular healthcare centres in Eastern Slovakia: study design and rationale of the AWATAR study. BMC Public Health. 2018;18(1):454. doi: 10.1186/s12889-018-5349-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0009] 9. Zhao G, Li C, Ford ES, et al. Leisure-time aerobic physical activity, muscle-strengthening activity and mortality risks among US adults: the NHANES linked mortality study. Br J Sports Med. 2014;48(3):244–249. doi: 10.1136/bjsports-2013-092731 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0010] 10. Zhao MV, Magnussen CG, Xi B. Recommended physical activity and all cause and cause specific mortality in US adults: prospective cohort study. BMJ. 2020;370:m2031. doi: 10.1136/bmj.m2031 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0011] 11. Schoenborn CA, Stommel M. Adherence to the 2008 adult physical activity guidelines and mortality risk. Am J Prev Med. 2011;40(5):514–521. doi: 10.1016/j.amepre.2010.12.029 [DOI] [PubMed] [Google Scholar]

[CIT0012] 12. Kamada M, Shiroma EJ, Buring JE, Miyachi M, Lee IM. Strength training and all-cause, cardiovascular disease, and cancer mortality in older women: a cohort study. J Am Heart Assoc. 2017;6(11):e007677. doi: 10.1161/JAHA.117.007677 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0013] 13. Liu Z, Kuo PL, Horvath S, Crimmins E, Ferrucci L, Levine M. A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: a cohort study. PLoS Med. 2018;15(12):e1002718. doi: 10.1371/journal.pmed.1002718 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0014] 14. Wu JW, Yaqub A, Ma Y, et al. Biological age in healthy elderly predicts aging-related diseases including dementia. Sci Rep. 2021;11(1):15929. doi: 10.1038/s41598-021-95425-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0015] 15. Lopez-Otin C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging. Cell. 2013;153(6):1194–1217. doi: 10.1016/j.cell.2013.05.039 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0016] 16. Shay JW. Role of telomeres and telomerase in aging and cancer. Cancer Discov. 2016;6(6):584–593. doi: 10.1158/2159-8290.CD-16-0062 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0017] 17. Sanders JL, Newman AB. Telomere length in epidemiology: a biomarker of aging, age-related disease, both, or neither? Epidemiol Rev. 2013;35:112–131. doi: 10.1093/epirev/mxs008 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0018] 18. Mons U, Muezzinler A, Schottker B, et al. Leukocyte telomere length and all-cause, cardiovascular disease, and cancer mortality: results from individual-participant-data meta-analysis of 2 large prospective cohort studies. Am J Epidemiol. 2017;185(12):1317–1326. doi: 10.1093/aje/kww210 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0019] 19. Bischoff C, Petersen HC, Graakjaer J, et al. No association between telomere length and survival among the elderly and oldest old. Epidemiology. 2006;17(2):190–194. doi: 10.1097/01.ede.0000199436.55248.10 [DOI] [PubMed] [Google Scholar]

[CIT0020] 20. Li C, Stoma S, Lotta LA, et al. Genome-wide association analysis in humans links nucleotide metabolism to leukocyte telomere length. Am J Hum Genet. 2020;106(3):389–404. doi: 10.1016/j.ajhg.2020.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0021] 21. Biobank U. Data Providers and Dates of Data Availability . https://biobank.ctsu.ox.ac.uk/crystal/exinfo.cgi?src=Data_providers_and_dates

[CIT0022] 22. Welsh S, Peakman T, Sheard S, Almond R. Comparison of DNA quantification methodology used in the DNA extraction protocol for the UK Biobank cohort. BMC Genomics. 2017;18(1):26. doi: 10.1186/s12864-016-3391-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0023] 23. Codd V, Denniff M, Swinfield C, et al. Measurement and initial characterization of leukocyte telomere length in 474,074 participants in UK Biobank. Nat Aging. 2022;2(2):170–179. doi: 10.1038/s43587-021-00166-9 [DOI] [PubMed] [Google Scholar]

[CIT0024] 24. Thygesen SK, Christiansen CF, Christensen S, Lash TL, Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson Comorbidity Index conditions in the population-based Danish National Registry of Patients. BMC Med Res Methodol. 2011;11:83. doi: 10.1186/1471-2288-11-83 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0025] 25. Jannasch F, Riordan F, Andersen LF, Schulze MB. Exploratory dietary patterns: a systematic review of methods applied in pan-European studies and of validation studies. Br J Nutr. 2018;120(6):601–611. doi: 10.1017/s0007114518001800 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0026] 26. White T, Westgate K, Hollidge S, et al. Estimating energy expenditure from wrist and thigh accelerometry in free-living adults: a doubly labelled water study. Int J Obes (Lond). 2019;43(11):2333–2342. doi: 10.1038/s41366-019-0352-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0027] 27. Ying K, Zhai R, Pyrkov TV, et al. Genetic and phenotypic analysis of the causal relationship between aging and COVID-19. Commun Med. 2021;1(1):35. doi: 10.1038/s43856-021-00033-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0028] 28. Parker DC, Bartlett BN, Cohen HJ, et al. Association of blood chemistry quantifications of biological aging with disability and mortality in older adults. J Gerontol A Biol Sci Med Sci. 2020;75(9):1671–1679. doi: 10.1093/gerona/glz219 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0029] 29. Hastings WJ, Shalev I, Belsky DW. Comparability of biological aging measures in the National Health and Nutrition Examination Study, 1999-2002. Psychoneuroendocrinology. 2019;106:171–178. doi: 10.1016/j.psyneuen.2019.03.012 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0030] 30. Asmundson GJG, Paluszek MM, Landry CA, Rachor GS, McKay D, Taylor S. Do pre-existing anxiety-related and mood disorders differentially impact COVID-19 stress responses and coping? J Anxiety Disord. 2020;74:102271. doi: 10.1016/j.janxdis.2020.102271 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0031] 31. Florido R, Kwak L, Lazo M, et al. Six-year changes in physical activity and the risk of incident heart failure: ARIC study. Circulation. 2018;137(20):2142–2151. doi: 10.1161/circulationaha.117.030226 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0032] 32. Yoon M, Yang PS, Jin MN, et al. Association of physical activity level with risk of dementia in a nationwide cohort in Korea. JAMA Netw Open. 2021;4(12):e2138526. doi: 10.1001/jamanetworkopen.2021.38526 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0033] 33. Pisani S, Mueller C, Huntley J, Aarsland D, Kempton MJ. A meta-analysis of randomised controlled trials of physical activity in people with Alzheimer’s disease and mild cognitive impairment with a comparison to donepezil. Int J Geriatr Psychiatry. 2021;36(10):1471–1487. doi: 10.1002/gps.5581 [DOI] [PubMed] [Google Scholar]

[CIT0034] 34. Erickson KI, Voss MW, Prakash RS, et al. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci USA. 2011;108(7):3017–3022. doi: 10.1073/pnas.1015950108 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0035] 35. Arem H, Moore SC, Patel A, et al. Leisure time physical activity and mortality: a detailed pooled analysis of the dose-response relationship. JAMA Intern Med. 2015;175(6):959–967. doi: 10.1001/jamainternmed.2015.0533 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0036] 36. Thuc DC. The influence of physical activities on biological age parameters of females from 17 to 18 years old. J Sports Med Ther. 2018;3:75–79. doi: 10.29328/journal.jsmt.1001030 [DOI] [Google Scholar]

[CIT0037] 37. Prysiazhniuk SP, Iermakov S, Oleniev D, Eider J, Mazurok N. Influence of weekly physical exercises on indicators of biological age of student’s youth. Phys Educ Stud. 2021;25(1):58–66. doi: 10.15561/20755279.2021.0108 [DOI] [Google Scholar]

[CIT0038] 38. Valente C, Andrade R, Alvarez L, Rebelo-Marques A, Stamatakis E, Espregueira-Mendes J. Effect of physical activity and exercise on telomere length: systematic review with meta-analysis. J Am Geriatr Soc. 2021;69(11):3285–3300. doi: 10.1111/jgs.17334 [DOI] [PubMed] [Google Scholar]

[CIT0039] 39. Werner CM, Hecksteden A, Morsch A, et al. Differential effects of endurance, interval, and resistance training on telomerase activity and telomere length in a randomized, controlled study. Eur Heart J. 2019;40(1):34–46. doi: 10.1093/eurheartj/ehy585 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0040] 40. Werner C, Hanhoun M, Widmann T, et al. Effects of physical exercise on myocardial telomere-regulating proteins, survival pathways, and apoptosis. J Am Coll Cardiol. 2008;52(6):470–482. doi: 10.1016/j.jacc.2008.04.034 [DOI] [PubMed] [Google Scholar]

[CIT0041] 41. Ludlow AT, Witkowski S, Marshall MR, et al. Chronic exercise modifies age-related telomere dynamics in a tissue-specific fashion. J Gerontol A Biol Sci Med Sci. 2012;67(9):911–926. doi: 10.1093/gerona/gls002 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0042] 42. Barnes RP, Fouquerel E, Opresko PL. The impact of oxidative DNA damage and stress on telomere homeostasis. Mech Ageing Dev. 2019;177:37–45. doi: 10.1016/j.mad.2018.03.013 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0043] 43. Kadi F, Ponsot E. The biology of satellite cells and telomeres in human skeletal muscle: effects of aging and physical activity. Scand J Med Sci Sports. 2010;20(1):39–48. doi: 10.1111/j.1600-0838.2009.00966.x [DOI] [PubMed] [Google Scholar]

[CIT0044] 44. Ahola K, Sirén I, Kivimäki M, et al. Work-related exhaustion and telomere length: a population-based study. PLoS One. 2012;7(7):e40186. doi: 10.1371/journal.pone.0040186 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0045] 45. Freni-Sterrantino A, Fiorito G, D’Errico A, et al. Work-related stress and well-being in association with epigenetic age acceleration: a Northern Finland birth cohort 1966 study. Aging. 2022;14(3):1128–1156. doi: 10.18632/aging.203872 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0046] 46. Harvey SB, Modini M, Joyce S, et al. Can work make you mentally ill? A systematic meta-review of work-related risk factors for common mental health problems. Occup Environ Med. 2017;74(4):301–310. doi: 10.1136/oemed-2016-104015 [DOI] [PubMed] [Google Scholar]

[CIT0047] 47. Virtanen M, Ferrie JE, Gimeno D, et al. Long working hours and sleep disturbances: the Whitehall II prospective cohort study. Sleep. 2009;32(6):737–745. doi: 10.1093/sleep/32.6.737 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0048] 48. Virtanen M, Magnusson Hansson L, Goldberg M, et al. Long working hours, anthropometry, lung function, blood pressure and blood-based biomarkers: cross-sectional findings from the CONSTANCES study. J Epidemiol Community Health. 2019;73(2):130–135. doi: 10.1136/jech-2018-210943 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0049] 49. Levine ME, Lu AT, Quach A, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging. 2018;10(4):573–591. doi: 10.18632/aging.101414 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CIT0050] 50. Sudlow C, Gallacher J, Allen N, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12(3):e1001779. doi: 10.1371/journal.pmed.1001779 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

The Association of Physical Activity Behaviors and Patterns With Aging Acceleration: Evidence From the UK Biobank

Jianwei Zhu, PhD

Yao Yang, MM

Yu Zeng, MM

Xin Han, PhD

Wenwen Chen, PhD

Yao Hu, MS

Yuanyuan Qu, MA

Huazhen Yang, MM

Unnur A Valdimarsdóttir, PhD

Fang Fang, PhD

Huan Song, MD, PhD